Working T13 Draft 1321D

Working T13 Draft 1321D
Working
Draft
T13
1321D
Revision 3
29 February 2000
Information Technology AT Attachment
with Packet Interface - 5
(ATA/ATAPI-5)
This is an internal working document of T13, a Technical Committee of Accredited Standards Committee
NCITS. As such, this is not a completed standard and has not been approved. The contents may be modified
by the T13 Technical Committee. This document is made available for review and comment only.
Permission is granted to members of NCITS, its technical committees, and their associated task groups to
reproduce this document for the purposes of NCITS standardization activities without further permission,
provided this notice is included. All other rights are reserved. Any commercial or for-profit replication or
republication is prohibited.
T13 Technical Editor:
Peter T. McLean
Maxtor Corporation
2190 Miller Drive
Longmont, CO 80501-6744
USA
Tel:
303-678-2149
Fax: 303-682-4811
Email: pete_mclean@maxtor.com
Reference number
ANSI NCITS.*** - xxxx
Printed February, 29, 2000 10:06AM
T13/1321D revision 3
Other Points of Contact:
T13 Chair
Gene Milligan
Seagate Technology
OKM 151
10321 West Reno
Oklahoma City, OK 73127-9705
P.O. Box 12313
Oklahoma City, OK 73157-2313
Tel:
405-324-3070
Fax:
405-324-3794
T13 Vice-Chair
Pete McLean
Maxtor Corporation
2190 Miller Drive
Longmont, CO 80501
Tel:
303-678-2149
Fax:
303-682-4811
NCITS Secretariat
Administrator Standards Processing
1250 Eye Street, NW Suite 200
Washington, DC 20005
Tel:
202-737-8888
Fax:
202-638-4922
Email: NCITS@ITIC.NW.DC.US
T13 Reflector
Internet address for subscription to the T13 reflector: majordomo@t13.org
Send email to above account and include in BODY of text, on a line by itself the following:
"subscribe T13"
Internet address for unsubscribing to the T13 reflector: majordomo@t13.org
Send email to above account and include in BODY of text, on a line by itself the following:
"unsubscribe T13"
Internet address for distribution via T13 reflector: T13@t13.org
T13 Web Site
http://www.t13.org
T13 Anonymous FTP Site
fission.dt.wdc.com
T13 directory is: "/ t13 "
T13 mailings
Global Engineering
15 Inverness Way East
Englewood, CO 80112-5704
Tel:
303-792-2181 or 800-854-7179
Fax:
303-792-2192
T13/1321D revision 3
DOCUMENT STATUS
Revision 0 - 25 August 1998
Document created from ATA/ATAPI-4-revision 17 (T13/1153Dr17). Added:
D97150R2 Power-up in Standby as amended at the 8/18/98 plenary.
D98101R2 State diagram proposal as amended at the 8/18/98 plenary.
D98102R2 Set multiple proposal as requested at the 4/14-16/98 plenary.
D98104R0 Master password revision as requested at the 6/16-18/98 plenary.
D98105R2 Clarification of nIEN and pending interrupt as amended at the 8/18/98 plenary.
D98111R0 Editorial changes requested at the 4/14-16/98 plenary.
D98112R2 New identify word as requested at the 8/18/98 plenary.
D98121R0 Proposed ATA/ATAPI-4 amendment as requested at the 8/18/98 plenary.
D98122R0 Cable, mechanical, and connector proposal as amended at the 8/18/98 plenary.
Replaced “should” with “shall in 6.8.2 and removed parenthetical phrase per 8/18/98 plenary.
Made ATA-1 bit obsolete in IDENTIFY DEVICE word 80 per 8/18/98 plenary.
Revision 0a - 25 September 1998
Made editorial changes requested during the changebar review at the 9/22-23/98 working group meeting.
Revision 0b - 18 December 1998
Added proposal D98113R3 as requested at the 10/27-29/98 plenary meeting.
Changed Power Management flow chart to a state diagram in clause 6.9.
Changed security flow chart to a state diagram in clause 6.11.
Added ANSI ATA/ATAPI-4 editor’s editorial changes.
Added proposal D98134R0 as modified at the 12/8-11/98 plenary.
Added proposal D98135R2 per 12/8-11/98 plenary.
Added proposal D98132R1 per 12/8-11/98 plenary.
Clarified tag values and queue depth per 12/8-11/98 plenary.
Clarified required registers in clause 8 per 12/8-11/98 plenary.
Made changes to clause 9 and annex A per review at 12/8-11/98 plenary.
Revision 0c - 5 March 1999
Added proposals:
D99101R0 Cache flush mandatory proposal
D99105R0 Self-test log modification
D99107R0 Checksum definition
D99108R0 Optional pointer on self-test log
Added insertion force into 4-pin connector description
Made changes per 2/22-26/99 plenary change bar review
Revision 1 - 11 May 1999
Added proposals:
D98133R4 Ultra DMA 66 timing
D98144R2 CFA SET FEATURES codes
D99104R1 Error log depth extension
D99112R1 Multiword DMA description
Changes requested at 4/27-30/99 plenary meeting.
T13/1321D revision 3
Revision 1a - 26 May 1999
Made editorial changes requested at the May 19-20, 1999 working group meeting.
Revision 1b - 7 July 1999
Added proposal D99122R0, new security state diagram and changed all state diagram text to 9 pt.
Added proposal D98139R3, Protected area locking proposal.
Made changes requested during June 22-25, 1999 document review.
Revision 1c – 31 August 1999
Added proposal D99126R0, CFA code reservations.
Added proposal D99127R0, Byte count = 0 handling.
Made changes requested during August 24-27, 1999 change bar review.
Revision 2 - 13 December 1999
Added proposal D99132R0, Multiword DMA timing diagram.
Added proposal D98109R6, Implementation guide – UMDA as annex D.
Added proposal D99125R0, 8-bit CFA PIO transfers with timing diagram changes.
Made changes requested at the 19-22 October 1999 meeting.
Made changes requested at the 30 November – 2 December 1999 meeting.
Revision 3 - 29 February 2000
Made changes indicated in E00102R1 ATA/ATAPI-5 Letter Ballot comment resolution and other editorial
chnges requested at the 2/22-25/00 plenary meeting.
T13/1321D revision 3
ANSI®
NCITS.***-xxxx
American National Standard
for Information Systems 
AT Attachment
with Packet Interface - 5  (ATA/ATAPI-5)
Secretariat
Information Technology Industry Council
Approved mm dd yy
American National Standards Institute, Inc.
Abstract
This standard specifies the AT Attachment Interface between host systems and storage devices. It provides a
common attachment interface for systems manufacturers, system integrators, software suppliers, and suppliers
of intelligent storage devices. It includes the Packet Command feature set implemented by devices commonly
known as ATAPI devices.
This standard maintains a high degree of compatibility with the AT Attachment Interface with Packet Interface
Extensions (ATA/ATAPI-4), NCITS 317-1998, and while providing additional functions, is not intended to require
changes to presently installed devices or existing software.
T13/1321D revision 3
American
National
Standard
Approval of an American National Standard requires verification by ANSI that the
requirements for due process, consensus, and other criteria for approval have been
met by the standards developer. Consensus is established when, in the judgment of
the ANSI Board of Standards Review, substantial agreement has been reached by
directly and materially affected interests. Substantial agreement means much more
than a simple majority, but not necessarily unanimity. Consensus requires that all
views and objections be considered, and that effort be made towards their resolution.
The use of American National Standards is completely voluntary; their existence
does not in any respect preclude anyone, whether he has approved the standards or
not, from manufacturing, marketing, purchasing, or using products, processes, or
procedures not conforming to the standards.
The American National Standards Institute does not develop standards and will in no
circumstances give interpretation on any American National Standard. Moreover, no
person shall have the right or authority to issue an interpretation of an American
National Standard in the name of the American National Standards Institute.
Requests for interpretations should be addressed to the secretariat or sponsor whose
name appears on the title page of this standard.
CAUTION NOTICE: This American National Standard may be revised or withdrawn at
any time. The procedures of the American National Standards Institute require that
action be taken periodically to reaffirm, revise, or withdraw this standard. Purchasers
of American National Standards may receive current information on all standards by
calling or writing the American National Standards Institute.
CAUTION: The developers of this standard have requested that holders of patents that may be
required for the implementation of the standard, disclose such patents to the publisher. However,
neither the developers nor the publisher have undertaken a patent search in order to identify
which, if any, patents may apply to this standard.
As of the date of publication of this standard and following calls for the identification of patents that
may be required for the implementation of the standard, notice of one or more such claims has
been received.
By publication of this standard, no position is taken with respect to the validity of this claim or of
any rights in connection therewith. The patent holders have, however, filed a statement of
willingness to grant a license under these rights on reasonable and nondiscriminatory terms and
conditions to applicants desiring to obtain such a license. Details may be obtained from the
publisher.
No further patent search is conducted by the developer or the publisher in respect to any standard
it processes. No representation is made or implied that licenses are not required to avoid
infringement in the use of this standard.
Published by
American National Standards Institute
11 West 42nd Street, New York, New York 10036
Copyright nnnn by American National Standards Institute
All rights reserved.
T13/1321D revision 3
Contents
Page
Foreword ........................................................................................................................................vi
Introduction.....................................................................................................................................viii
1
Scope .....................................................................................................................................1
2
Normative references ................................................................................................................1
2.1
Approved references........................................................................................................1
2.2
References under development.........................................................................................2
2.3
Other references .............................................................................................................2
3
Definitions, abbreviations, and conventions .................................................................................2
3.1
Definitions and abbreviations ............................................................................................2
3.2
Conventions ...................................................................................................................5
4
Interface physical and electrical requirements .............................................................................10
4.1
Cable configuration .........................................................................................................10
4.2
Electrical characteristics .................................................................................................10
5
Interface signal assignments and descriptions ............................................................................14
5.1
Signal summary .............................................................................................................14
5.2
Signal descriptions .........................................................................................................16
6
General operational requirements...............................................................................................20
6.1
Command delivery ..........................................................................................................20
6.2
Register delivered data transfer command sector addressing ..............................................20
6.3
Interrupts .......................................................................................................................22
6.4
General feature set .........................................................................................................22
6.5
Multiword DMA...............................................................................................................24
6.6
Ultra DMA feature set......................................................................................................25
6.7
Host determination of cable type by detecting CBLID- ........................................................27
6.8
PACKET Command feature set ........................................................................................29
6.9
Overlapped feature set.....................................................................................................30
6.10 Queued feature set .........................................................................................................31
6.11 Power Management feature set ........................................................................................32
6.12 Advanced Power Management feature set .........................................................................35
6.13 Security Mode feature set................................................................................................35
6.14 Self-monitoring, analysis, and reporting technology feature set............................................41
6.15 Host Protected Area feature set .......................................................................................42
6.16 CFA feature set ..............................................................................................................47
6.17 Removable Media Status Notification and Removable Media feature sets .............................47
6.18 Power-Up In Standby feature set ......................................................................................49
7
Interface register definitions and descriptions ..............................................................................50
7.1
Device addressing considerations.....................................................................................50
7.2
I/O register descriptions ..................................................................................................56
7.3
Alternate Status register..................................................................................................57
7.4
Command register ..........................................................................................................57
7.5
Cylinder High register......................................................................................................58
7.6
Cylinder Low register.......................................................................................................59
7.7
Data port........................................................................................................................59
7.8
Data register ..................................................................................................................60
7.9
Device Control register ....................................................................................................61
7.10 Device/Head register .......................................................................................................62
7.11 Error register ..................................................................................................................62
7.12 Features register ............................................................................................................63
7.13 Sector Count register ......................................................................................................64
7.14 Sector Number register ...................................................................................................64
7.15 Status register................................................................................................................65
8
Command descriptions .............................................................................................................68
8.1
CFA ERASE SECTORS..................................................................................................69
8.2
CFA REQUEST EXTENDED ERROR CODE .....................................................................71
Page i
T13/1321D revision 3
9
8.3
CFA TRANSLATE SECTOR ............................................................................................ 73
8.4
CFA WRITE MULTIPLE WITHOUT ERASE....................................................................... 75
8.5
CFA WRITE SECTORS WITHOUT ERASE....................................................................... 77
8.6
CHECK POWER MODE ................................................................................................. 79
8.7
DEVICE RESET............................................................................................................. 81
8.8
DOWNLOAD MICROCODE ............................................................................................. 82
8.9
EXECUTE DEVICE DIAGNOSTIC .................................................................................... 84
8.10 FLUSH CACHE .............................................................................................................. 86
8.11 GET MEDIA STATUS ..................................................................................................... 87
8.12 IDENTIFY DEVICE ......................................................................................................... 89
8.13 IDENTIFY PACKET DEVICE ........................................................................................... 109
8.14 IDLE.............................................................................................................................. 120
8.15 IDLE IMMEDIATE ........................................................................................................... 122
8.16 INITIALIZE DEVICE PARAMETERS ................................................................................. 124
8.17 MEDIA EJECT ............................................................................................................... 126
8.18 MEDIA LOCK................................................................................................................. 128
8.19 MEDIA UNLOCK ............................................................................................................ 130
8.20 NOP.............................................................................................................................. 132
8.21 PACKET........................................................................................................................ 133
8.22 READ BUFFER.............................................................................................................. 139
8.23 READ DMA ................................................................................................................... 141
8.24 READ DMA QUEUED..................................................................................................... 143
8.25 READ MULTIPLE ........................................................................................................... 147
8.26 READ NATIVE MAX ADDRESS....................................................................................... 150
8.27 READ SECTOR(S)......................................................................................................... 151
8.28 READ VERIFY SECTOR(S) ............................................................................................ 154
8.29 SECURITY DISABLE PASSWORD.................................................................................. 156
8.30 SECURITY ERASE PREPARE........................................................................................ 158
8.31 SECURITY ERASE UNIT ................................................................................................ 159
8.32 SECURITY FREEZE LOCK ............................................................................................. 162
8.33 SECURITY SET PASSWORD ......................................................................................... 164
8.34 SECURITY UNLOCK ...................................................................................................... 166
8.35 SEEK............................................................................................................................ 168
8.36 SERVICE ...................................................................................................................... 170
8.37 SET FEATURES ............................................................................................................ 171
8.38 SET MAX....................................................................................................................... 176
8.39 SET MULTIPLE MODE ................................................................................................... 187
8.40 SLEEP.......................................................................................................................... 189
8.41 SMART ......................................................................................................................... 191
8.42 STANDBY ..................................................................................................................... 218
8.43 STANDBY IMMEDIATE................................................................................................... 220
8.44 WRITE BUFFER ............................................................................................................ 221
8.45 WRITE DMA .................................................................................................................. 223
8.46 WRITE DMA QUEUED ................................................................................................... 225
8.47 WRITE MULTIPLE .......................................................................................................... 229
8.48 WRITE SECTOR(S)........................................................................................................ 232
Protocol .................................................................................................................................. 234
9.1
Power-on and hardware reset protocol .............................................................................. 237
9.2
Software reset protocol ................................................................................................... 241
9.3
Bus idle protocol ............................................................................................................ 245
9.4
Non-data command protocol............................................................................................ 256
9.5
PIO data-in command protocol......................................................................................... 258
9.6
PIO data-out command protocol....................................................................................... 262
9.7
DMA command protocol.................................................................................................. 266
9.8
PACKET command protocol............................................................................................ 269
9.9
READ/WRITE DMA QUEUED command protocol.............................................................. 282
Page ii
T13/1321D revision 3
9.10 EXECUTE DEVICE DIAGNOSTIC command protocol.........................................................286
9.11 DEVICE RESET command protocol .................................................................................290
9.12 Signature and persistence...............................................................................................291
9.13 Ultra DMA data-in commands ..........................................................................................292
9.14 Ultra DMA data-out commands ........................................................................................295
9.15 Ultra DMA CRC rules ......................................................................................................298
9.16 Single device configurations .............................................................................................299
10 Timing.....................................................................................................................................300
10.1 Deskewing .....................................................................................................................300
10.2 Transfer timing................................................................................................................301
Tables
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Page
Byte order..................................................................................................................................10
Byte order..................................................................................................................................10
DC characteristics ......................................................................................................................11
AC characteristics ......................................................................................................................11
Driver types and required termination ............................................................................................12
Typical series termination for Ultra DMA .......................................................................................14
Interface signal name assignments...............................................................................................15
Host detection of CBLID- .............................................................................................................29
Security mode command actions .................................................................................................40
Device repsonse to DOIW-/DOIR- ................................................................................................51
Device is not selected, DMACK- is not asserted ...........................................................................52
Device is selected, DMACK- is not asserted.................................................................................53
Device is selected, DMACK- is asserted (for Multiword DMA only)..................................................54
Device 1 is selected and Device 0 is responding for Device 1..........................................................55
Device is in Sleep mode, DEVICE RESET is not implemented, DMACK- is not asserted..................56
Device is in Sleep mode, DEVICE RESET is implemented, DMACK- is not asserted .......................56
Extended error codes .................................................................................................................72
CFA TRANSLATE SECTOR information.......................................................................................74
Diagnostic codes .......................................................................................................................85
IDENTIFY DEVICE information ....................................................................................................92
Minor revision number.................................................................................................................103
IDENTIFY PACKET DEVICE information ......................................................................................111
Automatic standby timer periods .................................................................................................121
Security password content..........................................................................................................158
SECURITY ERASE UNIT password .............................................................................................162
SECURITY SET PASSWORD data content..................................................................................165
Identifier and security level bit interaction......................................................................................166
SET FEATURES register definitions ............................................................................................173
Transfer/mode values..................................................................................................................174
Advanced power management levels............................................................................................175
SET MAX Features register values...............................................................................................177
SET MAX SET PASSWORD data content....................................................................................181
SMART Feature register values ...................................................................................................191
SMART EXECUTE OFF-LINE IMMEDIATE Sector Number register values ......................................199
Device SMART data structure .....................................................................................................203
Off-line data collection status byte values .....................................................................................203
Self-test execution status byte values ..........................................................................................204
Log address definition.................................................................................................................206
SMART log directory ..................................................................................................................208
SMART error log sector ..............................................................................................................208
Error log data structure...............................................................................................................209
Command data structure ............................................................................................................209
Page iii
T13/1321D revision 3
43
44
45
46
47
48
49
50
51
Error data structure.................................................................................................................... 210
State field values ....................................................................................................................... 210
Self-test log data structure.......................................................................................................... 211
Self-test log descriptor entry ....................................................................................................... 212
Equations for parallel generation of a CRC polynomial ................................................................... 299
Register transfer to/from device................................................................................................... 303
PIO data transfer to/from device .................................................................................................. 305
Multiword DMA data transfer....................................................................................................... 306
Ultra DMA data burst timing requirements.................................................................................... 312
Figures
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Page
State diagram convention ............................................................................................................ 8
Ultra DMA termination with pull-up or pull-down ............................................................................. 14
PDIAG- example using an 80-conductor cable assembly................................................................ 18
Cable select example ................................................................................................................. 19
Alternate cable select example.................................................................................................... 19
Example configuration of a system with a 40-conductor cable......................................................... 28
Example configuration of a system where the host detects a 40-conductor cable ............................. 28
Example configuration of a system where the host detects an 80-conductor cable............................ 29
Power management state diagram ............................................................................................... 33
Security mode state diagram ...................................................................................................... 37
SET MAX security state diagram................................................................................................. 45
Overall host protocol state sequence........................................................................................... 235
Overall device protocol state sequence ........................................................................................ 236
Host power-on or hardware reset state diagram ............................................................................ 237
Device power-on or hardware reset state diagram.......................................................................... 238
Host software reset state diagram ............................................................................................... 241
Device 0 software reset state diagram.......................................................................................... 242
Device 1 software reset state diagram.......................................................................................... 244
Host bus idle state diagram ........................................................................................................ 246
Additional host bus idle state diagram with overlap or overlap and queuing ...................................... 248
Device bus idle state diagram ..................................................................................................... 251
Additional device bus idle state diagram with overlap or overlap and queuing.................................... 253
Host non-data state diagram....................................................................................................... 257
Device non-data state diagram .................................................................................................... 257
Host PIO data-in state diagram ................................................................................................... 259
Device PIO data-in state diagram ................................................................................................ 261
Host PIO data-out state diagram ................................................................................................. 263
Device PIO data-out state diagram .............................................................................................. 265
Host DMA state diagram ............................................................................................................ 267
Device DMA state diagram ......................................................................................................... 268
Host PACKET non-data and PIO data command state diagram...................................................... 270
Device PACKET non-data and PIO data command state diagram................................................... 273
Host PACKET DMA command state diagram............................................................................... 276
Device PACKET DMA command state diagram ............................................................................ 279
Host DMA QUEUED state diagram ............................................................................................. 282
Device DMA QUEUED command state diagram ........................................................................... 284
Host EXECUTE DEVICE DIAGNOSTIC state diagram ................................................................... 286
Device 0 EXECUTE DEVICE DIAGNOSTIC state diagram.............................................................. 287
Device 1 EXECUTE DEVICE DIAGNOSTIC command state diagram .............................................. 289
Host DEVICE RESET command state diagram ............................................................................ 290
Device DEVICE RESET command state diagram.......................................................................... 291
Example parallel CRC generator.................................................................................................. 299
Register transfer to/from device................................................................................................... 302
PIO data transfer to/from device .................................................................................................. 304
Page iv
T13/1321D revision 3
45
46
47
48
49
50
51
52
53
54
55
56
57
58
Initiating a Multiword DMA data transfer .......................................................................................307
Sustaining a Multiword DMA data transfer ...................................................................................308
Device terminating a Multiword DMA data transfer........................................................................309
Host terminating a Multiword DMA data transfer...........................................................................310
Initiating an Ultra DMA data-in burst.............................................................................................313
Sustained Ultra DMA data-in burst...............................................................................................314
Host pausing an Ultra DMA data-in burst......................................................................................315
Device terminating an Ultra DMA data-in burst ..............................................................................316
Host terminating an Ultra DMA data-in burst.................................................................................317
Initiating an Ultra DMA data-out burst...........................................................................................318
Sustained Ultra DMA data-out burst.............................................................................................319
Device pausing an Ultra DMA data-out burst.................................................................................320
Host terminating an Ultra DMA data-out burst...............................................................................321
Device terminating an Ultra DMA data-out burst ............................................................................322
Annexes
A
B
C
D
E
F
Page
Connectors and cable assemblies ................................................................................................323
Device determination of cable type................................................................................................343
Identify device data for devices with more than 1024 logical cylinders ...............................................346
Signal integrity and UDMA implementation guide ...........................................................................349
Bibliography ................................................................................................................................392
ATA command set summary ........................................................................................................393
Page v
T13/1321D revision 3
Foreword
(This foreward is not part of American National Standard NCITS ***-****.)
This AT Attachment with Packet Interface -5 (ATA/ATAPI-5) standard is designed to maintain a high degree of
compatibility with the AT Attachment with Packet Interface (ATA/ATAPI-4) standard.
This standard was developed by the ATA ad hoc working group of Accredited Standards Committee NCITS
during 1998-99. The standards approval process started in 1999. This document includes annexes that are
informative and are not considered part of the standard.
Requests for interpretation, suggestions for improvement and addenda, or defect reports are welcome. They
should be sent to the NCITS Secretariat, Information Technology Industry Council, 1250 Eye Street, NW, Suite
200, Washington, DC 20005-3922.
This standard was processed and approved for submittal to ANSI by Accredited Standards Committee on
Information Processing Systems, NCITS. Committee approval of the standard does not necessarily imply that
all committee members voted for approval. At the time it approved this standard, the NCITS Committee had the
following members:
Karen Higginbottom, Chair
(Vacant), Vice-Chair
Monica Vago, Secretary
Organization Represented...................................................................... Name of Representative
AMP, Inc .............................................................................................. John Hill
Charles Brill (Alt.)
Apple Computer .................................................................................... David Michael
Jerry Kellenbenz (Alt.)
AT&T ................................................................................................... Thomas Frost
Paul Bartoli (Alt.)
Bull HN Information Systems, Inc. .......................................................... Patrick L. Harris
Compaq Computer Corporation............................................................... Steven Heil
Seve Park (Alt.)
Eastman Kodak .................................................................................... Michael Nier
Hewlett-Packard.................................................................................... Karen Higginbottom
Donald Loughry (Alt.)
Hitachi America, Ltd. ............................................................................. John Neumann
Kei Yamashita (Alt.)
Hughes Aircraft Company ...................................................................... Harold L. Zebrack
IBM Corporation .................................................................................... Ron Silletti
Joel Urman (Alt.)
Institute for Certification of Computer Professionals .................................. Kenneth M. Zemrowski
Tom Kurihara (Alt.)
Lucent Technologies, Inc. ...................................................................... Herbert Bertine
Tom Rutt (Alt.)
National Communications Systems ........................................................ Dennis Bodson
Frack McClelland (Alt.)
National Institute of Standards and Technology ........................................ Michael Hogan
Bruce K. Rosen (Alt.)
Panasonic Technologies, Inc.................................................................. Judson Hofmann
Terry J. Nelson (Alt.)
Share, Inc............................................................................................. David Thewlis
Gary Ainsworth (Alt.)
Sony Electronics, Inc. ........................................................................... Masataka Ogawa
Michael Deese (Alt.)
Page vi
T13/1321D revision 3
Organization Represented ......................................................................Name of Representative
Storage Technology Corporation .............................................................Joseph S. Zajaczkowski
Sun Microsystems, Inc. .........................................................................Gary Robinson
Sybase, Inc. .........................................................................................Donald Deutsch
Andrew Eisenberg (Alt.)
Texas Instruments, Inc...........................................................................Clyde Camp
Fritz Whittington (Alt.)
Unisys Corporation ................................................................................Arnold F. Winkler
Stephen P. Oksala (Alt.)
U.S. Department of Defense/DISA...........................................................Jerry L. Smith
C. J. Pasquariello (Alt.)
U.S. Department of Energy.....................................................................Carol Blackston
Bruce R. White (Alt.)
Xerox Corporation ..................................................................................John B. Flannery
Jean Baroness (Alt.)
Subcommittee T13 on ATA Interfaces, that reviewed this standard, had the following members:
Gene Milligan, Chairman
Pete McLean, Vice-Chairman
Dan Colegrove, Secretary
Amy Barton
Darrin Bulik
Litko Chan
Ben Chang
Dan Colegrove
Tom Colligan
David Dickson
Greg Elkins
Mark Evans
Tony Goodfellow
Tasuku Kasebayashi
Hale Landis
Ming Louie
Pete McLean
Gene Milligan
Masataka Ogawa
Darrell Redford
Ron Roberts
Matt Rooke
Bob Salem
Curtis Stevens
Tim Thompson
Anthony Yang
Ken Bovatsek [Alternate]
Tim Bradshaw [Alternate]
Andy Chen [Alternate]
Renee Depew [Alternate]
Tom Hanan [Alternate
Richard Harcourt [Alternate]
LeRoy Leach [Alternate]
Wen Lin [Alternate]
James McGrath [Alternate]
Kha Nguyen [Alternate]
Marc Noblitt [Alternate]
Yogi Schaffner [Alternate]
Paresh Sheth [Alternate]
Ron Stephens [Alternate]
Seiro Taniyama [Alternate]
Tokuyuki Totani [Alternate]
Tri Van [Alternate]
Quang Vuong [Alternate]
Sam Wong [Alternate]
ATA/ATAPI ad hoc Working Group, that developed this standard, had the following additional participants:
Charles Brill
Mike Christensen
Michael Eschmann
Jon Haines
Jonathan Hanmann
Jim Hatfield
Richard Kalish
Eric Kvamme
Lawrence Lamers
Raymond Liu
Kent Pryor
Paul Raikunen
Page vii
T13/1321D revision 3
Introduction
This standard encompasses the following:
Clause 1 describes the scope.
Clause 2 provides normative references.
Clause 3 provides definitions, abbreviations, and conventions used within this document.
Clause 4 contains the electrical and mechanical characteristics; covering the interface cabling
requirements of the interface and DC cables and connectors.
Clause 5 contains the signal descriptions of the AT Attachment Interface.
Clause 6 describes the general operating requirements of the AT Attachment Interface.
Clause 7 contains descriptions of the registers of the AT Attachment Interface.
Clause 8 contains descriptions of the commands of the AT Attachment Interface.
Clause 9 contains the protocol of the AT Attachment Interface.
Clause 10 contains the interface timing diagrams.
Page viii
T13/1321D revision 3
AMERICAN NATIONAL STANDARD
NCITS.***-nnnn
American National Standard
for Information Systems 
Information Technology 
AT Attachment with Packet Interface - 5  (ATA/ATAPI-5)
1 Scope
This standard specifies the AT Attachment Interface between host systems and storage devices. It provides a
common attachment interface for systems manufacturers, system integrators, software suppliers, and suppliers
of intelligent storage devices.
The application environment for the AT Attachment Interface is any host system that has storage devices
contained within the processor enclosure.
This standard defines the connectors and cables for physical interconnection between host and storage device,
as well as the electrical and logical characteristics of the interconnecting signals. It also defines the operational
registers within the storage device, and the commands and protocols for the operation of the storage device.
This standard maintains a high degree of compatibility with the AT Attachment with Packet Interface
Extensions standard (ATA/ATAPI-4), NCITS 317-1998, and while providing additional functions, is not intended
to require changes to presently installed devices or existing software.
2 Normative references
The following standards contain provisions that, through reference in the text, constitute provisions of this
standard. At the time of publication, the editions indicated were valid. All standards are subject to revision, and
parties to agreements based on this standard are encouraged to investigate the possibility of applying the most
recent editions of the standards listed below.
Copies of the following documents can be obtained from ANSI: Approved ANSI standards, approved and draft
international and regional standards (ISO, IEC, CEN/CENELEC, ITUT), and approved and draft foreign standards
(including BSI, JIS, and DIN). For further information, contact ANSI Customer Service Department at 212-6424900 (phone), 212-302-1286 (fax), or via the World Wide Web at http://www.ansi.org.
Additional availability contact information is provided below as needed.
2.1 Approved references
The following approved ANSI standards, approved international and regional standards (ISO, IEC,
CEN/CENELEC, ITUT), may be obtained from the international and regional organizations who control them.
SCSI-3 Primary Commands (SPC)
SCSI-3 Multimedia Commands (MMC)
[NCITS 301:1997] (PACKET command feature set device types)
[NCITS 304:1997] (PACKET command feature set sense codes)
Page 1
T13/1321D revision 3
To obtain copies of these documents, contact Global Engineering or NCITS.
2.2
References under development
At the time of publication, the following referenced standards were still under development. For information on
the current status of the document, or regarding availability, contact the relevant standards body or other
organization as indicated.
SCSI Primary Commands - 2 (SPC-2)
SCSI Primary Commands - 3 (SPC-3)
Multimedia Commands - 2 (MMC-2)
Multimedia Commands - 3 (MMC-3)
[T10/1236-D] (PACKET command feature set commands)
[T10/1416-D] (PACKET command feature set commands)
[T10/1228-D] (PACKET command feature set commands)
[T10/1363-D] (PACKET command feature set commands)
For more information on the current status of the above documents, contact NCITS. To obtain copies of these
documents, contact Global Engineering or NCITS.
2.3
Other references
The following standard and specifications were also referenced.
PC Card Standard , February 1995, PCMCIA (68-pin Connector)
For the PC Card Standard published by the Personal Computer Memory Card International Association,
contact PCMCIA at 408-433-2273.
CompactFlash Association Specification, Revision 1.4
For the CompactFlash Association Specification published by the CompactFlash Association, contact the
CompactFlash Association at http://www.compactflash.org.
3 Definitions, abbreviations, and conventions
3.1 Definitions and abbreviations
For the purposes of this standard, the following definitions apply:
3.1.1 ATA (AT Attachment): ATA defines the physical, electrical, transport, and command protocols for the
internal attachment of storage devices.
3.1.2 ATA-1 device: A device that complied with ANSI X3.221-1994, the AT Attachment Interface for Disk
Drives. ANSI X3.221-1994 has been withdrawn.
3.1.3 ATA-2 device: A device that complies with ANSI X3.279-1996, the AT Attachment Interface with
Extensions.
3.1.4 ATA-3 device: A device that complies with ANSI X3.298-1997, the AT Attachment-3 Interface.
3.1.5 ATA/ATAPI-4 device: A device that complies with ANSI NCITS 317-1998, AT Attachment Interface with
Packet Interface Extensions.
3.1.6 ATA/ATAPI-5 device: A device that complies with this standard.
3.1.7 ATAPI (AT Attachment Packet Interface) device: A device implementing the Packet Command feature
set.
Page 2
T13/1321D revision 3
3.1.8 bus release: For devices implementing overlap, the term bus release is the act of clearing both DRQ and
BSY to zero before the action requested by the command is completed to allow the host to select
the other device.
3.1.9 byte count: The value placed in the Byte Count register by the device to indicate the number of bytes to
be transferred under this DRQ assertion when executing a PACKET PIO data transfer command.
3.1.10 byte count limit: The value placed in the Byte Count register by the host as input to a PACKET PIO
data transfer command to indicate the maximum byte count that may be transferred under a single
DRQ assertion.
3.1.11 CFA: The CompactFlash Association that created the specification for compact flash memory that
uses the ATA interface.
3.1.12 check condition: For devices implementing the PACKET Command feature set, this indicates an error
or exception condition has occurred.
3.1.13 CHS (cylinder-head-sector): This term defines the addressing of the data on the device by cylinder
number, head number, and sector number.
3.1.14 command aborted: Command completion with ABRT set to one in the Error register and ERR set to
one in the Status register.
3.1.15 command acceptance: A command is considered accepted whenever the currently selected device
has the BSY bit cleared to zero in the Status register and the host writes to the Command register.
An exception exists for the EXECUTE DEVICE DIAGNOSTIC (see 8.9) and DEVICE RESET
commands (see 8.7).
3.1.16 Command Block registers: Interface registers used for delivering commands to the device or posting
status from the device.
3.1.17 command completion: Command completion is the completion by the device of the action requested
by the command or the termination of the command with an error, the placing of the appropriate error
bits in the Error register, the placing of the appropriate status bits in the Status register, the clearing
of both BSY and DRQ to zero, and the asserting of INTRQ if nIEN is cleared to zero and the
command protocol specifies that INTRQ be asserted.
3.1.18 command packet: A command packet is a data structure transmitted to the device during the
execution of a PACKET command that includes the command and command parameters.
3.1.19 command released: When a device supports overlap or queuing, a command is considered released
when a bus release occurs before the command is completed.
3.1.20 Control Block registers: Interface registers used for device control and to post alternate status.
3.1.21 CRC: Cyclical Redundancy Check used to check the validity of certain data transfers.
3.1.22 device: Device is a storage peripheral. Traditionally, a device on the interface has been a hard disk
drive, but any form of storage device may be placed on the interface provided the device adheres to
this standard.
3.1.23 device selection: A device is selected when the DEV bit of the Device/Head register is equal to the
device number assigned to the device by means of a Device 0/Device 1 jumper or switch, or use of
the CSEL signal.
Page 3
T13/1321D revision 3
3.1.24 DMA (direct memory access) data transfer: A means of data transfer between device and host
memory without host processor intervention.
3.1.25 don’t care: A term to indicate that a value is irrelevant for the particular function described.
3.1.26 driver: The active circuit inside a device or host that sources or sinks current to assert or negate a
signal on the bus.
3.1.27 DRQ data block: This term describes a unit of data words transferred during a single assertion of DRQ
when using PIO data transfer. A data block is transferred between the host and the device as a
complete unit. A data block is a sector, except for data blocks of READ MULTIPLE and WRITE
MULTIPLE commands. In the cases of READ MULTIPLE and WRITE MULTIPLE commands, the
size of the data block may be changed in multiples of sectors by the SET MULTIPLE MODE
command.
3.1.28 interrupt pending: Interrupt pending is an internal state of a device that exists when the device shall to
notify the host of an event by asserting INTRQ if nIEN is cleared to zero (see 6.3).
3.1.29 LBA (logical block address): This term defines the addressing of data on the device by the linear
mapping of sectors.
3.1.30 master: In ATA-1, Device 0 has also been referred to as the master. Throughout this document the term
Device 0 is used.
3.1.31 native max address: The highest address a device accepts in the factory default condition, that is, the
highest address that is accepted by the SET MAX ADDRESS command. The capacity defined by
native max address may be different in CHS and LBA translations.
3.1.32 overlap: Overlap is a protocol that allows devices that require extended command time to perform a bus
release so that commands may be executed by the other device on the bus.
3.1.33 packet delivered command: A command that is delivered to the device using the PACKET command
via a command packet that contains the command and the command parameters.
3.1.34 PIO (programmed input/output) data transfer: PIO data transfers are performed by the host
processor utilizing PIO register accesses to the Data register.
3.1.35 queued: Command queuing allows the host to issue concurrent commands to the same device. Only
commands included in the Overlapped feature set may be queued. In this standard, the queue
contains all commands for which command acceptance has occurred but command completion has
not occurred.
3.1.36 read command: A command that causes the device to read data from the media (e.g., READ
SECTOR(S), READ DMA, etc.).
3.1.37 register delivered command: A command that is delivered to the device by placing the command and
all of the parameters for the command in the device Command Block registers.
3.1.38 register transfers: Register transfers refer to the host reading and writing any device register except the
Data register. Register transfers are 8 bits wide.
3.1.39 released: Indicates that a signal is not being driven. For tri-state drivers, this means that the driver is in
the high impedance state. For open-collector drivers, the driver is not asserted.
3.1.40 sector: A uniquely addressable set of 256 words (512 bytes).
Page 4
T13/1321D revision 3
3.1.41 signature: A unique set of values placed in the Command Block registers by the device to allow the
host to distinguish between register delivered command devices and packet delivered command
devices.
3.1.42 slave: In ATA-1, Device 1 has also been referred to as the slave. Throughout this document the term
Device 1 is used.
3.1.43 SMART: Self-Monitoring, Analysis, and Reporting Technology for prediction of device degradation and/or
faults. Throughout this document this is noted as SMART.
3.1.44 Ultra DMA burst: An Ultra DMA burst is defined as the period from an assertion of DMACK- to the
subsequent negation of DMACK- when Ultra DMA has been enabled by the host.
3.1.45 unit attention condition: A state that a device implementing the PACKET Command feature set
maintains while the device has asynchronous status information to report to the host.
3.1.46 unrecoverable error: An unrecoverable error is defined as having occurred at any point when the device
sets either the ERR bit or the DF bit to one in the Status register at command completion.
3.1.47 VS (vendor specific): This term is used to describe bits, bytes, fields, and code values that are
reserved for vendor specific purposes. These bits, bytes, fields, and code values are not described in
this standard, and may vary among vendors. This term is also applied to levels of functionality whose
definition is left to the vendor.
NOTE − Industry practice could result in conversion of a Vendor Specific bit, byte, field, or
code value into a defined standard value in a future standard.
3.1.48 write command: A command that causes the device to write data to the media (e.g., WRITE
SECTOR(S), WRITE DMA, etc.).
3.2 Conventions
Lowercase is used for words having the normal English meaning. Certain words and terms used in this
standard have a specific meaning beyond the normal English meaning. These words and terms are defined
either in clause 3 or in the text where they first appear.
The names of abbreviations, commands, fields, and acronyms used as signal names are in all uppercase (e.g.,
IDENTIFY DEVICE). Fields containing only one bit are usually referred to as the "name" bit instead of the
"name" field. (see 3.2.6 for the naming convention used for naming bits.)
Names of device registers begin with a capital letter (e.g., Cylinder Low register).
3.2.1 Precedence
If there is a conflict between text, figures, and tables, the precedence shall be tables, figures, then text.
Page 5
T13/1321D revision 3
3.2.2 Lists
Ordered lists, those lists describing a sequence, are of the form:
a)
b)
1)
2)
c)
Unordered list are of the form:
1)
2)
a)
b)
3)
3.2.3 Keywords
Several keywords are used to differentiate between different levels of requirements and optionality.
3.2.3.1 expected: A keyword used to describe the behavior of the hardware or software in the design models
assumed by this standard. Other hardware and software design models may also be implemented.
3.2.3.2 mandatory: A keyword indicating items to be implemented as defined by this standard.
3.2.3.3 may: A keyword that indicates flexibility of choice with no implied preference.
3.2.3.4 obsolete: A keyword used to describe bits, bytes, fields, and code values that no longer have
consistent meaning or functionality from one implementation to another. However, some degree of
functionality may be required for items designated as “obsolete” to provide for backward compatibility.
An obsolete bit, byte, field, or command shall never be reclaimed for any other use in any future
standard.
Obsolete commands should not be used by the host. Commands defined as obsolete in previous
standards may be command aborted by devices conforming to this standard. However, if a device
does not command abort an obsolete command, the minimum that is required by the device in
response to the command is command completion.
3.2.3.5 optional: A keyword that describes features that are not required by this standard. However, if any
optional feature defined by the standard is implemented, the feature shall be implemented in the way
defined by the standard.
3.2.3.6 retired: A keyword indicating that the designated bits, bytes, fields, and code values that had been
defined in previous standards are not defined in this standard and may be reclaimed for other uses in
future standards. If retired bits, bytes, fields, or code values are utilized before they are reclaimed,
they shall have the meaning or functionality as described in previous standards.
3.2.3.7 reserved: A keyword indicating reserved bits, bytes, words, fields, and code values that are set aside
for future standardization. Their use and interpretation may be specified by future extensions to this
or other standards. A reserved bit, byte, word, or field shall be set to zero, or in accordance with a
future extension to this standard. The recipient shall not check reserved bits, bytes, words, or fields.
Receipt of reserved code values in defined fields shall be treated as a command parameter error and
reported by returning command aborted.
Page 6
T13/1321D revision 3
3.2.3.8 shall: A keyword indicating a mandatory requirement. Designers are required to implement all such
mandatory requirements to ensure interoperability with other standard conformant products.
3.2.3.9 should: A keyword indicating flexibility of choice with a strongly preferred alternative. Equivalent to the
phrase “it is recommended”.
3.2.4 Numbering
Numbers that are not immediately followed by a lowercase "b" or "h" are decimal values. Numbers that are
immediately followed by a lowercase "b" (e.g., 01b) are binary values. Numbers that are immediately followed
by a lowercase "h" (e.g., 3Ah) are hexadecimal values.
3.2.5 Signal conventions
Signal names are shown in all uppercase letters.
All signals are either high active or low active signals. A dash character (-) at the end of a signal name
indicates the signal is a low active signal. A low active signal is true when the signal is below ViL, and is false
when the signal is above ViH. No dash at the end of a signal name indicates the signal is a high active signal.
A high active signal is true when the signal is above ViH, and is false when the signal is below ViL.
Asserted means that the signal is driven by an active circuit to the true state. Negated means that the signal is
driven by an active circuit to the false state. Released means that the signal is not actively driven to any state
(see 4.2.1). Some signals have bias circuitry that pull the signal to either a true state or false state when no
signal driver is actively asserting or negating the signal.
Control signals that may be used for more than one mutually exclusive functions are identified with their
function names separated by a colon (e.g., DIOW-:STOP).
3.2.6 Bit conventions
Bit names are shown in all uppercase letters except where a lowercase n precedes a bit name. If there is no
preceding n, then when BIT is set to one the meaning of the bit is true, and when BIT is cleared to zero the
meaning of the bit is false. If there is a preceding n, then when nBIT is cleared to zero the meaning of the bit is
true and when nBIT is set to one the meaning of the bit is false.
True
False
True
False
TEST
Bit setting=1
Bit setting=0
nTEST
Bit setting=0
Bit setting=1
Page 7
T13/1321D revision 3
3.2.7 State diagram conventions
State diagrams shall be as shown in Figure 1.
State designator: State_name
Entry condition
Transition label
Transition action
State designator: State_name
Transition condition
Transition label
Exit condition
Transition label
Transition action
State_name
Transition action
Transition condition
Transition label
Transition action
State re-entry
BSY
v
DRQ
v
REL
v
SERV
v
C/D
v
I/O
v
INTRQ
V
DMARQ
V
PDIAGV
DASPV
Figure 1 − State diagram convention
Each state is identified by a state designator and a state name. The state designator is unique among all
states in all state diagrams in this document. The state designator consists of a set of letters that are
capitalized in the title of the figure containing the state diagram followed by a unique number. The state name is
a brief description of the primary action taken during the state, and the same state name may appear in other
state diagrams. If the same primary function occurs in other states in the same state diagram, they are
designated with a unique letter at the end of the name. Additional actions may be taken while in a state and
these actions are described in the state description text.
In device command protocol state diagrams, the state of bits and signals that change state during the
execution of this state diagram are shown under the state designator:state_name, and a table is included that
shows the state of all bits and signals throughout the state diagram as follows:
v = bit value changes.
1 = bit set to one.
0 = bit cleared to zero.
x = bit is don’t care.
V = signal changes.
A = signal is asserted.
N = signal is negated.
R = signal is released.
X = signal is don’t care.
Each transition is identified by a transition label and a transition condition. The transition label consists of the
state designator of the state from which the transition is being made followed by the state designator of the
state to which the transition is being made. In some cases, the transition to enter or exit a state diagram may
come from or go to a number of state diagrams, depending on the command being executed. In this case, the
state designator is labeled xx. The transition condition is a brief description of the event or condition that
causes the transition to occur and may include a transition action, indicated in italics, that is taken when the
transition occurs. This action is described fully in the transition description text.
Upon entry to a state, all actions to be executed in that state are executed. If a state is re-entered from itself,
all actions to be executed in the state are executed again.
Page 8
T13/1321D revision 3
It is assumed that all actions defined in a state are executed within the state and that transitions from state to
state are instantaneous.
3.2.8 Timing conventions
Certain symbols are used in the timing diagrams. These symbols and their respective definitions are listed
below.
or
- signal transition (asserted or negated)
or
- data transition (asserted or negated)
- data valid
- undefined but not necessarily released
- asserted, negated or released
- released
- the “other” condition if a signal is shown with no change
All signals are shown with the asserted condition facing to the top of the page. The negated condition is shown
towards the bottom of the page relative to the asserted condition.
The interface uses a mixture of negative and positive signals for control and data. The terms asserted and
negated are used for consistency and are independent of electrical characteristics.
In all timing diagrams, the lower line indicates negated, and the upper line indicates asserted. The following
illustrates the representation of a signal named TEST going from negated to asserted and back to negated,
based on the polarity of the signal.
Assert
Negate
Assert
Negate
TEST
> V iH
< V iL
TEST< V iL
> V iH
3.2.9 Byte ordering for data transfers
Data is transferred in blocks using either PIO or DMA protocols. PIO data transfers occur when the BSY bit is
cleared to zero and the DRQ bit is set to one. These transfers are usually 16-bit but CFA devices may
implement 8-bit PIO transfers. Data is transferred in blocks of one or more bytes known as a DRQ block. DMA
data transfers occur when the host asserts DMACK- in response to the device asserting DMARQ. DMA
transfers are always 16-bit. Each assertion of DMACK- by the host defines a DMA data burst. A DMA data
burst is two or more bytes.
Assuming a DRQ block or a DMA burst of data contains "n" bytes of information, the bytes are labeled Byte(0)
through Byte(n-1), where Byte(0) is first byte of the block, and Byte(n-1) is the last byte of the block. Table 1
shows the order the bytes shall be presented in when such a block of data is transferred on the interface using
Page 9
T13/1321D revision 3
16-bit PIO and DMA transfers. Table 2 shows the order the bytes shall be presented in when such a block or
burst of data is transferred on the interface using 8-bit PIO.
Table 1 − Byte order
DD
15
DD
14
First transfer
Second transfer
........
Last transfer
DD
13
DD
12
DD
11
DD
10
DD
9
DD
8
DD
7
DD
6
DD
5
DD
4
DD
3
Byte (1)
Byte (3)
Byte (0)
Byte (2)
Byte (n-1)
Byte (n-2)
DD
2
DD
1
DD
0
Table 2 − Byte order
DD
7
DD
6
DD
5
First transfer
Second transfer
........
Last transfer
DD
4
DD
3
DD
2
DD
1
DD
0
Byte (0)
Byte (1)
Byte (n-1)
NOTE − The above description is for data on the interface. Host systems and/or host adapters
may cause the order of data as seen in the memory of the host to be different.
4 Interface physical and electrical requirements
Connectors and cables are documented in annex A.
4.1 Cable configuration
This standard defines an interface containing a single host or host adapter and one or two devices. One device
is configured as Device 0 and the other device as Device 1.
The designation of a device as Device 0 or Device 1 may be made in a number of ways including but not limited
to:
−
−
a switch or a jumper on the device;
use of the Cable Select (CSEL) pin.
The host shall be placed at one end of the cable. It is recommended that for a single device configuration the
device be placed at the opposite end of the cable from the host. If a single device configuration is implemented
with the device not at the end of the cable, a cable stub results that may cause degradation of signals. Single
device configurations with the device not at the end of the cable shall not be used with Ultra DMA modes.
4.2 Electrical characteristics
Table 3 defines the DC characteristics of the interface signals. Table 4 defines the AC characteristics.
Page 10
T13/1321D revision 3
IoL
IoLDASP
IoH
IoHDMARQ
IZ
Table 3 − DC characteristics
Description
Driver sink current (see note 1)
Driver sink current for DASP (see note 1)
Driver source current (see note 2)
Driver source current for DMARQ (see note 2)
Device pull up current on DD (15:0) and
STROBE when high Z
Voltage input high
Voltage input low
Voltage output high at IoH min
Voltage output low at IoL min
Min
4 mA
12 mA
400 µA
500 µA
-10 µA
Max
200 µA
ViH
2.0 VDC
ViL
0.8 VDC
VoH
2.4 VDC
VoL
0.5 VDC
NOTES −
1 IoLDASP shall be 12 mA minimum to meet legacy timing and signal integrity.
2 IoH value at 400 µA is insufficient in the case of DMARQ that is pulled low by a 5.6 kΩ
resistor.
Table 4 − AC characteristics
Description
Min
Max
SRISE
Rising edge slew rate for any signal on AT interface (see
1.25
note)
V/ns
SFALL
Falling edge slew rate for any signal on AT interface (see
1.25
note)
V/ns
Chost
Host interface signal capacitance at the host connector
25 pf
Cdevice
Device interface signal capacitance at the device
20 pf
connector
NOTE – SRISE and SFALL shall meet this requirement when measured at the sender’s
connector from 10-90% of full signal amplitude with all capacitive loads from 15 pf through
40 pf where all signals have the same capacitive load value.
Page 11
T13/1321D revision 3
4.2.1 Driver types and required termination
Signal
Table 5 − Driver types and required termination
Source
Driver type
Host
Device
(see note 1)
(see note 2)
(see note 2)
Host
TP
Bidir
TS
Device
TS
5.6 kΩ PD
Host
TS
Notes
RESETDD (15:0)
3
DMARQ
DIOR-:HDMARDY:HSTROBE
DIOW-:STOP
Host
TS
IORDY:DDMARDYDevice
TS
1.0 kΩ PU
6,10
:DSTROBE
CSEL
Host
Ground
10 kΩ PU
4, 6
DMACKHost
TP
INTRQ
Device
TS
10 kΩ
5
DA (2:0)
Host
TP
PDIAG-:CBLIDDevice
TS
10 kΩ PU
2,6,7,8
CS0- CS1Host
TP
DASPDevice
OC
10 kΩ PU
6,9
NOTES −
1 TS=Tri-state; OC=Open Collector; TP=Totem-pole; PU=Pull-up; PD=Pull-down.
2 All resistor values are the minimum (lowest allowed) except for the 10kΩ PU on PDIAG-:CBLID- which
shall have a tolerance of ±5% or less.
3 Devices shall not have a pull-up resistor on DD7. The host shall have a 10 kΩ pull-down resistor and
not a pull-up resistor on DD7 to allow a host to recognize the absence of a device at power-up so that
a host shall detect BSY as being cleared when attempting to read the Status register of a device that
is not present.
4 When used as CSEL, this line is grounded at the Host and a 10 kΩ pull-up is required at both devices.
5 A 10 kΩ pull-down or pull-up, depending upon the level sensed, should be implemented at the host.
6 Pull-up values are based on +5 v Vcc.
7 Hosts that do not support Ultra DMA modes greater than mode 2 shall not connect to the PDIAG:CBLID- signal.
8 The 80-conductor cable assembly shall meet the following requirements: the PDIAG-:CBLID- signal
shall be connected to ground in the host connector of the cable assembly; the PDIAG-:CBLIDsignal shall not be connected between the host and the devices; and, the PDIAG-:CBLID- signal
shall be connected between the devices.
9 The host shall not drive DASP-. If the host connects to DASP- for any purpose, the host shall ensure
that the signal level detected on the interface for DASP- shall maintain VoH and VoL compatibility,
given the IoH and IoL requirements of the DASP- device drivers.
10 Values greater than 1 kΩ may improve noise margin.
4.2.2 Electrical characteristics for Ultra DMA
Hosts that support Ultra DMA transfer modes greater than mode 2 shall not share signals between primary and
secondary I/O ports. They shall provide separate drivers and separate receivers for each cable.
Page 12
T13/1321D revision 3
4.2.2.1 Cable configuration
The following table defines the host transceiver configurations for a dual cable system configuration for all
transfer modes.
Transfer
mode
All PIO and
Multiword DMA
Optional host
transceiver configuration
One transceiver may be used
for signals to both ports
Ultra DMA
0, 1, 2
One transceiver may be used
for signals to both ports
except DMACK-
Ultra DMA
3, 4
One transceiver may be used
for signals to both ports for
RESET-, INTRQ, DA(2:0),
CS0-, CS1-, and DASP-
Recommended
host
transceiver configuration
DIOR-, DIOW- and IORDY
should have a separate
transceiver for each port.
Mandatory host transceiver
configuration
Either DIOR-, DIOW- and
IORDY or CS0- and CS1shall have a separate
transceiver for each port.
DIOR-, DIOW- and IORDY Either DIOR-, DIOW- and
should have a separate IORDY or CS0- and CS1transceiver for each port.
shall have a separate
transceiver for each port.
DMACK- shall have a
separate transceiver for
each port
RESET-, INTRQ, DA(2:0), All signals shall have a
CS0-, CS1-, and DASP- separate transceiver for
should have a separate each port except for
transceiver for each port.
RESET-, INTRQ, DA(2:0),
CS0-, CS1-, and DASP-
The following table defines the system configuration for connection between devices and systems for all transfer
modes.
Transfer
mode
Single
device
direct
connection configuration (see
note 1)
May be used.
40-conductor
cable
connection configuration
(see note 2)
May be used.
80-conductor
cable
connection configuration
(see note 2)
May be used (see note 3)
All PIO and
Multiword DMA
Ultra DMA
May be used.
May be used.
May be used (see note 3)
0, 1, 2
Ultra DMA 3, 4
May be used (see note 4).
Shall not be used.
May be used (see note 4).
NOTES –
1 Direct connection is a direct point-to-point connection between the host connector and the device connector.
2 The 40-conductor cable assembly and the 80-conductor cable assembly are defined in Annex A.
3 80-conductor cable assemblies may be used in place of 40-conductor cable assemblies to improve signal
quality for data transfer modes that do not require an 80-conductor cable assembly.
4 Either a single device direct connection configuration or an 80-conductor cable connection configuration shall
be used for systems operating with Ultra DMA modes greater than 2.
4.2.2.2 Series termination required for Ultra DMA
Series termination resistors are required at both the host and the device for operation in any of the Ultra DMA
modes. Table 6 describes typical values for series termination at the host and the device.
Page 13
T13/1321D revision 3
Table 6 − Typical series termination for Ultra DMA
Signal
Host Termination
Device Termination
DIOR-:HDMARDY-:HSTROBE
22 ohm
82 ohm
DIOW-:STOP
22 ohm
82 ohm
CS0-, CS133 ohm
82 ohm
DA0, DA1, DA2
33 ohm
82 ohm
DMACK22 ohm
82 ohm
DD15 through DD0
33 ohm
33 ohm
DMARQ
82 ohm
22 ohm
INTRQ
82 ohm
22 ohm
IORDY:DDMARDY-:DSTROBE
82 ohm
22 ohm
RESET33 ohm
82 ohm
NOTE − Only those signals requiring termination are listed in this table. If a signal is not
listed, series termination is not required for operation in an Ultra DMA mode. Figure 2
shows signals also requiring a pull-up or pull-down resistor at the host. The actual
termination values should be selected to compensate for transceiver and trace
impedance to match the characteristic cable impedance.
VCC
IORDY
DMARQ
DD 7
Figure 2 − Ultra DMA termination with pull-up or pull-down
5 Interface signal assignments and descriptions
5.1 Signal summary
The physical interface consists of receivers and drivers communicating through a set of conductors using an
asynchronous interface protocol. Table 7 defines the signal names. For connector descriptions see annex A.
For driver and termination definition see 4.2.1. For signal protocol and timing see clause 9 and clause 10.
Page 14
T13/1321D revision 3
Table 7 − Interface signal name assignments
Description
Host
Dir
Dev
Acronym
Cable select
(see note)
CSEL
Chip select 0
→
CS0Chip select 1
→
CS1Data bus bit 0
↔
DD0
Data bus bit 1
↔
DD1
Data bus bit 2
↔
DD2
Data bus bit 3
↔
DD3
Data bus bit 4
↔
DD4
Data bus bit 5
↔
DD5
Data bus bit 6
↔
DD6
Data bus bit 7
↔
DD7
Data bus bit 8
↔
DD8
Data bus bit 9
↔
DD9
Data bus bit 10
↔
DD10
Data bus bit 11
↔
DD11
Data bus bit 12
↔
DD12
Data bus bit 13
↔
DD13
Data bus bit 14
↔
DD14
Data bus bit 15
↔
DD15
Device active or slave (Device 1) present
(see note)
DASPDevice address bit 0
→
DA0
Device address bit 1
→
DA1
Device address bit 2
→
DA2
DMA acknowledge
→
DMACKDMA request
←
DMARQ
Interrupt request
←
INTRQ
I/O read
→
DIORDMA ready during Ultra DMA data-in bursts
→
HDMARDYData strobe during Ultra DMA data-out bursts
→
HSTROBE
I/O ready
←
IORDY
DMA ready during Ultra DMA data-out bursts
←
DDMARDYData strobe during Ultra DMA data-in bursts
←
DSTROBE
I/O write
→
DIOWStop during Ultra DMA data bursts
→
STOP
Passed diagnostics
(see note)
PDIAGCable assembly type identifier
(see note)
CBLIDReset
→
RESETNOTE − See signal descriptions and annex A for information on source of these signals
Page 15
T13/1321D revision 3
5.2 Signal descriptions
5.2.1 CS (1:0)- (Chip select)
These are the chip select signals from the host used to select the Command Block or Control Block registers
(see 7.2). When DMACK- is asserted, CS0- and CS1- shall be negated and transfers shall be 16 bits wide.
5.2.2 DA (2:0) (Device address)
This is the 3-bit binary coded address asserted by the host to access a register or data port in the device (see
7.2).
5.2.3 DASP- (Device active, device 1 present)
This is a time-multiplexed signal that indicates that a device is active, or that Device 1 is present.
NOTE − The indication that the device is active may be unsynchronized with the execution of
the command.
5.2.4 DD (15:0) (Device data)
This is an 8- or 16-bit bi-directional data interface between the host and the device. The lower 8 bits are used
for 8-bit register transfers. Data transfers are 16-bits wide except for CFA device that implement 8-bit data
transfers.
5.2.5 DIOR-:HDMARDY-:HSTROBE (Device I/O read:Ultra DMA ready:Ultra DMA data strobe)
DIOR- is the strobe signal asserted by the host to read device registers or the Data port.
HDMARDY- is a flow control signal for Ultra DMA data-in bursts. This signal is asserted by the host to indicate
to the device that the host is ready to receive Ultra DMA data-in bursts. The host may negate HDMARDY- to
pause an Ultra DMA data-in burst.
HSTROBE is the data-out strobe signal from the host for an Ultra DMA data-out burst. Both the rising and
falling edge of HSTROBE latch the data from DD(15:0) into the device. The host may stop generating
HSTROBE edges to pause an Ultra DMA data-out burst.
5.2.6 DIOW-:STOP (Device I/O write:Stop Ultra DMA burst)
DIOW- is the strobe signal asserted by the host to write device registers or the Data port
DIOW- shall be negated by the host prior to initiation of an Ultra DMA burst. STOP shall be negated by the
host before data is transferred in an Ultra DMA burst. Assertion of STOP by the host during an Ultra DMA
burst signals the termination of the Ultra DMA burst.
5.2.7 DMACK- (DMA acknowledge)
This signal shall be used by the host in response to DMARQ to initiate DMA transfers.
5.2.8 DMARQ (DMA request)
This signal, used for DMA data transfers between host and device, shall be asserted by the device when the
device is ready to transfer data to or from the host. For Mulitword DMA transfers, the direction of data transfer
is controlled by DIOR- and DIOW-. This signal is used in a handshake manner with DMACK-, i.e., the device
Page 16
T13/1321D revision 3
shall wait until the host asserts DMACK- before negating DMARQ, and re-asserting DMARQ if there is more
data to transfer.
When a DMA operation is enabled, CS0- and CS1- shall not be asserted and transfers shall be 16 bits wide.
This signal shall be released when the device is not selected.
5.2.9 INTRQ (Device interrupt)
This signal is used by the selected device to interrupt the host system when interrupt pending is set. When the
nIEN bit is cleared to zero and the device is selected, INTRQ shall be enabled through a tri-state buffer. When
the nIEN bit is set to one or the device is not selected, the INTRQ signal shall be released.
When asserted, this signal shall be negated by the device within 400 ns of the negation of DIOR- that reads the
Status register to clear interrupt pending. When asserted, this signal shall be negated by the device within 400
ns of the negation of DIOW- that writes the Command register to clear interrupt pending.
When the device is selected by writing to the Device/Head register while interrupt pending is set, INTRQ shall
be asserted within 400 ns of the negation of DIOW- that writes the Device/Head register. When the device is
deselected by writing to the Device/Head register while interrupt pending is set, INTRQ shall be released within
400 ns of the negation of DIOW- that writes the Device/Head register.
For devices implementing the Overlapped feature set, if INTRQ assertion is being disabled using nIEN at the
same instant that the device asserts INTRQ, the minimum pulse width shall be at least 40 ns.
This signal shall be released when the device is not selected.
5.2.10 IORDY:DDMARDY-:DSTROBE (I/O channel ready:Ultra DMA ready:Ultra DMA data strobe)
This signal is negated to extend the host transfer cycle of any host register access (read or write) when the
device is not ready to respond to a data transfer request.
If the device requires that the host transfer cycle time be extended for PIO modes 3 and above, the device shall
utilize IORDY. Hosts that use PIO modes 3 and above shall support IORDY.
DDMARDY- is a flow control signal for Ultra DMA data-out bursts. This signal is asserted by the device to
indicate to the host that the device is ready to receive Ultra DMA data-out bursts. The device may negate
DDMARDY- to pause an Ultra DMA data-out burst.
DSTROBE is the data-in strobe signal from the device for an Ultra DMA data-in burst. Both the rising and
falling edge of DSTROBE latch the data from DD(15:0) into the host. The device may stop generating
DSTROBE edges to pause an Ultra DMA data-in burst.
This signal shall be released when the device is not selected.
5.2.11 PDIAG-:CBLID- (Passed diagnostics:Cable assembly type identifier)
PDIAG- shall be asserted by Device 1 to indicate to Device 0 that Device 1 has completed diagnostics (see
clause 9).
Page 17
T13/1321D revision 3
The host may sample CBLID- after a power-on or hardware reset in order to detect the presence or absence of
an 80-conductor cable assembly by performing the following steps:
a) The host shall wait until the power-on or hardware reset protocol is complete for all devices on the
cable;
b) If Device 1 is present, the host should issue IDENTIFY DEVICE or IDENTIFY PACKET DEVICE
and use the returned data to determine that Device 1 is compliant with ATA-3 or subsequent
standards. Any device compliant with ATA-3 or subsequent standards releases PDIAG- no later
than after the first command following a power-on or hardware reset sequence.
NOTE − Older devices not in compliance with ATA-3 or subsequent standards may continue to
assert this signal providing a false indication of the cable type. Issuing IDENTIFY DEVICE or
IDENTIFY PACKET DEVICE not only provides the host with the information required to verify
that the devices are compliant with these standards, but also provides a command resulting in
the release of this signal.
If the host detects that CBLID- is connected to ground, an 80-conductor cable assembly is installed in the
system. If the host detects that this signal is not connected to ground, an 80-conductor cable assembly is not
installed in the system. See Annex B for a description of the non-standard device determination of cable type.
Open
PDIAG- conductor
CBLID-
Host
Device 1
Device 0
NOTE − CBLID- is grounded in the 80-conductor cable assembly host connector for
the purpose of indicating to the host that the cable assembly being used is an
80-conductor assembly not a 40-conductor assembly.
Figure 3 − PDIAG- example using an 80-conductor cable assembly
5.2.12 RESET- (Hardware reset)
This signal, referred to as hardware reset, shall be used by the host to reset the device (see 9.1).
5.2.13 CSEL (Cable select)
The device is configured as either Device 0 or Device 1 depending upon the value of CSEL:
−
−
If CSEL is negated, the device number is 0;
If CSEL is asserted, the device number is 1.
The state of this signal may be sampled at any tine by the device to detect that the device is configured as
Device 0 or Device 1.
Page 18
T13/1321D revision 3
5.2.13.1 CSEL with 40-conductor cable
Special cabling may be used to selectively ground CSEL. CSEL of Device 0 is connected to the CSEL
conductor in the cable, and is grounded, thus allowing the device to recognize itself as Device 0. CSEL of
Device 1 is not connected to CSEL because the conductor is removed, thus the device recognizes itself as
Device 1. It should be recognized that if a single device is configured at the end of the cable using CSEL, a
device 1 only configuration results. See Figure 4 and Figure 5.
CSEL conductor
Open
Ground
Host
Device 0
Device 1
CSEL conductor
Open
Ground
Host
Device 1
Figure 4 − Cable select example
5.2.13.2 CSEL with 80-conductor cable
For designated cable assemblies (including all 80-conductor cable assemblies): these assemblies are
constructed so that CSEL is connected from the host connector to the connector at the opposite end of the
cable from the host (see Figure 5). Therefore, Device 0 shall be at the opposite end of the cable from the host.
Single device configurations with the device not at the end of the cable shall not be used with Ultra DMA
modes.
Page 19
T13/1321D revision 3
CSEL conductor
Open
Ground
Host
Device 1
Device 0
CSEL conductor
Open
Ground
Host
Device 0
Figure 5 − Alternate cable select example
6 General operational requirements
6.1 Command delivery
Commands may be delivered in two forms. For devices that do not implement the PACKET Command feature
set, all commands and command parameters are delivered by writing the device Command Block registers.
Such commands are defined as register delivered commands.
Devices that implement the PACKET Command feature set utilize packet delivered commands as well as some
register delivered commands.
All register delivered commands and the PACKET command are described in clause 8.
NOTE − The content of command packets delivered during execution of the PACKET
command are not described in this standard.
6.2 Register delivered data transfer command sector addressing
For register delivered data transfer commands all addressing of data sectors recorded on the device's media is
by a logical sector address. There is no implied relationship between logical sector addresses and the actual
physical location of the data sector on the media.
Devices shall support translations as described below:
− All devices shall support LBA translation.
− If the device’s capacity is greater than or equal to one sector and less than or equal to 16,514,064
−
−
−
sectors, then the device shall support CHS translation.
If the device’s capacity is greater than 16,514,064 sectors, the device may support CHS
translation.
If a device supports CHS translation, then, following a power-on or hardware reset, the CHS
translation enabled by the device shall be the default translation.
If a device supports CHS translation, a device may allow a host to use the INITIALIZE DEVICE
PARAMETERS command to select other CHS translations.
Page 20
T13/1321D revision 3
− If a device supports CHS translation, IDENTIFY DEVICE words 1,3, and 6 shall describe the default
translation, and words 53-58 shall describe the current translation.
A CHS address is made up of three fields: the sector number, the head number, and the cylinder number.
Sectors are numbered from 1 to the maximum value allowed by the current CHS translation but shall not
exceed 255. Heads are numbered from 0 to the maximum value allowed by the current CHS translation but
shall not exceed 15. Cylinders are numbered from 0 to the maximum value allowed by the current CHS
translation but shall not exceed 65,535.
When the host selects a CHS translation using the INITIALIZE DEVICE PARAMETERS command, the host
requests the number of sectors per logical track and the number of heads per logical cylinder. The device then
computes the number of logical cylinders available in the requested translation.
A device shall not change the addressing method specified by the command and shall return status information
utilizing the addressing method specified for the command.
1) The host may select either the currently selected CHS translation addressing or LBA addressing on a
command-by-command basis by using the LBA bit in the Device/Head register;
2) The device shall support LBA addressing for all media access commands;
3) Logical sectors on the device shall be linearly mapped with the first LBA addressed sector (sector 0)
being the same sector as the first logical CHS addressed sector (cylinder 0, head 0, sector 1).
Irrespective of the logical CHS translation currently in effect, the LBA address of a given logical sector
does not change. The following is always true for LBA numbers less than or equal to 16,514,064 for
devices supporting the current CHS translation:
LBA = (((cylinder_number ∗ heads_per_cylinder) + head_number) ∗ sectors_per_track) + sector_number - 1
Where: heads_per_cylinder and sectors_per_track are the current translation values.
cylinder_number, sector_number, and head_number are the values placed in the Cylinder High,
Cylinder Low, Sector Number, and Device/Head registers.
6.2.1 Definitions and value ranges of IDENTIFY DEVICE words (see 8.12)
1) Word 1 shall contain the number of user-addressable logical cylinders in the default CHS
translation. If the content of words (61:60) is less than 16,514,064 then the content of word 1 shall
be greater than or equal to one and less than or equal to 65,535. If the content of words (61:60) is
greater than or equal to 16,514,064 then the content of word 1 shall be equal to 16,383.
2) Word 3 shall contain the number of user-addressable logical heads in the default CHS translation.
The content of word 3 shall be greater than or equal to one and less than or equal to 16. For
compatibility with some BIOSs, the content of word 3 may be equal to 15 if the content of word 1 is
greater than 8192.
3) Word 6 shall contain the number of user-addressable logical sectors in the default CHS translation.
The content of word 6 shall be greater than or equal to one and less than or equal to 63.
4) [(The content of word 1) ∗ (the content of word 3) ∗ (the content of word 6)] shall be less than or
equal to 16,514,064.
5) Word 54 shall contain the number of user-addressable logical cylinders in the current CHS
translation. The content of word 54 shall be greater than or equal to one and less than or equal to
65,535. After power-on or after a hardware reset the content of word 54 shall be equal to the
content of word 1.
6) Word 55 shall contain the number of user-addressable logical heads in the current CHS translation.
The content of word 55 shall be greater than or equal to one and less than or equal to 16. After
power-on or after a hardware reset the content of word 55 shall be equal to the content of word 3.
7) Word 56 shall contain the number of user-addressable logical sectors in the current CHS
translation. The content of word 56 should be equal to 63 for compatibility with all BIOSs.
Page 21
T13/1321D revision 3
8)
9)
10)
11)
12)
However, the content of word 56 may be greater than or equal to one and less than or equal to 255.
At power-on or after a hardware reset the content of word 56 shall equal the content of word 6.
Words (58:57) shall contain the user-addressable capacity in sectors for the current CHS
translation. The content of words (58:57) shall equal [(the content of word 54) ∗ (the content of
word 55) ∗ (the content of word 56)]. The content of words (58:57) shall be less than or equal to
16,514,064. The content of words (58:57) shall be less than or equal to the content of words
(61:60).
The content of words 54, 55, 56, and (58:57) may be affected by the host issuing an INITIALIZE
DEVICE PARAMETERS command (see 8.16).
If the content of words (61:60) is greater than 16,514,064 and if the device does not support CHS
addressing, then the content of words 1, 3, 6, 54, 55, 56, and (58:57) shall equal zero. If the
content of word 1, word 3, or word 6 equals zero, then the content of words 1, 3, 6, 54, 55, 56, and
(58:57) shall equal zero.
Words (61:60) shall contain the total number of user-addressable sectors. The content of words
(61:60) shall be greater than or equal to one and less than or equal to 268,435,456.
The content of words 1, 54, (58:57), and (61:60) may be affected by the host issuing a SET MAX
ADDRESS command (see 8.38).
6.2.2 Addressing constraints and error reporting
1) Devices may access any address in the current CHS translation if [(the requested cylinder + 1) ∗
(the requested head + 1) ∗ (the requested sector)] is less than or equal to the content of words
(61:60).
2) Devices shall set IDNF to one or ABRT to one in the Error register and ERR to one in the Status
register in response to any command with a CHS address request where one or more of the
following are true:
a) The requested cylinder value is greater than [(the content of word 54) - 1]
b) The requested head value is greater than [(the content of word 55) - 1]
c) The requested sector value is zero or greater than the content of word 56
3) Devices shall sets IDNF to one or ABRT to one in the Error register and ERR to one in the Status
register in response to any command with an LBA address request where the requested LBA
number is greater than or equal to the content of words (61:60).
6.3 Interrupts
INTRQ is used by the selected device to notify the host of an event. The device internal interrupt pending state
is set when such an event occurs. If nIEN is cleared to zero, INTRQ is asserted (see 5.2.9).
The device shall enter the interrupt pending state when:
1)
2)
3)
4)
5)
6)
7)
8)
9)
any command except a PIO data-in command reaches command completion successfully;
any command reaches command completion with error;
the device is ready to send a data block during a PIO data-in command;
the device is ready to accept a data block after the first data block during a PIO data-out
command;
a device implementing the PACKET Command feature set is ready to receive the command packet
and bits 6-5 in word 0 of the IDENTIFY PACKET DEVICE response have the value 01b;
a device implementing the PACKET Command feature set is ready to transfer a DRQ data block
during a PIO transfer;
a device implementing the Overlap feature set performs a bus release if the Bus release interrupt is
enabled;
a device implementing the Overlap feature set has performed a Bus release and is now ready to
continue the command execution;
a device implementing the Overlap feature set is ready to transfer data after a SERVICE command
if the Service interrupt is enabled;
Page 22
T13/1321D revision 3
10) Device 0 completes an EXECUTE DEVICE DIAGNOSTIC command. Device 1 shall not enter the
interrupt pending state when completing an EXECUTE DEVICE DIAGNOSTIC command.
The device shall not exit the interrupt pending state as a result of the host changing the state of the DEV bit..
The device shall exit the interrupt pending state when:
1) the device is selected, BSY is cleared to zero, and the Status register is read;
2) the device is selected, both BSY and DRQ are cleared to zero, and the Command register is
written;
3) the RESET- signal is asserted;
4) the SRST bit is set to one.
6.4 General feature set
The General feature set defines the common commands implemented by devices.
6.4.1 General feature set for devices not implementing the PACKET command feature set
The following General feature set commands are mandatory for all devices that are capable of both reading and
writng their media and do not implementing the PACKET command feature set:
− EXECUTE DEVICE DIAGNOSTIC
− FLUSH CACHE
− IDENTIFY DEVICE
− INITIALIZE DEVICE PARAMETERS
− READ DMA
− READ MULTIPLE
− READ SECTOR(S)
− READ VERIFY SECTOR(S)
− SEEK
− SET FEATURES
− SET MULTIPLE MODE
− WRITE DMA
− WRITE MULTIPLE
− WRITE SECTOR(S)
The following General feature set commands are mandatory for all devices that are capable of only reading their
media and do not implementing the PACKET command feature set:
− EXECUTE DEVICE DIAGNOSTIC
− IDENTIFY DEVICE
− INITIALIZE DEVICE PARAMETERS
− READ DMA
− READ MULTIPLE
− READ SECTOR(S)
− READ VERIFY SECTOR(S)
− SEEK
− SET FEATURES
− SET MULTIPLE MODE
The following General feature set commands are optional for devices not implementing the PACKET command
feature set:
−
−
−
−
DOWNLOAD MICROCODE
NOP
READ BUFFER
WRITE BUFFER
Page 23
T13/1321D revision 3
The following General feature set command is prohibited for use by devices not implementing the PACKET
command feature set:
−
DEVICE RESET
The following resets are mandatory for devices not implementing the PACKET command feature set:
−
−
−
Power-on reset: Executed at power-on, the device executes a series of electrical circuitry
diagnostics, spins up the HDA, tests speed and other mechanical parametrics, and sets default
values (see 9.1).
Hardware reset: Executed in response to the assertion of the RESET- signal the device executes a
series of electrical circuitry diagnostics, and resets to default values (see 9.1).
Software reset: Executed in response to the setting of the SRST bit in the Device Control register
the device resets the interface circuitry (see 9.2).
6.4.2 General feature set for devices implementing the PACKET command feature set
The following General feature set commands are mandatory for all devices implementing the PACKET
command feature set:
−
−
−
−
−
−
−
−
−
DEVICE RESET
EXECUTE DEVICE DIAGNOSTIC
FLUSH CACHE
IDENTIFY DEVICE
IDENTIFY PACKET DEVICE
NOP
PACKET
READ SECTOR(S)
SET FEATURES
The following General command set commands are prohibited for use by devices implementing the PACKET
command feature set.
−
−
−
−
−
−
−
−
−
−
−
−
DOWNLOAD MICROCODE
INITIALIZE DEVICE PARAMETERS
READ BUFFER
READ DMA
READ MULTIPLE
READ VERIFY
SEEK
SET MULTIPLE MODE
WRITE BUFFER
WRITE DMA
WRITE MULTIPLE
WRITE SECTOR(S)
The following resets are mandatory for devices implementing the PACKET command feature set:
− Power-on reset: Executed at power-on, the device executes a series of electrical circuitry
diagnostics, spins up the HDA, tests speed and other mechanical parametrics, and sets default
values (see 9.1).
− Hardware reset: Executed in response to the assertion of the RESET- signal the device executes a
series of electrical circuitry diagnostics, and resets to default values (see 9.1).
− Software reset: Executed in response to the setting of the SRST bit in the Device Control register
the device resets the interface circuitry (see 9.2).
Page 24
T13/1321D revision 3
−
DEVICE RESET: Executed in response to the DEVICE RESET command the device resets the
interface circuitry (see 8.7).
6.5 Multiword DMA
Multiword DMA is a data transfer protocol used with the READ DMA, WRITE DMA, READ DMA QUEUED,
WRITE DMA QUEUED, and PACKET commands. When a Multiword DMA transfer is enabled as indicated by
IDENTIFY DEVICE (see 8.12) or IDENTIFY PACKET DEVICE (see 8.13) data, this data transfer protocol shall
be used for the data transfers associated with these commands. DMA transfer modes may be changed using
the SET FEATURES 03h subcommand (see 8.37.11). Signal timing for this protocol is described in 10.2.3.
The DMARQ and DMACK- signals are used to signify when a Multiword DMA transfer is to be executed. The
DMARQ and DMACK- signals are also used to control the data flow of a Multiword DMA data transfer.
When a device is ready to transfer data associated with a Multiword DMA transfer, the device shall assert
DMARQ. The host shall then respond by negating CS0- and CS1-, asserting DMACK-, and begin the data
transfer by asserting, then negating, DIOW- or DIOR- for each word transferred. CS0- and CS1- shall remain
negated as long as DMACK- is asserted. The host shall not assert DMACK- until DMARQ has been asserted
by the device. The host shall initiate DMA read or write cycles only when both DMARQ and DMACK- are
asserted. Having asserted DMARQ and DMACK-, these signals shall remain asserted until at least one word of
data has been transferred.
The device may pause the transfer for flow control purposes by negating DMARQ. The host shall negate
DMACK- in response to the negation of DMARQ. The device may then reassert DMARQ to continue the data
transfer when the device is ready to transfer more data and DMACK- has been negated by the host.
The host may pause the transfer for flow control purposes by either pausing the assertion of DIOW- or DIORpulses or by negating DMACK-. The device may leave DMARQ asserted if DMACK- is negated. The host may
then reassert DMACK- when DMARQ is asserted and begin asserting DIOW- or DIOR- pulses to continue the
data transfer.
When the Multiword DMA data transfer is complete, the device shall negate DMARQ and the host shall negate
DMACK- in response.
DMARQ shall be driven from the first assertion at the beginning of a DMA transfer until the negation after the
last word is transferred. This signal shall be released at all other times.
If the device detects an error before data transfer for the command is complete, the device may complete the
data transfer or may terminate the data transfer before completion and shall report the error in either case.
NOTE − If a data transfer is terminated before completion, the assertion of INTRQ should be
passed through to the host software driver regardless of whether all data requested by the
command has been transferred.
6.6 Ultra DMA feature set
6.6.1 Overview
Ultra DMA is a data transfer protocol used with the READ DMA, WRITE DMA, READ DMA QUEUED, WRITE
DMA QUEUED, and PACKET commands. When this protocol is enabled, the Ultra DMA protocol shall be
used instead of the Multiword DMA protocol when these commands are issued by the host. This protocol
applies to the Ultra DMA data burst only. When this protocol is used there are no changes to other elements
of the ATA protocol (e.g., Command Block Register access).
Page 25
T13/1321D revision 3
Several signal lines are redefined to provide different functions during an Ultra DMA burst. These lines assume
these definitions when:
1) an Ultra DMA mode is selected, and
2) a host issues a READ DMA, WRITE DMA, READ DMA QUEUED, WRITE DMA QUEUED, or a
PACKET command requiring data transfer, and
3) the host asserts DMACK-.
These signal lines revert back to the definitions used for non-Ultra DMA transfers upon the negation of DMACKby the host at the termination of an Ultra DMA burst. All of the control signals are unidirectional. DMARQ and
DMACK- retain their standard definitions.
With the Ultra DMA protocol, the control signal (STROBE) that latches data from DD(15:0) is generated by the
same agent (either host or device) that drives the data onto the bus. Ownership of DD(15:0) and this data
strobe signal are given either to the device during an Ultra DMA data-in burst or to the host for an Ultra DMA
data-out burst.
During an Ultra DMA burst a sender shall always drive data onto the bus, and, after a sufficient time to allow for
propagation delay, cable settling, and setup time, the sender shall generate a STROBE edge to latch the data.
Both edges of STROBE are used for data transfers so that the frequency of STROBE is limited to the maximum
frequency of the data. The highest fundamental frequency on the cable shall be 16.67 MHz which equates to
33.33 million transitions per second for any signal.
Words in the IDENTIFY DEVICE data indicate support of the Ultra DMA feature and the Ultra DMA modes the
device is capable of supporting. The Set transfer mode subcommand in the SET FEATURES command shall
be used by a host to select the Ultra DMA mode that the device shall use. The Ultra DMA mode selected by a
host shall be less than or equal to the fastest mode that the device supports. Only one Ultra DMA mode shall
be selected at any given time. All timing requirements for a selected Ultra DMA mode shall be satisfied.
Devices supporting Ultra DMA mode 4 shall also support Ultra DMA modes 3, 2, 1, and 0. Devices supporting
Ultra DMA mode 3 shall also support ultra DMA modes 2, 1, and 0. Devices supporting Ultra DMA mode 2
shall also support Ultra DMA modes 0 and 1. Devices supporting Ultra DMA mode 1 shall also support Ultra
DMA mode 0. Hosts supporting Ultra DMA may support any combination of modes.
An Ultra DMA capable device shall retain the previously selected Ultra DMA mode after executing a software
reset sequence or the sequence caused by receipt of a DEVICE RESET command. An Ultra DMA capable
device shall clear any previously selected Ultra DMA mode and revert to the default non-Ultra DMA modes after
executing a power-on or hardware reset.
Both the host and device perform a CRC function during an Ultra DMA burst. At the end of an Ultra DMA burst
the host sends its CRC data to the device. The device compares its CRC data to the data sent from the host.
If the two values do not match, the device reports an error in the error register. If an error occurs during one or
more Ultra DMA bursts for any one command, the device shall report the first error that occurred. If the device
detects that a CRC error has occurred before data transfer for the command is complete, the device may
complete the transfer and report the error or abort the command and report the error.
NOTE − If a data transfer is terminated before completion, the assertion of INTRQ should be
passed through to the host software driver regardless of whether all data requested by the
command has been transferred.
6.6.2 Phases of operation
An Ultra DMA data transfer is accomplished through a series of Ultra DMA data-in or data-out bursts. Each
Ultra DMA burst has three mandatory phases of operation: the initiation phase, the data transfer phase, and the
Ultra DMA burst termination phase. In addition, an Ultra DMA burst may be paused during the data transfer
phase (see 9.13 and 9.14 for the detailed protocol descriptions for each of these phases, 10.2.4 defines the
Page 26
T13/1321D revision 3
specific timing requirements). In the following rules DMARDY- is used in cases that could apply to either
DDMARDY- or HDMARDY-, and STROBE is used in cases that could apply to either DSTROBE or HSTROBE.
The following are general Ultra DMA rules.
1) An Ultra DMA burst is defined as the period from an assertion of DMACK- by the host to the subsequent
negation of DMACK-.
2) When operating in Ultra DMA modes 2, 1, or 0 a recipient shall be prepared to receive up to two data
words whenever an Ultra DMA burst is paused. When operating in Ultra DMA modes 4 or 3 a recipient
shall be prepared to receive up to three data words whenever an Ultra DMA burst is paused.
6.6.2.1 Ultra DMA burst initiation phase rules
1) An Ultra DMA burst initiation phase begins with the assertion of DMARQ by a device and ends when the
sender generates a STROBE edge to transfer the first data word.
2) An Ultra DMA burst shall always be requested by a device asserting DMARQ.
3) When ready to initiate the requested Ultra DMA burst, the host shall respond by asserting DMACK-.
4) A host shall never assert DMACK- without first detecting that DMARQ is asserted.
5) For Ultra DMA data-in bursts: a device may begin driving DD(15:0) after detecting that DMACK- is
asserted, STOP negated, and HDMARDY- is asserted.
6) After asserting DMARQ or asserting DDMARDY- for an Ultra DMA data-out burst, a device shall not negate
either signal until the first STROBE edge is generated.
7) After negating STOP or asserting HDMARDY- for an Ultra DMA data-in burst, a host shall not change the
state of either signal until the first STROBE edge is generated.
6.6.2.2 Data transfer phase rules
1) The data transfer phase is in effect from after Ultra DMA burst initiation until Ultra DMA burst termination.
2) A recipient pauses an Ultra DMA burst by negating DMARDY- and resumes an Ultra DMA burst by
reasserting DMARDY-.
3) A sender pauses an Ultra DMA burst by not generating STROBE edges and resumes by generating
STROBE edges.
4) A recipient shall not signal a termination request immediately when the sender stops generating STROBE
edges. In the absence of a termination from the sender the recipient shall always negate DMARDY- and
wait the required period before signaling a termination request.
5) A sender may generate STROBE edges at greater than the minimum period specified by the enabled Ultra
DMA mode. The sender shall not generate STROBE edges at less than the minimum period specified by
the enabled Ultra DMA mode. A recipient shall be able to receive data at the minimum period specified by
the enabled Ultra DMA mode.
6.6.2.3 Ultra DMA burst termination phase rules
1) Either a sender or a recipient may terminate an Ultra DMA burst.
2) Ultra DMA burst termination is not the same as command completion. If an Ultra DMA burst termination
occurs before command completion, the command shall be completed by initiation of a new Ultra DMA
burst at some later time or aborted by the host issuing a hardware or software reset to the device.
3) An Ultra DMA burst shall be paused before a recipient requests a termination.
4) A host requests a termination by asserting STOP. A device acknowledges a termination request by
negating DMARQ.
5) A device requests a termination by negating DMARQ. A host acknowledges a termination request by
asserting STOP.
6) Once a sender requests a termination, the sender shall not change the state of STROBE until the recipient
acknowledges the request. Then, if STROBE is not in the asserted state, the sender shall return STROBE
to the asserted state. No data shall be transferred on this transition of STROBE.
7) A sender shall return STROBE to the asserted state whenever the sender detects a termination request
from the recipient. No data shall be transferred nor CRC calculated on this edge of DSTROBE.
Page 27
T13/1321D revision 3
8) Once a recipient requests a termination, the responder shall not change DMARDY from the negated state
for the remainder of an Ultra DMA burst.
9) A recipient shall ignore a STROBE edge when DMARQ is negated or STOP is asserted.
6.7 Host determination of cable type by detecting CBLIDIn a system using a cable, hosts shall determine that an 80-conductor cable is installed in a system before
operating with transfer modes faster than Ultra DMA mode 2. Hosts shall detect that CBLID- is connected to
ground to determine the cable type. See Annex B.
For detecting that CBLID- is connected to ground, the host shall test to see if CBLID- is below VIL or above VIH.
If the signal is below VIL, then an 80-conductor cable assembly is installed in the system because this signal is
grounded in the 80-conductor cable assembly’s host connector. If the signal is above VIH, then a 40-conductor
cable assembly is installed because this signal is connected to the device(s) and is pulled up through a 10 kΩ
resistor at each device.
Host
connector
Device 0
connector
Host PCB
N/C
Device 1
connector
PDIAG-:CBLIDconductor
Device 0
PCB
Device 1
PCB
NOTES −
1 For this configuration hosts shall not set devices to operate at Ultra DMA modes greater
than 2.
2 N/C indicates that there is no connection from the host to this signal conductor.
Figure 6 – Example configuration of a system with a 40-conductor cable
Page 28
T13/1321D revision 3
Host
connector
Host PCB
Device 0
connector
Device 1
connector
PDIAG-:CBLIDconductor
Device 0
PCB
Device 1
PCB
Figure 7 – Example configuration of a system where the host detects a 40-conductor cable
Host
connector
Host PCB
N/C
Device 1
connector
Device 0
connector
PDIAG-:CBLIDconductor
Device 1
PCB
Device 0
PCB
NOTE − N/C indicates that there is no connection from the host to this signal conductor.
Figure 8 – Example configuration of a system where the host detects an 80-conductor cable
Table 8 – Host detection of CBLIDCable
Device 1
Electrical state
Host-determined Determination
assembly
releases
of CBLID- at
cable type
correct?
type
PDIAGhost
40-conductor
Yes
1
40-conductor
Yes
80-conductor
Yes
0
80-conductor
Yes
40-conductor
No
0
80-conductor
No (see note)
80-conductor
No
0
80-conductor
Yes
NOTE − Ultra DMA mode 3 or 4 may be set incorrectly resulting in ICRC errors.
6.8 PACKET Command feature set
The PACKET Command feature set provides for devices that require command parameters that are too
extensive to be expressed in the Command Block registers. Devices implementing the PACKET Command
feature set exhibit responses different from those exhibited by devices not implementing this feature set.
The commands unique to the PACKET Command feature set are:
Page 29
T13/1321D revision 3
−
−
−
PACKET
DEVICE RESET
IDENTIFY PACKET DEVICE
6.8.1 Identification of PACKET Command feature set devices
When executing a power-on, hardware, DEVICE RESET, or software reset, a device implementing the PACKET
Command feature set performs the same reset protocol as other devices but leaves the registers with a
signature unique to PACKET Command feature set devices (see 9.12).
In addition, the IDENTIFY DEVICE command shall not be executed but shall be command aborted and shall
return a signature unique to devices implementing the PACKET Command feature set. The IDENTIFY PACKET
DEVICE command is used by the host to get identifying parameter information for a device implementing the
PACKET Command feature set (see 8.12.5.2 and 8.13).
6.8.2 PACKET Command feature set resets
Devices implementing the PACKET Command feature set respond to power-on, hardware, and software resets
as any other device except for the resulting contents in the device registers as described above. However,
software reset should not be issued while a PACKET command is in progress. PACKET commands utilized by
some devices do not terminate if a software reset is issued.
The DEVICE RESET command is provided to allow the device to be reset without affecting the other device on
the bus.
6.8.3 The PACKET command
The PACKET command allows a host to send a command to the device via a command packet. The command
packet contains the command and command parameters that the device is to execute.
Upon receipt of the PACKET command the device sets BSY to one and prepares to receive the command
packet. When ready, the device sets DRQ to one and clears BSY to zero. The command packet is then
transferred to the device by PIO transfer. When the last word of the command packet is transferred, the device
sets BSY to one, and clears DRQ to zero (see 8.21 and 9.8).
6.9 Overlapped feature set
Overlap allows devices that require extended command time to perform a bus release so that the other device
on the bus may be used. To perform a bus release the device shall clear both DRQ and BSY to zero. When
selecting the other device during overlapped operations, the host shall disable assertion of INTRQ via the nIEN
bit on the currently selected device before writing the Device/Head register to select the other device.
The only commands that may be overlapped are:
−
−
−
−
−
NOP (with a subcommand code other than 00h)
PACKET
READ DMA QUEUED
SERVICE
WRITE DMA QUEUED
For the PACKET command, overlap is indicated by the OVL bit in the Features register when the PACKET
command is issued.
If the device supports PACKET command overlap, the OVL bit is set to one in the Features register and the
Release interrupt has been enabled via the SET FEATURES command, then the device shall perform a bus
Page 30
T13/1321D revision 3
release when the command packet has been received. This allows the host to select the other device to
execute commands. When the device is ready to continue the command, the device sets SERV to one, and
asserts INTRQ if selected and nIEN is cleared to zero. The host then issues the SERVICE command to
continue the execution of the command
If the device supports PACKET command overlap, the OVL bit is set to one in the Features register and the
Release interrupt has been disabled via the SET FEATURES command, then the device may or may not
perform a bus release. If the device is ready to complete execution of the command, the device may complete
the command immediately as described in the non-overlap case. If the device is not ready to complete
execution of the command, the device may perform a bus release and complete the command as described in
the previous paragraph.
For the READ DMA QUEUED and WRITE DMA QUEUED commands, the device may or may not perform a
bus release. If the device is ready to complete execution of the command, the device may complete the
command immediately. If the device is not ready to complete execution of the command, the device may
perform a bus release and complete the command via a service request.
If a device has an outstanding command that has been released, the device can only indicate that service is
required when the device is selected. This implies that the host has to poll each device to determine if a device
is requesting service. The polling can be performed at the host either by hardware or by a software routine. The
latter implies a considerable host processor overhead. Hardware polling is initiated by the NOP Auto Poll
command.
The NOP Poll command is a host adapter function and is ignored by the device. The host software can test for
the support of this feature by issuing the NOP Auto Poll subcommand and examining the Status register. If the
host adapter does not support this feature, the response received by the host will be from the device with the
ERR bit set to one. If the host adapter does support the command, the response will be from the host adapter
with the ERR bit cleared to zero. The only action taken by a device supporting the Overlapped feature set will
be to return the error indication in the Status register and to not abort any outstanding commands.
6.10 Queued feature set
Command queuing allows the host to issue concurrent commands to the same device. Only commands
included in the Overlapped feature set may be queued. The queue contains all commands for which command
acceptance has occurred but command completion has not occurred. If a queue exists when a non-queued
command is received, the non-queued command shall be command aborted and the commands in the queue
shall be discarded. The ending status shall be command aborted and the results are indeterminate.
The maximum queue depth supported by a device shall be indicated in word 75 of the IDENTIFY DEVICE or
IDENTIFY PACKET DEVICE response.
A queued command shall have a Tag provided by the host in the Sector Count register to uniquely identify the
command. When the device restores register parameters during the execution of the SERVICE command, this
Tag shall be restored so that the host may identify the command for which status is being presented. A Tag
value may be any value between 0 and 31, regardless of the queue depth supported. If a queued command is
issued with a Tag value that is identical to the Tag value for a command already in the queue, the entire queue
shall be aborted including the new command. The ending status shall be command aborted and the results are
indeterminate. If any error occurs, the command queue shall be aborted.
When the device is ready to continue the processing of a bus released command and BSY and DRQ are both
cleared to zero, the device requests service by setting SERV to one, setting a pending interrupt, and asserting
INTRQ if selected and if nIEN is cleared to zero. SERV shall remain set until all commands ready for service
have been serviced. A read of the Status register or a write of the Command register shall clear the interrupt
pending.
Page 31
T13/1321D revision 3
When the device is ready to continue the processing of a bus released command and BSY or DRQ is set to
one (i.e., the device is processing another command on the bus), the device requests service by setting SERV
to one. SERV shall remain set until all commands ready for service have been serviced. At command
completion of the current command processing (i.e., when both BSY and DRQ are cleared to zero), the device
shall process interrupt pending and INTRQ per the protocol for the command being completed. No additional
INTRQ assertion shall occur due to other commands ready for service until after the device’s SERV bit has
been cleared to zero.
When the device receives a new command while queued commands are ready for service, the device shall
execute the new command and process interrupt pending and INTRQ per the protocol for the new command. If
the queued commands ready for service still exist at command completion of this command, SERV remains
set to one but no additional INTRQ assertion shall occur due to commands ready for service.
When queuing commands, the host shall disable INTRQ assertion via the nIEN bit before writing a new
command to the Command register and may re-enable INTRQ assertion after writing the command. When
reading status at command completion of a command, the host shall check the SERV bit since the SERV bit
may be set because the device is ready for service associated with another command. The host receives no
additional INTRQ assertion to indicate that a queued command is ready for service.
6.11 Power Management feature set
A device shall implement power management. A device implementing the PACKET Command feature set may
implement the power management as defined by the packet command set implemented by the device.
Otherwise, the device shall implement the Power Management feature set as described in this standard.
The Power Management feature set permits a host to modify the behavior of a device in a manner that reduces
the power required to operate. The Power Management feature set provides a set of commands and a timer
that enable a device to implement low power consumption modes. A register delivered command device that
implements the Power Management feature set shall implement the following minimum set of functions:
−
−
−
−
−
−
−
A Standby timer
CHECK POWER MODE command
IDLE command
IDLE IMMEDIATE command
SLEEP command
STANDBY command
STANDBY IMMEDIATE command
A device that implements the PACKET Command feature set and implements the Power Management feature
set shall implement the following minimum set of functions:
−
−
−
−
CHECK POWER MODE command
IDLE IMMEDIATE command
SLEEP command
STANDBY IMMEDIATE command
6.11.1 Power management commands
The CHECK POWER MODE command allows a host to determine if a device is currently in, going to or leaving
Standby or Idle mode. The CHECK POWER MODE command shall not change the power mode or affect the
operation of the Standby timer.
Page 32
T13/1321D revision 3
The IDLE and IDLE IMMEDIATE commands move a device to Idle mode immediately from the Active or
Standby modes. The IDLE command also sets the Standby timer count and enables or disables the Standby
timer.
The STANDBY and STANDBY IMMEDIATE commands move a device to Standby mode immediately from the
Active or Idle modes. The STANDBY command also sets the Standby timer count and enables or disables the
Standby timer.
The SLEEP command moves a device to Sleep mode. The device's interface becomes inactive at command
completion of the SLEEP command. A hardware or software reset or DEVICE RESET command is required to
move a device out of Sleep mode.
6.11.2 Standby timer
The Standby timer provides a method for the device to automatically enter Standby mode from either Active or
Idle mode following a host programmed period of inactivity. If the Standby timer is enabled and if the device is
in the Active or Idle mode, the device waits for the specified time period and if no command is received, the
device automatically enters the Standby mode.
If the Standby timer is disabled, the device may not automatically enter Standby mode.
6.11.3 Power modes
Figure 9 shows the minimum set of mode transitions that shall be implemented.
Page 33
T13/1321D revision 3
PM0: Active
reset
Power-up with Power-up in Standby
feature not implemented or not enabled
PM0:PM0
SLEEP command
PM0:PM3
PM1: Idle
PM2: Standby
Power-up with
Power-up in
Standby
implemented
and enabled
STANDBY or STANDBY
IMMEDIATE command,
vendor specific
implementation, or
Standby timer expiration
PM0:PM2
reset
IDLE or IDLE
IMMEDIATE
command, or
vendor specific
implementation
PM1:PM1
PM0:PM1
Media access required
PM2:PM0
Media access required
reset
PM1:PM0
PM2:PM2
STANDBY or STANDBY
IMMEDIATE command, vendor
specific implementation, or
Standby timer expiration
PM1:PM2
IDLE or IDLE IMMEDIATE
command
PM2:PM1
PM3: Sleep
SLEEP command
SLEEP command
PM2:PM3
PM1:PM3
reset
PM3:PM2
Figure 9 − Power management state diagram
PM0: Active: This mode shall be entered when the device receives a media access command while in Idle
or Standby mode. This mode shall also be entered when the device is powered-up with the Power-Up In
Standby feature not implemented or not enabled (see 6.18).
In Active mode the device is capable of responding to commands. During the execution of a media access
command a device shall be in Active mode. Power consumption is greatest in this mode.
Transition PM0:PM0: When hardware reset, software rest, or DEVICE RESET command is received, the
device shall make a transition to the PM0: Active mode when the reset protocol is completed.
Transition PM0:PM1: When an IDLE or IDLE IMMEDIATE command is received or when a vendor specific
implementation determines a transition is required, then the device shall make a transition to the PM1:Idle
mode.
Page 34
T13/1321D revision 3
Transition PM0:PM2: When a STANDBY or STANDBY IMMEDIATE command is received, the Standby timer
expires, or a vendor specific implementation determines a transition is required, then the device shall make a
transition to the PM2:Standby mode.
Transition PM0:PM3: When a SLEEP command is received, the device shall make a transition to the
PM3:Sleep mode.
PM1: Idle: This mode shall be entered when the device receives an IDLE or IDLE IMMEDIATE command.
Some devices may perform vendor specific internal power management and make a transition to the Idle mode
without host intervention.
In Idle mode the device is capable of responding to commands but the device may take longer to complete
commands than when in the Active mode. Power consumption may be reduced from that of Active mode.
Transition PM1:PM0: When a media access is required, the device shall make a transition to the PM0:Active
mode.
Transition PM1:PM1: When hardware reset, software reset, or DEVICE RESET command is received, the
device shall make a transition to the PM1:Idle mode when the reset protocol is completed.
Transition PM1:PM2: When a STANDBY or STANDBY IMMEDIATE command is received, the Standby timer
expires, or a vendor specific implementation determines a transition is required, then the device shall make a
transition to the PM2:Standby mode.
Transition PM1:PM3: When a SLEEP command is received, the device shall make a transition to the
PM3:Sleep mode.
PM2: Standby: This mode shall be entered when the device receives a STANDBY command, a STANDBY
IMMEDIATE command, or the Standby timer expires. Some devices may perform vendor specific internal power
management and make a transition to the Standby mode without host intervention. This mode shall also be
entered when the device is powered-up with the Power-Up In Standby feature implemented and enabled.
In Standby mode the device is capable of responding to commands but the device may take longer to complete
commands than in the Idle mode. The time to respond could be as long as 30 s. Power consumption may be
reduced from that of Idle mode.
Transition PM2:PM0: When a media access is required, the device shall make a transition to the PM0:Active
mode.
Transition PM2:PM1: When an IDLE or IDLE IMMEDIATE command is received, or a vendor specific
implementation determines a transition is required, then the device shall make a transition to the PM1:Idle
mode.
Transition PM2:PM2: When hardware reset, software reset, or DEVICE RESET command is received, the
device shall make a transition to the PM2:Standby mode when the reset protocol is completed.
Transition PM2:PM3: When a SLEEP command is received, the device shall make a transition to the
PM3:Sleep mode.
PM3: Sleep: This mode shall be entered when the device receives a SLEEP command.
In Sleep mode the device requires a hardware or software reset or a DEVICE RESET command to be activated.
The time to respond could be as long as 30 s. Sleep mode provides the lowest power consumption of any
mode.
Page 35
T13/1321D revision 3
In Sleep mode, the device's interface is not active. The content of the Status register is invalid in this mode.
Transition PM3:PM2:, When hardware reset, software reset, or DEVICE RESET command is received the
device shall make a transition to the PM2:Standby mode.
6.12 Advanced Power Management feature set
The Advanced Power Management feature set is an optional feature set that allows the host to select a power
management level. The power management level is a scale from the lowest power consumption setting of 01h
to the maximum performance level of FEh. Device performance may increase with increasing power
management levels. Device power consumption may increase with increasing power management levels. A
device may implement one power management method for two or more contiguous power management levels.
For example, a device may implement one power management method from level 80h to A0h and a higher
performance, higher power consumption method from level A1h to FEh. Advanced power management levels
80h and higher do not permit the device to spin down to save power.
The Advanced Power Management feature set uses the following functions:
−
−
A SET FEATURES subcommand to enable Advanced Power Management
A SET FEATURES subcommand to disable Advanced Power Management
Advanced Power Management is independent of the Standby timer setting. If both Advanced Power
Management and the Standby timer are set, the device will go to the Standby state when the timer times out or
the device’s Advanced Power Management algorithm indicates that the Standby state should be entered.
The IDENTIFY DEVICE indicates that Advanced Power Management is supported, if Advanced Power
Management is enabled, and the current advanced power management level if Advanced Power Management is
enabled.
6.13 Security Mode feature set
The optional Security Mode feature set is a password system that restricts access to user data stored on a
device. The system has two passwords, User and Master and two security levels, High and Maximum. The
security system is enabled by sending a user password to the device with the SECURITY SET PASSWORD
command. When the security system is enabled, access to user data on the device is denied after a power
cycle until the User password is sent to the device with the SECURITY UNLOCK command.
A Master password may be set in a addition to the User password. The purpose of the Master password is to
allow an administrator to establish a password that is kept secret from the user, and which may be used to
unlock the device if the User password is lost. Setting the Master password does not enable the password
system.
The security level is set to High or Maximum with the SECURITY SET PASSWORD command. The security
level determines device behavior when the Master password is used to unlock the device. When the security
level is set to High the device requires the SECURITY UNLOCK command and the Master password to unlock.
When the security level is set to Maximum the device requires a SECURITY ERASE PREPARE command and
a SECURITY ERASE UNIT command with the master password to unlock. Execution of the SECURITY
ERASE UNIT command erases all user data on the device.
The SECURITY FREEZE LOCK command prevents changes to passwords until a following power cycle. The
purpose of the SECURITY FREEZE LOCK command is to prevent password setting attacks on the security
system.
A device that implements the Security Mode feature set shall implement the following minimum set of
commands:
Page 36
T13/1321D revision 3
−
−
−
−
−
−
SECURITY SET PASSWORD
SECURITY UNLOCK
SECURITY ERASE PREPARE
SECURITY ERASE UNIT
SECURITY FREEZE LOCK
SECURITY DISABLE PASSWORD
Support of the Security Mode feature set is indicated in IDENTIFY DEVICE word 128.
6.13.1 Security mode initial setting
When the device is shipped by the manufacturerer, the state of the Security Mode feature shall be disabled.
The initial Master password value is not defined by this standard.
If the Master Password Revision Code feature is supported, the Master Password Revision Code shall be set to
FFFEh by the manufacturer.
6.13.2 User password lost
If the User password sent to the device with the SECURITY UNLOCK command does not match the user
password previously set with the SECURITY SET PASSWORD command, the device shall not allow the user
to access data.
If the Security Level was set to High during the last SECURITY SET PASSWORD command, the device shall
unlock if the Master password is received.
If the Security Level was set to Maximum during the last SECURITY SET PASSWORD command, the device
shall not unlock if the Master password is received. The SECURITY ERASE UNIT command shall erase all
user data and unlock the device if the Master password matches the last Master password previously set with
the SECURITY SET PASSWORD command.
6.13.3 Attempt limit for SECURITY UNLOCK command
The device shall have an attempt limit counter. The purpose of this counter is to defeat repeated trial attacks.
After each failed User or Master password SECURITY UNLOCK command, the counter is decremented. When
the counter value reaches zero the EXPIRE bit (bit 4) of word 128 in the IDENTIFY DEVICE information is set to
one, and the SECURITY UNLOCK and SECURITY UNIT ERASE commands are command aborted until the
device is powered off or hardware reset. The EXPIRE bit shall be cleared to zero after power-on or hardware
reset. The counter shall be set to five after a power-on or hardware reset.
6.13.4 Security mode states
Figure 8 decribes security mode states and state transitions.
Page 37
T13/1321D revision 3
SEC2:Security disabled/Frozen
SEC0:Powered down /
Security disabled
Power-down
SEC2:SEC0
SEC1:Security disabled/not Frozen
RESET- asserted
SEC2:SEC1
Power-up
SEC0:SEC1
Power-down
SEC1:SEC0
SECURITY FREEZE
LOCK command
SEC1:SEC2
SEC5:Unlocked/not Frozen
RESET- asserted
SEC1:SEC1
SECURITY SET
PASSWORD command
SEC1:SEC5
SECURITY DISABLE
PASSWORD command
SEC5a:SEC1
SECURITY ERASE
UNIT command
SEC5b:SEC1
SEC4: Security enabled /
Locked
SEC3:Powered down /
Security enabled
Power-up
SEC3:SEC4
SECURITY ERASE SECURITY UNLOCK
command
UNIT command
SEC4:SEC5
SEC4:SEC1
RESET- asserted
SEC5:SEC4
Power-down
SEC4:SEC3
RESET- asserted
SEC4:SEC4
SEC6:Unlocked/Frozen
RESET- asserted
SEC6:SEC4
Power-down
SEC6:SEC3
SECURITY FREEZE
LOCK command
SEC5:SEC6
Power-down
SEC5:SEC3
Figure 10 − Security mode state diagram
SEC0: Powered down/Security disabled: This mode shall be entered when the device is powereddown with the Security Mode feature set disabled.
Transition SEC0:SEC1: When the device is powered-up, the device shall make a transition to the SEC1:
Security disabled/not Frozen state.
Page 38
T13/1321D revision 3
SEC1: Security disabled/not Frozen: This mode shall be entered when the device is powered-up or
a hardware reset is received with the Security Mode feature set disabled or when the Security Mode feature set
is disabled by a SECURITY DISABLE PASSWORD command.
In this state, the device is capable of responding to all commands (see Table 9 Unlocked column).
Transition SEC1:SEC0: When the device is powered-down, the device shall make a transition to the SEC0:
Powered down/Security disabled state.
Transition SEC1:SEC1: When the device receives a hardware reset, the device shall make a transition to the
SEC1: Security disabled/not Frozen state.
Transition SEC1:SEC2: When a SECURITY FREEZE LOCK command is received, the device shall make a
transition to the SEC2: Security disabled/Frozen state.
Transition SEC1:SEC5: When a SECURITY SET PASSWORD command is received, the device shall make a
transition to the SEC5: Unlocked/not frozen state
SEC2: Security disabled/ Frozen: This mode shall be entered when the device receives a SECURITY
FREEZE LOCK command while in Security disabled/not Frozen state.
In this state, the device is capable of responding to all commands except those indicated in Table 9 Frozen
column.
Transition SEC2:SEC0: When the device is powered-down, the device shall make a transition to the SEC0:
Powered down/Security disabled state.
Transition SEC2:SEC1: When the device receives a hardware reset, the device shall make a transition to the
SEC1: Security disabled/not Frozen state.
SEC3: Powered down/Security enabled: This mode shall be entered when the device is powereddown with the Security Mode feature set enabled.
Transition SEC3:SEC4: When the device is powered-up, the device shall make a transition to the SEC4:
Security enabled/locked state.
SEC4: Security enabled/locked: This mode shall be entered when the device is powered-up with the
Security Mode feature set enabled.
In this state, the device shall only respond to commands that do not access data in the user data area of the
media (see Table 9 Locked column).
Transition SEC4:SEC3: When the device is powered-down, the device shall make a transition to the SEC3:
Powered down/Security enabled state.
Transition SEC4:SEC4: When the device receives a hardware reset, the device shall make a transition to the
SEC4: Security enabled/locked state.
Transition SEC4:SEC5: When a valid SECURITY UNLOCK command is received, the device shall make a
transition to the SEC5: Unlocked/not Frozen state.
Transition SEC4:SEC1: When a SECURITY ERASE PREPARE command is received and is followed by a
SECURITY ERASE UNIT command, the device shall make a transition to the SEC1: Security disabled/not
Frozen state.
Page 39
T13/1321D revision 3
SEC5: Unlocked/not Frozen: This mode shall be entered when the device receives a SECURITY SET
PASSWORD command to enable the lock or a SECURITY UNLOCK command.
In this state, the device shall respond to all commands (see Table 9 Unlocked column).
Transition SEC5a:SEC1: When a valid SECURITY DISABLE PASSWORD command is received, the device
shall make a transition to the SEC1: Security disabled/not Frozen state.
Transition SEC5b:SEC1: When a SECURITY ERASE PREPARE command is received and is followed by a
SECURITY ERASE UNIT command, the device shall make a transition to the SEC1: Security disabled/not
Frozen state.
Transition SEC5:SEC6: When a SECURITY FREEZE LOCK command is received, the device shall make a
transition to the SEC6: Unlocked/Frozen state.
Transition SEC5:SEC3: When the device is powered-down, the device shall make a transition to the SEC3:
Powered down/Security enabled state.
Transition SEC5:SEC4: When the device receives a hardware reset, the device shall make a transition to the
SEC4: Security enabled/not Frozen state.
SEC6: Unlocked/ Frozen: This mode shall be entered when the device receives a SECURITY FREEZE
LOCK command while in Unlocked/not Frozen state.
In this state, the device is capable of responding to all commands except those indicated in Table 9 Frozen
column.
Transition SEC6:SEC3: When the device is powered-down, the device shall make a transition to the SEC3:
Powered down/Security enabled state.
Transition SEC6:SEC4: When the device receives a hardware reset, the device shall make a transition to the
SEC4: Security enabled/not Frozen state.
Page 40
T13/1321D revision 3
Table 9 − Security mode command actions
Command
CHECK POWER MODE
CFA REQUEST EXTENDED ERROR CODE
CFA ERASE SECTORS
CFA TRANSLATE SECTOR
CFA WRITE MULTIPLE WITHOUT ERASE
CFA WRITE SECTORS WITHOUT ERASE
DEVICE RESET
DOWNLOAD MICROCODE
EXECUTE DEVICE DIAGNOSTIC
FLUSH CACHE
GET MEDIA STATUS
IDENTIFY DEVICE
IDENTIFY PACKET DEVICE
IDLE
IDLE IMMEDIATE
INITIALIZE DEVICE PARAMETERS
MEDIA EJECT
MEDIA LOCK
MEDIA UNLOCK
NOP
PACKET
READ BUFFER
READ DMA
READ DMA QUEUED
READ MULTIPLE
READ NATIVE MAX ADDRESS
READ SECTORS
READ VERIFY SECTORS
SECURITY DISABLE PASSWORD
SECURITY ERASE PREPARE
SECURITY ERASE UNIT
SECURITY FREEZE LOCK
SECURITY SET PASSWORD
SECURITY UNLOCK
SEEK
SERVICE
SET FEATURES
SET MAX ADDRESS
SET MULTIPLE MODE
SLEEP
SMART DISABLE OPERATIONS
SMART ENABLE/DISABLE AUTOSAVE
SMART ENABLE OPERATIONS
SMART EXECUTE OFF-LINE IMMEDIATE
SMART READ DATA
SMART READ LOG
SMART RETURN STATUS
SMART SAVE ATTRIBUTE VALUES
SMART WRITE LOG
STANDBY
STANDBY IMMEDIATE
WRITE BUFFER
WRITE DMA
WRITE DMA QUEUED
WRITE MULTIPLE
WRITE SECTORS
Locked
Executable
Executable
Command aborted
Executable
Command aborted
Command aborted
Executable
Executable
Executable
Command aborted
Command aborted
Executable
Executable
Executable
Executable
Executable
Command aborted
Command aborted
Command aborted
Executable
Command aborted
Executable
Command aborted
Command aborted
Command aborted
Executable
Command aborted
Command aborted
Command aborted
Executable
Executable
Command aborted
Command aborted
Executable
Executable
Command aborted
Executable
Command aborted
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Command aborted
Command aborted
Command aborted
Command aborted
Unlocked
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Frozen
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Command aborted
Command aborted
Command aborted
Executable
Command aborted
Command aborted
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Executable
Page 41
T13/1321D revision 3
6.14 Self-monitoring, analysis, and reporting technology feature set
The intent of self-monitoring, analysis, and reporting technology (the SMART feature set) is to protect user data
and minimize the likelihood of unscheduled system downtime that may be caused by predictable degradation
and/or fault of the device. By monitoring and storing critical performance and calibration parameters, SMART
feature set devices attempt to predict the likelihood of near-term degradation or fault condition. Providing the
host system the knowledge of a negative reliability condition allows the host system to warn the user of the
impending risk of a data loss and advise the user of appropriate action. Support of this feature set is indicated
in the IDENTIFY DEVICE response.
Devices that implement the PACKET Command feature set shall not implement the SMART feature set as
described in this subclause. Devices that implement the PACKET Command feature set and SMART shall
implement SMART as defined by the command packet set implemented by the device.
6.14.1 Device SMART data structure
SMART feature set capability and status information for the device are stored in the device SMART data
structure. The off-line data collection capability and status data stored herein may be useful to the host if the
SMART EXECUTE OFF-LINE IMMEDIATE command is implemented (see 8.41.4).
6.14.2 On-line data collection
Collection of SMART data in an “on-line” mode shall have no impact on device performance. The SMART data
that is collected or the methods by which data is collected in this mode may be different than those in the offline data collection mode for any particular device and may vary from one device to another.
6.14.3 Off-line data collection
The device shall use off-line mode for data collection and self-test routines that have an impact on performance
if the device is required to respond to commands from the host while performing that data collection. This
impact on performance may vary from device to device. The data that is collected or the methods by which the
data is collected in this mode may be different than those in the on-line data collection mode for any particular
device and may vary from one device to another.
6.14.4 Threshold exceeded condition
This condition occurs when the device’s SMART reliability status indicates an impending degrading or fault
condition.
6.14.5 SMART feature set commands
These commands use a single command code and are differentiated from one another by the value placed in
the Features register (see 8.41).
If the SMART feature set is implemented, the following commands shall be implemented.
−
−
−
−
SMART DISABLE OPERATIONS
SMART ENABLE/DISABLE AUTOSAVE
SMART ENABLE OPERATIONS
SMART RETURN STATUS
Page 42
T13/1321D revision 3
If the SMART feature set is implemented, the following commands may be implemented.
−
−
−
−
SMART EXECUTE OFF-LINE IMMEDIATE
SMART READ DATA
SMART READ LOG SECTOR
SMART WRITE LOG SECTOR
6.14.6 SMART operation with power management modes
When used with a host that has implemented the Power Management feature set, a SMART enabled device
should automatically save the device accumulated SMART data upon receipt of an IDLE IMMEDIATE,
STANDBY IMMEDIATE, or SLEEP command or upon return to an Active or Idle mode from a Standby mode
(see 8.41.5).
If a SMART feature set enabled device has been set to utilize the Standby timer, the device should
automatically save the device accumulated SMART data prior to going from an Idle mode to the Standby mode
or upon return to an Active or Idle mode from a Standby mode.
A device shall not execute any routine to automatically save the device accumulated SMART data while the
device is in a Standby or Sleep mode.
6.14.7 SMART device error log reporting
Logging of reported errors is an optional SMART feature. If error logging is supported by a device, it is indicated
in byte 370 of the SMART READ DATA command response. If error logging is supported, the device shall
provide information on the last five errors that the device reported as described in the SMART READ LOG
SECTOR command (see 8.41.6). The device may also provide additional vendor specific information on these
reported errors.
If error logging is supported, it shall not be disabled when SMART is disabled. Error log information shall be
gathered at all times the device is powered-on except that logging of errors when in a reduced power mode is
optional. If errors are logged when in a reduced power mode, the reduced power mode shall not change.
Disabling SMART shall disable the delivering of error log information via the SMART READ LOG SECTOR
command.
If a device receives a firmware modification, all error log data shall be discarded and the device error count for
the life of the device shall be reset to zero.
6.15 Host Protected Area feature set
A reserved area for data storage outside the normal operating system file system is required for several
specialized applications. Systems may wish to store configuration data or save memory to the device in a
location that the operating systems cannot change. The Host Protected Area feature set allows a portion of the
device to be reserved for such an area when the device is initially configured. A device that implements the Host
Protected Area feature set shall implement the following minimum set of commands:
− READ NATIVE MAX ADDRESS
− SET MAX ADDRESS
Devices supporting this feature set shall set bit 10 of word 82 to one in the data returned by the IDENTIFY
DEVICE or IDENTIFY PACKET DEVICE command..
In addition a device supporting the Host Protected Area feature set may optionally include the security
extensions. The SET MAX commands use a single command code and are differentiated from one another by
the value placed in the Features register.
Page 43
T13/1321D revision 3
−
−
−
−
SET MAX SET PASSWORD
SET MAX LOCK
SET MAX FREEZE LOCK
SET MAX UNLOCK
Devices supporting these extensions shall set bit 10 of word 82 of the identify data to one and bit 8 of word 83
of the IDENTIFY DEVICE data shall be set to one.
If the Host Protected Area feature set is supported, the device shall indicate so in the IDENTIFY DEVICE
response.
The READ NATIVE MAX ADDRESS command allows the host to determine the maximum native address
space of the device even when a protected area has been allocated.
The SET MAX ADDRESS command allows the host to redefine the maximum address of the user accessible
address space. That is, when the SET MAX ADDRESS command is issued with a maximum address less than
the native maximum address, the device reduces the user accessible address space to the maximum set,
providing a protected area above that maximum address. The SET MAX ADDRESS command shall be
immediately preceded by a READ NATIVE MAX ADDRESS command. After the SET MAX ADDRESS
command has been issued, the device shall report only the reduced user address space in response to an
IDENTIFY DEVICE command in words 1, 54, 57, 58, 60, and 61. Any read or write command to an address
above the maximum address specified by the SET MAX ADDRESS command shall cause command
completion with the IDNF bit set to one and ERR set to one, or command aborted. A volatility bit in the Sector
Count register allows the host to specify if the maximum address set is preserved across power-on or hardware
reset cycles. On power-on or hardware reset the device maximum address returns to the last non-volatile
address setting regardless of subsequent volatile SET MAX ADDRESS commands. If the SET MAX ADDRESS
command is issued with a value that exceeds the native maximum address command aborted shall be
returned.
Typical use of these commands would be:
On reset
a) BIOS receives control after a system reset;
b) BIOS issues a READ NATIVE MAX ADDRESS command to find the max capacity of the device;
c) BIOS issues a SET MAX ADDRESS command to the values returned by READ NATIVE MAX
ADDRESS;
d) BIOS reads configuration data from the highest area on the disk;
e) BIOS issues a READ NATIVE MAX ADDRESS command followed by a SET MAX ADDRESS
command to reset the device to the size of the file system.
On save to disk
a) BIOS receives control prior to shut down;
b) BIOS issues a READ NATIVE MAX ADDRESS command to find the max capacity of the device;
c) BIOS issues a volatile SET MAX ADDRESS command to the values returned by READ NATIVE
MAX ADDRESS;
d) Memory is copied to the reserved area;
e) Shut down completes;
f) On power-on or hardware reset the device max address returns to the last non-volatile setting.
These commands are intended for use only by system BIOS or other low level boot time process. The process
should select a single addressing translation, CHS or LBA, for all SET MAX ADDRESS and READ NATIVE
MAX ADDRESS commands. Using these commands outside BIOS controlled boot or shutdown may result in
Page 44
T13/1321D revision 3
damage to file systems on the device. Devices should command aborted a second non-volatile SET MAX
ADDRESS command after a power-on or hardware reset
The SET MAX SET PASSWORD command allows the host to define the password to be used during the
current power-on cycle. The password does not persist over a power cycle but does persist over a hardware or
software reset. This password is not related to the password used for the Security Mode Feature set. When the
password is set the device is in the Set_Max_Unlocked mode.
The SET MAX LOCK command allows the host to disable the SET MAX commands (except SET MAX
UNLOCK) until the next power cycle or the issuance and acceptance of the SET MAX UNLOCK command.
When this command is accepted the device is in the Set_Max_Locked mode.
The SET MAX UNLOCK command changes the device from the Set_Max_Locked mode to the
Set_Max_Unlocked mode.
The SET MAX FREEZE LOCK command allows the host to disable the SET MAX commands (including SET
MAX UNLOCK) until the next power cycle. When this command is accepted the device is in the
Set_Max_Frozen mode.
6.15.1 BIOS determination of SET MAX security exension status
When the device is locked bit 8 of word 86 shall be set to one.
6.15.2 BIOS locking SET MAX
To allow for multiple BIOSs to gain access to the protected area the host BIOS should only lock the protected
area immediately prior to booting the operating system.
Page 45
T13/1321D revision 3
SM1:Set_Max_Unlocked
SM0:Set_Max_Security_Inactive
SET MAX ADDRESS
command
SET MAX SET
PASSWORD command
SM0:SM1
SET MAX SET
PASSWORD command
SM1b:SM1
SM0a:SM0
SET MAX ADDRESS
command
SET MAX LOCK, SET
MAX UNLOCK, or
SET MAX FREEZE
LOCK command
SM0b:SM0
SET MAX UNLOCK
command
SM1a:SM1
SM1c:SM1
Power on
SET MAX LOCK
command
SM1:SM2
SM2:Set_Max_Locked
SET MAX ADDRESS
or SET MAX SET
PASSWORD
command
SM2a:SM2
SET MAX LOCK
command
SET MAX UNLOCK
command
SM2:SM1
SET MAX FREEZE
LOCK command
(optional)
SM2:SM3
SM2b:SM2
SET MAX FREEZE
LOCK command
SM1:SM3
SM3:Set_Max_Frozen
SET MAX ADDRESS, SET
MAX SET PASSWORD, SET
MAX UNLOCK, SET MAX
FREEZE LOCK or SET MAX
LOCK command
SM3:SM3
Figure 11 − SET MAX security state diagram
SM0: Set_Max_Security_Inactive: This state shall be entered when the device is powered-on.
When in this state, SET MAX security is disabled.
Transition SM0a:SM0: When a SET MAX ADDRESS command is received, the command shall be executed
and the device shall make a transition to the SM0: Set_MAX_Security_Inactive state.
Transition SM0b:SM0: When a SET MAX LOCK, SET MAX UNLOCK, or SET MAX FREEZE LOCK command
is received, the device shall abort the command and make a transition to the SM0: Set_MAX_Security_Inactive
state.
Transition SM0:SM1: When a SET MAX-SET PASSWORD command is received, the device shall make a
transition to the SM1: Set_Max_Unlocked state.
SM1: Set_Max_Unlocked: This state is entered when a SET MAX SET PASSWORD or a SET MAX
UNLOCK command is received.
When in this state, a SET MAX security password has been established and the SET MAX security is
unlocked. Bit 8 of word 86 of the identify device data shall be set to one.
Page 46
T13/1321D revision 3
Transition SM1a:SM1: When a SET MAX ADDRESS command is received, the command shall be executed
and the device shall make a transition to the SM1: Set_MAX_Unlocked state.
Transition SM1b:SM1: When a SET MAX SET PASSWORD is received, the password stored by the device
shall be changed to the new value and the device shall make a transition to the SM1: Set_MAX_Unlocked
state.
Transition SM1c:SM1: When a SET MAX UNLOCK command is received, the command shall not be executed
and the device shall make a transition to the SM1: Set_MAX_Unlocked state.
Transition SM1:SM2: When a SET MAX LOCK command is received, the device shall make a transition to the
SM2: Set_Max_Locked state.
Transition SM1:SM3: When a SET MAX FREEZE LOCK command is received, the device shall make a
transition to the SM3: Set_Max_Frozen state.
SM2: Set_Max_Locked: This state is entered when a SET MAX LOCK command is received.
When in this state, a SET MAX security password has been established and the SET MAX security is locked.
Bit 8 of word 86 of the identify device data shall be set to one.
Transition SM2a:SM2: When a SET MAX ADDRESS or SET MAX SET PASSWORD command is received,
the command shall be aborted and the device shall make a transition to the SM2: Set_Max_Locked state
Transition SM2b:SM2: When a SET MAX LOCK command is received, the command shall be executed and
the device shall make a transition to the SM2: Set_Max_Locked state.
Transition SM2:SM1: When a SET MAX UNLOCK command is received, the device shall make a transition to
the SM1: Set Max Unlocked state.
Transition SM2:SM3: When a SET MAX FREEZE LOCK command is received, the device may make a
transition to the SM3: Set_Max_Frozen state. Hosts should not issue the SET MAX FREEZE LOCK command
when in this state. T13 intends to remove this transition in ATA/ATAPI-6.
SM3: Set_Max_Frozen: This state is entered when a SET MAX FREEZE LOCK command is received.
In this state, the device may not transition to any other state except by a power cycling. When in this mode bit
8 of word 86 of the identify device data shall be set to one.
Transition SM3:SM3: When a SET MAX ADDRESS, SET MAX SET PASSWORD, SET MAX UNLOCK, SET
MAX FREEZE LOCK, or SET MAX LOCK command is received, the command shall be aborted and the device
shall make a transition to the SM3: Set_Max_Frozen state.
6.15.3 Host protected area orphan sectors
Issuing a SET MAX ADDRESS command with an LBA value may create orphan sectors just as an INITIALIZE
DEVICE PARAMETERS command may create such sectors (see B.3). If the SET MAX ADDRESS LBA value
does not correspond to a cylinder boundary, orphan sectors are created. The device shall report the CHS
boundary just below the requested LBA value in word 1. Sectors above this cylinder boundary are orphan
sectors and the device may or may not allow access to them in CHS translation.
Page 47
T13/1321D revision 3
6.16 CFA feature set
The CompactFlash Association (CFA) feature set provides support for solid state memory devices. A device
that implements the CFA feature set shall implement the following minimum set of commands:
−
−
−
−
−
−
CFA REQUEST EXTENDED ERROR CODE
CFA WRITE SECTORS WITHOUT ERASE
CFA ERASE SECTORS
CFA WRITE MULTIPLE WITHOUT ERASE
CFA TRANSLATE SECTOR
SET FEATURES Enable/Disable 8-bit transfer
Devices reporting the value 848Ah in IDENTIFY DEVICE data word 0 or devices having bit 2 of IDENTIFY
DEVICE data word 83 set to one shall support the CFA feature Set. If the CFA feature set is implemented, all
five commands shall be implemented.
Support of DMA commands is optional for devices that support the CFA feature set.
The CFA ERASE SECTORS command preconditions the sector for a subsequent CFA WRITE SECTORS
WITHOUT ERASE or CFA WRITE MULTIPLE WITHOUT ERASE command to achieve higher performance
during the write operation. The CFA TRANSLATE SECTOR command provides information about a sector such
as the number of write cycles performed on that sector and an indication of the sector’s erased precondition.
The CFA REQUEST EXTENDED ERROR CODE command provides more detailed error information.
Command codes B8h through BFh are reserved for assignment by the CompactFlash Association.
6.17 Removable Media Status Notification and Removable Media feature sets
This section describes two feature sets that secure the media in removable media storage devices using the
ATA/ATAPI interface protocols. First, the Removable Media Status Notification feature set is intended for use
in both devices implementing the PACKET command feature set and those not implementing the PACKET
command feature set. Second, the Removable Media feature set is intended for use only in devices not
implementing the PACKET command feature set. Only one of these feature sets is enabled at any time. If the
Removable Media Status Notification feature set is in use then the Removable Media feature set is disabled and
vice versa.
The reasons for implementing the Removable Media Status Notification feature Set or the Removable Media
feature set are:
− to prevent data loss caused by writing to new media while still referencing the previous media’s
−
−
information.
to prevent data loss by locking the media until completion of a cached write.
to prevent removal of the media by unauthorized persons.
6.17.1 Removable Media Status Notification feature set
The Removable Media Status Notification feature set is the preferred feature set for securing the media in
removable media storage devices. This feature set uses the SET FEATURES command to enable Removable
Media Status Notification. Removable Media Status Notification gives the host system maximum control of the
media. The host system determines media status by issuing the GET MEDIA STATUS command and controls
the device eject mechanism via the MEDIA EJECT command (for devices not implementing the PACKET
command feature set) or the START/STOP UNIT command (for devices implementing the PACKET command
feature set, see SCSI Primary Commands, NCITS 301-1997). While Removable Media Status Notification is
enabled devices not implementing the PACKET command feature set execute MEDIA LOCK and MEDIA
Page 48
T13/1321D revision 3
UNLOCK commands without changing the media lock state (no-operation). While Removable Media Status
Notification is enabled the eject button does not eject the media.
Removable Media Status Notification is persistent through medium removal and insertion and is only disabled
via the SET FEATURES command, hardware reset, software reset, the DEVICE RESET command, the
EXECUTE DEVICE DIAGNOSTIC command, or power-on reset. Removable Media Status Notification shall be
re-enabled after any of the previous reset conditions occur. All media status is reset when Removable Media
Status Notification is disabled because a reset condition occurred. Any pending media change or media
change request is cleared when the Removable Media Status Notification reset condition occurs.
The following task file commands are defined to implement the Removable Media Status Notification feature
set.
−
−
−
−
GET MEDIA STATUS
MEDIA EJECT
SET FEATURES (Enable media status notification)
SET FEATURES (Disable media status notification)
NOTE − Devices implementing the PACKET command feature set control the media eject
mechanism via the START/STOP UNIT packet command.
The preferred sequence of events to use the Removable Media Status Notification feature set is as follows:
a) Host system checks whether or not the device implements the PACKET command feature set via the
device signature in the task file registers.
b) Host system issues the IDENTIFY DEVICE command or the IDENTIFY PACKET DEVICE command
and checks that the device is a removable media device and that the Removable Media Status
Notification feature set is supported.
c) Host system uses the SET FEATURES command to enable Media Status Notification that gives
control of the media to the host. At this time the host system checks the Cylinder High register to
determine if :
− the device is capable of locking the media.
− the device is capable of power ejecting the media.
− Media Status Notification was enabled prior to this command.
d) Host system periodically checks media status using the GET MEDIA STATUS command to determine
if any of the following events occurred:
− no media is present in the device (NM).
− media was changed since the last command (MC).
− a media change request has occurred (MCR).
− media is write protected (WP).
6.17.2 Removable Media feature set
The Removable Media feature set is intended only for devices not implementing the PACKET command feature
set. This feature set operates with Media Status Notification disabled. The MEDIA LOCK and MEDIA UNLOCK
commands are used to secure the media and the MEDIA EJECT command is used to remove the media.
While the media is locked the eject button does not eject the disk. Media status is determined by checking
the media status bits returned by the MEDIA LOCK and MEDIA UNLOCK commands.
Power-on reset, hardware reset, and the EXECUTE DEVICE DIAGNOSTIC command clear the Media Lock
(LOCK) state and the Media Change Request (MCR) state. Software reset clears the Media Lock (LOCK) state,
clears the Media Change Request (MCR) state, and preserves the Media Change (MC) state.
The following commands are defined to implement the Removable Media feature set.
Page 49
T13/1321D revision 3
−
−
−
MEDIA EJECT
MEDIA LOCK
MEDIA UNLOCK
The preferred sequence of events to use the Removable Media feature set is as follows:
a) Host system checks whether or not the device implements the PACKET command feature set via the
device signature in the task file registers.
b) Host system issues the IDENTIFY DEVICE command and checks that the device is a removable
media device and and that the Removable Media feature set is supported.
c) Host system periodically issues MEDIA LOCK commands to determine if:
− no media is present in the device (NM) – media is locked if present.
− a media change request has occurred (MCR).
6.18 Power-Up In Standby feature set
The optional Power-Up In Standby feature set allows devices to be powered-up into the Standby power
management state to minimize inrush current at power-up and to allow the host to sequence the spin-up of
devices. This optional feature set may be enabled or disabled via the SET FEATURES command or may be
enabled by use of a jumper or similar means, or both. When enabled by a jumper, the feature set shall not be
disabled via the SET FEATURES command. The IDENTIFY DEVICE or IDENTIFY PACKET DEVICE response
indicates whether this feature set is implemented and/or enabled.
The enabling of this feature set shall be persistent after power-down and power-up. When this feature set is
enabled, the device shall power-up into Standby.
A device may implement a SET FEATURES subcommand that notifies the device to spin-up to the Active state
when the device has powered-up into Standby. If the device implements this SET FEATURES subcommand
and power-up into Standby is enabled, the device shall remain in Standby until the SET FEATURES
subcommand is received. If the device implements this SET FEATURES subcommand, the fact that the feature
is implemented is reported in the IDENTIFY DEVICE or IDENTIFY PACKET DEVICE response.
If the device:
−
−
−
implements this SET FEATURES subcommand
power-up into Standby is enabled,and
an IDENTIFY DEVICE or IDENTIFY PACKET DEVICE is received while the device is in Standby as a
result of powering up into Standby
the device shall respond to the command remaining in Standby (without spinning-up).
If the device has IDENTIFY DEVICE or IDENTIFY PACKET DEVICE that requires access to the media, the
device shall set word 0 bit 2 to one to indicate that the response is incomplete. At a minimum, words 0 and 2
shall be correctly reported. Those fields that cannot be provided shall be filled with zero. Once the full IDENTIFY
DEVICE or IDENTIFY PACKET DEVICE response data has been accessed, a full response shall be returned
until the next power-down/power-up sequence has taken place.
If the device does not implement the SET FEATURES subcommand to spin-up the device after power-up and
power-up into Standby is enabled, the device shall spin-up upon receipt of the first command that requires the
device to access the media.
Page 50
T13/1321D revision 3
7
Interface register definitions and descriptions
7.1 Device addressing considerations
In traditional controller operation, only the selected device receives commands from the host following selection.
In this standard, when a register is written the value is written to the register of both devices. The host
discriminates between the two by using the DEV bit in the Device/Head register.
Data is transferred in parallel either to or from host memory to the device’s buffer under the direction of
commands previously transferred from the host. The device performs all of the operations necessary to properly
write data to, or read data from, the media. Data read from the media is stored in the device’s buffer pending
transfer to the host memory and data is transferred from the host memory to the device’s buffer to be written to
the media.
The devices using this interface shall be programmed by the host computer to perform commands and return
status to the host at command completion. When two devices are connected on the cable, commands are
written in parallel to both devices, and for all except the EXECUTE DEVICE DIAGNOSTIC command, only the
selected device executes the command. Both devices shall execute an EXECUTE DEVICE DIAGNOSTIC
command regardless of which device is selected, and Device 1 shall post status to Device 0 via PDIAG-.
When the Device Control register is written, both devices respond to the write regardless of which device is
selected (see 7.9.5).
Devices are selected by the DEV bit in the Device/Head register (see 7.10). When the DEV bit is cleared to
zero, Device 0 is selected. When the DEV bit is set to one, Device 1 is selected. When two devices are
connected to the cable, one shall be set as Device 0 and the other as Device 1.
For register access protocols and timing see clauses 9 and 10.
When the host initiates a register or Data port read or write cycle by asserting then negating either DIOW- or
DIOR-, the device(s) on the ATA interface shall determine how to respond and what action(s), if any, are to be
taken. The following text and tables describe this decision process.
The device response begins with these steps:
1) For a device that is not in Sleep mode, see Table 10.
2) If DMACK- is asserted, a device in Sleep mode shall ignore all DIOW-/DIOR- activity. If DMACK- is not
asserted, a device in Sleep mode shall respond as described in Table 15 if the device does not
implement the PACKET Command feature set or Table 16 if the device does implement the PACKET
Command feature set.
Page 51
T13/1321D revision 3
Is the device selected?
(see note 1)
No
No
Table 10 − Device response to DIOW-/DIORIs DMACK- asserted?
Action/Response
No
Yes
See Table 11
DIOW-/DIOR- cycle is ignored
(possible DMA transfer with the
other device)
See Table 12
See Table 13 (see note 2)
See Table 14 and 9.16.1
Yes
No
Yes
Yes
Device 1 is selected but there
No
is no Device 1 and Device 0
responds for Device 1.
Device 1 is selected but there
Yes
DIOW-/DIOR- cycle is ignored
is no Device 1 and Device 0
(possible malfunction of the host)
responds for Device 1.
NOTES −
1 Device selected means that the DEV bit in the Device/Head register matches the logical device
number of the device.
2 Applicable only to Multiword DMA, not applicable to Ultra DMA.
Page 52
T13/1321D revision 3
Table 11 − Device is not selected, DMACK- is not asserted
CS0N
N
N
N
CS1N
A
A
A
DA2
X
N
A
A
DA1
X
X
N
A
DA0
X
X
X
N
DIOxX
X
X
W
DMARQ
Z
Z
Z
Z
BSY
X
X
X
X
DRQ
X
X
X
X
N
N
A
A
A
A
A
A
A
N
N
N
N
N
A
A
N
N
N
N
N
A
A
N
N
N
N
A
N
A
N
A
A
A
N
R
X
X
W
W
R
W
Z
Z
Z
Z
Z
Z
Z
X
X
X
0
1
X
0
X
X
X
X
X
X
X
A
A
A
N
N
N
N
N
N
A
A
A
N
N
A
W
R
W
Z
Z
Z
1
X
0
X
X
X
A
A
A
N
N
N
N
N
A
A
A
N
A
A
N
W
R
W
Z
Z
Z
1
X
0
X
X
X
A
A
A
N
N
N
A
A
A
N
N
N
N
N
A
W
R
W
Z
Z
Z
1
X
0
X
X
X
A
A
A
N
N
N
A
A
A
N
N
A
A
A
N
W
R
W
Z
Z
Z
1
X
0
X
X
X
A
A
A
N
N
N
A
A
A
A
A
A
N
N
A
W
R
W
Z
Z
Z
1
X
0
X
X
X
Device Response
DIOW-/DIOR- cycle is ignored.
Place new data into the Device Control
register and respond to the new values of the
nIEN and SRST bits.
DIOW-/DIOR- cycle is ignored.
Place new data into the Feature register.
DIOW-/DIOR- cycle is ignored.
Place new data into the Sector Count
register.
DIOW-/DIOR- cycle is ignored.
Place new data into the Sector Number
register.
DIOW-/DIOR- cycle is ignored.
Place new data into the Cylinder Low
register.
DIOW-/DIOR- cycle is ignored.
Place new data into the Cylinder High
register.
DIOW-/DIOR- cycle is ignored.
Place new data into the Device/Head
register. Respond to the new value of the
DEV bit.
DIOW-/DIOR- cycle is ignored.
Place new data into the Command register.
Do not respond unless the command is
EXECUTE DEVICE DIAGNOSTICS.
DIOW-/DIOR- cycle is ignored.
A
N
A
A
A
W
Z
1
X
A
N
A
A
A
R
Z
X
X
A
A
X
X
X
X
Z
X
X
DIOW-/DIOR- cycle is ignored.
NOTE −
1. Except in the DIOx- column, A = asserted, N = negated, X = don’t care.
2. In the DIOx- column, R = DIOR- asserted, W = DIOW- asserted, X = either DIOR- or DIOW- is asserted.
3. Device selected means that the DEV bit in the Device/Head register matches the logical device number
of the device.
Page 53
T13/1321D revision 3
Table 12 − Device is selected, DMACK- is not asserted
CS0N
N
N
N
CS1N
A
A
A
DA2
X
N
A
A
DA1
X
X
N
A
DA0
X
X
X
N
DIOxX
X
X
W
DMARQ
X
X
X
X
BSY
X
X
X
X
DRQ
X
X
X
X
N
A
A
A
N
R
X
X
X
N
A
A
A
N
N
A
N
N
A
N
N
A
N
N
X
X
X
X
X
X
X
0
0
X
0
1
A
A
A
N
N
N
N
N
N
N
N
N
N
A
A
X
W
W
X
X
X
1
0
0
X
0
1
A
A
N
N
N
N
N
N
A
A
W
R
X
X
1
0
X
X
A
N
N
N
A
R
X
1
X
A
N
N
A
N
W
X
0
0
A
N
N
A
N
W
X
0
1
A
A
N
N
N
N
A
A
N
N
W
R
X
X
1
0
X
X
A
N
N
A
N
R
X
1
X
A
N
N
A
A
W
X
0
0
A
N
N
A
A
W
X
0
1
A
A
N
N
N
N
A
A
A
A
W
R
X
X
1
0
X
X
A
N
N
A
A
R
X
1
X
A
N
A
N
N
W
X
0
0
A
N
A
N
N
W
X
0
1
A
A
N
N
A
A
N
N
N
N
W
R
X
X
1
0
X
X
A
N
A
N
N
R
X
1
X
A
N
A
N
A
W
X
0
0
A
N
A
N
A
W
X
0
1
Device Response
DIOW-/DIOR- cycle is ignored.
Place new data into the Device Control
register and respond to the new values of the
nIEN and SRST bits.
Place Status register contents on the data
bus (do not change the Interrupt Pending
state).
DIOW-/DIOR- cycle is ignored.
PIO data transfer for this device, a 16-bit data
word is transferred via the Data register.
Result of DIOW-/DIOR- cycle is indeterninate.
Place new data into the Features register.
DIOW- is ignored, this is a malfunction of the
host.
Result of DIOW-/DIOR- cycle is indeterninate.
Place the contents of the Error register on the
data bus.
Place the contents of the Status register on
the data bus.
Place new data into the Sector Count
register.
DIOW- is ignored, this is a malfunction of the
host.
Result of DIOW-/DIOR- cycle is indeterninate.
Place the contents of the Sector Count
register on the data bus.
Place the contents of the Status register on
the data bus.
Place new data into the Sector Number
register.
DIOW- is ignored, this is a malfunction of the
host.
Result of DIOW-/DIOR- cycle is indeterninate.
Place the contents of the Sector Number
register on the data bus.
Place the contents of the Status register on
the data bus.
Place new data into the Cylinder Low
register.
DIOW- is ignored, this is a malfunction of the
host.
Result of DIOW-/DIOR- cycle is indeterninate.
Place the contents of the Cylinder Low
register on the data bus.
Place the contents of the Status register on
the data bus.
Place new data into the Cylinder High
register.
DIOW- is ignored, this is a malfunction of the
host.
(continued)
Page 54
T13/1321D revision 3
Table 12 − Device is selected, DMACK- is not asserted (continued)
CS0A
A
CS1N
N
DA2
A
A
DA1
N
N
DA0
A
A
DIOxW
R
DMARQ
X
X
BSY
1
0
DRQ
X
X
A
N
A
N
A
R
X
1
X
A
N
A
A
N
W
X
0
0
A
N
A
A
N
W
X
0
1
A
A
N
N
A
A
A
A
N
N
W
R
X
X
1
0
X
X
A
N
A
A
N
R
X
1
X
A
N
A
A
A
W
X
0
0
A
A
N
N
A
A
A
A
A
A
W
W
X
X
0
1
1
X
A
N
A
A
A
R
X
X
X
Device Response
Result of DIOW-/DIOR- cycle is indeterninate.
Place the contents of the Cylinder High
register on the data bus.
Place the contents of the Status register on
the data bus.
Place new data into the Device/Head register.
Respond to the new value of the DEV bit.
DIOW- is ignored, this is a malfunction of the
host.
Result of DIOW-/DIOR- cycle is indeterninate.
Place the contents of the Device/Head
register on the data bus.
Place the contents of the Status register on
the data bus.
Place new data into the Command register
and respond to the new command (exit the
interrupt pending State).
Result of DIOW-/DIOR- cycle is
indeterninate., unless the device supports
DEVICE RESET. If the device supports the
DEVICE RESET command, exit the interrupt
pending state.
Place contents of Status register on the data
bus and exit the interrupt pending state.
DIOW-/DIOR- cycle is ignored.
A
A
X
X
X
X
X
X
X
NOTE −
1. Except in the DIOx- column, A = asserted, N = negated, X = don’t care.
2. In the DIOx- column, R = DIOR- asserted, W = DIOW- asserted, X = either DIOR- or DIOW- is asserted.
3. Device selected means that the DEV bit in the Device/Head register matches the logical device number
of the device.
(concluded)
Table 13 − Device is selected, DMACK- is asserted (for Multiword DMA only)
CS0X
X
X
X
N
N
CS1X
X
X
X
N
N
DA2
X
X
X
X
X
X
DA1
X
X
X
X
X
X
DA0
X
X
X
X
X
X
DIOxX
X
X
X
X
X
DMARQ
Z
Z
Z
N
N
N
BSY
0
1
0
0
1
0
DRQ
0
X
1
0
X
X
Device Response
DIOW-/DIOR- cycle is ignored(possible
malfunction of the host).
This could be the final DIOW-/DIOR- of a
Multiword DMA transfer burst, or a possible
malfunction of the host that is ignored.
DMA transfer for this device, a 16-bit word of
data is transferred via the Data Port.
DIOW-/DIOR- cycle is ignored(possible
Malfunction of the host).
N
N
X
X
X
X
A
1
X
N
N
X
X
X
X
A
0
1
X
A
X
X
X
X
X
X
X
A
X
X
X
X
X
X
X
X
NOTE −
1. Except in the DIOx- column, A = asserted, N = negated, X = don’t care.
2. In the DIOx- column, R = DIOR- asserted, W = DIOW- asserted, X = either DIOR- or DIOW- is asserted.
3. Device selected means that the DEV bit in the Device/Head register matches the logical device number
of the device.
Page 55
T13/1321D revision 3
Table 14 − Device 1 is selected and Device 0 is responding for Device 1
CS0N
N
N
N
CS1N
A
A
A
DA2
X
N
A
A
DA1
X
X
N
A
DA0
X
X
X
N
DIOxX
X
X
W
DMARQ
Z
Z
X
X
BSY
0
0
0
0
DRQ
0
0
0
0
N
N
A
A
A
A
N
N
A
A
N
N
A
A
N
N
N
A
N
A
R
X
X
W
X
X
X
X
0
0
0
0
0
0
0
0
A
N
N
N
A
R
X
0
0
A
N
N
A
N
W
X
0
0
A
N
N
A
N
R
X
0
0
A
N
N
A
A
W
X
0
0
A
N
N
A
A
R
X
0
0
A
N
A
N
N
W
X
0
0
A
N
A
N
N
R
X
0
0
A
N
A
N
A
W
X
0
0
A
N
A
N
A
R
X
0
0
A
N
A
A
N
W
X
0
0
A
N
A
A
N
R
X
0
0
A
N
A
A
A
W
X
0
0
Device Response
DIOW-/DIOR- cycle is ignored.
Place new data into the Device 0 Device
Control register and respond to the new
values of the nIEN and SRST bits.
Place 00H on the data bus.
DIOW-/DIOR- cycle is ignored.
Place new data into the Device 0 Feature
register.
Place the contents of the Device 0 Error
register on the data bus.
Place new data into Device 0 Sector Count
register.
Place the contents of the Device 0 Sector
Count register on the data bus.
Place new data into Device 0 Sector Number
register.
Place the contents of the Device 0 Sector
Number register on the data bus.
Place new data into Device 0 Cylinder Low
register.
Place the contents of the Device 0 Cylinder
Low register on the data bus.
Place new data into Device 0 Cylinder High
register.
Place the contents of the Device 0 Cylinder
High register on the data bus.
Place new data into the Device 0
Device/Head register. Respond to the new
value of the DEV bit.
Place the contents of the Device 0
Device/Head register, with DEV bit set to
1,on the data bus.
Place new data into the Command register
of Device 0. Do not respond unless the
command is EXECUTE DEVICE
DIAGNOSTICS.
Place 00H on the data bus.
DIOW-/DIOR- cycle is ignored.
A
N
A
A
A
R
X
0
0
A
A
X
X
X
X
X
0
0
NOTE −
1. Except in the DIOx- column, A = asserted, N = negated, X = don’t care.
2. In the DIOx- column, R = DIOR- asserted, W = DIOW- asserted, X = either DIOR- or DIOW- is asserted.
3. Device selected means that the DEV bit in the Device/Head register matches the logical device number
of the device.
Page 56
T13/1321D revision 3
Table 15 − Device is in Sleep mode, DEVICE RESET is not implemented, DMACK- is not asserted
CS0N
N
N
N
CS1N
A
A
A
DA2
X
N
A
A
DA1
X
X
N
A
DA0
X
X
X
N
DIOxX
X
X
W
DMARQ
Z
Z
Z
Z
BSY
X
X
X
X
DRQ
X
X
X
X
Device Response
DIOW-/DIOR- cycle is ignored.
Place new data into the Device Control
register SRST bit and respond only if SRST
bit is 1.
DIOW-/DIOR- cycle is ignored.
N
A
A
A
N
R
Z
X
X
N
A
A
A
A
X
Z
X
X
A
N
X
X
X
X
Z
X
X
A
A
X
X
X
X
Z
X
X
NOTE −
1. Except in the DIOx- column, A = asserted, N = negated, X = don’t care.
2. In the DIOx- column, R = DIOR- asserted, W = DIOW- asserted, X = either DIOR- or DIOW- is asserted.
3. Device selected means that the DEV bit in the Device/Head register matches the logical device number
of the device.
Table 16 − Device is in Sleep mode, DEVICE RESET is implemented, DMACK- is not asserted
CS0N
N
N
N
CS1N
A
A
A
DA2
X
N
A
A
DA1
X
X
N
A
DA0
X
X
X
N
DIOxX
X
X
W
DMARQ
Z
Z
Z
Z
BSY
X
X
X
X
DRQ
X
X
X
X
N
N
A
A
A
A
A
N
N
N
A
A
N
A
A
A
A
X
N
A
N
A
X
X
N
R
X
X
X
W
Z
Z
Z
Z
Z
X
X
X
X
X
X
X
X
X
X
A
A
N
N
A
A
A
A
N
A
R
W
Z
Z
X
X
X
X
Device Response
DIOW-/DIOR- cycle is ignored.
Place new data into the Device Control
register SRST bit and respond only if SRST
bit is 1.
DIOW-/DIOR- cycle is ignored.
Place new data into the Device/Head register
DEV bit.
DIOR- cycle is ignored.
DIOW- cycle is ignored unless the device is
selected and the command is DEVICE
RESET.
DIOR- cycle is ignored.
DIOW-/DIOR- cycle is ignored
A
N
A
A
A
R
Z
X
X
A
A
X
X
X
X
Z
X
X
NOTE −
1. Except in the DIOx- column, A = asserted, N = negated, X = don’t care.
2. In the DIOx- column, R = DIOR- asserted, W = DIOW- asserted, X = either DIOR- or DIOW- is asserted.
3. Device selected means that the DEV bit in the Device/Head register matches the logical device number of the
device.
7.2 I/O register descriptions
Communication to or from the device is through registers addressed by the signals from the host (CS0-, CS1-,
DA (2:0), DIOR-, and DIOW-). CS0- and CS1- both asserted or negated is an invalid (not used) address except
when both are negated during a DMA data transfer. When CS0- and CS1- are both asserted or both negated
and a DMA transfer is not in progress, the device shall hold DD (15:0) in the released state and ignore
transitions on DIOR- and DIOW-. When CS0- is negated and CS1- is asserted only DA (2:0) with a value of 6h
is valid. During invalid combinations of assertion and negation of CS0-, CS1-, DA0, DA1, and DA2, a device
shall keep DD(15:0) in the high impedance state and ignore transitions on DIOR- and DIOW-. Valid register
addresses are described in the clauses defining the registers.
The Command Block registers are used for sending commands to the device or posting status from the device.
These registers include the Cylinder High, Cylinder Low, Device/Head, Sector Count, Sector Number,
Page 57
T13/1321D revision 3
Command, Status, Features, Error, and Data registers. The Control Block registers are used for device control
and to post alternate status. These registers include the Device Control and Alternate Status registers.
Each register description in the following clauses contain the following format:
Address – the CS and DA address of the register.
Direction – indicates if the register is read/write, read only, or write only from the host.
Access restrictions – indicates when the register may be accessed.
Effect – indicates the effect of accessing the register.
Functional description – describes the function of the register.
Field/bit description – describes the content of the register.
7.3 Alternate Status register
7.3.1 Address
CS1
CS0
DA2
DA1
A
N
A
A
A = asserted, N = negated
DA0
N
7.3.2 Direction
This register is read only. If this address is written to by the host, the Device Control register is written.
7.3.3 Access restrictions
When the BSY bit is set to one, the other bits in this register shall not be used. The entire contents of this
register are not valid while the device is in Sleep mode.
7.3.4 Effect
Reading this register shall not clear a pending interrupt.
7.3.5 Functional description
This register contains the same information as the Status register in the command block.
See 7.15 for definitions of the bits in this register.
7.4 Command register
7.4.1 Address
CS1
CS0
DA2
DA1
N
A
A
A
A = asserted, N = negated
Page 58
DA0
A
T13/1321D revision 3
7.4.2 Direction
This register is write only. If this address is read by the host, the Status register is read.
7.4.3 Access restrictions
For all commands except DEVICE RESET, this register shall only be written when BSY and DRQ are both
cleared to zero and DMACK- is not asserted. If written when BSY or DRQ is set to one, the results of writing
the Command register are indeterminate except for the DEVICE RESET command. For a device in the Sleep
mode, writing of the Command register shall be ignored except for writing of the DEVICE RESET command to a
device that implements the PACKET Command feature set.
7.4.4 Effect
Command processing begins when this register is written. The content of the Command Block registers
become parameters of the command when this register is written. Writing this register clears any pending
interrupt condition.
7.4.5 Functional description
This register contains the command code being sent to the device. Command execution begins immediately
after this register is written. The executable commands, the command codes, and the necessary parameters
for each command are summarized in the tables in informative annex E.
7.4.6 Field/bit description
7
6
5
4
3
Command Code
2
1
0
7.5 Cylinder High register
7.5.1 Address
CS1
CS0
DA2
DA1
N
A
A
N
A = asserted, N = negated
DA0
A
7.5.2 Direction
This register is read/write.
7.5.3 Access restrictions
This register shall be written only when both BSY and DRQ are cleared to zero and DMACK- is not asserted.
The contents of this register are valid only when BSY is cleared to zero. If this register is written when BSY or
DRQ is set to one, the result is indeterminate. The contents of this register are not valid while a device is in the
Sleep mode.
7.5.4 Effect
The content of this register becomes a command parameter when the Command register is written.
Page 59
T13/1321D revision 3
7.5.5 Functional description
The content of this register is command dependent (see clause 8).
7.6 Cylinder Low register
7.6.1 Address
CS1
CS0
DA2
DA1
N
A
A
N
A = asserted, N = negated
DA0
N
7.6.2 Direction
This register is read/write.
7.6.3 Access restrictions
This register shall be written only when both BSY and DRQ are cleared to zero and DMACK- is not asserted.
The contents of this register are valid only when BSY is cleared to zero. If this register is written when BSY or
DRQ is set to one, the result is indeterminate. The contents of the this register are not valid while a device is in
the Sleep mode.
7.6.4 Effect
The content of this register becomes a command parameter when the Command register is written.
7.6.5 Functional description
The content of this register is command dependent (see clause 8).
7.7 Data port
7.7.1 Address
When DMACK- is asserted, CS0- and CS1- shall be negated and transfers shall be 16-bits wide.
CS1
CS0
DA2
DA1
DA0
N
N
X
X
X
A = asserted, N = negated, X = don’t care
7.7.2 Direction
This port is read/write.
7.7.3 Access restrictions
This port shall be accessed for host DMA data transfers only when DMACK- and DMARQ are asserted.
Page 60
T13/1321D revision 3
7.7.4 Effect
DMA out data transfers are processed by a series of reads to this port, each read transferring the data that
follows the previous read. DMA in data transfers are processed by a series of writes to this port, each write
transferring the data that follows the previous write. The results of a read during a DMA in or a write during a
DMA out are indeterminate.
7.7.5 Functional description
The data port is 16-bits in width.
7.7.6 Field/bit description
15
14
13
7
6
5
12
11
Data(15:8)
4
3
Data(7:0)
10
9
8
2
1
0
7.8 Data register
7.8.1 Address
CS1
CS0
DA2
DA1
N
A
N
N
A = asserted, N = negated
DA0
N
7.8.2 Direction
This register is read/write.
7.8.3 Access restrictions
This register shall be accessed for host PIO data transfer only when DRQ is set to one and DMACK- is not
asserted. The contents of this register are not valid while a device is in the Sleep mode.
7.8.4 Effect
PIO out data transfers are processed by a series of reads to this register, each read transferring the data that
follows the previous read. PIO in data transfers are processed by a series of writes to this register, each write
transferring the data that follows the previous write. The results of a read during a PIO in or a write during a PIO
out are indeterminate.
7.8.5 Functional description
The data register is 16-bits wide. When a CFA device is in 8-bit PIO data transfer mode this register is 8-bits
wide using only DD7 to DD0.
Page 61
T13/1321D revision 3
7.8.6 Field/bit description
15
14
13
7
6
5
12
11
Data(15:8)
4
3
Data(7:0)
10
9
8
2
1
0
7.9 Device Control register
7.9.1 Address
CS1
CS0
DA2
DA1
A
N
A
A
A = asserted, N = negated
DA0
N
7.9.2 Direction
This register is write only. If this address is read by the host, the Alternate Status register is read.
7.9.3 Access restrictions
This register shall only be written when DMACK- is not asserted.
7.9.4 Effectiveness
The content of this register shall take effect when written.
7.9.5 Functional description
This register allows a host to software reset attached devices and to enable or disable the assertion of the
INTRQ signal by a selected device. When the Device Control register is written, both devices respond to the
write regardless of which device is selected. When the SRST bit is set to one, both devices shall perform the
software reset protocol. The device shall respond to the SRST bit when in the SLEEP mode.
7.9.6 Field/bit description
7
r
6
r
5
r
4
r
3
r
2
SRST
1
nIEN
0
0
− Bits 7 through 3 are reserved.
− SRST is the host software reset bit (see 9.2).
− nIEN is the enable bit for the device Assertion of INTRQ to the host. When the nIEN bit is cleared to
zero, and the device is selected, INTRQ shall be enabled through a tri-state buffer and shall be
asserted or negated by the device as appropriate. When the nIEN bit is set to one, or the device is not
selected, the INTRQ signal shall be in a high impedance state.
− Bit 0 shall be cleared to zero.
Page 62
T13/1321D revision 3
7.10 Device/Head register
7.10.1 Address
CS1
CS0
DA2
DA1
N
A
A
A
A = asserted, N = negated
DA0
N
7.10.2 Direction
This register is read/write.
7.10.3 Access restrictions
This register shall be written only when both BSY and DRQ are cleared to zero and DMACK- is not asserted.
The contents of this register are valid only when BSY is cleared to zero. If this register is written when BSY or
DRQ is set to one, the result is indeterminate. For devices not implementing the PACKET Command feature
set, the contents of this register are not valid while a device is in the Sleep mode. For devices implementing the
PACKET Command feature set, the contents of this register are valid while the device is in Sleep mode.
7.10.4 Effect
The DEV bit becomes effective when this register is written by the host or the signature is set by the device. All
other bits in this register become a command parameter when the Command register is written.
7.10.5 Functional description
Bit 4, DEV, in this register selects the device. Other bits in this register are command dependent (see clause
8).
7.10.6 Field/bit description
7
Obsolete
−
6
#
5
Obsolet
e
4
DEV
3
#
2
#
1
#
0
#
Obsolete – These bits are obsolete.
NOTE − Some hosts set these bits to one. Devices shall ignore these bits.
−
−
# - The content of these bits is command dependent (see clause 8).
DEV – Device select. Cleared to zero selects Device 0. Set to one selects Device 1.
7.11 Error register
7.11.1 Address
CS1
CS0
DA2
DA1
N
A
N
N
A = asserted, N = negated
DA0
A
Page 63
T13/1321D revision 3
7.11.2 Direction
This register is read only. If this address is written to by the host, the Features register is written.
7.11.3 Access restrictions
The contents of this register shall be valid when BSY and DRQ equal zero and ERR equals one. The contents
of this register shall be valid upon completion of power-on, or after a hardware or software reset, or after
command completion of an EXECUTE DEVICE DIAGNOSTICS or DEVICE RESET command. The contents of
this register are not valid while a device is in the Sleep mode.
7.11.4 Effect
None.
7.11.5 Functional description
This register contains status for the current command.
Following a power-on, a hardware or software reset (see 9.1), or command completion of an EXECUTE DEVICE
DIAGNOSTIC (see 8.9) or DEVICE RESET command (see 8.7), this register contains a diagnostic code .
At command completion of any command except EXECUTE DEVICE DIAGNOSTIC, the contents of this
register are valid when the ERR bit is set to one in the Status register.
7.11.6 Field/bit description
7
#
6
#
5
#
4
#
3
#
2
ABRT
1
#
0
#
− Bit 2 – ABRT (command aborted) is set to one to indicate the requested command has been command
aborted because the command code or a command parameter is invalid or some other error has
occurred.
− # -The content of this bit is command dependent (see clause 8).
7.12 Features register
7.12.1 Address
CS1
CS0
DA2
DA1
N
A
N
N
A = asserted, N = negated
DA0
A
7.12.2 Direction
This register is write only. If this address is read by the host, the Error register is read.
7.12.3 Access restrictions
This register shall be written only when BSY and DRQ equal zero and DMACK- is not asserted. If this register
is written when BSY or DRQ is set to one, the result is indeterminate.
Page 64
T13/1321D revision 3
7.12.4 Effect
The content of this register becomes a command parameter when the Command register is written.
7.12.5 Functional description
The content of this register is command dependent (see clause 8).
7.13 Sector Count register
7.13.1 Address
CS1
CS0
DA2
DA1
N
A
N
A
A = asserted, N = negated
DA0
N
7.13.2 Direction
This register is read/write.
7.13.3 Access restrictions
This register shall be written only when both BSY and DRQ are zero and DMACK- is not asserted. The
contents of this register are valid only when both BSY and DRQ are zero. If this register is written when BSY or
DRQ is set to one, the result is indeterminate. The contents of the this register are not valid while a device is in
the Sleep mode.
7.13.4 Effect
The content of this register becomes a command parameter when the Command register is written.
7.13.5 Functional description
The content of this register is command dependent (see clause 8).
7.14 Sector Number register
7.14.1 Address
CS1
CS0
DA2
DA1
N
A
N
A
A = asserted, N = negated
DA0
A
7.14.2 Direction
This register is read/write.
7.14.3 Access restrictions
This register shall be written only when both BSY and DRQ are zero and DMACK- is not asserted. The
contents of this register are valid only when both BSY and DRQ are zero. If this register is written when BSY or
DRQ is set to one, the result is indeterminate. The contents of this register are not valid while a device is in the
Sleep mode.
Page 65
T13/1321D revision 3
7.14.4 Effect
The content of this register becomes a command parameter when the Command register is written.
7.14.5 Functional description
The content of this register is command dependent (see clause 8).
7.15 Status register
7.15.1 Address
CS1
CS0
DA2
DA1
N
A
A
A
A = asserted, N = negated
DA0
A
7.15.2 Direction
This register is read only. If this address is written to by the host, the Command register is written.
7.15.3 Access restrictions
The contents of this register, except for BSY, shall be ignored when BSY is set to one. BSY is valid at all
times. The contents of this register are not valid while a device is in the Sleep mode.
7.15.4 Effect
Reading this register when an interrupt is pending causes the interrupt pending to be cleared (see 5.2.9). The
host should not read the Status register when an interrupt is expected as this may clear the interrupt pending
before the INTRQ can be recognized by the host.
7.15.5 Functional description
This register contains the device status. The contents of this register are updated to reflect the current state of
the device and the progress of any command being executed by the device.
7.15.6 Field/bit description
7
BSY
6
DRDY
5
#
4
#
3
DRQ
2
1
Obsolet
e
Obsolet
e
0
ERR
7.15.6.1 BSY (Busy)
BSY is set to one to indicate that the device is busy. After the host has written the Command register the
device shall have either the BSY bit set to one, or the DRQ bit set to one, until command completion or the
device has performed a bus release for an overlapped command.
The BSY bit shall be set to one by the device:
1) after either the negation of RESET- or the setting of the SRST bit to one in the Device
Control register;
2) after writing the Command register if the DRQ bit is not set to one;
3) between blocks of a data transfer during PIO data-in commands before the DRQ bit is
cleared to zero;
Page 66
T13/1321D revision 3
4) after the transfer of a data block during PIO data-out commands before the DRQ bit is
cleared to zero;
5) during the data transfer of DMA commands either the BSY bit , the DRQ bit, or both shall
be set to one;
6) after the command packet is received during the execution of a PACKET command.
NOTE − The BSY bit may be set to one and then cleared to zero so quickly, that host
detection of the BSY bit being set to one is not certain.
When BSY is set to one, the device has control of the Command Block Registers and:
1) a write to a Command Block register by the host shall be ignored by the device except for
writing DEVICE RESET command;
2) a read from a Command Block register by the host will most likely yield invalid contents
except for the BSY bit itself.
The BSY bit shall be cleared to zero by the device:
1)
2)
3)
4)
after setting DRQ to one to indicate the device is ready to transfer data;
at command completion;
upon releasing the bus for an overlapped command;
when the device is ready to accept commands that do not require DRDY during a poweron, hardware or software reset.
When BSY is cleared to zero, the host has control of the Command Block registers, the device shall:
1)
2)
3)
4)
not set DRQ to one;
not change ERR bit;
not change the content of any other Command Block register;
set the SERV bit to one when ready to continue an overlapped command that has been
bus released.
7.15.6.2 DRDY (Device ready)
The DRDY bit shall be cleared to zero by the device:
1) when power-on, hardware, or software reset or DEVICE RESET or EXECUTE DEVICE
DIAGNOSTIC commands for devices implementing the PACKET command feature set.
When the DRDY bit is cleared to zero, the device shall accept and attempt to execute as described in
clause 8.
The DRDY bit shall be set to one by the device:
1) when the device is capable of accepting all commands for devices not implementing the
PACKET command feature set;
2) prior to command completion except the DEVICE RESET or EXECUTE DEVICE
DIAGNOSTIC command for devices implementing the PACKET feature set.
When the DRDY bit is set to one:
1) the device shall accept and attempt to execute all implemented commands;
2) devices that implement the Power Management feature set shall maintain the DRDY bit
set to one when they are in the Idle or Standby modes.
Page 67
T13/1321D revision 3
7.15.6.3 Command dependent
The use of bits marked with # are command dependent (see clause 8). Bit 4 was formerly the DSC (Device
Seek Complete) bit.
7.15.6.4 DRQ (Data request)
DRQ indicates that the device is ready to transfer a word of data between the host and the device. After the
host has written the Command register the device shall either set the BSY bit to one or the DRQ bit to one,
until command completion or the device has performed a bus release for an overlapped command.
The DRQ bit shall be set to one by the device:
1) when BSY is set to one and data is ready for PIO transfer;
2) during the data transfer of DMA commands either the BSY bit , the DRQ bit, or both shall
be set to one.
When the DRQ bit is set to one, the host may:
1) transfer data via PIO mode;
2) transfer data via DMA mode if DMARQ and DMACK- are asserted.
The DRQ bit shall be cleared to zero by the device:
1) when the last word of the data transfer occurs;
2) when the last word of the command packet transfer occurs for a PACKET command.
When the DRQ bit is cleared to zero, the host may:
1) transfer data via DMA mode if DMARQ and DMACK- are asserted and BSY is set to one.
7.15.6.5 Obsolete bits
Some bits in this register were defined in previous ATA standards but have been declared obsolete in this
standard. These bits are labeled “obsolete”.
7.15.6.6 ERR (Error)
ERR indicates that an error occurred during execution of the previous command. For the PACKET and
SERVICE commands, this bit is defined as CHK and indicates that an exception conditions exists.
The ERR bit shall be set to one by the device:
1) when BSY or DRQ is set to one and an error occurs in the executing command.
When the ERR bit is set to one:
1) the bits in the Error register shall be valid;
2) the device shall not change the contents of the following registers until a new command
has been accepted, the SRST bit is set to one or RESET- is asserted:
− Error register;
− Cylinder High/Low registers;
− Sector Count register;
− Sector Number register;
− Device/Head register.
Page 68
T13/1321D revision 3
The ERR bit shall be cleared to zero by the device:
1) when a new command is written to the Command register;
2) when the SRST bit is set to one;
3) when the RESET- signal is asserted.
When the ERR bit is cleared to zero at the end of a command:
1) the content of the Error register shall be ignored by the host.
8 Command descriptions
Commands are issued to the device by loading the required registers in the command block with the needed
parameters and then writing the command code to the Command register. Required registers are those
indicated by a specific content in the Inputs table for the command, i.e., not noted as na or obs.
Each command description in the following clauses contains the following subclauses:
Command code – Indicates the command code for this command.
Feature set – Indicates feature set and if the command is:
− Mandatory – Required to be implemented by devices as specified.
− Optional – Implementation is optional but if implemented shall be implemented as specified.
Protocol – Indicates which protocol is used by the command (see clause 9).
Inputs – Describes the Command Block register data that the host shall supply.
Register
7
6
5
4
3
2
1
0
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
Command Code
NOTE − na indicates the content of a bit or field is not applicable to the particular
command. Obs indicates that the use of this bit is obsolete.
Normal outputs – Describes the Command Block register data returned by the device at the end of a command.
Register
7
6
5
4
3
2
1
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
NOTE − na indicates the content of a bit or field is not applicable to the particular
command. Obs indicates that the use of this bit is obsolete.
0
Page 69
T13/1321D revision 3
Error outputs – Describes the Command Block register data that shall be returned by the device at command
completion with an unrecoverable error.
Register
7
6
5
4
3
2
1
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
NOTE − na indicates the content of a bit or field is not applicable to the particular
command. Obs indicates that the use of this bit is obsolete.
0
Prerequisites – Any prerequisite commands or conditions that shall be met before the command is issued.
Description – The description of the command function(s).
8.1 CFA ERASE SECTORS
8.1.1 Command code
C0h
8.1.2 Feature set
CFA feature set.
− If the CFA feature set is implemented this command shall be implemented.
This command code is Vendor Specific for devices not implementing the CFA feature Set.
8.1.3 Protocol
Non-data command (see 9.4).
8.1.4 Inputs
The Cylinder Low, Cylinder High, Device/Head, and Sector Number specify the starting sector address to be
erased. The Sector Count register specifies the number of sectors to be erased.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
obs
6
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
C0h
Sector Count –
number of sectors to be erased. A value of 00h indicates that 256 sectors are to be erased.
Sector Number –
starting sector number or LBA address bits (7:0).
Cylinder Low –
Page 70
T13/1321D revision 3
starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High –
starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head –
bit 6 set to one if LBA address, cleared to zero if CHS address starting head number or LBA address
bits (27:24).
8.1.5 Normal outputs
Register
Error
Sector Count
Sector
Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
Status register
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be cleared to zero.
8.1.6 Error outputs
The device shall return command aborted if the command is not supported. An unrecoverable error encountered
during execution of this command results in the termination of the command. The command block registers
contain the address of the sector where the first unrecovered error occurred.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
na
na
DRDY
5
na
4
IDNF
3
na
2
ABRT
1
na
0
MED
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
na
na
na
ERR
Error Register –
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
ABRT shall be set to one if the command is not supported. ABRT may be set to one if the device is not
able to complete the action requested by the command. ABRT shall be set to one if an
address outside of the range of user-accessible addresses is requested if IDNF is not set to
one.
MED shall be set to one if a media error is detected.
Sector Number, Cylinder Low, Cylinder High, Device/Head –
shall be written with the address of first unrecoverable error.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
Page 71
T13/1321D revision 3
DF (Device Fault) shall be set to one if a device fault has occurred.
ERR shall be set to one if an Error register bit is set to one.
8.1.7 Prerequisites
DRDY set to one.
8.1.8 Description
This command pre-erases and conditions from 1 to 256 sectors as specified in the Sector Count register. This
command should be issued in advance of a CFA WRITE SECTORS WITHOUT ERASE or a CFA WRITE
MULTIPLE WITHOUT ERASE command to increase the execution speed of the write operation.
8.2 CFA REQUEST EXTENDED ERROR CODE
8.2.1 Command code
03h
8.2.2 Feature set
CFA feature set.
− If the CFA feature set is implemented this command shall be implemented.
8.2.3 Protocol
Non-data command (see 9.4).
8.2.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
obs
na
obs
DEV
03h
na
8.2.5 Normal outputs
The extended error code written into the Error register is an 8-bit code. Table 17 defines these values.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
obs
BSY
na
DRDY
Error register –
Extended error code.
Page 72
5
4
3
2
1
Extended error code
Vendor specific
Vendor specific
Vendor specific
Vendor specific
obs
DEV
Vendor specific
na
na
na
na
na
0
ERR
T13/1321D revision 3
Sector Number, Cylinder Low, Cylinder High, Device/Head –
May contain additional information.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be cleared to zero.
Extended error code
00h
01h
03h
05h
09h
0Bh
0Ch
0D-0Fh
10h
11h
14h
18h
1Dh, 1Eh
1Fh
20h
21h
22-23h
27h
2Fh
30-34h
35h, 36h
37h, 3Eh
38h
39h
3Ah
3Bh 3Ch, 3Fh
3Dh
All other values
Table 17 − Extended error codes
Description
No error detected / no additional information
Self-test passed
Write / Erase failed
Self-test or diagnostic failed
Miscellaneous error
Vendor specific
Corrupted media format
Vendor specific
ID Not Found / ID Error
Uncorrectable ECC error
ID Not Found
Corrected ECC error
Vendor specific
Data transfer error / command aborted
Invalid command
Invalid address
Vendor specific
Write protect violation
Address overflow (address too large)
Self-test or diagnostic failed
Supply or generated voltage out of tolerance
Self-test or diagnostic failed
Corrupted media format
Vendor specific
Spare sectors exhausted
Corrupted media format
Vendor specific
Reserved
8.2.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
na
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error Register –
ABRT shall be set to one if the command is not supported. ABRT may be set to one if the device is not
able to complete the action requested by the command.
Status register –
Page 73
T13/1321D revision 3
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
ERR shall be set to one if an Error register bit is set to one.
8.2.7 Prerequisites
DRDY set to one.
8.2.8 Description
This command provides an extended error code which identifies the cause of an error condition in more detail
than is available with Status and Error register values. The CFA REQUEST EXTENDED ERROR CODE
command shall return an extended error code if the previous command completed with an error or a no error
detected extended error code if the previous command completed without error.
8.3 CFA TRANSLATE SECTOR
8.3.1 Command code
87h
8.3.2 Feature set
CFA feature set.
− If the CFA feature set is implemented this command shall be implemented.
This command code is Vendor Specific for devices not implementing the CFA feature Set.
8.3.3 Protocol
PIO data-in command (see 9.5).
8.3.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
87h
Sector Number –
sector number or LBA address bits (7:0).
Cylinder Low –
cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High –
cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head –
bit 6 set to one if LBA address, cleared to zero if CHS address head number or LBA address bits
(27:24).
Page 74
T13/1321D revision 3
8.3.5 Normal outputs
A 512 byte information table is transferred to the host. Table 18 defines these values.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
na
na
1
0
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be cleared to zero.
Byte
00h
01h
02h
03h
04h
05h
06h
07-12h
13h
14-17h
18h
19h
1Ah
1B-FFh
Table 18 − CFA TRANSLATE SECTOR Information
Description
Cylinder number MSB
Cylinder number LSB
Head number
Sector number
LBA bits (23:16)
LBA bits (15:8)
LBA bits (7:0)
Reserved
Sector erased flag (FFh = erased; 00h = not erased)
Reserved
Sector write cycles count bits (23:16)
Sector write cycles count bits (15:8)
Sector write cycles count bits (7:0)
Reserved
8.3.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
na
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error Register –
ABRT shall be set to one if the command is not supported. ABRT may be set to one if the device is not
able to complete the action requested by the command.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
Page 75
T13/1321D revision 3
ERR shall be set to one if an Error register bit is set to one.
8.3.7 Prerequisites
DRDY set to one.
8.3.8 Description
This command provides information related to a specific sector. The data indicates the erased or not erased
status of the sector, and the number of erase and write cycles performed on that sector. Devices may return
zero in fields that do not apply or that are not supported by the device.
8.4 CFA WRITE MULTIPLE WITHOUT ERASE
8.4.1 Command code
CDh
8.4.2 Feature set
CFA feature set.
− If the CFA feature set is implemented this command shall be implemented.
8.4.3 Protocol
PIO data-out command (see 9.6).
8.4.4 Inputs
The Cylinder Low, Cylinder High, Device/Head, and Sector Number specify the starting sector address to be
written. The Sector Count register specifies the number of sectors to be transferred.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
CDh
Sector Count –
number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number –
starting sector number or LBA address bits (7:0).
Cylinder Low –
starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High –
starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head –
bit 6 set to one if LBA address, cleared to zero if CHS address starting head number or LBA address
bits (27:24).
Page 76
T13/1321D revision 3
8.4.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be cleared to zero.
8.4.6 Error outputs
The device shall return command aborted if the command is not supported. An unrecoverable error encountered
during execution of this command results in the termination of the command. The command block registers
contain the address of the sector where the first unrecovered error occurred. The amount of data transferred is
indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
na
na
DRDY
5
na
4
IDNF
3
na
2
ABRT
1
na
0
MED
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
DRQ
na
na
ERR
Error Register –
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
ABRT shall be set to one if the command is not supported. ABRT may be set to one if the device is not
able to complete the action requested by the command. ABRT shall be set to one if an
address outside of the range of user-accessible addresses is requested if IDNF is not set to
one.
MED shall be set to one if a media error is detected
Sector Number, Cylinder Low, Cylinder High, Device/Head –
shall be written with the address of first unrecoverable error.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
Page 77
T13/1321D revision 3
8.4.7 Prerequisites
DRDY set to one. If bit 8 of IDENTIFY DEVICE word 59 is cleared to zero, a successful SET MULTIPLE MODE
command shall precede a CFA WRITE MULTIPLE WITHOUT ERASE command.
8.4.8 Description
This command is similar to the WRITE MULTIPLE command. Interrupts are not generated on every sector, but
on the transfer of a block that contains the number of sectors defined by the SET MULTIPLE MODE.
Command execution is identical to the WRITE MULTIPLE operation except that the sectors are written without
an implied erase operation. The sectors should be pre-erased by a preceding CFA ERASE SECTORS
command.
8.5 CFA WRITE SECTORS WITHOUT ERASE
8.5.1 Command code
38h
8.5.2 Feature set
CFA feature set.
− If the CFA feature set is implemented this command shall be implemented.
8.5.3 Protocol
PIO data-out command (see 9.6).
8.5.4 Inputs
The Cylinder Low, Cylinder High, Device/Head, and Sector Number specify the starting sector address to be
written. The Sector Count register specifies the number of sectors to be transferred.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
38h
Sector Count –
number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number –
starting sector number or LBA address bits (7:0).
Cylinder Low –
starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High –
starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head –
bit 6 set to one if LBA address, cleared to zero if CHS address starting head number or LBA address
bits (27:24).
Page 78
T13/1321D revision 3
8.5.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be cleared to zero.
8.5.6 Error outputs
The device shall return command aborted if the command is not supported. An unrecoverable error encountered
during execution of this command results in the termination of the command. The command block registers
contain the address of the sector where the first unrecovered error occurred. The amount of data transferred is
indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
na
na
DRDY
5
na
4
IDNF
3
na
2
ABRT
1
na
0
MED
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
DRQ
na
na
ERR
Error Register –
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
ABRT shall be set to one if the command is not supported. ABRT may be set to one if the device is not
able to complete the action requested by the command. ABRT shall be set to one if an
address outside of the range of user-accessible addresses is requested if IDNF is not set to
one.
MED shall be set to one if a media error is detected
Sector Number, Cylinder Low, Cylinder High, Device/Head –
shall be written with the address of first unrecoverable error.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
Page 79
T13/1321D revision 3
8.5.7 Prerequisites
DRDY set to one.
8.5.8 Description
This command is similar to the WRITE SECTORS command. Command execution is identical to the WRITE
SECTORS operation except that the sectors are written without an implied erase operation. The sectors should
be pre-erased by a preceding CFA ERASE SECTORS command.
8.6 CHECK POWER MODE
8.6.1 Command code
E5h
8.6.2 Feature set
Power Management feature set.
− Power Management feature set is mandatory when power management is not implemented by a
PACKET power management feature set.
− This command is mandatory when the Power Management feature set is implemented.
8.6.3 Protocol
Non-data command (see 9.4).
8.6.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
E5h
Device/Head register –
DEV shall indicate the selected device.
8.6.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
obs
na
DRDY
obs
DF
BSY
4
3
na
Result value
na
na
na
DEV
na
na
DRQ
Status register –
BSY shall be cleared to zero indicating command completion.
Page 80
2
1
0
na
na
na
na
na
ERR
T13/1321D revision 3
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
Device/Head register –
DEV shall indicate the selected device.
Sector Count result value –
00h – device is in Standby mode.
80h – device is in Idle mode.
FFh – device is in Active mode or Idle mode.
8.6.6 Error outputs
The device shall return command aborted if the device does not support the Power Management feature set.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register –
ABRT shall be set to one if Power Management feature set is not supported. ABRT may be set to one
if the device is not able to complete the action requested by the command.
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.6.7 Prerequisites
DRDY set to one.
8.6.8 Description
The CHECK POWER MODE command allows the host to determine the current power mode of the device. The
CHECK POWER MODE command shall not cause the device to change power or affect the operation of the
Standby timer.
Page 81
T13/1321D revision 3
8.7 DEVICE RESET
8.7.1 Command code
08h
8.7.2 Feature set
General feature set
− Use prohibited when the PACKET Command feature set is not implemented.
− Mandatory when the PACKET Command feature set is implemented.
8.7.3 Protocol
Device reset (see 9.11).
8.7.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
1
0
0
0
na
na
na
na
na
obs
na
obs
DEV
08h
Device/Head register –
DEV shall indicate the selected device.
8.7.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
0
0
5
4
3
2
Diagnostic results
signature
signature
signature
signature
DEV
0
0
0
see 9.11
Error register –
The diagnostic code as described in 8.9 is placed in this register.
Sector Count, Sector Number, Cylinder low, Cylinder High –
Signature (see 9.12).
Device/Head register –
DEV shall indicate the selected device.
Status register –
see 9.11.
8.7.6 Error outputs
If supported, this command shall not end in an error condition. If this command is not supported and the device
has the BSY bit or the DRQ bit set to one when the command is written, the results of this command are
Page 82
T13/1321D revision 3
indeterminate. If this command is not supported and the device has the BSY bit and the DRQ bit cleared to
zero when the command is written, the device shall respond command aborted.
8.7.7 Prerequisites
This command shall be accepted when BSY or DRQ is set to one, DRDY is cleared to zero, or DMARQ is
asserted. This command shall be accepted when in Sleep mode.
8.7.8 Description
The DEVICE RESET command enables the host to reset an individual device without affecting the other device.
8.8 DOWNLOAD MICROCODE
8.8.1 Command code
92h
8.8.2 Feature set
General feature set
− Optional for devices not implementing the PACKET Command feature set.
− Use prohibited for devices implementing the PACKET Command feature set.
8.8.3 Protocol
PIO data-out (see 9.6).
8.8.4 Inputs
The head bits of the Device/Head register shall always be cleared to zero. The Cylinder High and Low registers
shall be cleared to zero. The Sector Number and Sector Count registers are used together as a 16-bit sector
count value. The Feature register specifies the subcommand code.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
Subcommand code
Sector count (low order)
Sector count (high order)
00h
00h
obs DEV
0
0
92h
1
0
0
0
Device/Head register –
DEV shall indicate the selected device.
Page 83
T13/1321D revision 3
8.8.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.8.6 Error outputs
The device shall return command aborted if the device does not support this command or did not accept the
microcode data. The device shall return command aborted if subcommand code is not a supported value.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register –
ABRT shall be set to one if the device does not support this command or did not accept the microcode
data. ABRT may be set to one if the device is not able to complete the action requested by the
command.
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.8.7 Prerequisites
DRDY set to one.
Page 84
T13/1321D revision 3
8.8.8 Description
This command enables the host to alter the device’s microcode. The data transferred using the DOWNLOAD
MICROCODE command is vendor specific.
All transfers shall be an integer multiple of the sector size. The size of the data transfer is determined by the
contents of the Sector Number register and the Sector Count register. The Sector Number register shall be
used to extend the Sector Count register to create a sixteen bit sector count value. The Sector Number
register shall be the most significant eight bits and the Sector Count register shall be the least significant eight
bits. A value of zero in both the Sector Number register and the Sector Count register shall indicate no data is
to be transferred. This allows transfer sizes from 0 bytes to 33,553,920 bytes, in 512 byte increments.
The Features register shall be used to determine the effect of the DOWNLOAD MICROCODE command. The
values for the Features register are:
−
−
01h – download is for immediate, temporary use.
07h – save downloaded code for immediate and future use.
Either or both values may be supported. All other values are reserved.
8.9 EXECUTE DEVICE DIAGNOSTIC
8.9.1 Command code
90h
8.9.2 Feature set
General feature set
− Mandatory for all devices.
8.9.3 Protocol
Device diagnostic (see 9.10).
8.9.4 Inputs
None. The device selection bit in the Device/Head register is ignored.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
na
90h
Device/Head register –
DEV shall indicate the selected device.
Page 85
T13/1321D revision 3
8.9.5 Normal outputs
The diagnostic code written into the Error register is an 8-bit code. Table 19 defines these values. The values of
the bits in the Error register are not as defined in 7.11.6.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
Diagnostic code
Signature
Signature
Signature
Signature
Signature
see 9.10
2
1
0
Error register –
Diagnostic code.
Sector Count, Sector number, Cylinder Low, Cylinder High, Device/Head registers –
device signature (see 9.12).
Device/Head register –
DEV shall indicate the selected device.
Status register –
see 9.10.
Table 19 − Diagnostic codes
Code (see note 1)
Description
When this code is in the Device 0 Error register
01h
Device 0 passed, Device 1 passed or not present
00h, 02h-7Fh
Device 0 failed, Device 1 passed or not present
81h
Device 0 passed, Device 1 failed
80h, 82h-FFh
Device 0 failed, Device 1 failed
When this code is in the Device 1 Error register
01h
Device 1 passed (see note 2)
00h,02h-7Fh
Device 1 failed (see note 2)
NOTE −
1 Codes other than 01h and 81h may indicate additional information about
the failure(s).
2 If Device 1 is not present, the host may see the information from Device 0
even though Device 1 is selected.
8.9.6 Error outputs
Table 19 shows the error information that is returned as a diagnostic code in the Error register.
8.9.7 Prerequisites
This command shall be accepted regardless of the state of DRDY.
8.9.8 Description
This command shall perform the internal diagnostic tests implemented by the device. The DEV bit in the
Device/Head register is ignored. Both devices, if present, shall execute this command regardless of which
device is selected.
Page 86
T13/1321D revision 3
If the host issues an EXECUTE DEVICE DIAGNOSTIC command while a device is in or going to a power
management mode except Sleep, then the device shall execute the EXECUTE DEVICE DIAGNOSTIC
sequence.
8.10 FLUSH CACHE
8.10.1 Command code
E7h
8.10.2 Feature set
General feature set
− Mandatory for all devices.
8.10.3 Protocol
Non-data command (see 9.4).
8.10.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
E7h
Device/Head register –
DEV shall indicate the selected device.
8.10.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
Page 87
T13/1321D revision 3
8.10.6 Error outputs
An unrecoverable error encountered during execution of writing data results in the termination of the command
and the Command Block registers contain the sector address of the sector where the first unrecoverable error
occurred. The sector is removed from the cache. Subsequent FLUSH CACHE commands continue the process
of flushing the cache.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
na
na
DRDY
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
DRQ
na
na
ERR
Error register –
ABRT may be set to one if the device is not able to complete the action requested by the command.
Sector Number, Cylinder Low, Cylinder High, Device/Head –
shall be written with the address of the first unrecoverable error.
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.10.7 Prerequisites
DRDY set to one.
8.10.8 Description
This command is used by the host to request the device to flush the write cache. If the write cache is to be
flushed, all data cached shall be written to the media. The BSY bit shall remain set to one until all data has
been successfully written or an error occurs. The device should use all error recovery methods available to
ensure the data is written successfully. The flushing of write cache may take several seconds to complete
depending upon the amount of data to be flushed and the success of the operation.
NOTE − This command may take longer than 30 s to complete.
8.11 GET MEDIA STATUS
8.11.1 Command code
DAh
Page 88
T13/1321D revision 3
8.11.2 Feature set
Removable Media Status Notification feature set
− Mandatory for devices implementing the Removable Media Status Notification feature set.
Removable Media feature set
− Optional for devices implementing the Removable Media feature set.
8.11.3 Protocol
Non-data command (see 9.4).
8.11.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
DAh
na
na
na
na
na
na
na
na
obs
na
obs
DEV
Device/Head register –
DEV shall indicate the selected device.
8.11.5 Normal outputs
Normal outputs are returned if Media Status Notification is disabled or if no bits are set to one in the Error
register.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.11.6 Error outputs
If the device does not support this command, the device shall return command aborted.
Page 89
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
WP
5
MC
4
na
3
MCR
2
ABRT
1
NM
0
obs
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register –
ABRT shall be set to one if device does not support this command. ABRT may be set to one if the
device is not able to complete the action requested by the command.
NM (No Media) shall be set to one if no media is present in the device. This bit shall be set to one for
each execution of GET MEDIA STATUS until media is inserted into the device.
MCR (Media Change Request) shall be set to one if the eject button is pressed by the user and
detected by the device. The device shall reset this bit after each execution of the GET MEDIA
STATUS command and only set the bit again for subsequent eject button presses.
MC (Media Change) shall be set to one when the device detects media has been inserted. The device
shall reset this bit after each execution of the GET MEDIA STATUS command and only set the
bit again for subsequent media insertions.
WP (Write Protect) shall be set to one for each execution of GET MEDIA STATUS while the media is
write protected.
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.11.7 Prerequisites
DRDY set to one.
8.11.8 Description
This command returns media status bits WP, MC, MCR, and NM, as defined above. When Media Status
Notification is disabled this command returns zeros in the WP, MC, MCR, and NM bits.
8.12 IDENTIFY DEVICE
8.12.1 Command code
ECh
8.12.2 Feature set
General feature set
−
−
Mandatory for all devices.
Devices implementing the PACKET Command feature set (see 8.12.5.2).
Page 90
T13/1321D revision 3
8.12.3 Protocol
PIO data-in (see 9.5).
8.12.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
ECh
na
na
na
na
na
na
na
na
obs
na
DEV
obs
Device/Head register –
DEV shall indicate the selected device.
8.12.5 Outputs
8.12.5.1 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register –
DEV shall indicate the selected device.
Status register –
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.12.5.2 Outputs for PACKET Command feature set devices
In response to this command, devices that implement the PACKET Command feature set shall post command
aborted and place the PACKET Command feature set signature in the Command Block registers (see 9.12).
8.12.6 Error outputs
Devices not implementing the PACKET Command feature set shall not report an error.
8.12.7 Prerequisites
DRDY set to one.
Page 91
T13/1321D revision 3
8.12.8 Description
The IDENTIFY DEVICE command enables the host to receive parameter information from the device.
Some devices may have to read the media in order to complete this command.
When the command is issued, the device sets the BSY bit to one, prepares to transfer the 256 words of device
identification data to the host, sets the DRQ bit to one, clears the BSY bit to zero, and asserts INTRQ if nIEN
is cleared to zero. The host may then transfer the data by reading the Data register. Table 20 defines the
arrangement and meaning of the parameter words in the buffer. All reserved bits or words shall be zero.
Some parameters are defined as a 16-bit value. A word that is defined as a 16-bit value places the most
significant bit of the value on bit DD15 and the least significant bit on bit DD0.
Some parameters are defined as 32-bit values (e.g., words 57 and 58). Such fields are transferred using two
successive word transfers. The device shall first transfer the least significant bits, bits 15 through 0 of the
value, on bits DD (15:0) respectively. After the least significant bits have been transferred, the most significant
bits, bits 31 through 16 of the value, shall be transferred on DD (15:0) respectively.
Some parameters are defined as a string of ASCII characters. ASCII data fields shall contain only graphic
codes (i.e., code values 20h through 7Eh). For the string “Copyright”, the character “C” is the first byte, the
character “o” is the second byte, etc. When such fields are transferred, the order of transmission is:
the 1st character (“C”) is on bits DD (15:8) of the first word,
the 2nd character (“o”) is on bits DD (7:0) of the first word,
the 3rd character (“p”) is on bits DD (15:8) of the second word,
the 4th character (“y”) is on bits DD (7:0) of the second word,
the 5th character (“r”) is on bits DD (15:8) of the third word,
the 6th character (“i”) is on bits DD (7:0) of the third word,
the 7th character (“g”) is on bits DD (15:8) of the fourth word,
the 8th character (“h”) is on bits DD (7:0) of the fourth word,
the 9th character (“t”) is on bits DD (15:8) of the fifth word,
the 10th character (“space”) is on bits DD (7:0) of the fifth word,
Page 92
T13/1321D revision 3
Table 20 − IDENTIFY DEVICE information
Word
0
1
2
3
4-5
6
7-8
9
10-19
20-21
22
23-26
27-46
47
48
49
F/V
F
F
F
F
F
V
F
F
V
V
F
F
F
V
F
F
F
F
F
F
X
R
F
R
R
F
R
F
50
51-52
53
F
R
R
X
F
F
R
F
F
V
54
V
General configuration bit-significant information:
15
0 = ATA device
14-8
Retired
7
1 = removable media device
6
1 = not removable controller and/or device
5-3
Retired
2
Response incomplete
1
Retired
0
Reserved
Number of logical cylinders
Specific configuration
Number of logical heads
Retired
Number of logical sectors per logical track
Reserved for assignment by the CompactFlash Association
Retired
Serial number (20 ASCII characters)
Retired
Obsolete
Firmware revision (8 ASCII characters)
Model number (40 ASCII characters)
15-8
80h
7-0
00h = Reserved
01h-FFh = Maximum number of sectors that shall be transferred per interrupt on
READ/WRITE MULTIPLE commands
Reserved
Capabilities
15-14
Reserved for the IDENTIFY PACKET DEVICE command.
13
1 = Standby timer values as specified in this standard are supported
0 = Standby timer values shall be managed by the device
12
Reserved for the IDENTIFY PACKET DEVICE command.
11
1 = IORDY supported
0 = IORDY may be supported
10
1 = IORDY may be disabled
9
Shall be set to one. Utilized by IDENTIFY PACKET DEVICE command.
8
Shall be set to one. Utilized by IDENTIFY PACKET DEVICE command.
7-0
Retired
Capabilities
15
Shall be cleared to zero.
14
Shall be set to one.
13-1
Reserved.
0
Shall be set to one to indicate a device specific Standby timer value minimum.
Obsolete
15-3
Reserved
2
1 = the fields reported in word 88 are valid
0 = the fields reported in word 88 are not valid
1
1 = the fields reported in words 64-70 are valid
0 = the fields reported in words 64-70 are not valid
0
1 = the fields reported in words 54-58 are valid
0 = the fields reported in words 54-58 are not valid
Number of current logical cylinders
(continued)
Page 93
T13/1321D revision 3
Table 20 − IDENTIFY DEVICE information (continued)
Word
55
56
57-58
59
F/V
V
V
V
R
V
V
60-61
62
63
F
F
R
V
V
V
64
R
F
F
F
R
F
65
F
66
F
67
F
68
69-70
71-74
75
F
R
R
F
76-79
R
Number of current logical heads
Number of current logical sectors per track
Current capacity in sectors
15-9
Reserved
8
1 = Multiple sector setting is valid
7-0
Xxh = Current setting for number of sectors that shall be transferred per interrupt on
R/W Multiple command
Total number of user addressable sectors (LBA mode only)
Obsolete
15-11
Reserved
10
1 = Multiword DMA mode 2 is selected
0 = Multiword DMA mode 2 is not selected
9
1 = Multiword DMA mode 1 is selected
0 = Multiword DMA mode 1 is not selected
8
1 = Multiword DMA mode 0 is selected
0 = Multiword DMA mode 0 is not selected
7-3
Reserved
2
1 = Multiword DMA mode 2 and below are supported
1
1 = Multiword DMA mode 1 and below are supported
0
1 = Multiword DMA mode 0 is supported
15-8
Reserved
7-0
Advanced PIO modes supported
Minimum Multiword DMA transfer cycle time per word
15-0
Cycle time in nanoseconds
Manufacturer’s recommended Multiword DMA transfer cycle time
15-0
Cycle time in nanoseconds
Minimum PIO transfer cycle time without flow control
15-0
Cycle time in nanoseconds
Minimum PIO transfer cycle time with IORDY flow control
15-0
Cycle time in nanoseconds
Reserved (for future command overlap and queuing)
Reserved for IDENTIFY PACKET DEVICE command.
Queue depth
15-5
Reserved
4-0
Maximum queue depth – 1
Reserved
(continued)
Page 94
T13/1321D revision 3
Table 20 − IDENTIFY DEVICE information (continued)
Word
80
F/V
F
81
F
82
F
83
F
Major version number
0000h or FFFFh = device does not report version
15
Reserved
14
Reserved for ATA/ATAPI-14
13
Reserved for ATA/ATAPI-13
12
Reserved for ATA/ATAPI-12
11
Reserved for ATA/ATAPI-11
10
Reserved for ATA/ATAPI-10
9
Reserved for ATA/ATAPI-9
8
Reserved for ATA/ATAPI-8
7
Reserved for ATA/ATAPI-7
6
Reserved for ATA/ATAPI-6
5
1 = supports ATA/ATAPI-5
4
1 = supports ATA/ATAPI-4
3
1 = supports ATA-3
2
1 = supports ATA-2
1
Obsolete
0
Reserved
Minor version number
0000h or FFFFh = device does not report version
0001h-FFFEh = see 8.12.44
Command set supported.
15
Obsolete
14
1 = NOP command supported
13
1 = READ BUFFER command supported
12
1 = WRITE BUFFER command supported
11
Obsolete
10
1 = Host Protected Area feature set supported
9
1 = DEVICE RESET command supported
8
1 = SERVICE interrupt supported
7
1 = release interrupt supported
6
1 = look-ahead supported
5
1 = write cache supported
4
Shall be cleared to zero
3
1 = supports Power Management feature set
2
1 = supports Removable Media feature set
1
1 = supports Security Mode feature set
0
1 = supports SMART feature set
Command sets supported.
15
Shall be cleared to zero
14
Shall be set to one
13-9
Reserved
8
1 = SET MAX security extension supported
7
Reserved for project 1407DT Address Offset Reserved Area Boot
6
1 = SET FEATURES subcommand required to spinup after power-up
5
1 = Power-Up In Standby feature set supported
4
1 = Removable Media Status Notification feature set supported
3
1 = Advanced Power Management feature set supported
2
1 = CFA feature set supported
1
1 = READ/WRITE DMA QUEUED supported
0
1 = DOWNLOAD MICROCODE command supported
(continued)
Page 95
T13/1321D revision 3
Table 20 − IDENTIFY DEVICE information (continued)
Word
84
F/V
F
85
V
86
V
87
V
88
R
V
V
V
V
V
R
F
F
F
F
F
Command set/feature supported extension.
15
Shall be cleared to zero
14
Shall be set to one
13-0
Reserved
Command set/feature enabled.
15
Obsolete
14
1 = NOP command enabled
13
1 = READ BUFFER command enabled
12
1 = WRITE BUFFER command enabled
11
Obsolete
10
1 = Host Protected Area feature set enabled
9
1 = DEVICE RESET command enabled
8
1 = SERVICE interrupt enabled
7
1 = release interrupt enabled
6
1 = look-ahead enabled
5
1 = write cache enabled
4
Shall be cleared to zero
3
1 = Power Management feature set enabled
2
1 = Removable Media feature set enabled
1
1 = Security Mode feature set enabled
0
1 = SMART feature set enabled
Command set/feature enabled.
15-9
Reserved
8
1 = SET MAX security extension enabled by SET MAX SET PASSWORD
7
Reserved for project 1407DT Address Offset Reserved Area Boot
6
1 = SET FEATURES subcommand required to spin-up after power-up
5
1 = Power-Up In Standby feature set enabled
4
1 = Removable Media Status Notification feature set enabled
3
1 = Advanced Power Management feature set enabled
2
1 = CFA feature set enabled
1
1 = READ/WRITE DMA QUEUED command supported
0
1 = DOWNLOAD MICROCODE command supported
Command set/feature default.
15
Shall be cleared to zero
14
Shall be set to one
13-0
Reserved
15-13
Reserved
12
1 = Ultra DMA mode 4 is selected
0 = Ultra DMA mode 4 is not selected
11
1 = Ultra DMA mode 3 is selected
0 = Ultra DMA mode 3 is not selected
10
1 = Ultra DMA mode 2 is selected
0 = Ultra DMA mode 2 is not selected
9
1 = Ultra DMA mode 1 is selected
0 = Ultra DMA mode 1 is not selected
8
1 = Ultra DMA mode 0 is selected
0 = Ultra DMA mode 0 is not selected
7-5
Reserved
4
1 = Ultra DMA mode 4 and below are supported
3
1 = Ultra DMA mode 3 and below are supported
2
1 = Ultra DMA mode 2 and below are supported
1
1 = Ultra DMA mode 1 and below are supported
0
1 = Ultra DMA mode 0 is supported
(continued)
Page 96
T13/1321D revision 3
Table 20 − IDENTIFY DEVICE information (continued)
Word
89
90
91
92
93
F/V
F
F
V
V
V
94-126
127
R
F
Time required for security erase unit completion
Time required for Enhanced security erase completion
Current advanced power management value
Master Password Revision Code
Hardware reset result. The contents of bits 12-0 of this word shall change only during the
execution of a hardware reset.
15
Shall be cleared to zero.
14
Shall be set to one.
13
1 = device detected CBLID- above ViH
0 = device detected CBLID- below ViL
12-8
Device 1 hardware reset result. Device 0 shall clear these bits to zero. Device 1 shall
set these bits as follows:
12
Reserved.
11
0 = Device 1 did not assert PDIAG-.
1 = Device 1 asserted PDIAG-.
10-9
These bits indicate how Device 1 determined the device number:
00 = Reserved.
01 = a jumper was used.
10 = the CSEL signal was used.
11 = some other method was used or the method is unknown.
8
Shall be set to one.
7-0
Device 0 hardware reset result. Device 1 shall clear these bits to zero. Device 0 shall
set these bits as follows:
7
Reserved.
6
0 = Device 0 does not respond when Device 1 is selected.
1 = Device 0 responds when Device 1 is selected.
5
0 = Device 0 did not detect the assertion of DASP-.
1 = Device 0 detected the assertion of DASP-.
4
0 = Device 0 did not detect the assertion of PDIAG-.
1 = Device 0 detected the assertion of PDIAG-.
3
0 = Device 0 failed diagnostics.
1 = Device 0 passed diagnostics.
2-1
These bits indicate how Device 0 determined the device number:
00 = Reserved.
01 = a jumper was used.
10 = the CSEL signal was used.
11 = some other method was used or the method is unknown.
0
Shall be set to one.
Reserved
Removable Media Status Notification feature set support
15-2
Reserved
1-0
00 = Removable Media Status Notification feature set not supported
01 = Removable Media Status Notification feature supported
10 = Reserved
11 = Reserved
(continued)
Page 97
T13/1321D revision 3
Table 20 − IDENTIFY DEVICE information (concluded)
Word
128
F/V
V
129-159
160
X
V
161-175
176-254
255
R
R
F/V
Security status
15-9
Reserved
8
Security level 0 = High, 1 = Maximum
7-6
Reserved
5
1 = Enhanced security erase supported
4
1 = Security count expired
3
1 = Security frozen
2
1 = Security locked
1
1 = Security enabled
0
1 = Security supported
Vendor specific
CFA power mode 1
15
Word 160 supported
14
Reserved
13
CFA power mode 1 is required for one or more commands implemented by the
device
12
CFA power mode 1 disabled
11-0
Maximum current in ma
Reserved for assignment by the CompactFlash Association
Reserved
Integrity word
15-8
Checksum
7-0
Signature
Key:
F = the content of the word is fixed and does not change. For removable media devices, these values may
change when media is removed or changed.
V = the contents of the word is variable and may change depending on the state of the device or the
commands executed by the device.
X = the content of the word is vendor specific and may be fixed or variable.
R = the content of the word is reserved and shall be zero.
8.12.9 Word 0: General configuration
Devices that conform to this standard shall clear bit 15 to zero.
If bit 2 is set to one it indicates that the content of the IDENTIFY DEVICE response is incomplete. This will
occur if the device supports the Power-up in Standby feature set and required data is contained on the device
media. In this case the content of at least words 0 and 2 shall be valid.
Devices reporting a value of 848Ah in this word shall support the CFA feature set.
8.12.10 Word 1: Number of cylinders
This word contains the number of user-addressable logical cylinders in the default CHS translation (see 6.2).
8.12.11 Word 2: Specific configuration.
Word 2 shall be set as follows:
Page 98
T13/1321D revision 3
Value
37C8h
738Ch
8C73h
C837h
All other values
Description
Device requires SET FEATURES subcommand to spin-up after power-up and
IDENTIFY DEVICE response is incomplete (see 6.18).
Device requires SET FEATURES subcommand to spin-up after power-up and
IDENTIFY DEVICE response is complete (see 6.18).
Device does not require SET FEATURES subcommand to spin-up after
power-up and IDENTIFY DEVICE response is incomplete (see 6.18).
Device does not require SET FEATURES subcommand to spin-up after
power-up and IDENTIFY DEVICE response is complete (see 6.18).
Reserved.
8.12.12 Word 3: Number of logical heads
This word contains the number of user-addressable logical heads per logical cylinder in the default CHS
translation (see 6.2).
8.12.13 Word 4-5: Retired.
8.12.14 Word 6: Number of logical sectors per logical track
This word contains the number of user-addressable logical sectors per logical track in the default CHS
translation (see 6.2).
8.12.15
Words 7-8: Reserved for assignment by the CompactFlash  Association
8.12.16 Word 9: Retired.
8.12.17 Words 10-19: Serial number
This field contains the serial number of the device. The contents of this field is an ASCII character string of
twenty bytes. The device shall pad the character string with spaces (20h), if necessary, to ensure that the
string is the proper length. The combination of Serial number (words 10-19) and Model number (words 27-46)
shall be unique for a given manufacturer.
8.12.18 Word 20-21: Retired.
8.12.19 Word 22: Obsolete.
8.12.20 Word 23-26: Firmware revision
This field contains the firmware revision number of the device. The contents of this field is an ASCII character
string of eight bytes. The device shall pad the character string with spaces (20h), if necessary, to ensure that
the string is the proper length.
8.12.21 Words 27-46: Model number
This field contains the model number of the device. The contents of this field is an ASCII character string of forty
bytes. The device shall pad the character string with spaces (20h), if necessary, to ensure that the string is the
proper length. The combination of Serial number (words 10-19) and Model number (words 27-46) shall be unique
for a given manufacturer.
Page 99
T13/1321D revision 3
8.12.22 Word 47: READ/WRITE MULTIPLE support.
Bits 7-0 of this word define the maximum number of sectors per block that the device supports for
READ/WRITE MULTIPLE commands.
8.12.23 Word 48: Reserved.
8.12.24 Word 49-50: Capabilities
Bits 15 and 14 of word 49 are reserved for use in the IDENTIFY PACKET DEVICE command response.
Bit 13 of word 49 is used to determine whether a device utilizes the Standby timer values as defined in this
standard. Table 23 specifies the Standby timer values utilized by the device if bit 13 is set to one. If bit 13 is
cleared to zero, the timer values shall be vendor specific.
Bit 12 of word 49 is reserved for use in the IDENTIFY PACKET DEVICE command response.
Bit 11 of word 49 is used to determine whether a device supports IORDY. If this bit is set to one, then the
device supports IORDY operation. If this bit is zero, the device may support IORDY. This ensures backward
compatibility. If a device supports PIO mode 3 or higher, then this bit shall be set to one.
Bit 10 of word 49 is used to indicate a device’s ability to enable or disable the use of IORDY. If this bit is set to
one, then the device supports the disabling of IORDY. Disabling and enabling of IORDY is accomplished using
the SET FEATURES command.
Bits 9 and 8 of word 49 shall be set to one for backward compatibility. These bits are defined for use in the
IDENTIFY PACKET DEVICE command response.
Bits 7 through 0 of word 49 are retired.
Bit 15 of word 50 shall be cleared to zero to indicate that the contents of word 50 are valid.
Bit 14 of word 50 shall be set to one to indicate that the contents of word 50 are valid.
Bits 13 through 1 of word 50 are reserved.
Bit 0 of word 50 set to one indicates that the device has a minimum Standby timer value that is device specific.
8.12.25 Words 51 and 52: Obsolete
8.12.26 Word 53: Field validity
If bit 0 of word 53 is set to one, the values reported in words 54 through 58 are valid. If this bit is cleared to
zero, the values reported in words 54 through 58 are not valid. If bit 1 of word 53 is set to one, the values
reported in words 64 through 70 are valid. If this bit is cleared to zero, the values reported in words 64-70 are
not valid. Any device that supports PIO mode 3 or above, or supports Multiword DMA mode 1 or above, shall
set bit 1 of word 53 to one and support the fields contained in words 64 through 70. If the device supports Ultra
DMA and the values reported in word 88 are valid, then bit 2 of word 53 shall be set to one. If the device does
not support Ultra DMA and the values reported in word 88 are not valid, then this bit is cleared to zero.
8.12.27 Word 54: Number of current logical cylinders
This field contains the number of user-addressable logical cylinders in the current CHS translation (see 6.2).
Page 100
T13/1321D revision 3
8.12.28 Word 55: Number of current logical heads
This field contains the number of user-addressable logical heads per logical cylinder in the current CHS
translation (see 6.2).
8.12.29 Word 56: Number of current logical sectors per logical track
This field contains the number of user-addressable logical sectors per logical track in the current CHS
translation (see 6.2).
8.12.30 Word (58:57): Current capacity in sectors
This field contains the current capacity in sectors in the current CHS translation (see 6.2).
8.12.31 Word 59: Multiple sector setting
If bit 8 is set to one, bits 7-0 reflect the number of sectors currently set to transfer on a READ/WRITE
MULTIPLE command. This field may default to the preferred value for the device.
8.12.32 Word (61:60): Total number of user addressable sectors
This field contains the total number of user addressable sectors (see 6.2).
8.12.33 Word 62: Obsolete
8.12.34 Word 63: Multiword DMA transfer
Word 63 identifies the Multiword DMA transfer modes supported by the device and indicates the mode that is
currently selected. Only one DMA mode shall be selected at any given time. If an Ultra DMA mode is enabled,
then no Multiword DMA mode shall be enabled. If a Multiword DMA mode is enabled then no Ultra DMA mode
shall be enabled.
8.12.34.1 Reserved
Bits 15 though 11 of word 63 are reserved.
8.12.34.2 Multiword DMA mode 2 selected
If bit 10 of word 63 is set to one, then Multiword DMA mode 2 is selected. If this bit is cleared to zero, then
Multiword DMA mode 2 is not selected. If bit 9 is set to one or if bit 8 is set to one, then this bit shall be
cleared to zero.
8.12.34.3 Multiword DMA mode 1 selected
If bit 9 of word 63 is set to one, then Multiword DMA mode 1 is selected. If this bit is cleared to zero then
Multiword DMA mode 1 is not selected. If bit 10 is set to one or if bit 8 is set to one, then this bit shall be
cleared to zero.
8.12.34.4 Multiword DMA mode 0 selected
If bit 8 of word 63 is set to one, then Multiword DMA mode 0 is selected. If this bit is cleared to zero then
Multiword DMA mode 0 is not selected. If bit 10 is set to one or if bit 9 is set to one, then this bit shall be
cleared to zero.
Page 101
T13/1321D revision 3
8.12.34.5 Reserved
Bits 7 through 3 of word 63 are reserved.
8.12.34.6 Multiword DMA mode 2 supported
If bit 2 of word 63 is set to one, then Multiword DMA modes 2 and below are supported. If this bit is cleared to
zero, then Multiword DMA mode 2 is not supported. If Multiword DMA mode 2 is supported, then Multiword
DMA modes 1 and 0 shall also be supported. If this bit is set to one, bits 1 and 0 shall be set to one.
8.12.34.7 Multiword DMA mode 1 supported
If bit 1 of word 63 is set to one, then Multiword DMA modes 1 and below are supported. If this bit is cleared to
zero, then Multiword DMA mode 1 is not supported. If Multiword DMA mode 1 is supported, then Multiword
DMA mode 0 shall also be supported. If this bit is set to one, bit 0 shall be set to one.
8.12.34.8 Multiword DMA mode 0 supported
If bit 0 of word 63 is set to one, then Multiword DMA mode 0 is supported.
8.12.35 Word 64: PIO transfer modes supported
Bits 7 through 0 of word 64 of the Identify Device parameter information is defined as the advanced PIO data
transfer supported field. If this field is supported, bit 1 of word 53 shall be set to one. This field is bit significant.
Any number of bits may be set to one in this field by the device to indicate the advanced PIO modes the device
is capable of supporting.
Of these bits, bits 7 through 2 are Reserved for future advanced PIO modes. Bit 0, if set to one, indicates that
the device supports PIO mode 3. Bit 1, if set to one, indicates that the device supports PIO mode 4.
8.12.36 Word 65: Minimum Multiword DMA transfer cycle time per word
Word 65 of the parameter information of the IDENTIFY DEVICE command is defined as the minimum Multiword
DMA transfer cycle time per word. This field defines, in nanoseconds, the minimum cycle time that the device
supports when performing Multiword DMA transfers on a per word basis.
If this field is supported, bit 1 of word 53 shall be set to one. Any device that supports Multiword DMA mode 1
or above shall support this field, and the value in word 65 shall not be less than the minimum cycle time for the
fastest DMA mode supported by the device.
If bit 1 of word 53 is set to one because a device supports a field in words 64-70 other than this field and the
device does not support this field, the device shall return a value of zero in this field.
8.12.37 Word 66: Device recommended Multiword DMA cycle time
Word 66 of the parameter information of the IDENTIFY DEVICE command is defined as the device
recommended Multiword DMA transfer cycle time. This field defines, in nanoseconds, the minimum cycle time
per word during a single sector host transfer while performing a multiple sector READ DMA or WRITE DMA
command for any location on the media under nominal conditions. If a host runs at a faster cycle rate by
operating at a cycle time of less than this value, the device may negate DMARQ for flow control. The rate at
which DMARQ is negated could result in reduced throughput despite the faster cycle rate. Transfer at this rate
does not ensure that flow control will not be used, but implies that higher performance may result.
If this field is supported, bit 1 of word 53 shall be set to one. Any device that supports Multiword DMA mode 1
or above shall support this field, and the value in word 66 shall not be less than the value in word 65.
Page 102
T13/1321D revision 3
If bit 1 of word 53 is set to one because a device supports a field in words 64-70 other than this field and the
device does not support this field, the device shall return a value of zero in this field.
8.12.38 Word 67: Minimum PIO transfer cycle time without flow control
Word 67 of the parameter information of the IDENTIFY DEVICE command is defined as the minimum PIO
transfer without flow control cycle time. This field defines, in nanoseconds, the minimum cycle time that, if
used by the host, the device guarantees data integrity during the transfer without utilization of flow control.
If this field is supported, Bit 1 of word 53 shall be set to one.
Any device that supports PIO mode 3 or above shall support this field, and the value in word 67 shall not be
less than the value reported in word 68.
If bit 1 of word 53 is set to one because a device supports a field in words 64-70 other than this field and the
device does not support this field, the device shall return a value of zero in this field.
8.12.39 Word 68: Minimum PIO transfer cycle time with IORDY
Word 68 of the parameter information of the IDENTIFY DEVICE command is defined as the minimum PIO
transfer with IORDY flow control cycle time. This field defines, in nanoseconds, the minimum cycle time that
the device supports while performing data transfers while utilizing IORDY flow control.
If this field is supported, Bit 1 of word 53 shall be set to one.
Any device that supports PIO mode 3 or above shall support this field, and the value in word 68 shall be the
fastest defined PIO mode supported by the device.
If bit 1 of word 53 is set to one because a device supports a field in words 64-70 other than this field and the
device does not support this field, the device shall return a value of zero in this field.
8.12.40 Words 69-74: Reserved
8.12.41 Word 75: Queue depth
Bits 4 through 0 of word 75 indicate the maximum queue depth supported by the device. The queue depth
includes all commands for which command acceptance has occured and command completion has not
occurred. The value in this field is the maximum queue depth - 1, e.g., a value of 0 indicates a queue depth of 1,
a value of 31 indicates a queue depth of 32. If bit 1 of word 83 is cleared to zero indicating that the device does
not support READ/WRITE DMA QUEUED commands, the value in this field shall be 0. A device may support
READ/WRITE DMA QUEUED commands to provide overlap only (i.e., queuing not supported), in this case, bit
1 of word 83 shall be set to one and the queue depth shall be set to 0.
8.12.42 Words 76-79: Reserved
8.12.43 Word 80: Major version number
If not 0000h or FFFFh, the device claims compliance with the major version(s) as indicated by bits 2 through 5
being set to one. Values other than 0000h and FFFFh are bit significant. Since ATA standards maintain
downward compatibility, a device may set more than one bit.
Page 103
T13/1321D revision 3
8.12.44 Word 81: Minor version number
If an implementor claims that the revision of the standard they used to guide their implementation does not
need to be reported or if the implementation was based upon a standard prior to the ATA-3 standard, word 81
shall be 0000h or FFFFh.
Table 21 defines the value that may optionally be reported in word 81 to indicate the revision of the standard
that guided the implementation.
Value
0001h
0002h
0003h
0004h
0005h
0006h
0007h
0008h
0009h
000Ah
000Bh
000Ch
000Dh
000Eh
000Fh
0010h
0011h
0012h
0013h
0014h
0015h
0016h
0017h
0018h-FFFFh
Table 21 − Minor version number
Minor revision
Obsolete
Obsolete
Obsolete
ATA-2 published, ANSI X3.279-1996
ATA-2 X3T10 948D prior to revision 2k
ATA-3 X3T10 2008D revision 1
ATA-2 X3T10 948D revision 2k
ATA-3 X3T10 2008D revision 0
ATA-2 X3T10 948D revision 3
ATA-3 published, ANSI X3.298-199x
ATA-3 X3T10 2008D revision 6
ATA-3 X3T13 2008D revision 7 and 7a
ATA/ATAPI-4 X3T13 1153D revision 6
ATA/ATAPI-4 T13 1153D revision 13
ATA/ATAPI-4 X3T13 1153D revision 7
ATA/ATAPI-4 T13 1153D revision 18
ATA/ATAPI-4 T13 1153D revision 15
ATA/ATAPI-4 published, ANSI NCITS 317-1998
ATA/ATAPI-5 T13 1321D revision 3
ATA/ATAPI-4 T13 1153D revision 14
ATA/ATAPI-5 T13 1321D revision 1
Reserved
ATA/ATAPI-4 T13 1153D revision 17
Reserved
8.12.45 Words 82-84: Features/command sets supported
Words 82, 83, and 84 shall indicate features/command sets supported. If bit 14 of word 83 is set to one and bit
15 of word 83 is cleared to zero, the contents of words 82 and 83 contain valid support information. If not,
support information is not valid in these words. If bit 14 of word 84 is set to one and bit 15 of word 84 is cleared
to zero, the contents of word 84 contains valid support information. If not, support information is not valid in this
word.
If bit 0 of word 82 is set to one, the SMART feature set is supported.
If bit 1 of word 82 is set to one, the Security Mode feature set is supported.
If bit 2 of word 82 is set to one, the Removable Media feature set is supported.
If bit 3 of word 82 is set to one, the Power Management feature set is supported.
Bit 4 of word 82 shall be cleared to zero.
Page 104
T13/1321D revision 3
If bit 5 of word 82 is set to one, write cache is supported.
If bit 6 of word 82 is set to one, look-ahead is supported.
If bit 7 of word 82 is set to one, release interrupt is supported.
If bit 8 of word 82 is set to one, SERVICE interrupt is supported.
If bit 9 of word 82 is set to one, the DEVICE RESET command is supported.
If bit 10 of word 82 is set to one, the Host Protected Area feature set is supported.
Bit 11 of word 82 is obsolete.
If bit 12 of word 82 is set to one, the device supports the WRITE BUFFER command.
If bit 13 of word 82 is set to one, the device supports the READ BUFFER command.
If bit 14 of word 82 is set to one, the device supports the NOP command.
Bit 15 of word 82 is obsolete.
If bit 0 of word 83 is set to one, the device supports the DOWNLOAD MICROCODE command.
If bit 1 of word 83 is set to one, the device supports the READ DMA QUEUED and WRITE DMA QUEUED
commands.
If bit 2 of word 83 is set to one, the device supports the CFA feature set.
If bit 3 of word 83 is set to one, the device supports the Advanced Power Management feature set.
If bit 4 of word 83 is set to one, the device supports the Removable Media Status feature set.
If bit 5 of word 83 is set to one, the device supports the Power-Up In Standby feature set.
If bit 6 of word 83 is set to one, the device requires the SET FEATURES subcommand to spin-up after power-up
if the Power-Up In Standby feature set is enabled (see 8.37.15).
Bit 7 is reserved for project 1407DT Address Offset Reserved Area Boot.
If bit 8 of word 83 is set to one, the device supports the SET MAX security extension.
8.12.46 Words 85-87: Features/command sets enabled
Words 85, 86, and 87 shall indicate features/command sets enabled. If bit 14 of word 87 is set to one and bit
15 of word 87 is cleared to zero, the contents of words 85, 86, and 87 contain valid information. If not,
information is not valid in these words.
If bit 0 of word 85 is set to one, the SMART feature set has been enabled via the SMART ENABLE
OPERATIONS command.
If bit 1 of word 85 is set to one, the Security Mode feature set has been enabled via the SECURITY SET
PASSWORD command.
If bit 2 of word 85 is set to one, the Removable Media feature set is supported.
Page 105
T13/1321D revision 3
If bit 3 of word 85 is set to one, the Power Management feature set is supported.
Bit 4 of word 85 shall be cleared to zero.
If bit 5 of word 85 is set to one, write cache has been enabled via the SET FEATURES command (see 8.37.10).
If bit 6 of word 85 is set to one, look-ahead has been enabled via the SET FEATURES command (see 8.37.17).
If bit 7 of word 85 is set to one, release interrupt has been enabled via the SET FEATURES command (see
8.37.18).
If bit 8 of word 85 is set to one, SERVICE interrupt has been enabled via the SET FEATURES command (see
8.37.19).
If bit 9 of word 85 is set to one, the DEVICE RESET command is supported.
If bit 10 of word 85 is set to one, the Host Protected Area feature set is supported.
Bit 11 of word 85 is obsolete.
If bit 12 of word 85 is set to one, the device supports the WRITE BUFFER command.
If bit 13 of word 85 is set to one, the device supports the READ BUFFER command.
If bit 14 of word 85 is set to one, the device supports the NOP command.
Bit 15 of word 85 is obsolete.
If bit 0 of word 86 is set to one, the device supports the DOWNLOAD MICROCODE command.
If bit 1 of word 86 is set to one, the device supports the READ DMA QUEUED and WRITE DMA QUEUED
commands.
If bit 2 of word 86 is set to one, the device supports the CFA feature set.
If bit 3 of word 86 is set to one, the Advanced Power Management feature set has been enabled via the SET
FEATURES command.
If bit 4 of word 86 is set to one, the Removable Media Status feature set has been enabled via the SET
FEATURES command.
If bit 5 of word 86 is set to one, the Power-Up In Standby feature set has been enabled via the SET FEATURES
command (see 8.37.13).
If bit 6 of word 86 is set to one, the device requires the SET FEATURES subcommand to spin-up after power-up
(see 8.37.15).
Bit 7 is reserved for project 1407DT Address Offset Reserved Area Boot.
If bit 8 is set to one, the device has had the SET MAX security extension enabled via a SET MAX SET
PASSWORD command.
8.12.47 Word 88: Ultra DMA modes
Word 88 identifies the Ultra DMA transfer modes supported by the device and indicates the mode that is
currently selected. Only one DMA mode shall be selected at any given time. If an Ultra DMA mode is enabled,
Page 106
T13/1321D revision 3
then no Multiword DMA mode shall be enabled. If a Multiword DMA mode is enabled then no Ultra DMA mode
shall be enabled.
8.12.47.1 Reserved
Bits 15 though 13 of word 88 are reserved.
8.12.47.2 Ultra DMA mode 4 selected
If bit 12 of word 88 is set to one, then Ultra DMA mode 4 is selected. If this bit is cleared to zero, then Ultra
DMA mode 4 is not selected. If bit 11 or bit 10 or bit 9 or bit 8 is set to one, then this bit shall be cleared to
zero.
8.12.47.3 Ultra DMA mode 3 selected
If bit 11 of word 88 is set to one, then Ultra DMA mode 3 is selected. If this bit is cleared to zero, then Ultra
DMA mode 3 is not selected. If bit 12 or bit 10 or bit 9 or bit 8 is set to one, then this bit shall be cleared to
zero.
8.12.47.4 Ultra DMA mode 2 selected
If bit 10 of word 88 is set to one, then Ultra DMA mode 2 is selected. If this bit is cleared to zero, then Ultra
DMA mode 2 is not selected. If bit 12 or bit 11 or bit 9 or bit 8 is set to one, then this bit shall be cleared to
zero.
8.12.47.5 Ultra DMA mode 1 selected
If bit 9 of word 88 is set to one, then Ultra DMA mode 1 is selected. If this bit is cleared to zero then Ultra DMA
mode 1 is not selected. If bit 12 or bit 11 or bit 10 or bit 8 is set to one, then this bit shall be cleared to zero.
8.12.47.6 Ultra DMA mode 0 selected
If bit 8 of word 88 is set to one, then Ultra DMA mode 0 is selected. If this bit is cleared to zero then Ultra DMA
mode 0 is not selected. If bit 12 or bit 11 or bit 10 or bit 9 is set to one, then this bit shall be cleared to zero.
8.12.47.7 Reserved
Bits 7 through 5 of word 88 are reserved.
8.12.47.8 Ultra DMA mode 4 supported
If bit 4 of word 88 is set to one, then Ultra DMA modes 4 and below are supported. If this bit is cleared to zero,
then Ultra DMA mode 4 is not supported. If Ultra DMA mode 4 is supported, then Ultra DMA modes 3, 2, 1 and
0 shall also be supported. If this bit is set to one, then bits 3, 2, 1 and 0 shall be set to one.
8.12.47.9 Ultra DMA mode 3 supported
If bit 3 of word 88 is set to one, then Ultra DMA modes 3 and below are supported. If this bit is cleared to zero,
then Ultra DMA mode 3 is not supported. If Ultra DMA mode 3 is supported, then Ultra DMA modes 2, 1 and 0
shall also be supported. If this bit is set to one, then bits 2, 1 and 0 shall be set to one.
Page 107
T13/1321D revision 3
8.12.47.10 Ultra DMA mode 2 supported
If bit 2 of word 88 is set to one, then Ultra DMA modes 2 and below are supported. If this bit is cleared to zero,
then Ultra DMA mode 2 is not supported. If Ultra DMA mode 2 is supported, then Ultra DMA modes 1 and 0
shall also be supported. If this bit is set to one, bits 1 and 0 shall be set to one.
8.12.47.11 Ultra DMA mode 1 supported
If bit 1 of word 88 is set to one, then Ultra DMA modes 1 and below are supported. If this bit is cleared to zero,
then Ultra DMA mode 1 is not supported. If Ultra DMA mode 1 is supported, then Ultra DMA mode 0 shall also
be supported. If this bit is set to one, bit 0 shall be set to one.
8.12.47.12 Ultra DMA mode 0 supported
If bit 0 of word 88 is set to one, then Ultra DMA mode 0 is supported. If this bit is cleared to zero, then Ultra
DMA is not supported.
8.12.48 Word 89: Time required for Security erase unit completion
Word 89 specifies the time required for the SECURITY ERASE UNIT command to complete.
Value
0
1-254
255
Time
Value not specified
(Value∗2) minutes
>508 minutes
8.12.49 Word 90: Time required for Enhanced security erase unit completion
Word 90 specifies the time required for the ENHANCED SECURITY ERASE UNIT command to complete.
Value
0
1-254
255
Time
Value not specified
(Value∗2) minutes
>508 minutes
8.12.50 Word 91: Advanced power management level value
Bits 7-0 of word 91 contain the current Advanced Power Management level setting.
8.12.51 Word 92: Master Password Revision Code
Word 92 contains the value of the Master Password Revision Code set when the Master Password was last
changed. Valid values are 0001h through FFFEh. A value of 0000h or FFFFh indicates that the Master
Password Revision is not supported.
8.12.52 Word 93: Hardware configuration test results
During hardware reset execution, Device 0 shall clear bits 13-8 of this word to zero and shall set bits 7-0 of the
word as indicated to show the result of the hardware reset execution. During hardware reset execution, Device
1 shall clear bits 7-0 of this word to zero and shall set bits 13-8 as indicated to show the result of the hardware
reset execution.
Bit 13 shall be set or cleared by the selected device to indicate whether the device detected CBLID- above VIH
or below VIL at any time during execution of each IDENTIFY DEVICE routine after receiving the command from
Page 108
T13/1321D revision 3
the host but before returning data to the host. This test may be repeated as desired by the device during
command execution (see Annex B).
8.12.53 Words 94-126: Reserved
8.12.54 Word 127: Removable Media Status Notification feature set support
If bit 0 of word 127 is set to one and bit 1 of word 127 is cleared to zero, the device supports the Removable
Media Status Notification feature set. Bits 15 through 2 shall be cleared to zero.
8.12.55 Word 128: Security status
Bit 8 of word 128 indicates the security level. If security mode is enabled and the security level is high, bit 8
shall be cleared to zero. If security mode is enabled and the security level is maximum, bit 8 shall be set to
one. When security mode is disabled, bit 8 shall be cleared to zero.
Bit 5 of word 128 indicates the Enhanced security erase unit feature is supported. If bit 5 is set to one, the
Enhanced security erase unit feature set is supported.
Bit 4 of word 128 indicates that the security count has expired. If bit 4 is set to one, the security count is
expired and SECURITY UNLOCK and SECURITY ERASE UNIT are command aborted until a power-on reset or
hardware reset.
Bit 3 of word 128 indicates security Frozen. If bit 3 is set to one, the security is Frozen.
Bit 2 of word 128 indicates security locked. If bit 2 is set to one, the security is locked.
Bit 1 of word 128 indicates security enabled. If bit 1 is set to one, the security is enabled.
Bit 0 of word 128 indicates the Security Mode feature set supported. If bit 0 is set to one, security is supported.
8.12.56 Words 129-159: Vendor specific.
8.12.57 Word 160: CFA power mode
Word 160 indicates the presence and status of a CFA feature set device that supports CFA Power Mode 1.
If bit 13 of word 160 is set to one then the device must be in CFA Power Mode 1 to perform one or more
commands implemented by the device.
If bit 12 of word 160 is set to one the device is in CFA Power Mode 0 (see 8.37.14).
Bits 11-0 indicate the maximum average RMS current in Milliamperes required during 3.3V or 5V device
operation in CFA Power Mode 1.
8.12.58 Words 161-175: Reserved for assignment by the CompactFlash  Association
8.12.59 Words 176-254: Reserved.
8.12.60 Word 255: Integrity word
The use of this word is optional. If bits 7:0 of this word contain the signature A5h, bits 15:8 contain the data
structure checksum. The data structure checksum is the two’s complement of the sum of all bytes in words 0
through 254 and the byte consisting of bits 7:0 in word 255. Each byte shall be added with unsigned arithmetic,
and overflow shall be ignored. The sum of all 512 bytes is zero when the checksum is correct.
Page 109
T13/1321D revision 3
8.13 IDENTIFY PACKET DEVICE
8.13.1 Command code
A1h
8.13.2 Feature set
PACKET Command feature set
− Use prohibited for devices not implementing the PACKET Command feature set.
− Mandatory for devices implementing the PACKET Command feature set.
8.13.3 Protocol
PIO data-in (see 9.5).
8.13.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
A1h
Device/Head register DEV shall indicate the selected device.
8.13.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
Page 110
T13/1321D revision 3
8.13.6 Error outputs
The device shall return command aborted if the device does not implement this command, otherwise, the device
shall not report an error.
8.13.7 Prerequisites
This command shall be accepted regardless of the state of DRDY.
8.13.8 Description
The IDENTIFY PACKET DEVICE command enables the host to receive parameter information from a device
that implements the PACKET Command feature set.
Some devices may have to read the media in order to complete this command.
When the command is issued, the device sets the BSY bit to one, prepares to transfer the 256 words of device
identification data to the host, sets the DRQ bit to one, clears the BSY bit to zero, and asserts INTRQ if nIEN
is cleared to zero. The host may then transfer the data by reading the Data register. Table 22 defines the
arrangement and meanings of the parameter words in the buffer. All reserved bits or words shall be zero.
Some parameters are defined as a group of bits. A word that is defined as a set of bits is transmitted with
indicated bits on the respective data bus bit (e.g., bit 15 appears on DD15).
Some parameters are defined as a 16-bit value. A word that is defined as a 16-bit value places the most
significant bit of the value on bit DD15 and the least significant bit on bit DD0.
Some parameters are defined as 32-bit values (e.g., words 57 and 58). Such fields are transferred using two
word transfers. The device shall first transfer the least significant bits, bits 15 through 0 of the value, on bits DD
(15:0) respectively. After the least significant bits have been transferred, the most significant bits, bits 31
through 16 of the value, shall be transferred on DD (15:0) respectively.
Some parameters are defined as a string of ASCII characters. For the string “Copyright”, the character “C” is
the first byte, the character “o” is the second byte, etc. When such fields are transferred, the order of
transmission is:
the 1st character (“C”) is on bits DD (15:8) of the first word,
the 2nd character (“o”) is on bits DD (7:0) of the first word,
the 3rd character (“p”) is on bits DD (15:8) of the second word,
the 4th character (“y”) is on bits DD (7:0) of the second word,
the 5th character (“r”) is on bits DD (15:8) of the third word,
the 6th character (“i”) is on bits DD (7:0) of the third word,
the 7th character (“g”) is on bits DD (15:8) of the fourth word,
the 8th character (“h”) is on bits DD (7:0) of the fourth word,
the 9th character (“t”) is on bits DD (15:8) of the fifth word,
the 10th character (“space”) is on bits DD (7:0) of the fifth word,
Page 111
T13/1321D revision 3
Table 22 − IDENTIFY PACKET DEVICE information (continued)
Word
0
F/V
F
R
F
F
F
R
V
F
1
2
3-9
10-19
20-22
23-26
27-46
47-48
49
50
51-52
53
R
V
R
F
R
F
F
R
F
F
F
F
F
F
F
F
X
R
F
R
F
F
V
54-62
R
General configuration bit-significant information:
15-14
10 = ATAPI device
11 = Reserved
13
Reserved
12-8
Field indicates command packet set used by device
7
1 = removable media device
6-5
00 = Device shall set DRQ to one within 3 ms of receiving PACKET command.
01 = Obsolete.
10 = Device shall set DRQ to one within 50 µs of receiving PACKET command.
11 = Reserved
4-3
Reserved
2
Incomplete response
1-0
00 = 12 byte command packet
01 = 16 byte command packet
1x = Reserved
Reserved
Unique configuration
Reserved
Serial number (20 ASCII characters)
Reserved
Firmware revision (8 ASCII characters)
Model number (40 ASCII characters)
Reserved
Capabilities
15
1 = interleaved DMA supported
14
1 = command queuing supported
13
1 = overlap operation supported
12
1 = ATA software reset required (Obsolete)
11
1 = IORDY supported
10
1 = IORDY may be disabled
9
1 = LBA supported
8
1 = DMA supported
7-0
Vendor specific
Reserved
Obsolete
15-3
Reserved
2
1 = the fields reported in word 88 are valid
0 = the fields reported in word 88 are not valid
1
1 = the fields reported in words 64-70 are valid
0 = the fields reported in words 64-70 are not valid
0
1 = the fields reported in words 54-58 are valid
0 = the fields reported in words 54-58 are not valid
Reserved
(continued)
Page 112
T13/1321D revision 3
Table 22 − IDENTIFY PACKET DEVICE information (continued)
Word
63
F/V
R
V
V
V
64
R
F
F
F
R
F
65
F
66
F
67
F
68
69-70
71
72
73-74
75
F
R
F
F
R
F
76-79
80
R
F
81
F
15-11
10
Reserved
1 = Multiword DMA mode 2 is selected
0 = Multiword DMA mode 2 is not selected
9
1 = Multiword DMA mode 1 is selected
0 = Multiword DMA mode 1 is not selected
8
1 = Multiword DMA mode 0 is selected
0 = Multiword DMA mode 0 is not selected
7-3
Reserved
2
1 = Multiword DMA mode 2 and below are supported
1
1 = Multiword DMA mode 1 and below are supported
0
1 = Multiword DMA mode 0 is supported Multiword DMA mode selected
15-8
Reserved
7-0
Advanced PIO transfer modes supported
Minimum Multiword DMA transfer cycle time per word
15-0
Cycle time in nanoseconds
Manufacturer’s recommended Multiword DMA transfer cycle time
15-0
Cycle time in nanoseconds
Minimum PIO transfer cycle time without flow control
15-0
Cycle time in nanoseconds
Minimum PIO transfer cycle time with IORDY flow control
15-0
Cycle time in nanoseconds
Reserved (for future command overlap and queuing)
Typical time in ns from receipt of PACKET command to bus release.
Typical time in ns from receipt of SERVICE command to BSY cleared to zero
Reserved
Queue depth
15-5
Reserved
4-0
Maximum queue depth supported - 1
Reserved
Major version number
0000h or FFFFh = device does not report version
15
Reserved
14
Reserved for ATA/ATAPI-14
13
Reserved for ATA/ATAPI-13
12
Reserved for ATA/ATAPI-12
11
Reserved for ATA/ATAPI-11
10
Reserved for ATA/ATAPI-10
9
Reserved for ATA/ATAPI-9
8
Reserved for ATA/ATAPI-8
7
Reserved for ATA/ATAPI-7
6
Reserved for ATA/ATAPI-6
5
1 = supports ATA/ATAPI-5
4
1 = supports ATA/ATAPI-4
3
1 = supports ATA-3
2
1 = supports ATA-2
1
Obsolete
0
Reserved
Minor version number
0000h or FFFFh=device does not report version
0001h-FFFEh=see 8.12.44
(continued)
Page 113
T13/1321D revision 3
Table 22 − IDENTIFY PACKET DEVICE information (continued)
Word
82
F/V
F
83
F
84
F
85
V
Command set supported. If words 82 and 83 = 0000h or FFFFh command set notification not
supported.
15
Obsolete
14
1 = NOP command supported
13
1 = READ BUFFER command supported
12
1 = WRITE BUFFER command supported
11
Obsolete
10
1 = Host Protected Area feature set supported
9
1 = DEVICE RESET command supported
8
1 = SERVICE interrupt supported
7
1 = release interrupt supported
6
1 = look-ahead supported
5
1 = write cache supported
4
Shall be set to one
3
1 = supports Power Management feature set
2
1 = supports Removable Media feature set
1
1 = supports Security Mode feature set
0
1 = supports SMART feature set
Command sets supported. If words 82 and 83 = 0000h or FFFFh command set notification not
supported.
15
Shall be cleared to zero
14
Shall be set to one
13-9
Reserved
8
1 = SET MAX security extension supported
7
Reserved for project 1407DT Address Offset Reserved Area Boot
6
1 = SET FEATURES subcommand required to spinup after power-up
5
1 = Power-Up In Standby feature set supported
4
1 = Removable Media Status Notification feature set supported
3-1
Reserved
0
1 = DOWNLOAD MICROCODE command supported
Command set/feature supported extension. If words 82, 83, and 84 = 0000h or FFFFh
command set notification extension is not supported.
15
Shall be cleared to zero
14
Shall be set to one
13-0
Reserved
Command set/feature enabled. If words 85, 86, and 87 = 0000h or FFFFh command set
enabled notification is not supported.
15
Obsolete
14
1 = NOP command enabled
13
1 = READ BUFFER command enabled
12
1 = WRITE BUFFER command enabled
11
Obsolete
10
1 = Host Protected Area feature set enabled
9
1 = DEVICE RESET command enabled
8
1 = SERVICE interrupt enabled
7
1 = release interrupt enabled
6
1 = look-ahead enabled
5
1 = write cache enabled
4
Shall be set to one
3
1 = Power Management feature set enabled
2
1 = Removable Media feature set enabled
1
1 = Security Mode feature set enabled
0
1 = SMART feature set enabled
(continued)
Page 114
T13/1321D revision 3
Table 22 − IDENTIFY PACKET DEVICE information (continued)
Word
86
F/V
V
87
V
88
R
V
V
V
V
V
89-92
R
F
F
F
F
F
R
Command set/feature enabled. If words 85, 86, and 87 = 0000h or FFFFh command set
enabled notification is not supported.
15-9
Reserved
8
1 = SET MAX security extension enabled by a SET MAX SET PASSWORD
7
Reserved for project 1407DT Address Offset Reserved Area Boot
6
1 = SET FEATURES subcommand required to spinup after power-up
5
1 = Power-Up In Standby feature set enabled
4
1 = Removable Media Status Notification feature set enabled via the SET FEATURES
command.
3-1
Reserved
0
1 = DOWNLOAD MICROCODE command enabled
Command set/feature default. If words 85, 86, and 87 = 0000h or FFFFh command set default
notification is not supported.
15
Shall be cleared to zero
14
Shall be set to one
13-0
Reserved
15-13
Reserved
12
1 = Ultra DMA mode 4 is selected
0 = Ultra DMA mode 4 is not selected
11
1 = Ultra DMA mode 3 is selected
0 = Ultra DMA mode 3 is not selected
10
1 = Ultra DMA mode 2 is selected
0 = Ultra DMA mode 2 is not selected
9
1 = Ultra DMA mode 1 is selected
0 = Ultra DMA mode 1 is not selected
8
1 = Ultra DMA mode 0 is selected
0 = Ultra DMA mode 0 is not selected
7-5
Reserved
4
1 = Ultra DMA mode 4 and below are supported
3
1 = Ultra DMA mode 3 and below are supported
2
1 = Ultra DMA mode 2 and below are supported
1
1 = Ultra DMA mode 1 and below are supported
0
1 = Ultra DMA mode 0 is supported
Reserved
(continued)
Page 115
T13/1321D revision 3
Table 22 − IDENTIFY PACKET DEVICE information (continued)
Word
93
F/V
V
94-125
126
127
R
F
F
Hardware reset result. The contents of bits 12-0 of this word shall change only during the
execution of a hardware reset.
15
Shall be cleared to zero.
14
Shall be set to one.
13
1 = device detected CBLID- above ViH
0 = device detected CBLID- below ViL
12-8
Device 1 hardware reset result. Device 0 shall clear these bits to zero. Device 1 shall
set these bits as follows:
12
Reserved.
11
0 = Device 1 did not assert PDIAG-.
1 = Device 1 asserted PDIAG-.
10-9
These bits indicate how Device 1 determined the device number:
00 = Reserved.
01 = a jumper was used.
10 = the CSEL signal was used.
11 = some other method was used or the method is unknown.
8
Shall be set to one.
7-0
Device 0 hardware reset result. Device 1 shall clear these bits to zero. Device 0 shall
set these bits as follows:
7
Reserved.
6
0 = Device 0 does not respond when Device 1 is selected.
1 = Device 0 responds when Device 1 is selected.
5
0 = Device 0 did not detect the assertion of DASP-.
1 = Device 0 detected the assertion of DASP-.
4
0 = Device 0 did not detect the assertion of PDIAG-.
1 = Device 0 detected the assertion of PDIAG-.
3
0 = Device 0 failed diagnostics.
1 = Device 0 passed diagnostics.
2-1
These bits indicate how Device 0 determined the device number:
00 = Reserved.
01 = a jumper was used.
10 = the CSEL signal was used.
11 = some other method was used or the method is unknown.
0
Shall be set to one.
Reserved
ATAPI byte count = 0 behavior
Removable Media Status Notification feature set support
15-2
Reserved
1-0
00 = Removable Media Status Notification feature set not supported
01 = Removable Media Status Notification feature set supported
10 = Reserved
11 = Reserved
(continued)
Page 116
T13/1321D revision 3
Table 22 − IDENTIFY PACKET DEVICE information (concluded)
Word
128
F/V
V
129-159
160-175
176-254
255
X
R
R
F/V
Security status
15-9
Reserved
8
Security level 0 = High, 1 = Maximum
7-6
Reserved
5
1 = Enhanced security erase supported
4
1 = Security count expired
3
1 = Security frozen
2
1 = Security locked
1
1 = Security enabled
0
1 = Security supported
Vendor specific
Reserved for assignment by the CompactFlash Association
Reserved
Integrity word
15-8
Checksum
7-0
Signature
Key:
F = the content of the word is fixed and does not change. For removable media devices, these values may
change when media is removed or changed.
V = the contents of the word is variable and may change depending on the state of the device or the
commands executed by the device.
X = the content of the word is vendor specific and may be fixed or variable.
R = the content of the word is reserved and shall be zero.
8.13.9 Word 0: General configuration
Bits 15 and 14 of word 0 indicate the type of device. If bit 15 is cleared to zero the device does not implement
the PACKET Command feature set. If bit 15 is set to one and bit 14 is cleared to zero, the device implements
the PACKET Command feature set. The value bit 15 and bit 14 both set to one is reserved.
Bits 12 through 8 of word 0 indicate the command packet set implemented by the device. This value follows the
peripheral device type value as defined in SCSI Primary Commands - 2 (SPC-2) T10/1236D.
Value
00h
01h
02h
03h
04h
05h
06h
07h
08h
09h
0A-0Bh
0Ch
0Dh
0Eh
0Fh
10-1Eh
1Fh
Description
Direct-access device
Sequential-access device
Printer device
Processor device
Write-once device
CD-ROM device
Scanner device
Optical memory device
Medium changer device
Communications device
Reserved for ACS IT8 (Graphic arts pre-press devices)
Array controller device
Enclosure services device
Reduced block command devices
Optical card reader/writer device
Reserved
Unknown or no device type
Page 117
T13/1321D revision 3
Bit 7 if set to one indicates that the device has removable media.
Bits 6 and 5 of word 0 indicates the DRQ response time when a PACKET command is received. A value of 00b
indicates a maximum time of 3 ms from receipt of PACKET to the setting of DRQ to one. A value of 10b
indicates a maximum time of 50 µs from the receipt of PACKET to the setting of DRQ to one. The value 11b is
reserved.
If bit 2 is set to one it indicates that the content of the IDENTIFY DEVICE response is incomplete. This will
occur if the device supports the Power-up in Standby feature set and required data is contained on the device
media. In this case the content of at least words 0 and 2 shall be valid.
Bits 1 and 0 of word 0 indicate the packet size the device supports. A value of 00b indicates that a 12 byte
packet is supported; a value of 01b indicates a 16 byte packet. The values 10b and 11b are reserved.
8.13.10 Word 1: Reserved
8.13.11 Word 2: Specific configuration
Word 2 shall have the same content described for word 2 of the IDENTIFY DEVICE command.
8.13.12 Words 3-9: Reserved
8.13.13 Words 10-19: Serial number
The use of these words is optional. If not implemented, the content shall be zeros. If implemented, the content
shall be as described in words 10-19 of the IDENTIFY DEVICE command (see 8.12).
8.13.14 Words 20-22: Reserved
8.13.15 Words 23-26: Firmware revision
Words 23 through 26 shall have the content described for words 23 through 26 of the IDENTIFY DEVICE
command.
8.13.16 Words 27-46: Model number
Words 27 through 46 shall have the content described for words 27 through 46 of the IDENTIFY DEVICE
command.
8.13.17 Words 47-48: Reserved
8.13.18 Word 49: Capabilities
Bit 15 of word 49 is used to indicated that the device supports interleaved DMA data transfer for overlapped
DMA commands.
Bit 14 of word 49 is used to indicated that the device supports command queuing for overlapped commands. If
bit 14 is set to one, bit 13 shall be set to one.
Bit 13 of word 49 is used to indicated that the device supports command overlap operation.
Bit 12 of word 49 indicates that the device requires a software reset to reset the device when BSY is set to one.
Some devices produced before this standard are unable to process a DEVICE RESET when the BSY bit is set
to one. The use of this bit is obsolete.
Page 118
T13/1321D revision 3
Bit 11 of word 49 is used to determine whether a device supports IORDY. If this bit is set to one, then the
device supports IORDY operation. If this bit is zero, the device may support IORDY. This ensures backward
compatibility. If a device supports PIO mode 3 or higher, then this bit shall be set to one.
Bit 10 of word 49 is used to indicate a device’s ability to enable or disable the use of IORDY. If this bit is set to
one, then the device supports the disabling of IORDY. Disabling and enabling of IORDY is accomplished using
the SET FEATURES command.
Bit 9 of word 49 indicates that an LBA translation is supported
Bits 8 of word 49 indicates that DMA is supported.
8.13.19 Word 50: Reserved
8.13.20 Word 51: Obsolete
8.13.21 Word 52: Reserved
8.13.22 Word 53: Field validity
Word 53 shall have the content described for word 53 of the IDENTIFY DEVICE command.
8.13.23 Words 54-62: Reserved
8.13.24 Word 63: Multiword DMA transfer
Word 63 shall have the content described for word 63 of the IDENTIFY DEVICE command.
8.13.25 Word 64: PIO transfer mode supported
Word 64 shall have the content described for word 64 of the IDENTIFY DEVICE command.
8.13.26 Word 65: Minimum multiword DMA transfer cycle time per word
Word 65 shall have the content described for word 65 of the IDENTIFY DEVICE command.
8.13.27 Word 66: Device recommended multiword DMA cycle time
Word 66 shall have the content described for word 66 of the IDENTIFY DEVICE command.
8.13.28 Word 67: Minimum PIO transfer cycle time without flow control
Word 67 shall have the content described for word 67 of the IDENTIFY DEVICE command.
8.13.29 Word 68: Minimum PIO transfer cycle time with IORDY
Word 68 shall have the content described for word 68 of the IDENTIFY DEVICE command.
8.13.30 Word 69-70: Reserved
8.13.31 Word 71: PACKET to bus release time
Word 71 shall contain the time (for 99.7 % of the occurances) in microseconds from the receipt of a PACKET
command until the device performs a bus release.
Page 119
T13/1321D revision 3
8.13.32 Word 72: SERVICE to bus release time
Word 72 shall contain the time (for 99.7% of the occurances) in microseconds from the receipt of a SERVICE
command until the device performs a bus release.
8.13.33 Word 73-74: Reserved
8.13.34 Word 75: Queue depth
Bits 4 through 0 of word 75 shall have the content described for word 75 of the IDENTIFY DEVICE command.
8.13.35 Words 76-79: Reserved
8.13.36 Word 80: Major revision number
Word 80 shall have the content described for word 80 of the IDENTIFY DEVICE command.
8.13.37 Word 81: Minor revision number
Word 81 shall have the content described for word 81 of the IDENTIFY DEVICE command.
8.13.38 Words 82-84: Features/command sets supported
Words 82, 83, and 84 shall have the content described for words 82, 83, and 84 of the IDENTIFY DEVICE
command except that bit 4 of word 82 shall be set to one.
8.13.39 Words 85-87: Features/command sets enabled
Words 85, 86, and 87 shall have the content described for words 85, 86, and 87 of the IDENTIFY DEVICE
command except that bit 4 of word 85 shall be set to one.
8.13.40 Word 88:Ultra DMA modes
Word 88 shall have the content described for word 88 of the IDENTIFY DEVICE command.
8.13.41 Word 89: Time required for Security erase unit completion
Word 89 shall have the content described for word 89 of the IDENTIFY DEVICE command.
8.13.42 Word 90: Time required for Enhanced security erase unit completion
Word 90 shall have the content described for word 90 of the IDENTIFY DEVICE command.
8.13.43 Word 91-92: Reserved
8.13.44 Word 93: Hardware reset results
Word 93 shall have the content described for word 93 of the IDENTIFY DEVICE command.
8.13.45 Word 94-125: Reserved
8.13.46 Word 126: ATAPI byte count=0 behavior
If the contents of word 126 are 0000h and the byte count limit is set to a non-zero value, the device shall return
command aborted.
Page 120
T13/1321D revision 3
If the contents of word 126 are non-zero and the byte count limit is set to zero, the device shall use the
contents of word 126 as the actual byte count limit for the current command and shall not abort.
The device may be reconfigured to report a new value. However, after the device is reconfigured, the content of
word 126 reported shall not change until after the next hardware reset or power-on reset event.
8.13.47 Word 127: Removable Media Status Notification feature set support
Word 127 shall have the content described for word 127 of the IDENTIFY DEVICE command.
8.13.48 Word 128: Security status
Word 128 shall have the content described for word 128 of the IDENTIFY DEVICE command.
8.13.49 Words 129-160: Reserved
8.13.50 Words 161-175: Reserved for assignment by the CompactFlash  Association
8.13.51 Words 176-254: Reserved
8.13.52 Word 255: Integrity Word
Word 255 shall have the content described for word 255 of the IDENTIFY DEVICE command.
8.14 IDLE
8.14.1 Command code
E3h
8.14.2 Feature set
Power Management feature set.
− Power Management feature set is mandatory when power management is not implemented by a
PACKET power management feature set.
− This command is mandatory when the Power Management feature set is implemented and the
PACKET Command feature set is not implemented.
8.14.3 Protocol
Non-data command (see 9.4).
8.14.4 Inputs
Values other than zero in the Sector Count register when the IDLE command is issued shall determine the time
period programmed into the Standby timer. Table 23 defines these values.
Page 121
T13/1321D revision 3
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
na
Timer period value
na
na
na
obs DEV na
na
E3h
1
0
na
na
Device/Head register DEV shall indicate the selected device.
Table 23 − Automatic Standby timer periods
Sector Count register
Corresponding timeout period
contents
0
(00h)
Timeout disabled
1-240
(01h-F0h)
(value ∗ 5) s
241-251
(F1h-FBh)
((value - 240) ∗30) min
252
(FCh)
21 min
253
(FDh)
Period between 8 and 12 hrs
254
(FEh)
Reserved
255
(FFh)
21 min 15 s
NOTE − Times are approximate.
8.14.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.14.6 Error outputs
The device shall return command aborted if the device does not support the Power Management feature set.
Page 122
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register ABRT shall be set to one if Power Management feature set is not supported. ABRT may be set to one
if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.14.7 Prerequisites
DRDY set to one.
8.14.8 Description
The IDLE command allows the host to place the device in the Idle mode and also set the Standby timer. INTRQ
may be asserted even though the device may not have fully transitioned to Idle mode.
If the Sector Count register is non-zero then the Standby timer shall be enabled. The value in the Sector Count
register shall be used to determine the time programmed into the Standby timer (see 6.11). If the Sector Count
register is zero then the Standby timer is disabled.
8.15 IDLE IMMEDIATE
8.15.1 Command code
E1h
8.15.2 Feature set
Power Management feature set.
−
−
Power Management feature set is mandatory when power management is not implemented by a
PACKET power management feature set.
This command is mandatory when the Power Management feature set is implemented.
8.15.3 Protocol
Non-data command (see 9.4).
Page 123
T13/1321D revision 3
8.15.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
DEV
obs
E1h
Device/Head register DEV shall indicate the selected device.
8.15.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.15.6 Error outputs
The device shall return command aborted if the device does not support the Power Management feature set.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register ABRT shall be set to one if Power Management feature set is not supported. ABRT may be set to one
if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
Page 124
T13/1321D revision 3
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.15.7 Prerequisites
DRDY set to one.
8.15.8 Description
The IDLE IMMEDIATE command allows the host to immediately place the device in the Idle mode. INTRQ may
be asserted even though the device may not have fully transitioned to Idle mode (see 6.11).
8.16 INITIALIZE DEVICE PARAMETERS
8.16.1 Command code
91h
8.16.2 Feature set
General feature set
−
−
−
Mandatory for devices not implementing the PACKET Command feature set if a CHS translation is
supported.
Not mandatory for devices not implementing the PACKET command feature set if the device
capacity is greater then 8 Gbytes and only LBA translation is supported.
Use prohibited for devices implementing the PACKET Command feature set.
8.16.3 Protocol
Non-data (see 9.4).
8.16.4 Inputs
The Sector Count register specifies the number of logical sectors per logical track, and the Device/Head
register specifies the maximum head number.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
obs
6
5
4
3
2
1
na
Logical sectors per logical track
na
na
na
na
obs DEV
Max head
91h
0
Device/Head register DEV shall indicate the selected device.
Page 125
T13/1321D revision 3
8.16.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
3
na
2
ABRT
1
na
0
na
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
na
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.16.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
na
na
na
na
obs
BSY
na
na
obs
DF
DEV
na
Error register ABRT shall be set to one if the device does not support the requested CHS translation. ABRT may be
set to one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.16.7 Prerequisites
This command shall be accepted regardless of the state of DRDY.
8.16.8 Description
This command enables the host to set the number of logical sectors per track and the number of logical heads
minus 1, per logical cylinder for the current CHS translation mode.
If the capacity of the device is less than 16,514,064 sectors, a device shall support the CHS translation
described in words 1, 3, and, 6 of the IDENTIFY DEVICE information. Support of other CHS translations is
optional.
Page 126
T13/1321D revision 3
If the host requests a CHS translation that is not supported by the device, the device shall return command
aborted. The device shall also clear bit 0 of word 53 in the IDENTIFY DEVICE data to zero, and the content of
words 54, 55, 56, and, (58:57) may be zero until a supported translation is requested by the host.
If the requested CHS translation is not supported, the device shall fail all media access commands with an ID
Not Found error until a valid CHS translation is established.
After a successful INITIALIZE DEVICE PARAMETERS command the content of all IDENTIFY DEVICE words
shall comply with 6.2.1 in addition to the following:
1) The content of words 1, 3, 6, and (61:60) shall be unchanged.
2) The content of word 55 shall equal (Max head value requested by the host + 1).
3) The content of word 56 shall equal (Logical sectors per logical track value requested by the host).
4) If the content of word (61:60) is less than or equal to 16,514,064, then word 54 shall equal the whole
number result of [[(content of words (61:60)) ÷ [(new content of word 55 as determined by the successful
INITIALIZE DEVICE PARAMETERS command) ∗ (new content of word 56 as determined by the successful
INITIALIZE DEVICE PARAMETERS command)]], or 65,535 whichever is less.
5) If the content of word (61:60) is greater than 16,514,064, then word 54 shall equal the whole number result
of [[(16,514,064) ÷ [(new content of word 55 as determined by the successful INITIALIZE DEVICE
PARAMETERS command) ∗ (new content of word 56 as determined by the successful INITIALIZE DEVICE
PARAMETERS command)]] or 65,535 whichever is less.
6) Words (58:57) shall equal [(new content of word 54) ∗ (new content of word 55) ∗ (new content of word 56)].
8.17 MEDIA EJECT
8.17.1 Command code
EDh
8.17.2 Feature set
Removable Media Status Notification feature set
−
−
Mandatory for devices not implementing the PACKET command feature set and implementing the
Removable Media Status Notification feature set.
Prohibited for devices implementing the PACKET command feature set.
Removable Media feature set
− Mandatory for devices not implementing the PACKET command feature set and implementing the
Removable Media feature set.
− Prohibited for devices implementing the PACKET command feature set.
8.17.3 Protocol
Non-data command (see 9.4).
Page 127
T13/1321D revision 3
8.17.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
EDh
na
na
na
na
na
na
na
na
obs
na
obs
DEV
Device/Head register DEV shall indicate the selected device.
8.17.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.17.6 Error outputs
If the device does not support this command, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
NM
0
obs
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register ABRT shall be set to one if device does not support this command. ABRT may be set to one if the
device is not able to complete the action requested by the command.
NM (No Media) shall be set to one if no media is present in the device.
Device/Head register DEV shall indicate the selected device.
Status register Page 128
T13/1321D revision 3
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.17.7 Prerequisites
DRDY set to one.
8.17.8 Description
This command causes any pending operations to complete, spins down the device if needed, unlocks the
media if locked, and ejects the media. The device keeps track of only one level of media lock.
8.18 MEDIA LOCK
8.18.1 Command code
DEh
8.18.2 Feature set
Removable Media Status Notification feature set
−
−
Optional for devices not implementing the PACKET command feature set and implementing the
Removable Media Status Notification feature set.
Prohibited for device implementing the PACKET command feature set.
Removable Media feature set
− Mandatory for devices not implementing the PACKET command feature set and implementing the
Removable Media feature set.
− Prohibited for devices implementing the PACKET command feature set.
8.18.3 Protocol
Non-data command (see 9.4).
8.18.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
DEh
na
na
na
na
na
na
na
na
obs
na
obs
DEV
Device/Head register DEV shall indicate the selected device.
Page 129
T13/1321D revision 3
8.18.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.18.6 Error outputs
If the device does not support this command, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
MCR
2
ABRT
1
NM
0
obs
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register ABRT shall be set to one if device does not support this command. ABRT may be set to one if the
device is not able to complete the action requested by the command.
NM (No Media) shall be set to one if no media is present in the device.
MCR (Media Change Request) shall be set to one if the device is locked and a media change request
has been detected by the device.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.18.7 Prerequisites
DRDY set to one.
Page 130
T13/1321D revision 3
8.18.8 Description
This command shall be used to lock the media, if Media Status Notification is disabled. If Media Status
Notification is enabled, this command shall return good status (no ERR bit in the Status register) and perform
no action.
If the media is unlocked and media is present, the media shall be set to the LOCKED state and no Error
register bit shall be set to one. The device keeps track of only one level of media lock. Subsequent MEDIA
LOCK commands, while the media is in the LOCKED state, do not set additional levels of media locks.
If the media is locked, the status returned shall indicate whether a media change request has been detected by
the device. If a media change request has been detected, the MCR bit in the Error register and the ERR bit in
the Status register shall be set to one.
When media is in the LOCKED state, the device shall respond to the media change request button, by setting
the MCR bit in the Error register and the ERR bit in the Status register to one, until the media LOCKED
condition is cleared.
8.19 MEDIA UNLOCK
8.19.1 Command code
DFh
8.19.2 Feature set
Removable Media Status Notification feature set
−
−
Optional for devices not implementing the PACKET command feature set and implementing the
Removable Media Status Notification feature set.
Prohibited for devices implementing the PACKET command feature set.
Removable Media feature set
− Mandatory for devices not implementing the PACKET command feature set and implementing the
Removable Media feature set.
− Prohibited for devices implementing the PACKET command feature set.
8.19.3 Protocol
Non-data command (see 9.4).
8.19.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
DFh
na
na
na
na
na
na
na
na
obs
na
obs
DEV
Device/Head register DEV shall indicate the selected device.
Page 131
T13/1321D revision 3
8.19.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.19.6 Error outputs
If the device does not support this command, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
NM
0
obs
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register ABRT shall be set to one if device does not support this command. ABRT may be set to one if the
device is not able to complete the action requested by the command.
NM (No Media) shall be set to one if no media is present in the device.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.19.7 Prerequisites
DRDY set to one.
Page 132
T13/1321D revision 3
8.19.8 Description
This command can be used to unlock the device, if Media Status Notification is disabled. If Media Status
Notification is enabled, this command will return good status (no ERR bit in the Status register) and perform no
action.
If the media is present, the media shall be set to the UNLOCKED state and no Error register bit shall be set to
one. The device keeps track of only one level of media lock. A single MEDIA UNLOCK command unlocks the
media.
If a media change request has been detected by the device prior to the issuance of this command, the media
shall be ejected at MEDIA UNLOCK command completion.
8.20 NOP
8.20.1 Command code
00h
8.20.2 Feature set
General feature set
−
−
−
Optional for devices not implementing the PACKET Command feature set.
Mandatory for devices implementing the PACKET Command feature set.
Mandatory for devices implementing the Overlapped feature set.
8.20.3 Protocol
Non-data (see 9.4).
8.20.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
Subcommand code
na
na
na
na
obs DEV na
na
00h
1
0
na
na
Features register Subcommand code
00h
Description
NOP
01h
NOP Auto Poll
02h-FFh
Reserved
Action
Return command aborted and abort any
outstanding queued commands.
Return command aborted and do not abort any
outstanding queued commands.
Return command aborted and do not abort any
outstanding queued commands.
Device/Head register DEV shall indicate the selected device.
Page 133
T13/1321D revision 3
8.20.5 Normal outputs
This command always fails with an error.
8.20.6 Error outputs
The Command Block registers, other than the Error and Status registers, are not changed by this command.
This command always fails with the device returning command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
BSY
DRDY
DF
4
3
na
na
Initial value
Initial value
Initial value
Initial value
Initial value
na
DRQ
2
ABRT
1
na
0
na
na
na
ERR
Error register ABRT shall be set to one.
Sector Count, Sector Number, Cylinder Low, Cylinder High, Device/Head value set by host is not changed.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one.
8.20.7 Prerequisites
DRDY set to one.
8.20.8 Description
The device shall respond with command aborted. For devices implementing the Overlapped feature set,
subcommand code 00h in the Features register shall abort any outstanding queue. Subcommand codes 01h
through FFh in the Features register shall not affect the status of any outstanding queue.
8.21 PACKET
8.21.1 Command code
A0h
8.21.2 Feature set
PACKET Command feature set
−
−
Use prohibited for devices not implementing the PACKET Command feature set.
Mandatory for devices implementing the PACKET Command feature set.
Page 134
T13/1321D revision 3
8.21.3 Protocol
Packet (see 9.8).
8.21.4 Inputs
Register
Features
Sector Count
Sector Number
Byte count low
(Cylinder Low)
Byte count high
(Cylinder High)
Device/Head
Command
7
na
6
na
5
na
Tag
4
na
3
na
2
na
1
0
OVL
DMA
na
na
Byte count limit (7-0)
Byte count limit (15-8)
obs
na
obs
DEV
na
A0h
na
na
na
Features register OVL - This bit is set to one to inform the device that the PACKET command is to be overlapped.
DMA - This bit is set to one to inform the device that the data transfer (not the command packet
transfer) associated with this command is via DMA or Ultra DMA mode.
Sector Count register Tag - If the device supports command queuing, this field contains the command Tag for the command
being delivered. A Tag may have any value between 0 and 31 regardless of the queue depth
supported. If queuing is not supported, this field is not applicable.
Byte count low and Byte count high registers These registers are written by the host with the maximum byte count that is to be transferred in any
single DRQ assertion for PIO transfers. The byte count does not apply to the command PACKET
transfer. If the PACKET command does not transfer data, the byte count is ignored.
If the PACKET command results in a data transfer:
1) the host should not set the byte count limit to zero. If the host sets the byte count limit to
zero, the contents of IDENTIFY PACKET DEVICE word 126 determines the expected behavior;
2) the value set into the byte count limit shall be even if the total requested data transfer length is
greater than the byte count limit;
3) the value set into the byte count limit may be odd if the total requested data transfer length is
equal to or less than the byte count limit;
4) the value FFFFh is interpreted by the device as though the value were FFFEh.
Device/Head register DEV shall indicate the selected device.
8.21.5 Normal outputs
8.21.5.1 Awaiting command
When the device is ready to accept the command packet from the host the register content shall be as shown
below.
Page 135
T13/1321D revision 3
Register
Error
Interrupt reason
(Sector Count)
Sector Number
Byte count low
(Cylinder Low)
Byte count high
(Cylinder High)
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
na
na
na
na
CHK
na
Tag
na
Byte count (7:0)
Byte count (15:8)
obs
BSY
na
DRDY
obs
DMRD
DEV
SERV
na
DRQ
Byte count High/Low - shall reflect the value set by the host when the command was issued.
Interrupt reason register Tag - If the device supports command queuing and overlap is enabled, this field contains the command
Tag for the command. A Tag value may be any value between 0 and 31 regardless of the queue
depth supported. If the device does not support command queuing or overlap is disabled, this field
is not applicable.
REL - Shall be cleared to zero.
I/O - Shall be cleared to zero indicating transfer to the device.
C/D - Shall be set to one indicating the transfer of a command packet.
Device/Head register DEV shall indicate the selected device.
Status register BSY - Shall be cleared to zero.
DRDY - na.
DMRD (DMA ready) - Shall be cleared to zero.
SERV (Service) - Shall be set to one if another command is ready to be serviced. If overlap is not
supported, this bit is command specific.
DRQ - Shall be set to one.
CHK - Shall be cleared to zero.
8.21.5.2 Data transmission
If overlap is not supported or not indicated by the command, data transfer shall occur after the receipt of the
command packet. If overlap is supported and the command indicates that the command may be overlapped,
data transfer may occur after receipt of the command packet or may occur after the receipt of a SERVICE
command. When the device is ready to transfer data requested by a data transfer command, the device sets
the following register content to initiate the data transfer.
Register
Error
Interrupt reason
(Sector Count)
Sector Number
Byte count low
(Cylinder Low)
Byte count high
(Cylinder High)
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
na
na
na
na
CHK
na
Tag
na
Byte count (7:0)
Byte count (15:8)
obs
BSY
na
DRDY
obs
DMRD
DEV
SERV
na
DRQ
Byte count High/Low - If the transfer is to be in PIO mode, the byte count of the data to be transferred for this
DRQ assertion shall be presented.
Page 136
T13/1321D revision 3
Valid byte count values are as follows:
1)
2)
3)
4)
5)
the byte count shall be less than or equal to the byte count limit value from the host;
the byte count shall not be zero;
the byte count shall be less than or equal to FFFEh;
the byte count shall be even except for the last transfer of a command;
if the byte count is odd, the last valid byte transferred is on DD[7:0] and the data on DD[15:8]
is a pad byte of undefined value;
6) if the last transfer of a command has a pad byte, the byte count shall be odd.
Interrupt reason register Tag - If the device supports command queuing and overlap is enabled, this field contains the command
Tag for the command. A Tag value may be any value between 0 and 31 regardless of the queue
depth supported. If the device does not support command queuing or overlap is disabled, this field
is not applicable.
REL - Shall be cleared to zero.
I/O - Shall be cleared to zero if the transfer is to the device. Shall be set to one if the transfer is to the
host.
C/D - Shall be cleared to zero indicating the transfer of data.
Device/Head register DEV shall indicate the selected device.
Status register BSY - Shall be cleared to zero.
DRDY - na.
DMRD (DMA ready) - Shall be set to one if the transfer is to be a DMA or Ultra DMA transfer and the
device supports overlap DMA.
SERV (Service) - Shall be set to one if another command is ready to be serviced. If overlap is not
supported, this bit is command specific.
DRQ - Shall be set to one.
CHK - Shall be cleared to zero.
8.21.5.3 Bus release (overlap feature set only)
After receiving the command packet, the device sets BSY to one and clears DRQ to zero. If the command
packet requires a data transfer, the OVL bit is set to one, and the device is not prepared to immediately transfer
data, the device may perform a bus release by placing the following register content. If the command packet
requires a data transfer, the OVL bit is set to one, and the Release interrupt is enabled, the device shall perform
a bus release by setting the register content as follows.
Register
Error
Interrupt reason
(Sector Count)
Sector Number
Byte count low
(Cylinder Low)
Byte count high
(Cylinder High)
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
na
na
na
na
CHK
na
Tag
na
na
na
obs
BSY
na
DRDY
obs
DMRD
DEV
SERV
na
DRQ
Byte count High/Low - na.
Interrupt reason register -
Page 137
T13/1321D revision 3
Tag - If the device supports command queuing and overlap is enabled, this field contains the command
Tag for the command. A Tag value may be any value between 0 and 31 regardless of the queue
depth supported. If the device does not support command queuing or overlap is disabled, this field
is not applicable.
REL - Shall be set to one.
I/O - Shall be cleared to zero.
C/D - Shall be cleared to zero.
Device/Head register DEV shall indicate the selected device.
Status register BSY - Shall be cleared to zero indicating bus release.
DRDY - na.
DMRD (DMA ready) - Shall be cleared to zero.
SERV (Service) - Shall be set to one if another command is ready to be serviced. If overlap is not
supported, this bit is command specific.
DRQ - Shall cleared to zero.
CHK - Shall be cleared to zero.
8.21.5.4 Service request (overlap feature set only)
When the device is ready to transfer data or complete a command after the command has performed a bus
release, the device shall set the SERV bit and not change the state of any other register bit (see 6.9). When
the SERVICE command is received, the device shall set outputs as described in data transfer, successful
command completion, or error outputs depending on the service the device requires.
8.21.5.5
Successful command completion
When the device has command completion without error, the device sets the following register content.
Register
Error
Interrupt reason
(Sector Count)
Sector Number
Byte count low
(Cylinder Low)
Byte count high
(Cylinder High)
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
na
na
na
na
CHK
na
Tag
na
na
na
obs
BSY
na
DRDY
obs
DMRD
DEV
SERV
na
DRQ
Byte count High/Low -na.
Interrupt reason register Tag - If the device supports command queuing and overlap is enabled, this field contains the command
Tag for the command. A Tag value may be any value between 0 and 31 regardless of the queue
depth supported. If the device does not support command queuing or overlap is disabled, this
field is not applicable.
REL - Shall be cleared to zero.
I/O - Shall be set to one.
C/D - Shall be set to one.
Device/Head register DEV shall indicate the selected device.
Status register BSY - Shall be cleared to zero indicating command completion.
DRDY - Shall be set to one.
Page 138
T13/1321D revision 3
DMRD (DMA ready) - na.
SERV (Service) - Shall be set to one if another command is ready to be serviced. If overlap is not
supported, this bit is command specific.
DRQ - Shall be cleared to zero.
CHK - Shall be cleared to zero.
8.21.6 Error outputs
The device shall not terminate the PACKET command with an error before the last byte of the command packet
has been written (see 9.8).
Register
Error
Interrupt reason
(Sector Count)
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
Sense key
Tag
4
3
na
2
ABRT
REL
1
EOM
I/O
0
ILI
C/D
na
DRQ
na
na
na
na
na
CHK
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
SERV
Error register Sense Key is a command packet set specific error indication.
ABRT shall be set to one if the requested command has been command aborted because the
command code or a command parameter is invalid. ABRT may be set to one if the device is
not able to complete the action requested by the command.
EOM - the meaning of this bit is command set specific. See the appropriate command set standard for
the definition of this bit.
ILI - the meaning of this bit is command set specific. See the appropriate command set standard for
the definition of this bit.
Interrupt reason register Tag - If the device supports command queuing and overlap is enabled, this field contains the command
Tag for the command. A Tag value may be any value between 0 and 31 regardless of the queue
depth supported. If the device does not support command queuing or overlap is disabled, this
field is not applicable.
REL - Shall be cleared to zero.
I/O - Shall be set to one.
C/D - Shall be set to one.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
SERV (Service) - Shall be set to one if another command is ready to be serviced. If overlap is not
supported, this bit is command specific.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
CHK shall be set to one if an Error register sense key or code bit is set.
8.21.7 Prerequisites
This command shall be accepted regardless of the state of DRDY.
Page 139
T13/1321D revision 3
8.21.8 Description
The PACKET command is used to transfer a device command via a command packet. If the native form of the
encapsulated command is shorter than the packet size reported in bits 1 and 0 of word 0 of the IDENTIFY
PACKET DEVICE response, the encapsulated command shall begin at byte 0 of the packet. Packet bytes
beyond the end of the encapsulated command are reserved.
If the device supports overlap, the OVL bit is set to one in the Features register and the Release interrupt has
been disabled via the SET FEATURES command, the device may or may not perform a bus release. If the
device is ready for the data transfer, the device may begin the transfer immediately as described in the nonoverlapped protocol (see 9.8). If the data is not ready, the device may perform a bus release and complete the
transfer after the execution of a SERVICE command.
8.22 READ BUFFER
8.22.1 Command code
E4h
8.22.2 Feature set
General feature set
−
−
Optional for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.22.3 Protocol
PIO data-in (see 9.5).
8.22.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
Device/Head register DEV shall indicate the selected device.
Page 140
6
na
obs
DEV
E4h
T13/1321D revision 3
8.22.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
3
na
2
ABRT
1
na
0
na
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.22.6 Error outputs
The device shall return command aborted if the command is not supported.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Error register ABRT shall be set to one if this command is not supported. ABRT may be set to one if the device is
not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.22.7 Prerequisites
DRDY set to one. The command prior to a READ BUFFER command shall be a WRITE BUFFER command.
8.22.8 Description
The READ BUFFER command enables the host to read the current contents of the device’s sector buffer.
Page 141
T13/1321D revision 3
The READ BUFFER and WRITE BUFFER commands shall be synchronized such that sequential WRITE
BUFFER and READ BUFFER commands access the same 512 bytes within the buffer.
8.23 READ DMA
8.23.1 Command code
C8h
8.23.2 Feature set
General feature set
−
−
Mandatory for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.23.3 Protocol
DMA (see 9.7).
8.23.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
C8h
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
Page 142
T13/1321D revision 3
8.23.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.23.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred. The amount of data transferred is indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
ICRC
obs
BSY
6
UNC
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
obs
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
na
DF
na
DRQ
na
ERR
Error register ICRC shall be set to one if an interface CRC error has occurred during an Ultra DMA data transfer. The
content of this bit is not applicable for Multiword DMA transfers.
UNC shall be set to one if data is uncorrectable
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
Page 143
T13/1321D revision 3
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.23.7 Prerequisites
DRDY set to one. The host shall initialize the DMA channel.
8.23.8 Description
The READ DMA command allows the host to read data using the DMA data transfer protocol.
8.24 READ DMA QUEUED
8.24.1 Command code
C7h
8.24.2 Feature set
Overlapped feature set
− Mandatory for devices implementing the Overlapped feature set but not implementing the PACKET
command feature set.
− Use prohibited for devices implementing the PACKET command feature set.
8.24.3 Protocol
DMA QUEUED (see 9.9).
8.24.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
obs
6
LBA
5
4
3
Sector Count
2
1
0
Tag
na
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
C7h
Features number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector count if the device supports command queuing, bits (7:3) contain the Tag for the command being delivered. A
Tag value may be any value between 0 and 31 regardless of the queue depth supported. If queuing is
not supported, this field is not applicable.
Sector number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High Page 144
T13/1321D revision 3
starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
8.24.5 Normal outputs
8.24.5.1 Data transmission
Data transfer may occur after receipt of the command or may occur after the receipt of a SERVICE command.
When the device is ready to transfer data requested by a data transfer command, the device sets the following
register content to initiate the data transfer.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
na
na
na
na
CHK
na
Tag
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
SERV
na
DRQ
Interrupt reason register Tag -This field contains the command Tag for the command. A Tag value may be any value between 0
and 31 regardless of the queue depth supported. If the device does not support command queuing
or overlap is disabled, this field is not applicable.
REL - Shall be cleared to zero.
I/O - Shall be set to one indicating the transfer is to the host.
C/D - Shall be cleared to zero indicating the transfer of data.
Device/Head register DEV shall indicate the selected device.
Status register BSY - Shall be cleared to zero.
DRDY - Shall be set to one.
DF (Device Fault) - Shall be cleared to zero
SERV (Service) - Shall be set to one if another command is ready to be serviced.
DRQ - Shall be set to one.
CHK - Shall be cleared to zero.
8.24.5.2 Release
If the device performs a bus release before transferring data for this command, the register content upon
performing a bus release shall be as shown below.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
ERR
na
Tag
obs
BSY
na
DRDY
obs
DF
na
na
na
DEV
SERV
na
DRQ
na
Page 145
T13/1321D revision 3
Sector Count register Tag - If the device supports command queuing, this field shall contain the Tag of the command being
bus released. If the device does not support command queuing, this field shall be zeros.
REL shall be set to one.
I/O shall be zero.
C/D shall be zero.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating bus release.
DRDY shall be set to one.
SERV (Service) shall be cleared to zero when no other queued command is ready for service. SERV
shall be set to one when another queued command is ready for service. SERV shall be set to
one when the device has prepared this command for service.
DF (Device Fault) shall be cleared to zero
DRQ bit shall be cleared to zero.
ERR bit shall be cleared to zero.
8.24.5.3 Service request
When the device is ready to transfer data or complete a command after the command has performed a bus
release, the device shall set the SERV bit and not change the state of any other register bit (see 6.9). When
the SERVICE command is received, the device shall set outputs as described in data transfer, command
completion, or error outputs depending on the service the device requires.
8.24.5.4 Command completion
When the transfer of all requested data has occurred without error, the register content shall be as shown
below.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
00h
3
Tag
obs
BSY
na
DRDY
obs
DF
na
na
na
DEV
SERV
2
1
0
REL
I/O
C/D
na
ERR
na
DRQ
na
Sector Count register Tag - If the device supports command queuing, this field shall contain the Tag of the completed
command. If the device does not support command queuing, this field shall be zeros.
REL shall be cleared to zero.
I/O shall be set to one.
C/D shall be set to one.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
SERV (Service) shall be cleared to zero when no other queued command is ready for service. SERV
shall be set to one when another queued command is ready for service.
DF (Device Fault) shall be cleared to zero.
DRQ bit shall be cleared to zero.
ERR bit shall be cleared to zero.
Page 146
T13/1321D revision 3
8.24.6 Error outputs
The Sector Count register contains the Tag for this command if the device supports command queuing. The
device shall return command aborted if the command is not supported or if the device has not had overlapped
interrupt enabled. The device shall return command aborted if the device supports command queuing and the
Tag is invalid. An unrecoverable error encountered during the execution of this command results in the
termination of the command and the Command Block registers contain the sector where the first unrecoverable
error occurred. If a queue existed, the unrecoverable error shall cause the queue to abort.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
ICRC
7
6
UNC
obs
BSY
na
DRDY
5
4
3
2
1
0
IDNF
MC
MCR ABRT NM
obs
Tag
REL
I/O
C/D
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
SERV
DRQ
na
na
ERR
Error register ICRC shall be set to one if an interface CRC error has occurred during an Ultra DMA data transfer. The
content of this bit is not applicable for Multiword DMA transfers.
UNC shall be set to one if data is uncorrectable.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if ABRT is not set to one.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Count register Tag - If the device supports command queuing, this field shall contain the Tag of the completed
command. If the device does not support command queuing, this field shall be zeros.
REL shall be cleared to zero.
I/O shall be set to one.
C/D shall be set to one.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
SERV (Service) shall be cleared to zero when no other queued command is ready for service. SERV
shall be set to one when another queued command is ready for service.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
Page 147
T13/1321D revision 3
8.24.7 Prerequisites
DRDY set to one. The host shall initialize the DMA channel.
8.24.8 Description
This command executes in a similar manner to a READ DMA command. The device may perform a bus release
or may execute the data transfer without performing a bus release if the data is ready to transfer.
8.25 READ MULTIPLE
8.25.1 Command code
C4h
8.25.2 Feature set
General feature set
−
−
Mandatory for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.25.3 Protocol
PIO data-in (see 9.5).
8.25.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
C4h
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
Page 148
T13/1321D revision 3
8.25.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.25.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred. The amount of data transferred is indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
UNC
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
obs
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
na
DF
na
DRQ
na
ERR
Error register UNC shall be set to one if data is uncorrectable.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Status register Page 149
T13/1321D revision 3
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.25.7 Prerequisites
DRDY set to one. If bit 8 of IDENTIFY DEVICE word 59 is cleared to zero, a successful SET MULTIPLE MODE
command shall precede a READ MULTIPLE command.
8.25.8 Description
The READ MULTIPLE command performs the same as the READ SECTOR(S) command except that when the
device is ready to transfer data for a block of sectors, the device clears BSY, sets DRDY and DRQ (and sets
DF, ERR, and the bits in the Error register, as required), and enters the interrupt pending state only before the
data transfer for the first sector of the block sectors. The remaining sectors for the block are transferred without
the device asserting INTRQ. In addition, the DRQ qualification of the transfer is required only before the first
sector of a block, not before each sector of the block.
The number of sectors per block is defined by a successful SET MULTIPLE command. If no successful SET
MULTIPLE command has been issued, the block is defined by the device’s default value for number of sectors
per block as defined in bits 0-7 in word 47 in the IDENTIFY DEVICE information.
If bit 8 equals 1 and bits 0-7 are cleared to zero in word 59 in the IDENTIFY DEVICE information, then the block
count of sectors to be transferred without intervening interrupts shall be programmed by the SET MULTIPLE
MODE command before issuing the first READ MULTIPLE or WRITE MULTIPLE command after a power-on or
hardware reset. If bit 8 in word 1 is set to one and bits 0-7 are not cleared to zero in word 59 in the IDENTIFY
DEVICE information, then the block count of sectors to be transferred without intervening interrupts may be
reprogrammed by the SET MULTIPLE MODE command before issuing the next READ MULTIPLE or WRITE
MULTIPLE command.
When the READ MULTIPLE command is issued, the Sector Count register contains the number of sectors (not
the number of blocks) requested.
If the number of requested sectors is not evenly divisible by the block count, as many full blocks as possible
are transferred, followed by a final, partial block transfer. The partial block transfer shall be for n sectors, where
n = remainder (sector count/ block count).
If the READ MULTIPLE command is received when READ MULTIPLE commands are disabled, the READ
MULTIPLE operation shall be rejected with command aborted.
Device errors encountered during READ MULTIPLE commands are posted at the beginning of the block or
partial block transfer, but the DRQ bit is still set to one and the data transfer shall take place, including transfer
of corrupted data, if any. The contents of the Command Block Registers following the transfer of a data block
that had a sector in error are undefined. The host should retry the transfer as individual requests to obtain valid
error information.
Subsequent blocks or partial blocks are transferred only if the error was a correctable data error. All other
errors cause the command to stop after transfer of the block that contained the error.
Page 150
T13/1321D revision 3
8.26 READ NATIVE MAX ADDRESS
8.26.1 Command code
F8h
8.26.2 Feature set
Host Protected Area feature set.
−
−
Mandatory when the Host Protected Area feature set is implemented.
Use prohibited when Removable feature set is implemented.
8.26.3 Protocol
Non-data command (see 9.4).
8.26.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
obs
LBA
obs
DEV
na
F8h
Device/Head If LBA is set to one, the maximum address shall be reported as an LBA value.
If LBA is cleared to zero, the maximum address shall be reported as a CHS value.
DEV shall indicate the selected device.
8.26.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
obs
BSY
6
5
4
3
2
1
0
na
na
Native max address sector number or LBA
Native max address cylinder low or LBA
Native max address cylinder high or LBA
na
obs
DEV
Native max address head or LBA
DRDY
DF
na
DRQ
na
na
ERR
Sector Number maximum native sector number (IDENTIFY DEVICE word 6) or LBA bits (7:0) for native max address on
the device.
Cylinder Low maximum native cylinder number low or LBA bits (15:8) for native max address on the device.
Cylinder High maximum native cylinder number high or LBA bits (23:16) for native max address on device.
Device/Head maximum native head number (IDENTIFY DEVICE word 3 minus one) or LBA bits (27:24) for native
max address on the device.
Page 151
T13/1321D revision 3
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.26.6 Error outputs
If this command is not supported the device shall return command aborted. The device shall return command
aborted if a CHS address is requested and the device does not support a CHS translation.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
na
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
Error register ABRT shall be set to one if this command is not supported. ABRT may be set to one if the device is
not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be set to one if an Error register bit is set to one.
8.26.7 Prerequisites
DRDY set to one.
8.26.8 Description
This command returns the native maximum address. The native maximum address is the highest address
accepted by the device in the factory default condition. The native maximum address is the maximum address
that is valid when using the SET MAX ADDRESS command.
8.27 READ SECTOR(S)
8.27.1 Command code
20h
Page 152
T13/1321D revision 3
8.27.2 Feature set
General feature set
− Mandatory for all devices.
−
− PACKET Command feature set devices (see 8.27.5.2).
8.27.3 Protocol
PIO data-in (see 9.5).
8.27.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
20h
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
8.27.5 Outputs
8.27.5.1 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
Page 153
T13/1321D revision 3
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.27.5.2 Outputs for PACKET Command feature set devices
In response to this command, devices that implement the PACKET Command feature set shall post command
aborted and place the PACKET Command feature set signature in the Cylinder High and the Cylinder Low
register (see 9.12).
8.27.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred. The amount of data transferred is indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
UNC
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
obs
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
na
DF
na
DRQ
na
ERR
Error register UNC shall be set to one if data is uncorrectable.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.27.7 Prerequisites
DRDY set to one.
Page 154
T13/1321D revision 3
8.27.8 Description
This command reads from 1 to 256 sectors as specified in the Sector Count register. A sector count of 0
requests 256 sectors. The transfer shall begin at the sector specified in the Sector Number register.
The DRQ bit is always set to one prior to data transfer regardless of the presence or absence of an error
condition.
8.28 READ VERIFY SECTOR(S)
8.28.1 Command code
40h
8.28.2 Feature set
General feature set
−
−
Mandatory for all devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.28.3 Protocol
Non-data (see 9.4).
8.28.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
40h
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
Page 155
T13/1321D revision 3
8.28.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.28.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
UNC
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
obs
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
na
DF
na
DRQ
na
ERR
Error register UNC shall be set to one if data is uncorrectable.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Status register Page 156
T13/1321D revision 3
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.28.7 Prerequisites
DRDY set to one.
8.28.8 Description
This command is identical to the READ SECTOR(S) command, except that the DRQ bit is never set to one,
and no data is transferred to the host.
8.29 SECURITY DISABLE PASSWORD
8.29.1 Command code
F6h
8.29.2 Feature set
Security Mode feature set.
−
Mandatory when the Security Mode feature set is implemented.
8.29.3 Protocol
PIO data-out (see 9.6).
8.29.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
F6h
Device/Head register DEV shall indicate the selected device.
Page 157
T13/1321D revision 3
8.29.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.29.6 Error outputs
The device shall return command aborted if the command is not supported, the device is in Locked mode, or
the device is in Frozen mode.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
na
Error register ABRT shall be set to one if this command is not supported. ABRT may be set to one if the device is
not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.29.7 Prerequisites
DRDY set to one. Device shall be in Unlocked mode.
8.29.8 Description
The SECURITY DISABLE PASSWORD command requests a transfer of a single sector of data from the host.
Table 24 defines the content of this sector of information. If the password selected by word 0 matches the
Page 158
T13/1321D revision 3
password previously saved by the device, the device disables the Lock mode. This command does not change
the Master password that may be reactivated later by setting a User password (see 6.13).
Word
0
1-16
17-255
Table 24 − Security password content
Content
Control word
Bit 0
Identifier
0=compare User password
1=compare Master password
Bit 1-15 Reserved
Password (32 bytes)
Reserved
8.30 SECURITY ERASE PREPARE
8.30.1 Command code
F3h
8.30.2 Feature set
Security Mode feature set.
−
Mandatory when the Security Mode feature set is implemented.
8.30.3 Protocol
Non-data (see 9.4).
8.30.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
F3h
Device/Head register DEV shall indicate the selected device.
8.30.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register Page 159
T13/1321D revision 3
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.30.6 Error outputs
The device shall return command aborted if the command is not supported or the device is in Frozen mode.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported or device is in Frozen mode. ABRT may be
set to one if the device is not able to complete the action requested by the command.
NOTE − In a previous revision of this standard, there were conflicting descriptions of the
handling of this command when in the Frozen mode.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.30.7 Prerequisites
DRDY set to one.
8.30.8 Description
The SECURITY ERASE PREPARE command shall be issued immediately before the SECURITY ERASE UNIT
command to enable device erasing and unlocking. This command prevents accidental erase of the device.
8.31 SECURITY ERASE UNIT
8.31.1 Command code
F4h
Page 160
T13/1321D revision 3
8.31.2 Feature set
Security Mode feature set.
−
Mandatory when the Security Mode feature set is implemented.
8.31.3 Protocol
PIO data-out (see 9.6).
8.31.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
F4h
Device/Head register DEV shall indicate the selected device.
8.31.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.31.6 Error outputs
The device shall return command aborted if the command is not supported, the device is in Frozen mode, not
preceded by a SECURITY ERASE PREPARE command, or if the data area is not successfully overwritten.
Page 161
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported, device is in Frozen mode, not preceded by
a SECURITY ERASE PREPARE command, or if the data area is not successfully overwritten.
ABRT may be set to one if the device is not able to complete the action requested by the
command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.31.7 Prerequisites
DRDY set to one. This command shall be immediately preceded by a SECURITY ERASE PREPARE
command.
8.31.8 Description
This command requests transfer of a single sector of data from the host. Table 25 defines the content of this
sector of information. If the password does not match the password previously saved by the device, the device
rejects the command with command aborted.
The SECURITY ERASE PREPARE command shall be completed immediately prior to the SECURITY ERASE
UNIT command. If the device receives a SECURITY ERASE UNIT command without an immediately prior
SECURITY ERASE PREPARE command, the device command aborts the SECURITY ERASE UNIT
command.
When normal erase mode is selected, the SECURITY ERASE UNIT command writes binary zeroes to all user
data areas. The enhanced erase mode is optional. When enhanced erase mode is selected, the device writes
predetermined data patterns to all user data areas. In enhanced mode, all previously written user data is
overwritten, including sectors that are no longer in use due to reallocation.
This command disables the device Lock mode, however, the Master password is still stored internally within the
device and may be reactivated later when a new User password is set.
Page 162
T13/1321D revision 3
Word
0
1-16
17-255
Table 25 − SECURITY ERASE UNIT password
Content
Control word
Bit 0
Identifier
0=compare User password
1=compare Master password
Bit 1
Erase mode
0=Normal erase
1=Enhanced erase
Bit 2-15 Reserved
Password (32 bytes)
Reserved
8.32 SECURITY FREEZE LOCK
8.32.1 Command code
F5h
8.32.2 Feature set
Security Mode feature set.
−
Mandatory when the Security Mode feature set is implemented.
8.32.3 Protocol
Non-data (see 9.4).
8.32.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
F5h
Device/Head register DEV shall indicate the selected device.
8.32.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Page 163
T13/1321D revision 3
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.32.6 Error outputs
The device shall return command aborted if the command is not supported, or the device is in Locked mode.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported or device is in locked mode. ABRT may be
set to one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.32.7 Prerequisites
DRDY set to one.
8.32.8 Description
The SECURITY FREEZE LOCK command sets the device to Frozen mode. After command completion any
other commands that update the device Lock mode are rejected. Frozen mode is disabled by power-off or
hardware reset. If SECURITY FREEZE LOCK is issued when the device is in Frozen mode, the command
executes and the device remains in Frozen mode.
Commands disabled by SECURITY FREEZE LOCK are:
−
−
−
−
−
SECURITY SET PASSWORD
SECURITY UNLOCK
SECURITY DISABLE PASSWORD
SECURITY ERASE PREPARE
SECURITY ERASE UNIT
Page 164
T13/1321D revision 3
8.33 SECURITY SET PASSWORD
8.33.1 Command code
F1h
8.33.2 Feature set
Security Mode feature set.
−
Mandatory when the Security Mode feature set is implemented.
8.33.3 Protocol
PIO data-out (see 9.6).
8.33.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
F1h
Device/Head register DEV shall indicate the selected device.
8.33.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
Page 165
T13/1321D revision 3
8.33.6 Error outputs
The device shall return command aborted if the command is not supported, the device is in Locked mode, or
the device is in Frozen mode.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported, if device is in Frozen mode, or if device is
in locked mode. ABRT may be set to one if the device is not able to complete the action
requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.33.7 Prerequisites
DRDY set to one.
8.33.8 Description
This command requests a transfer of a single sector of data from the host. Table 26 defines the content of this
sector of information. The data transferred controls the function of this command. Table 27 defines the
interaction of the identifier and security level bits.
The revision code field is returned in the IDENTIFY DEVICE word 92. The valid revision codes are 0001h through
FFFEh. A value of 0000h or FFFFh indicates that the Master Password Revision Code is not supported.
Word
0
1-16
17
18-255
Page 166
Table 26 − SECURITY SET PASSWORD data content
Content
Control word
Bit 0
Identifier
0=set User password
1=set Master password
Bits 1-7
Reserved
Bit 8
Security level
0=High
1=Maximum
Bits 9-15
Reserved
Password (32 bytes)
Master Password Revision Code (valid if word 0 bit 0 = 1)
Reserved
T13/1321D revision 3
Identifier
User
Level
High
User
Maximum
Master
High or
Maximum
Table 27 − Identifier and security level bit interaction
Command result
The password supplied with the command shall be saved as the new User password.
The Lock mode shall be enabled from the next power-on or hardware reset. The device
shall then be unlocked by either the User password or the previously set Master
password.
The password supplied with the command shall be saved as the new User password.
The Lock mode shall be enabled from the next power-on or hardware reset. The device
shall then be unlocked by only the User password. The Master password previously set
is still stored in the device but shall not be used to unlock the device.
This combination shall set a Master password but shall not enable or disable the Lock
mode. The security level is not changed. Master password revision code set to the
value in Master Password Revision Code field.
8.34 SECURITY UNLOCK
8.34.1 Command code
F2h
8.34.2 Feature set
Security Mode feature set.
− Mandatory when the Security Mode feature set is implemented.
8.34.3 Protocol
PIO data-out (see 9.6).
8.34.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
F2h
Device/Head register DEV shall indicate the selected device.
8.34.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Page 167
T13/1321D revision 3
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.34.6 Error outputs
The device shall return command aborted if the command is not supported, or the device is in Frozen mode.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported or if device is in Frozen mode. ABRT may
be set to one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.34.7 Prerequisites
DRDY set to one.
8.34.8 Description
This command requests transfer of a single sector of data from the host. Table 24 defines the content of this
sector of information.
If the Identifier bit is set to Master and the device is in high security level, then the password supplied shall be
compared with the stored Master password. If the device is in maximum security level then the unlock shall be
rejected.
If the Identifier bit is set to user then the device compares the supplied password with the stored User
password.
If the password compare fails then the device returns command aborted to the host and decrements the unlock
counter. This counter is initially set to five and is decremented for each password mismatch when SECURITY
UNLOCK is issued and the device is locked. When this counter reaches zero then SECURITY UNLOCK and
SECURITY ERASE UNIT commands are command aborted until a power-on reset or a hardware reset.
SECURITY UNLOCK commands issued when the device is unlocked have no effect on the unlock counter.
Page 168
T13/1321D revision 3
8.35 SEEK
8.35.1 Command code
70h
8.35.2 Feature set
General feature set
−
−
Mandatory for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.35.3 Protocol
Non-data (see 9.4).
8.35.4 Inputs
The Cylinder High register, the Cylinder Low register, the head portion of Device/Head register, and the Sector
Number register contain the address of a sector that the host may request in a subsequent command.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
70h
Sector Number sector number or LBA address bits (7:0).
Cylinder Low cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) head number or LBA address bits (27:24).
8.35.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
DSC
Page 169
T13/1321D revision 3
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DSC (Device Seek Complete) shall be set to one concurrent with or after the setting of DRDY to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.35.6 Error outputs
Some devices may not report IDNF because they do not range check the address values requested by the
host.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
MC
4
IDNF
3
MCR
2
ABRT
DRQ
na
1
NM
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.35.7 Prerequisites
DRDY set to one.
Page 170
T13/1321D revision 3
8.35.8 Description
This command allows the host to provide advanced notification that particular data may be requested by the
host in a subsequent command. DSC shall be set to one concurrent with or after the setting of DRDY to one
when updating the Status register for this command.
8.36 SERVICE
8.36.1 Command code
A2h
8.36.2 Feature set
Overlap and Queued feature sets
−
Mandatory when the PACKET, Overlapped feature set is implemented.
8.36.3 Protocol
PACKET or READ/WRITE DMA QUEUED (see 9.8 and 9.9).
8.36.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
obs
na
obs
DEV
na
A2h
Device/Head register DEV shall indicate the selected device.
8.36.5 Outputs
Outputs as a result of a SERVICE command are described in the command description for the command for
which SERVICE is being requested.
8.36.6 Prerequisites
The device shall have performed a bus release for a previous overlap PACKET, READ DMA QUEUED, or
WRITE DMA QUEUED command and shall have set the SERV bit to one to request the SERVICE command
be issued to continue data transfer and/or provide command status (see 8.37.19).
8.36.7 Description
The SERVICE command is used to provide data transfer and/or status of a command that was previously bus
released.
Page 171
T13/1321D revision 3
8.37 SET FEATURES
8.37.1 Command code
EFh
8.37.2 Feature set
General feature set
−
−
−
−
−
Mandatory for all devices.
Set transfer mode subcommand is mandatory.
Enable/disable write cache subcommands are mandatory when a write cache is implemented.
Enable/Disable Media Status Notification sub commands are mandatory if the Removable Media
feature set is implemented.
All other subcommands are optional.
8.37.3 Protocol
Non-data (see 9.4).
8.37.4 Inputs
Table 28 defines the value of the subcommand in the Feature register. Some subcommands use other
registers, such as the Sector Count register to pass additional information to the device.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
Subcommand code
Subcommand specific
Subcommand specific
Subcommand specific
Subcommand specific
obs DEV na
na
EFh
1
0
na
na
Device/Head register DEV shall indicate the selected device.
8.37.5 Normal outputs
See the subcommand descriptions.
8.37.6 Error outputs
If any subcommand input value is not supported or is invalid, the device shall return command aborted.
Page 172
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this subcommand is not supported or if value is invalid. ABRT may be set
to one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.37.7 Prerequisites
DRDY shall be set to one.
8.37.8 Description
This command is used by the host to establish parameters that affect the execution of certain device features.
Table 28 defines these features.
At power-on, or after a hardware reset, the default settings of the functions specified by the subcommands are
vendor specific.
Page 173
T13/1321D revision 3
Table 28 − SET FEATURES register definitions
Value
(see note)
01h
Enable 8-bit PIO transfer mode (CFA feature set only)
02h
Enable write cache
03h
Set transfer mode based on value in Sector Count register. Table 29 defines values.
04h
Obsolete
05h
Enable advanced power management
06h
Enable Power-Up In Standby feature set.
07h
Power-Up In Standby feature set device spin-up.
09h
Reserved for Address offset reserved area boot method technical report
0Ah
Enable CFA power mode 1
31h
Disable Media Status Notification
33h
Obsolete
44h
Obsolete
54h
Obsolete
55h
Disable read look-ahead feature
5Dh
Enable release interrupt
5Eh
Enable SERVICE interrupt
66h
Disable reverting to power-on defaults
77h
Obsolete
81h
Disable 8-bit PIO transfer mode (CFA feature set only)
82h
Disable write cache
84h
Obsolete
85h
Disable advanced power management
86h
Disable Power-Up In Standby feature set.
88h
Obsolete
89h
Reserved for Address offset reserved area boot method technical report
8Ah
Disable CFA power mode 1
95h
Enable Media Status Notification
99h
Obsolete
9Ah
Obsolete
AAh
Enable read look-ahead feature
ABh
Obsolete
BBh
Obsolete
CCh
Enable reverting to power-on defaults
DDh
Disable release interrupt
DEh
Disable SERVICE interrupt
F0h-FFh
Reserved for assignment by the CompactFlash Association
NOTE − All values not shown are reserved for future definition.
8.37.9 Enable/disable 8-bit PIO data transfer
Devices implementing the CFA feature set shall support 8-bit PIO data transfers. Devices not implementing the
CFA feature set shall not support 8-bit PIO data transfers. When 8-bit PIO data transfer is enabled the Data
register is 8-bits wide using only DD7 to DD0.
8.37.10 Enable/disable write cache
Subcommand codes 02h and 82h allow the host to enable or disable write cache in devices that implement
write cache. When the subcommand disable write cache is issued, the device shall initiate the sequence to
flush cache to non-volatile memory before command completion (see 8.10).
Page 174
T13/1321D revision 3
8.37.11 Set transfer mode
A host selects the transfer mechanism by Set Transfer Mode, subcommand code 03h, and specifying a value
in the Sector Count register. The upper 5 bits define the type of transfer and the low order 3 bits encode the
mode value. The host may change the selected modes by the SET FEATURES command.
Table 29 − Transfer mode values
Mode
Bits (7:3)
PIO default mode
00000b
PIO default mode, disable IORDY
00000b
PIO flow control transfer mode
00001b
Retired
00010b
Multiword DMA mode
00100b
Ultra DMA mode
01000b
Reserved
10000b
mode = transfer mode number
Bits (2:0)
000b
001b
mode
na
mode
mode
na
If a device supports this standard, and receives a SET FEATURES command with a Set Transfer Mode
parameter and a Sector Count register value of “00000000b”, the device shall set the default PIO mode. If the
value is “00000001b” and the device supports disabling of IORDY, then the device shall set the default PIO
mode and disable IORDY. A device shall support all PIO modes below the highest mode supported, e.g., if PIO
mode 1 is supported PIO mode 0 shall be supported.
Support of IORDY is mandatory when PIO mode 3 or above is the current mode of operation.
Devices reporting support for Multiword DMA mode 1 shall also support Multiword DMA mode 0. A device shall
support all Multiword DMA modes below the highest mode supported, e.g., if Multiword DMA mode 1 is
supported Multiword DMA mode 0 shall be supported.
A device shall support all Ultra DMA modes below the highest mode supported, e.g., if Ultra DMA mode 1 is
supported Ultra DMA mode 0 shall be supported.
If an Ultra DMA mode is enabled any previously enabled Multiword DMA mode shall be disabled by the device.
If a Multiword DMA mode is enabled any previously enabled Ultra DMA mode shall be disabled by the device.
For systems using a cable assembly, the host shall detect that an 80-conductor cable assembly is connecting
the host with the device(s) before enabling any Ultra DMA mode greater than 2 in the device(s) (see Annex B).
8.37.12 Enable/disable advanced power management
Subcommand code 05h allows the host to enable Advanced Power Management. To enable Advanced Power
Management, the host writes the Sector Count register with the desired advanced power management level and
then executes a SET FEATURES command with subcommand code 05h. The power management level is a
scale from the lowest power consumption setting of 01h to the maximum performance level of FEh. Table 30
shows these values.
Page 175
T13/1321D revision 3
Table 30 − Advanced power management levels
Level
Sector Count value
Maximum performance
FEh
Intermediate power management levels without Standby
81h-FDh
Minimum power consumption without Standby
80h
Intermediate power management levels with Standby
02h-7Fh
Minimum power consumption with Standby
01h
Reserved
FFh
Reserved
00h
Device performance may increase with increasing power management levels. Device power consumption may
increase with increasing power management levels. The power management levels may contain discrete bands.
For example, a device may implement one power management method from 80h to A0h and a higher
performance, higher power consumption method from level A1h to FEh. Advanced power management levels
80h and higher do not permit the device to spin down to save power.
Subcommand code 85h disables Advanced Power Management. Subcommand 85h may not be implemented
on all devices that implement SET FEATURES subcommand 05h.
8.37.13 Enable/disable Power-Up In Standby feature set
Subcommand code 06h enables the Power-Up In Standby feature set. When this feature set is enabled, the
device shall power-up into Standby mode, i.e., the device shall be ready to receive commands but shall not
spinup (see 6.18). Having been enabled, this feature shall remain enabled through power-down, hardware reset
and software rest.
Subcommand code 86h disables the Power-Up In Standby feature set. When this feature set is disabled, the
device shall power-up into Active mode. The factory default for this feature set shall be disabled.
8.37.14 Enable/disable CFA power mode 1
Subcommand code 0Ah enables CFA Power Mode 1. CFA devices may consume up to 500 mA maximum
average RMS current for either 3.3V or 5V operation in Power Mode 1. CFA devices revert to Power Mode 1 on
hardware or power-on reset. CFA devices revert to Power Mode 1 on software reset except when Set Features
disable reverting to power-on defaults is set (see 8.12.57). Enabling CFA Power Mode 1 does not spin up
rotating media devices.
Subcommand 8Ah disables CFA Power Mode 1, placing the device to CFA Power Mode 0. CFA devices may
consume up to 75 mA maximum average RMS current for 3.3V or 100 mA maximum average RMS current for
5V operation in Power Mode 0.
A device in Power Mode 0 the device shall accept the following commands:
−
−
−
−
−
−
−
−
IDENTIFY DEVICE
SET FEATURES (function codes 0Ah and 8Ah)
STANDBY
STANDBY IMMEDIATE
SLEEP
CHECK POWER MODE
EXECUTE DEVICE DIAGNOSTICS
CFA REQUEST EXTENDED ERROR
A device in Power Mode 0 may accept any command that the device is capable of executing within the Power
Mode 0 current restrictions. Commands that require more current than specified for Power Mode 0 shall be
rejected with an abort error.
Page 176
T13/1321D revision 3
8.37.15 Power-Up In Standby feature set device spin-up
Subcommand code 07h shall cause a device that has powered-up into Standby to go to the Active state (see
6.18 and Figure 9).
8.37.16 Enable/disable Media Status Notification
Subcommand code 31h disables Media Status Notification and leaves the media in an unlocked state. If Media
Status Notification is disabled when this subcommand is received, the subcommand has no effect.
Subcommand code 95h enables Media Status Notification and clears any previous media lock state. This
subcommand returns the device capabilities for media eject, media lock, previous state of Media Status
Notification and the current version of Media Status Notification supported in the Cylinder Low and Cylinder High
registers as described below.
Register
Cylinder Low
Cylinder High
7
6
5
4
3
2
1
0
r
PEJ
LOCK
PENA
VER
r
r
r
r
Cylinder Low register VER shall contain the Media Status Notification version supported by the device (currently 0x00h)
Cylinder High register PENA shall be set to one if Media Status Notification was enabled prior to the receipt of this command,
LOCK shall be set to one if the device is capable of locking the media preventing manual ejection.
PEJ shall be set to one if the device has a power eject mechanism that is capable of physically
ejecting the media when a MEDIA EJECT command is sent to the device. This bit must be set to
zero if the device only unlocks the media when the device receives a MEDIA EJECT command.
r (reserved) shall be cleared to zero.
8.37.17 Enable/disable read look-ahead
Subcommand codes AAh and 55h allow the host to request the device to enable or disable read look-ahead.
Error recovery performed by the device is vendor specific.
8.37.18 Enable/disable release interrupt
Subcommand codes 5Dh and DDh allow a host to enable or disable the asserting of interrupt pending when a
device releases the bus for an overlapped PACKET command.
8.37.19 Enable/disable SERVICE interrupt
Subcommand codes 5Eh and DEh allow a host to enable or disable the asserting of an interrupt pending when
DRQ is set to one in response to a SERVICE command.
8.37.20 Enable/disable reverting to defaults
Subcommand codes CCh and 66h allow the host to enable or disable the device from reverting to power-on
default values. A setting of 66h allows settings that may have been modified since power-on to remain at the
same setting after a software reset.
8.38 SET MAX
Individual SET MAX commands are identified by the value placed in the Features register. Table 31 shows
these Features register values.
Page 177
T13/1321D revision 3
Table 31 − SET MAX Features register values
Value
Command
00h
Obsolete
01h
SET MAX SET PASSWORD
02h
SET MAX LOCK
03h
SET MAX UNLOCK
04h
SET MAX FREEZE LOCK
05h-FFh
Reserved
8.38.1 SET MAX ADDRESS
8.38.1.1 Command code
F9h with the content of the Features register equal to 00h.
8.38.1.2 Feature set
Host Protected Area feature set.
−
−
Mandatory when the Host Protected Area feature set is implemented.
Use prohibited when the Removable feature set is implemented.
8.38.1.3 Protocol
Non-data command (see 9.4).
8.38.1.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
obs
na
VV
Native max address sector number or SET MAX LBA
SET MAX cylinder low or LBA
SET MAX cylinder high or LBA
DEV
LBA
obs
Native max address head number or
SET MAX LBA
F9h
Sector Count V V (Value volatile). If bit 0 is set to one, the device shall preserve the maximum values over power-up
or hardware reset. If bit 0 is cleared to zero, the device shall revert to the most recent non-volatile
maximum address value setting over power-up or hardware reset.
Sector Number contains the native max address sector number (IDENTIFY DEVICE word 6) or LBA bits (7:0) value to
be set.
Cylinder Low contains the maximum cylinder low or LBA bits (15:8) value to be set.
Cylinder High contains the maximum cylinder high or LBA bits (23:16) value to be set.
Device/Head if LBA is set to one, the maximum address value is an LBA value.
If LBA is cleared to zero, the maximum address value is a CHS value.
DEV shall indicate the selected device.
Page 178
T13/1321D revision 3
Bits (3:0) contain the native max address head number (IDENTIFY DEVICE word 3 minus one) or LBA
bits (27:24) value to be set.
8.38.1.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
obs
BSY
6
5
4
3
2
1
0
na
na
Native max sector number or max LBA
Max cylinder low or LBA
Max cylinder high or LBA
na
obs
DEV
Native max head or max LBA
DRDY
DF
na
DRQ
na
na
ERR
Sector Number maximum native sector number or LBA bits (7:0) set on the device.
Cylinder Low maximum cylinder number low or LBA bits (15:8) set on the device.
Cylinder High maximum cylinder number high or LBA bits (23:16) set on device.
Device/Head DEV shall indicate the selected device.
maximum native head number or LBA bits (27:24) set on the device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.38.1.6 Error outputs
If this command is not supported, the maximum value to be set exceeds the capacity of the device, or the
device is in the Set_Max_Locked or Set_Max_Frozen state, then the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
IDNF
3
na
2
ABRT
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
na
na
Error register ABRT shall be set to one if this command is not supported, maximum value requested exceeds the
device capacity, the SET MAX cylinder number is greater than 16,383, or the command is not
immediately preceded by a READ NATIVE MAX ADDRESS command. ABRT may be set to
one if the device is not able to complete the action requested by the command.
IDNF shall be set to one if the command was the second non-volatile SET MAX ADDRESS command
after power-on or hardware reset.
Device/Head register DEV shall indicate the selected device.
Status register Page 179
T13/1321D revision 3
BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be set to one if an Error register bit is set to one.
8.38.1.7 Prerequisites
DRDY set to one. A successful READ NATIVE MAX ADDRESS command shall immediately precede a SET
MAX ADDRESS command.
8.38.1.8 Description
After successful command completion, all read and write access attempts to addresses greater than specified
by the successful SET MAX ADDRESS command shall be rejected with an IDNF error. IDENTIFY DEVICE
response words 1, 54, 57, 60, and 61 shall reflect the maximum address set with this command.
Hosts should not issue more than one non-volatile SET MAX ADDRESS command after a power-on or hardware
reset. Devices should report an IDNF error upon receiving a second non-volatile SET MAX ADDRESS
command after a power-on or hardware reset.
The contents of IDENTIFY DEVICE words and the max address shall not be changed if a SET MAX ADDRESS
command fails.
After a successful SET MAX ADDRESS command using a new maximum cylinder number value the content of
all IDENTIFY DEVICE words shall comply with 6.2.1 in addition to the following:
1) The content of words 3, 6, 55, and 56 are unchanged
2) The content of word 1 shall equal (the new SET MAX cylinder number + 1) or 16,383, whichever is less
3) The content of words (61:60) shall equal [(the new content of word 1 as determined by the successful SET
MAX ADDRESS command) ∗ (the content of word 3) ∗ (the content of word 6)]
4) If the content of words (61:60) as determined by a successful SET MAX ADDRESS command is less than
16,514,064, then the content of word 54 shall be equal to [(the content of words (61:60)) ÷ ((the content of
IDENTIFY DEVICE word 55) ∗ (the content of word 56)] or 65,535, whichever is less
5) If the content of word (61:60) as determined by a successful SET MAX ADDRESS command is greater
than 16,514,064, then word 54 shall equal the whole number result of [[(16,514,064) ÷ [(the content of word
55) ∗ (the content of word 56)]] or 65,535 whichever is less) The content of words (58:57) shall be equal
to [(the new content of word 54 as determined by the successful SET MAX ADDRESS command) ∗ (the
content of word 55) ∗ (the content of word 56)]
After a successful SET MAX ADDRESS command using a new maximum LBA address the content of all
IDENTIFY DEVICE words shall comply with 6.2.1 in addition to the following:
− The content of words (61:60) shall be equal to the new Maximum LBA address + 1.
− If the content of words (61:60) is greater than 16,514,064 and if the device does not support CHS
addressing, then the content of words 1, 3, 6, 54, 55, 56, and (58:57) shall equal zero.
If the device supports CHS addressing:
− The content of words 3, 6, 55, and 56 are unchanged.
− If the new content of words (61:60) is less than 16,514,064, then the content of word 1 shall be equal to
−
−
[(the new content of words (61:60)) ÷ [(the content of word 3) ∗ (the content of word 6)]] or 65,535,
whichever is less.
If the new content of words (61:60) is greater than or equal to 16,514,064, then the content of word 1 shall
be equal to 16,383.
If the new content of words (61:60) is less than 16,514,064, then the content of word 54 shall be equal to
[(the new content of words (61:60)) ÷ [(the content of word 55) ∗ (the content of word 56)]].
Page 180
T13/1321D revision 3
− If the new content of words (61:60) is greater than or equal to 16,514,064, then the content of word 54 shall
−
be equal to 16,383.
Words (58:57) shall be equal to [(the content of word 54) ∗ (the content of word 55) ∗ (the content of word
56).
8.38.2 SET MAX SET PASSWORD
8.38.2.1 Command code
F9h with the content of the Features register equal to 01h.
8.38.2.2 Feature set
Host Protected Area feature set.
−
−
Mandatory when the Host Protected Area feature set security extensions are implemented.
Use prohibited when the Removable feature set is implemented.
8.38.2.3 Protocol
PIO data-out (see 9.6).
8.38.2.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
01h
na
na
na
na
obs
na
obs
DEV
na
F9h
Device/Head DEV shall indicate the selected device.
8.38.2.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
na
Device/Head DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
Page 181
T13/1321D revision 3
ERR shall be cleared to zero.
8.38.2.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
IDNF
3
na
2
ABRT
na
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
Error register ABRT shall be set to one if this command is not supported. or the device is in the Set_Max_Locked or
Set_Max_Frozen state. ABRT may be set to one if the device is not able to complete the
action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be set to one if an Error register bit is set to one.
8.38.2.7 Prerequisites
DRDY set to one. This command shall not be immediately preceded by a READ NATIVE MAX ADDRESS
command. If this command is immediately preceded by a READ NATIVE MAX ADDRESS command, it shall
be interpreted as a SET MAX ADDRESS command.
8.38.2.8 Description
This command requests a transfer of a single sector of data from the host. Table 32 defines the content of this
sector of information. The password is retained by the device until the next power cycle. When the device
accepts this command the device is in Set_Max_Unlocked state.
Table 32 − SET MAX SET PASSWORD data content
Word
Content
0
Reserved
1-16
Password (32 bytes)
17-255
Reserved
8.38.3 SET MAX LOCK
8.38.3.1 Command code
F9h with the content of the Features register equal to 02h.
8.38.3.2 Feature set
Host Protected Area feature set.
−
−
Mandatory when the Host Protected Area feature set security extensions are implemented.
Use prohibited when the Removable feature set is implemented.
Page 182
T13/1321D revision 3
8.38.3.3 Protocol
Non-data command (see 9.4).
8.38.3.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
02h
na
na
na
na
obs
na
obs
DEV
na
F9h
Device/Head DEV shall indicate the selected device.
8.38.3.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
na
ERR
3
na
2
ABRT
1
na
0
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
Device/Head DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.38.3.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
IDNF
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
Error register ABRT shall be set to one if this command is not supported or the device is not in the Set_Max_Locked
state. ABRT may be set to one if the device is not able to complete the action requested by
the command.
Device/Head register Page 183
T13/1321D revision 3
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be set to one if an Error register bit is set to one.
8.38.3.7 Prerequisites
DRDY set to one. This command shall not be immediately preceded by a READ NATIVE MAX ADDRESS
command. If this command is immediately preceded by a READ NATIVE MAX ADDRESS command, it shall
be interpreted as a SET MAX ADDRESS command.
8.38.3.8 Description
The SET MAX LOCK command sets the device into Set_Max_Locked state. After this command is completed
any other SET MAX commands except SET MAX UNLOCK and SET MAX FREEZE LOCK are rejected. The
device remains in this state until a power cycle or the acceptance of a SET MAX UNLOCK or SET MAX
FREEZE LOCK command.
8.38.4 SET MAX UNLOCK
8.38.4.1 Command code
F9h with the content of the Features register equal to 03h.
8.38.4.2 Feature set
Host Protected Area feature set.
−
−
Mandatory when the Host Protected Area feature set security extensions are implemented.
Use prohibited when the Removable feature set is implemented.
8.38.4.3 Protocol
PIO data-out (see 9.6).
8.38.4.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
3
2
1
03h
na
na
na
na
obs
na
obs
Device/Head DEV shall indicate the selected device.
Page 184
4
DEV
na
F9h
0
T13/1321D revision 3
8.38.4.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
DRQ
na
na
ERR
2
ABRT
1
na
0
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Device/Head DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.38.4.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
IDNF
3
na
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
na
na
Error register ABRT shall be set to one if this command is not supported or the device is not in the Set_Max_Locked
state. ABRT may be set to one if the device is not able to complete the action requested by
the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be set to one if an Error register bit is set to one.
8.38.4.7 Prerequisites
DRDY set to one. This command shall not be immediately preceded by a READ NATIVE MAX ADDRESS
command. If this command is immediately preceded by a READ NATIVE MAX ADDRESS command, it shall
be interpreted as a SET MAX ADDRESS command.
8.38.4.8 Description
This command requests a transfer of a single sector of data from the host. Table 32 defines the content of this
sector of information.
The password supplied in the sector of data transferred shall be compared with the stored SET MAX password.
Page 185
T13/1321D revision 3
If the password compare fails, then the device returns command aborted and decrements the unlock counter.
On the acceptance of the SET MAX LOCK command, this counter is set to a value of five and shall be
decremented for each password mismatch when SET MAX UNLOCK is issued and the device is locked. When
this counter reaches zero, then the SET MAX UNLOCK command shall return command aborted until a power
cycle.
If the password compare matches, then the device shall make a transition to the Set_Max_Unlocked state and
all SET MAX commands shall be accepted.
8.38.5 SET MAX FREEZE LOCK
8.38.5.1 Command code
F9h with the content of the Features register equal to 04h.
8.38.5.2 Feature set
Host Protected Area feature set.
−
−
Mandatory when the Host Protected Area feature set security extensions are implemented.
Use prohibited when the Removable feature set is implemented.
8.38.5.3 Protocol
Non-data command (see 9.4).
8.38.5.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
04h
na
na
na
na
obs
na
obs
DEV
na
F9h
Device/Head DEV shall indicate the selected device.
8.38.5.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
4
3
2
1
0
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
Device/Head DEV shall indicate the selected device.
Page 186
5
obs
DF
DEV
na
na
DRQ
na
T13/1321D revision 3
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.38.5.6 Error outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
na
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
na
DEV
na
na
Error register ABRT shall be set to one if this command is not supported or the device is in the Set_Max_Unlocked
state. ABRT may be set to one if the device is not able to complete the action requested by
the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
ERR shall be set to one if an Error register bit is set to one.
8.38.5.7 Prerequisites
DRDY set to one. A SET MAX SET PASSWORD command shall previously have been successfully
completed. This command shall not be immediately preceded by a READ NATIVE MAX ADDRESS command.
If this command is immediately preceded by a READ NATIVE MAX ADDRESS command, it shall be
interpreted as a SET MAX ADDRESS command.
8.38.5.8 Description
The SET MAX FREEZE LOCK command sets the device to Set_Max_Frozen state. After command completion
any subsequent SET MAX commands are rejected.
Commands disabled by SET MAX FREEZE LOCK are:
−
−
−
−
SET MAX ADDRESS
SET MAX SET PASSWORD
SET MAX LOCK
SET MAX UNLOCK
8.39 SET MULTIPLE MODE
8.39.1 Command code
C6h
Page 187
T13/1321D revision 3
8.39.2 Feature set
General feature set
Mandatory for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.39.3 Protocol
Non-data (see 9.4).
8.39.4 Inputs
If the content of the Sector Count register is not zero, then the Sector Count register contains the number of
sectors per block for the device to be used on all following READ/WRITE MULTIPLE commands. The content
of the Sector Count register shall be less than or equal to the value in bits 0-7 in word 47 in the IDENTIFY
DEVICE information. The host should set the content of the Sector Count register to 1, 2, 4, 8, 16, 32, 64 or
128.
If the content of the Sector Count register is zero and the SET MULTIPLE command completes without error,
then the device shall respond to any subsequent READ MULTIPLE or WRITE MULTIPLE command with
command aborted until a subsequent successful SET MULTIPLE command completion where the Sector Count
register is not set to zero.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
na
Sectors per block
na
na
na
obs DEV na
na
C6h
1
0
na
na
Device/Head register DEV shall indicate the selected device.
8.39.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
Page 188
T13/1321D revision 3
8.39.6 Error outputs
If a block count is not supported, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if the block count is not supported. ABRT may be set to one if the device is
not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.39.7 Prerequisites
DRDY set to one.
8.39.8 Description
This command establishes the block count for READ MULTIPLE and WRITE MULTIPLE commands.
Devices shall support the block size specified in the IDENTIFY DEVICE parameter word 47, bits 7 through 0,
and may also support smaller values.
Upon receipt of the command, the device checks the Sector Count register. If the content of the Sector Count
register is not zero, the Sector Count register contains a valid value, and the block count is supported, then the
value in the Sector Count register is used for all subsequent READ MULTIPLE and WRITE MULTIPLE
commands and their execution is enabled. If the content of the Sector Count register is zero, the device may:
1) disable multiple mode and respond with command aborted to all subsequent READ MULTIPLE and
WRITE MULTIPLE commands;
2) respond with command aborted to the SET MULTIPLE MODE command;
3) retain the previous multiple mode settings.
After a successful SET MULTIPLE command the device shall report the valid value set by that command in bits
0-7 in word 59 in the IDENTIFY DEVICE information.
8.40 SLEEP
8.40.1 Command code
E6h
Page 189
T13/1321D revision 3
8.40.2 Feature set
Power Management feature set.
−
−
Power Management feature set is mandatory when power management is not implemented by a
PACKET power management feature set.
This command is mandatory when the Power Management feature set is implemented.
8.40.3 Protocol
Non-data command (see 9.4).
8.40.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
E6h
Device/Head register DEV shall indicate the selected device.
8.40.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.40.6 Error outputs
The device shall return command aborted if the device does not support the Power Management feature set.
Page 190
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if the device does not support the Power Management feature set. ABRT may
be set to one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.40.7 Prerequisites
DRDY set to one.
8.40.8 Description
This command is the only way to cause the device to enter Sleep mode.
This command causes the device to set the BSY bit to one, prepare to enter Sleep mode, clear the BSY bit to
zero and assert INTRQ. The host shall read the Status register in order to clear the interrupt pending and allow
the device to enter Sleep mode. In Sleep mode, the device only responds to the assertion of the RESET- signal
and the writing of the SRST bit in the Device Control register and releases the device driven signal lines. The
host shall not attempt to access the Command Block registers while the device is in Sleep mode.
Because some host systems may not read the Status register and clear the interrupt pending, a device may
automatically release INTRQ and enter Sleep mode after a vendor specific time period of not less than 2 s.
The only way to recover from Sleep mode is with a software reset, a hardware reset, or a DEVICE RESET
command.
A device shall not power-on in Sleep mode nor remain in Sleep mode following a reset sequence.
8.41 SMART
Individual SMART commands are identified by the value placed in the Feature register. Table 33 shows these
Feature register values.
Page 191
T13/1321D revision 3
Value
00h-CFh
D0h
D1h
D2h
D3h
D4h
D5h
D6h
D7h
D8h
D9h
DAh
DBh
DCh-DFh
E0h-FFh
Table 33 − SMART Feature register values
Command
Reserved
SMART READ DATA
Obsolete
SMART ENABLE/DISABLE ATTRIBUTE AUTOSAVE
SMART SAVE ATTRIBUTE VALUES
SMART EXECUTE OFF-LINE IMMEDIATE
SMART READ LOG
SMART WRITE LOG
Obsolete
SMART ENABLE OPERATIONS
SMART DISABLE OPERATIONS
SMART RETURN STATUS
Obsolete
Reserved
vendor specific
8.41.1 SMART DISABLE OPERATIONS
8.41.1.1 Command code
B0h with a Feature register value of D9h.
8.41.1.2 Feature set
SMART feature set.
−
−
Mandatory when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.1.3 Protocol
Non-data command (see 9.4).
8.41.1.4 Inputs
The Features register shall be set to D9h. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
5
4
3
2
1
0
na
na
na
na
D9h
na
na
4Fh
C2h
obs
Device/Head register DEV shall indicate the selected device.
Page 192
6
na
obs
DEV
B0h
T13/1321D revision 3
8.41.1.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.1.6 Error outputs
If the device does not support this command, if SMART is not enabled, or if the values in the Features, Cylinder
Low, or Cylinder High registers are invalid, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
na
Error register ABRT shall be set to one if this command is not supported, if SMART is not enabled, or if input register
values are invalid. ABRT may be set to one if the device is not able to complete the action
requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.41.1.7 Prerequisites
DRDY set to one. SMART enabled.
Page 193
T13/1321D revision 3
8.41.1.8 Description
This command disables all SMART capabilities within the device including any and all timer and event count
functions related exclusively to this feature. After receipt of this command the device shall disable all SMART
operations. SMART data shall no longer be monitored or saved by the device. The state of SMART (either
enabled or disabled) shall be preserved by the device across power cycles.
After receipt of this command by the device, all other SMART commands (including SMART DISABLE
OPERATIONS commands), with the exception of SMART ENABLE OPERATIONS, are disabled and invalid and
shall be command aborted by the device.
8.41.2 SMART ENABLE/DISABLE ATTRIBUTE AUTOSAVE
8.41.2.1 Command code
B0h with a Feature register value of D2h.
8.41.2.2 Feature set
SMART feature set.
−
−
Mandatory when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.2.3 Protocol
Non-data command (see 9.4).
8.41.2.4 Inputs
The Features register shall be set to D2h. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h. The Sector Count register is set to 00h to disable attribute autosave and a value of
F1h is set to enable attribute autosave.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
obs
na
obs
Device/Head register DEV shall indicate the selected device.
Page 194
4
3
D2h
00h or F1h
na
4Fh
C2h
DEV
na
B0h
2
1
0
na
na
na
T13/1321D revision 3
8.41.2.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.2.6 Error outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Cylinder
Low, or Cylinder High registers are invalid, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
na
Error register ABRT shall be set to one if this command is not supported, if SMART is disabled, or if the input
register values are invalid. ABRT may be set to one if the device is not able to complete the
action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.41.2.7 Prerequisites
DRDY set to one. SMART enabled.
Page 195
T13/1321D revision 3
8.41.2.8 Description
This command enables and disables the optional attribute autosave feature of the device. Depending upon the
implementation, this command may either allow the device, after some vendor specified event, to automatically
save the device updated attribute values to non-volatile memory; or this command may cause the autosave
feature to be disabled. The state of the attribute autosave feature (either enabled or disabled) shall be preserved
by the device across power cycles.
A value of zero written by the host into the device’s Sector Count register before issuing this command shall
cause this feature to be disabled. Disabling this feature does not preclude the device from saving SMART data
to non-volatile memory during some other normal operation such as during a power-on or power-off sequence or
during an error recovery sequence.
A value of F1h written by the host into the device’s Sector Count register before issuing this command shall
cause this feature to be enabled. Any other meaning of this value or any other non-zero value written by the
host into this register before issuing this command may differ from device to device. The meaning of any nonzero value written to this register at this time shall be preserved by the device across power cycles.
If this command is not supported by the device, the device shall return command aborted upon receipt from the
host.
During execution of the autosave routine the device shall not set BSY to one nor clear DRDY to zero. If the
device receives a command from the host while executing the autosave routine the device shall respond to the
host within two seconds.
8.41.3 SMART ENABLE OPERATIONS
8.41.3.1 Command code
B0h with a Feature register value of D8h.
8.41.3.2 Feature set
SMART feature set.
−
−
Mandatory when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.3.3 Protocol
Non-data command (see 9.4).
8.41.3.4 Inputs
The Features register shall be set to D8h. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h.
Page 196
T13/1321D revision 3
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
D8h
na
na
4Fh
C2h
obs
na
obs
DEV
B0h
Device/Head register DEV shall indicate the selected device.
8.41.3.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.3.6 Error outputs
If the device does not support this command or if the values in the Features, Cylinder Low, or Cylinder High
registers are invalid, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported or if the input register values are invalid.
ABRT may be set to one if the device is not able to complete the action requested by the
command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
Page 197
T13/1321D revision 3
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.41.3.7 Prerequisites
DRDY set to one.
8.41.3.8 Description
This command enables access to all SMART capabilities within the device. Prior to receipt of this command
SMART data are neither monitored nor saved by the device. The state of SMART (either enabled or disabled)
shall be preserved by the device across power cycles. Once enabled, the receipt of subsequent SMART
ENABLE OPERATIONS commands shall not affect any SMART data or functions.
8.41.4 SMART EXECUTE OFF-LINE IMMEDIATE
8.41.4.1 Command code
B0h with the content of the Features register equal to D4h
8.41.4.2 Feature set
SMART feature set.
−
−
Optional when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.4.3 Protocol
Non-data command (see 9.4).
8.41.4.4 Inputs
The Features register shall be set to D4h. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h. Table 34 defines the subcommand that shall be executed based on the value in the
Sector Number register.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
Device/Head register DEV shall indicate the selected device.
Page 198
5
4
3
2
D4h
na
Subcommand specific
4Fh
C2h
obs
DEV
B0h
1
na
0
T13/1321D revision 3
8.41.4.5 Normal Outputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
obs
BSY
na
DRDY
obs
DF
4
3
na
na
na
na or 4Fh
na or C2h
DEV
na
DRQ
2
1
0
na
ERR
na
na
Cylinder Low na when the subcommand specified an off-line routine (including an off-line self-test routine).
4Fh when the subcommand specified a captive self-test routine (see 8.41.4.8.2) that has executed
without failure.
Cylinder High na when the subcommand specified an off-line routine (including an off-line self-test routine).
C2h when the subcommand specified a captive self-test routine that has executed without failure.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.4.6 Error Outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Cylinder
Low, or Cylinder High registers are invalid, the device shall return command aborted. When a failure occurs
while executing a test in captive mode, the device shall return command aborted with the Cylinder Low register
value of F4h and the Cylinder High value of 2Ch.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
na
na
DRDY
5
na
obs
DF
4
IDNF
3
na
na
na
na or 4Fh or F4h
na or C2h or 2Ch
DEV
na
DRQ
2
ABRT
1
na
0
obs
na
ERR
na
na
Error register IDNF shall be set to one if SMART data sector’s ID field could not be found.
ABRT shall be set to one if this command is not supported, if SMART is not enabled, if register values
are invalid, or if a self-test fails while executing a sequence in captive mode. ABRT may be set
to one if the device is not able to complete the action requested by the command.
Cylinder Low register –
na when the subcommand specified an off-line routine (including an off-line self-test routine).
4Fh when the subcommand specified a captive self-test routine and some error other than a self-test
routine failure occurred (i.e., if the sub-command is not supported or register values are invalid)
F4h when the subcommand specified a captive self-test routine which has failed during execution.
Cylinder High register –
Page 199
T13/1321D revision 3
na when the subcommand specified an off-line routine (including an off-line self-test routine).
2Ch when the subcommand specified a captive self-test routine which has failed during execution.
C2h when the subcommand specified a captive self-test routine and some error other than a self-test
routine failure occurred (i.e., if the sub-command is not supported or register values are invalid)
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be set to one indicating that a device fault has occurred.
DRQ shall be cleared to zero indicating that there is no data to be transferred.
ERR shall be set to one if any Error register bit is set to one.
8.41.4.7 Prerequisites
DRDY set to one. SMART enabled.
8.41.4.8 Description
This command causes the device to immediately initiate the optional set of activities that collect SMART data
in an off-line mode and then save this data to the device's non-volatile memory, or execute a self-diagnostic test
routine in either captive or off-line mode.
Table 34 − SMART EXECUTE OFF-LINE IMMEDIATE Sector Number register values
Value
Description of subcommand to be executed
0
Execute SMART off-line routine immediately in off-line mode
1
Execute SMART Short self-test routine immediately in off-line mode
2
Execute SMART Extended self-test routine immediately in off-line mode
3-63
Reserved
64-126
Vendor specific
127
Abort off-line mode self-test routine
128
Reserved
129
Execute SMART Short self-test routine immediately in captive mode
130
Execute SMART Extended self-test routine immediately in captive mode
131-191 Reserved
192-255 Vendor specific
8.41.4.8.1 Off-line mode
The following describes the protocol for executing a SMART EXECUTE OFF-LINE IMMEDIATE subcommand
routine (including a self-test routine) in the off-line mode.
a) The device shall execute command completion before executing the subcommand routine.
b) After clearing BSY to zero and setting DRDY to one after receiving the command, the device shall not set
BSY nor clear DRDY during execution of the subcommand routine.
c) If the device is in the process of performing the subcommand routine and is interrupted by any new
command from the host except a SLEEP, SMART DISABLE OPERATIONS, SMART EXECUTE OFF-LINE
IMMEDIATE, or STANDBY IMMEDIATE command, the device shall suspend or abort the subcommand
routine and service the host within two seconds after receipt of the new command. After servicing the
interrupting command from the host the device may immediately re-initiate or resume the subcommand
routine without any additional commands from the host (see 8.41.5.8.4).
d) If the device is in the process of performing a subcommand routine and is interrupted by a SLEEP
command from the host, the device may abort the subcommand routine and execute the SLEEP
command. If the device is in the process of performing any self-test routine and is interrupted by a SLEEP
Page 200
T13/1321D revision 3
command from the host, the device shall abort the subcommand routine and execute the SLEEP
command.
e) If the device is in the process of performing the subcommand routine and is interrupted by a SMART
DISABLE OPERATIONS command from the host, the device shall suspend or abort the subcommand
routine and service the host within two seconds after receipt of the command. Upon receipt of the next
SMART ENABLE OPERATIONS command the device may, either re-initiate the subcommand routine or
resume the subcommand routine from where it had been previously suspended.
f) If the device is in the process of performing the subcommand routine and is interrupted by a SMART
EXECUTE OFF-LINE IMMEDIATE command from the host, the device shall abort the subcommand routine
and service the host within two seconds after receipt of the command. The device shall then service the
new SMART EXECUTE OFF-LINE IMMEDIATE subcommand.
g) If the device is in the process of performing the subcommand routine and is interrupted by a STANDBY
IMMEDIATE or IDLE IMMEDIATE command from the host, the device shall suspend or abort the
subcommand routine, and service the host within two seconds after receipt of the command. After
receiving a new command that causes the device to exit a power saving mode, the device shall initiate or
resume the subcommand routine without any additional commands from the host unless these activities
were aborted by the host (see 8.41.5.8).
h) While the device is performing the subcommand routine it shall not automatically change power states
(e.g., as a result of its Standby timer expiring).
i) If a test failure occurs while a device is performing a self-test routine the device may discontinue the testing
and place the test results in the Self-test execution status byte.
8.41.4.8.2 Captive mode
When executing a self-test in captive mode, the device sets BSY to one and executes the self-test routine after
receipt of the command. At the end of the routine the device places the results of this routine in the Self-test
execution status byte and executes command completion. If an error occurs while a device is performing the
routine the device may discontinue its testing, place the results of this routine in the Self-test execution status
byte, and complete the command.
8.41.4.8.3 SMART off-line routine
This routine shall only be performed in the off-line mode. The results of this routine are placed in the Off-line
data collection status byte (see Table 36).
8.41.4.8.4 SMART Short self-test routine
Depending on the value in the Sector Number register, this self-test routine may be performed in either the
captive or the off-line or mode. This self-test routine should take on the order of ones of minutes to complete
(see 8.41.5.8).
8.41.4.8.5 SMART Extended self-test routine
Depending on the value in the Sector Number register, this self-test routine may be performed in either the
captive or the off-line or mode. This self-test routine should take on the order of tens of minutes to complete
(see 8.41.5.8).
8.41.5 SMART READ DATA
8.41.5.1 Command code
B0h with the content of the Features register equal to D0h.
Page 201
T13/1321D revision 3
8.41.5.2 Feature set
SMART feature set.
− Optional when the SMART feature set is implemented.
− Use prohibited when the PACKET Command feature set is implemented.
8.41.5.3 Protocol
PIO data-in (see 9.5).
8.41.5.4 Inputs
The Features register shall be set to D0h. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
1
0
na
ERR
D0h
na
na
4Fh
C2h
obs
na
obs
DEV
na
B0h
Device/Head register DEV shall indicate the selected device.
8.41.5.5 Normal outputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.5.6 Error outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Cylinder
Low, or Cylinder High registers are invalid, the device shall return command aborted.
Page 202
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
UNC
5
na
4
IDNF
3
na
2
ABRT
DRQ
na
1
na
0
obs
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register UNC shall be set to one if SMART data is uncorrectable.
IDNF shall be set to one if SMART data sector’s ID field could not be found or data structure checksum
occurred.
ABRT shall be set to one if this command is not supported, if SMART is not enabled, or if register
values are invalid. ABRT may be set to one if the device is not able to complete the action
requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be set to one indicating that a device fault has occurred.
DRQ shall be cleared to zero indicating that there is no data to be transferred.
ERR shall be set to one if any Error register bit is set to one.
8.41.5.7 Prerequisites
DRDY set to one. SMART enabled.
8.41.5.8 Description
This command returns the Device SMART data structure to the host.
Table 35 defines the 512 bytes that make up the Device SMART data structure. All multi-byte fields shown in
this structure follow the byte ordering described in 3.2.9.
Page 203
T13/1321D revision 3
Byte
0-361
362
363
364-365
366
367
368-369
370
F/V
X
V
X
V
X
F
F
F
Table 35 − Device SMART data structure
Descriptions
Vendor specific
Off-line data collection status
Self-test execution status byte
Total time in seconds to complete off-line data collection activity
Vendor specific
Off-line data collection capability
SMART capability
Error logging capability
7-1
Reserved
0
1=Device error logging supported
Vendor specific
Short self-test routine recommended polling time (in minutes)
Extended self-test routine recommended polling time (in minutes)
Reserved
Vendor specific
Data structure checksum
371
X
372
F
373
F
374-385
R
386-510
X
511
V
Key:
F=the content of the byte is fixed and does not change.
V=the content of the byte is variable and may change depending on the state of the
device or the commands executed by the device.
X=the content of the byte is vendor specific and may be fixed or variable.
R=the content of the byte is reserved and shall be zero.
8.41.5.8.1 Off-line collection status byte
The value of the off-line data collection status byte defines the current status of the off-line activities of the
device. Table 36 lists the values and their respective definitions.
Value
00h or 80h
01h
02h or 82h
03h
04h or 84h
05h or 85h
06h or 86h
07h-3Fh
40h-7Fh
81h
83h
87h-BFh
C0h-FFh
Table 36 − Off-line data collection status byte values
Definition
Off-line data collection activity was never started.
Reserved
Off-line data collection activity was completed without error.
Reserved
Off-line data collection activity was suspended by an interrupting command from host.
Off-line data collection activity was aborted by an interrupting command from host.
Off-line data collection activity was aborted by the device with a fatal error.
Reserved
Vendor specific
Reserved
Reserved
Reserved
Vendor specific
8.41.5.8.2 Self-test execution status byte
The self-test execution status byte reports the execution status of the self-test routine.
−
Page 204
Bits 0-3 (Percent Self-Test Remaining) The value in these bits indicates an approximation of the
percent of the self-test routine remaining until completion in ten percent increments. Valid values
T13/1321D revision 3
−
are 0 through 9. A value of 0 indicates the self-test routine is complete. A value of 9 indicates
90% of total test time remaining.
Bits 4-7 (Self-test Execution Status) The value in these bits indicates the current Self-test
Execution Status (see Table 37).
Table 37 − Self-test execution status values
Value
0
Description
The previous self-test routine completed without error or no self-test has ever been run
1
The self-test routine was aborted by the host
2
The self-test routine was interrupted by the host with a hardware or software reset
3
A fatal error or unknown test error occurred while the device was executing its self-test
routineand the device was unable to complete the self-test routine.
The previous self-test completed having a test element that failed and the test element
that failed is not known.
The previous self-test completed having the electrical element of the test failed.
The previous self-test completed having the servo (and/or seek) test element of the test
failed.
The previous self-test completed having the read element of the test failed.
Reserved.
4
5
6
7
8-14
15
Self-test routine in progress.
8.41.5.8.3 Total time to complete off-line data collection
The total time in seconds to complete off-line data collection activity word specifies how many seconds the
device requires to complete the sequence of off-line data collection activity. Valid values for this word are from
0001h to FFFFh.
8.41.5.8.4 Off-line data collection capabilities
The following describes the definition for the off-line data collection capability bits. If the value of all of these bits
is cleared to zero, then no off-line data collection is implemented by this device.
−
Bit 0 (EXECUTE OFF-LINE IMMEDIATE implemented bit) - If this bit is set to one, then the SMART
EXECUTE OFF-LINE IMMEDIATE command is implemented by this device. If this bit is cleared to zero,
then the SMART EXECUTE OFF-LINE IMMEDIATE command is not implemented by this device.
−
Bit 1 (vendor specific).
−
Bit 2 (abort/restart off-line by host bit) - If this bit is set to one, then the device shall abort all off-line data
collection activity initiated by an SMART EXECUTE OFF-LINE IMMEDIATE command upon receipt of a
new command within 2 seconds of receiving the new command. If this bit is cleared to zero, the device
shall suspend off-line data collection activity after an interrupting command and resume off-line data
collection activity after some vendor-specified event.
−
Bit 3 (off-line read scanning implemented bit) - If this bit is cleared to zero, the device does not support offline read scanning. If this bit is set to one, the device supports off-line read scanning.
−
Bit 4 (self-test implemented bit) – If this bit is cleared to zero, the device does not implement the Short and
Extended self-test routines. If this bit is set to one, the device implements the Short and Extended selftest routines.
−
Bits 5-7 (Reserved).
Page 205
T13/1321D revision 3
8.41.5.8.5 SMART capablilities
The following describes the definition for the SMART capabilities bits. If all of these bits are cleared to zero,
then automatic saving of SMART data is not implemented by this device.
−
Bit 0 (power mode SMART data saving capability bit) - If this bit is set to one, the device saves SMART
data prior to going into a power saving mode (Idle, Standby, or Sleep) or immediately upon return to Active
or Idle mode from a Standby mode. If this bit is cleared to zero, the device does not save SMART data
prior to going into a power saving mode (Idle, Standby, or Sleep) or immediately upon return to Active or
Idle mode from a Standby mode.
−
Bit 1 (SMART data autosave after event capability bit) - This bit is set to one for devices complying with this
standard.
−
Bits 2-15 (Reserved).
8.41.5.8.6 Self-test routine recommended polling time
The self-test routine recommended polling time shall be equal to the number of minutes that is the minimum
recommended time before which the host should first poll for test completion status. Actual test time could be
several times this value. Polling before this time could extend the self-test execution time or abort the test
depending on the state of bit 2 of the off-line data capability bits.
8.41.5.8.7 Data structure checksum
The data structure checksum is the two's complement of the sum of the first 511 bytes in the data structure.
Each byte shall be added with unsigned arithmetic, and overflow shall be ignored. The sum of all 512 bytes will
be zero when the checksum is correct. The checksum is placed in byte 511.
8.41.6 SMART READ LOG
8.41.6.1 Command code
B0h with the content of the Features register equal to D5h.
8.41.6.2 Feature set
SMART feature set.
−
−
Optional when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.6.3 Protocol
PIO data-in (see 9.5).
8.41.6.4 Inputs
The Features register shall be set to D5h. The Sector Count register shall indicate the number of sectors to be
read from the log number indicated by the Sector Number register. The Cylinder Low register shall be set to
4Fh. The Cylinder High register shall be set to C2h.
Page 206
T13/1321D revision 3
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
D5h
Number of sectors to be read
Log address
4Fh
C2h
obs
DEV
B0h
1
0
na
Sector count – Indicates the number of sectors to be read from the indicated log. The log transferred by the
drive shall start at the first sector in the indicated log, regardless of the sector count requested.
Sector number - Indicates the log to be returned as described in Table 38. If this command is implemented, all
address values for which the contents are defined shall be implemented and all address values defined as
host vendor specific shall be implemented. The host vendor specific logs may be used by the host to
store any data desired. If a host vendor specific log has never been written by the host, when read the
content of the log shall be zeros. Device vendor specific logs may be used by the device vendor to store
any data and need only be implemented if used.
Table 38 − Log address definition
Content
Log directory
SMART error log
Reserved
SMART self-test log
Reserved
Host vendor specific
Device vendor specific
Reserved
Log address
00h
01h
02h-05h
06h
07h-7Fh
80h-9Fh
A0h-BFh
C0h-FFh
R/W
RO
RO
Reserved
RO
Reserved
R/W
VS
Reserved
Key −
RO - Log is read only by the host.
R/W - Log is read or written by the host.
VS - Log is vendor specific thus read/write abiltiy is vendor specific.
Device/Head register DEV shall indicate the selected device.
8.41.6.5 Normal outputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
DRQ
na
1
0
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be cleared to zero.
Page 207
T13/1321D revision 3
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.6.6 Error outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Sector
Number, Sector Count, Cylinder Low, or Cylinder High registers are invalid, the device shall return command
aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
UNC
5
na
4
IDNF
3
na
2
ABRT
DRQ
na
1
na
0
obs
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Na
Error register UNC shall be set to one if SMART log sector is uncorrectable.
IDNF shall be set to one if SMART log sector’s ID field was not found or data structure checksum error
occurred.
ABRT shall be set to one if this command is not supported, if SMART is not enabled, if the log sector
address is not implemented, or if other register values are invalid. ABRT may be set to one if
the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be set to one indicating that a device fault has occurred.
DRQ shall be cleared to zero indicating that there is no data to be transferred.
ERR shall be set to one if any Error register bit is set to one.
8.41.6.7 Prerequisites
DRDY set to one. SMART enabled.
8.41.6.8 Description
This command returns the indicated log to the host.
8.41.6.8.1 SMART Log Directory
Table 39 defines the 512 bytes that make up the SMART Log Directory, which is optional. If implemented, the
SMART Log Directory is SMART Log address zero, and is defined as one sector long.
Page 208
T13/1321D revision 3
Byte
0-1
2
3
4
5
…
510
511
Table 39 − SMART Log Directory
Descriptions
SMART Logging Version
Number of sectors in the log at log address 1
Reserved
Number of sectors in the log at log address 2
Reserved
…
Number of sectors in the log at log address 255
Reserved
The value of the SMART Logging Version word shall be 01h if the drive supports multi-sector SMART logs. In
addition, if the drive supports multi-sector logs, then the logs at log addresses 80-9Fh shall each be defined as
16 sectors long.
If the drive does not support multi-sector SMART logs, then log number zero is defined as reserved, and the
drive shall return a command aborted response to the host’s request to read log number zero.
If the host issues a SMART READ LOG or SMART WRITE LOG command with a Sector Count value of zero,
the device shall return command aborted.
8.41.6.8.2 Error log sector
Table 40 defines the 512 bytes that make up the SMART error log sector. All multi-byte fields shown in this
structure follow the byte ordering described in 3.2.9. Error log data structures shall include UNC errors, IDNF
errors for which the address requested was valid, servo errors, write fault errors, etc. Error log data structures
shall not include errors attributed to the receipt of faulty commands such as command codes not implemented
by the device or requests with invalid parameters or invalid addresses.
Byte
0
1
2-91
92-181
182-271
272-361
362-451
452-453
454-510
511
Table 40 − SMART error log sector
Descriptions
SMART error log version
Error log index
First error log data structure
Second error log data structure
Third error log data structure
Fourth error log data structure
Fifth error log data structure
Device error count
Reserved
Data structure checksum
8.41.6.8.2.1 Error log version
The value of the SMART error log version byte shall be 01h.
8.41.6.8.2.2 Error log index
The error log index indicates the error log data structure representing the most recent error. Only values 1
through 5 are valid.
Page 209
T13/1321D revision 3
8.41.6.8.2.3 Error log data structure
An error log data structure shall be presented for each of the last five errors reported by the device. These error
log data structure entries are viewed as a circular buffer. That is, the first error shall create the first error log
data structure; the second error, the second error log structure; etc. The sixth error shall create an error log
data structure that replaces the first error log data structure; the seventh error replaces the second error log
structure, etc. The error log pointer indicates the most recent error log structure. If fewer than five errors have
occurred, the unused error log structure entries shall be zero filled. Table 41 describes the content of a valid
error log data structure.
Byte
n thru n+11
n+12 thru n+23
n+24 thru n+35
n+36 thru n+47
n+48 thru n+59
n+60 thru n+89
Table 41 − Error log data structure
Descriptions
First command data structure
Second command data structure
Third command data structure
Fourth command data structure
Fifth command data structure
Error data structure
8.41.6.8.2.3.1 Command data structure
The fifth command data structure shall contain the command or reset for which the error is being reported. The
fourth command data structure should contain the command or reset that preceded the command or reset for
which the error is being reported, the third command data structure should contain the command or reset
preceding the one in the fourth command data structure, etc. If fewer than four commands and resets preceded
the command or reset for which the error is being reported, the unused command data structures shall be zero
filled, for example, if only three commands and resets preceded the command or reset for which the error is
being reported, the first command data structure shall be zero filled. In some devices, the hardware
implementation may preclude the device from reporting the commands that preceded the command for which
the error is being reported or that preceded a reset. In this case, the command data structures are zero filled.
If the command data structure represents a command or software reset, the content of the command data
structure shall be as shown in Table 42. If the command data structure represents a hardware reset, the
content of byte n shall be FFh, the content of bytes n+1 through n+7 are vendor specific, and the content of
bytes n+8 through n+11 shall contain the timestamp.
Table 42 − Command data structure
Byte
n
n+1
n+2
n+3
n+4
n+5
n+6
n+7
n+8
n+9
n+10
n+11
Descriptions
Content of the Device Control register when the Command register was written.
Content of the Features register when the Command register was written.
Content of the Sector Count register when the Command register was written.
Content of the Sector Number register when the Command register was written.
Content of the Cylinder Low register when the Command register was written.
Content of the Cylinder High register when the Command register was written.
Content of the Device/Head register when the Command register was written.
Content written to the Command register.
Timestamp (least significant byte)
Timestamp (next least significant byte)
Timestamp (next most significant byte)
Timestamp (most significant byte)
Timestamp shall be the time since power-on in milliseconds when command acceptance occurred. This
timestamp may wrap around.
Page 210
T13/1321D revision 3
8.41.6.8.2.3.2 Error data structure
The error data structure shall contain the error description of the command for which an error was reported as
described in Table 43. If the error was logged for a hardware reset, the content of bytes n+1 through n+7 shall
be vendor specific and the remaining bytes shall be as defined in Table 43.
Table 43 − Error data structure
Byte
n
n+1
n+2
n+3
n+4
n+5
n+6
n+7
n+8 thru n+26
n+27
n+28
n+29
Descriptions
Reserved
Content of the Error register after command completion occurred.
Content of the Sector Count register after command completion occurred.
Content of the Sector Number register after command completion occurred.
Content of the Cylinder Low register after command completion occurred.
Content of the Cylinder High register after command completion occurred.
Content of the Device/Head register after command completion occurred.
Content written to the Status register after command completion occurred.
Extended error information
State
Life timestamp (least significant byte)
Life timestamp (most significant byte)
Extended error information shall be vendor specific.
State shall contain a value indicating the state of the device when command was written to the Command
register or the reset occurred as described in Table 44.
Table 44 − State field values
Value
State
x0h
Unknown
x1h
Sleep
x2h
Standby
x3h
Active/Idle with BSY cleared to zero
x4h
Executing SMART off-line or self-test
x5h-xAh
Reserved
xBh-xFh
Vendor unique
The value of x is vendor specific and may be different for each state.
Sleep indicates the reset for which the error is being reported was received when the device was in the Sleep
mode.
Standby indicates the command or reset for which the error is being reported was received when the device was
in the Standby mode.
Active/Idle with BSY cleared to zero indicates the command or reset for which the error is being reported was
received when the device was in the Active or Idle mode and BSY was cleared to zero.
Executing SMART off-line or self-test indicates the command or reset for which the error is being reported was
received when the device was in the process of executing a SMART off-line or self-test.
Life timestamp shall contain the power-on lifetime of the device in hours when command completion occurred.
Page 211
T13/1321D revision 3
8.41.6.8.2.4 Device error count
The device error count field shall contain the total number of errors attributable to the device that have been
reported by the device during the life of the device. These errors shall include UNC errors, IDNF errors for which
the address requested was valid, servo errors, write fault errors, etc. This count shall not include errors
attributed to the receipt of faulty commands such as commands codes not implemented by the device or
requests with invalid parameters or invalid addresses. If the maximum value for this field is reached, the count
shall remain at the maximum value when additional errors are encountered and logged.
8.41.6.8.2.5 Data structure checksum
The data structure checksum is the two's complement of the sum of the first 511 bytes in the data structure.
Each byte shall be added with unsigned arithmetic, and overflow shall be ignored. The sum of all 512 bytes will
be zero when the checksum is correct. The checksum is placed in byte 511.
8.41.6.8.3 Self-test log sector
Table 45 defines the 512 bytes that make up the SMART self-test log sector. All multi-byte fields shown in this
structure follow the byte ordering described in 3.2.9.
Byte
0-1
2-25
26-49
.....
482-505
506-507
508
509-510
511
Table 45 − Self-test log data structure
Descriptions
Self-test log data structure revision number
First descriptor entry
Second descriptor entry
............
Twenty-first descriptor entry
Vendor specific
Self-test index
Reserved
Data structure checksum
This log is viewed as a circular buffer. The first entry shall begin at byte 2, the second entry shall begin at byte
26, and so on until the twenty-second entry, that shall replace the first entry. Then, the twenty-third entry shall
replace the second entry, and so on. If fewer than 21 self-tests have been performed by the device, the unused
descriptor entries shall be filled with zeroes.
8.41.6.8.3.1 Self-test log data structure revision number
The value of the self-test log data structure revision number shall be 0001h.
8.41.6.8.3.2 Self-test log descriptor entry
The content of the self-test descriptor entry is shown in Table 46.
Page 212
T13/1321D revision 3
Byte
n
n+1
n+2
n+3
n+4
n+5
n+6
n+7
n+8
n+9 - n+23
Table 46 − Self-test log descriptor entry
Descriptions
Content of the Sector Number register.
Content of the self-test execution status byte.
Life timestamp (least significant byte).
Life timestamp (most significant byte).
Content of the self-test failure checkpoint byte.
Failing LBA (least significant byte).
Failing LBA (next least significant byte).
Failing LBA (next most significant byte).
Failing LBA (most significant byte).
Vendor specific.
Content of the Sector Number register shall be the content of the Sector Number register when the nth self-test
subcommand was issued (see 8.41.4.8).
Content of the self-test execution status byte shall be the content of the self-test execution status byte when
the nth self-test was completed (see 8.41.5.8.2).
Life timestamp shall contain the power-on lifetime of the device in hours when the nth self-test subcommand
was completed.
Content of the self-test failure checkpoint byte shall be the content of the self-test failure checkpoint byte when
the nth self-test was completed.
The failing LBA shall be the LBA of the uncorrectable sector that caused the test to fail. If the device
encountered more than one uncorrectable sector during the test, this field shall indicate the LBA of the first
uncorrectable sector encountered. If the test passed or the test failed for some reason other than an
uncorrectable sector, the value of this field is undefined.
8.41.6.8.3.3 Self-test index
The self-test index shall point to the most recent entry. Initially, when the log is empty, the index shall be set to
zero. It shall be set to one when the first entry is made, two for the second entry, etc., until the 22nd entry,
when the index shall be reset to one.
8.41.6.8.3.4 Data structure checksum
The data structure checksum is the two's complement of the sum of the first 511 bytes in the data structure.
Each byte shall be added with unsigned arithmetic, and overflow shall be ignored. The sum of all 512 bytes is
zero when the checksum is correct. The checksum is placed in byte 511.
8.41.7 SMART RETURN STATUS
8.41.7.1 Command code
B0h with a Feature register value of DAh.
8.41.7.2 Feature set
SMART feature set.
−
−
Mandatory when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
Page 213
T13/1321D revision 3
8.41.7.3 Protocol
Non-data command (see 9.4).
8.41.7.4 Inputs
The Features register shall be set to DAh. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
obs
na
obs
4
3
DAh
na
na
4Fh
C2h
DEV
na
B0h
2
1
0
na
na
na
Device/Head register DEV shall indicate the selected device.
8.41.7.5 Normal outputs
If the device has not detected a threshold exceeded condition, the device sets the Cylinder Low register to 4Fh
and the Cylinder High register to C2h. If the device has detected a threshold exceeded condition, the device
sets the Cylinder Low register to F4h and the Cylinder High register to 2Ch.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
obs
BSY
na
DRDY
obs
DF
4
3
na
na
na
4Fh or F4h
C2h or 2Ch
DEV
na
na
DRQ
2
1
0
na
na
na
na
na
ERR
Cylinder Low 4Fh if threshold not exceeded, F4h if threshold exceeded.
Cylinder High C2h if threshold not exceeded, 2Ch if threshold exceeded.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.7.6 Error outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Cylinder
Low, or Cylinder High registers are invalid, the device shall return command aborted.
Page 214
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported, if SMART is disabled, or if the input
register values are invalid. ABRT may be set to one if the device is not able to complete the
action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.41.7.7 Prerequisites
DRDY set to one. SMART enabled.
8.41.7.8 Description
This command is used to communicate the reliability status of the device to the host at the host’s request. If a
threshold exceeded condition is not detected by the device, the device shall set the Cylinder Low register to
4Fh and the Cylinder High register to C2h. If a threshold exceeded condition is detected by the device, the
device shall set the Cylinder Low register to F4h and the Cylinder High register to 2Ch.
8.41.8 SMART SAVE ATTRIBUTE VALUES
8.41.8.1 Command code
B0h with a Feature register value of D3h.
8.41.8.2 Feature set
SMART feature set.
−
−
Optional and not recommended when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.8.3 Protocol
Non-data command (see 9.4).
8.41.8.4 Inputs
The Features register shall be set to D3h. The Cylinder Low register shall be set to 4Fh. The Cylinder High
register shall be set to C2h.
Page 215
T13/1321D revision 3
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
D3h
na
na
4Fh
C2h
obs
na
obs
DEV
B0h
Device/Head register DEV shall indicate the selected device.
8.41.8.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.8.6 Error outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Cylinder
Low, or Cylinder High registers are invalid, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported, if SMART is disabled, or if the input
register values are invalid. ABRT may be set to one if the device is not able to complete the
action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
Page 216
T13/1321D revision 3
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero..
ERR shall be set to one if an Error register bit is set to one.
8.41.8.7 Prerequisites
DRDY set to one. SMART enabled.
8.41.8.8 Description
This command causes the device to immediately save any updated attribute values to the device’s non-volatile
memory regardless of the state of the attribute autosave timer.
8.41.9 SMART WRITE LOG
8.41.9.1 Command code
B0h with the content of the Features register equal to D6h.
8.41.9.2 Feature set
SMART feature set.
−
−
Optional when the SMART feature set is implemented.
Use prohibited when the PACKET Command feature set is implemented.
8.41.9.3 Protocol
PIO data-out (see 9.6).
8.41.9.4 Inputs
The Features register shall be set to D6h. The Sector Count register shall indicate the number of sectors that
shall be written to the log number indicated by the Sector Number register. The Cylinder Low register shall be
set to 4Fh. The Cylinder High register shall be set to C2h.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
D6h
Number of sectors to be written
Log sector address
4Fh
C2h
obs
DEV
na
B0h
1
0
Sector count – Indicates the number of sectors that shall be written to the indicated log. If the number is
greater than the number indicated in the “Log directory” (which is available in Log number zero), the drive
shall return command aborted. The log transferred to the drive shall be stored by the drive starting at the
first sector in the indicated log.
Sector number - Indicates the log to be written as described in Table 38. If this command is implemented, all
address values defined as host vendor specific shall be implemented. These host vendor specific logs may
be used by the host to store any data desired. Support for device vendor specific logs is optional. If the host
attempts to write to a read only (RO) log address, the device shall return command aborted.
Page 217
T13/1321D revision 3
Device/Head register DEV shall indicate the selected device.
8.41.9.5 Normal outputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
DRQ
na
1
0
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.41.9.6 Error outputs
If the device does not support this command, if SMART is disabled, or if the values in the Features, Sector
Number, Sector Count, Cylinder Low, or Cylinder High registers are invalid, the device shall return command
aborted. If the host attempts to write to a read only (RO) log address, the device shall return command aborted.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
IDNF
3
na
2
ABRT
DRQ
na
1
na
0
obs
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Na
Error register IDNF shall be set to one if SMART log sector’s ID field could not be found.
ABRT shall be set to one if this command is not supported, if SMART is not enabled, if the log sector
address is not implemented, or if other register values are invalid. ABRT may be set to one if
the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one indicating that the device is capable of receiving any command.
DF (Device Fault) shall be set to one indicating that a device fault has occurred.
DRQ shall be cleared to zero indicating that there is no data to be transferred.
ERR shall be set to one if any Error register bit is set to one.
Page 218
T13/1321D revision 3
8.41.9.7 Prerequisites
DRDY set to one. SMART enabled.
8.41.9.8 Description
This command writes an indicated number of 512 byte data sectors to the indicated log.
8.42 STANDBY
8.42.1 Command code
E2h
8.42.2 Feature set
Power Management feature set.
−
−
Power Management feature set is mandatory when power management is not implemented by a
PACKET power management feature set.
This command is mandatory when the Power Management feature set is implemented when the
PACKET Command feature set is not implemented.
8.42.3 Protocol
Non-data command (see 9.4).
8.42.4 Inputs
The value in the Sector Count register when the STANDBY command is issued shall determine the time period
programmed into the Standby timer. Table 23 defines these values.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
na
5
4
3
2
na
Time period value
na
na
na
obs DEV na
na
E2h
1
0
na
na
Device/Head register DEV shall indicate the selected device.
Page 219
T13/1321D revision 3
8.42.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.42.6 Error outputs
The device shall return command aborted if the device does not support the Power Management feature set.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
DRQ
na
Error register ABRT shall be set to one if the Power Management feature set is not supported. ABRT may be set to
one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.42.7 Prerequisites
DRDY set to one.
8.42.8 Description
This command causes the device to enter the Standby mode.
Page 220
T13/1321D revision 3
If the Sector Count register is non-zero then the Standby timer shall be enabled. The value in the Sector Count
register shall be used to determine the time programmed into the Standby timer (see Table 23).
If the Sector Count register is zero then the Standby timer is disabled.
8.43 STANDBY IMMEDIATE
8.43.1 Command code
E0h
8.43.2 Feature set
Power Management feature set.
−
−
Power Management feature set is mandatory when power management is not implemented by a
PACKET power management feature set.
This command is mandatory when the Power Management feature set is implemented.
8.43.3 Protocol
Non-data command (see 9.4).
8.43.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
E0h
Device/Head register DEV shall indicate the selected device.
8.43.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
Page 221
T13/1321D revision 3
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.43.6 Error outputs
The device shall return command aborted if the device does not support the Power Management feature set.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
3
na
2
ABRT
DRQ
na
1
na
0
na
na
ERR
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if the Power Management feature set is not supported. ABRT may be set to
one if the device is not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.43.7 Prerequisites
DRDY set to one.
8.43.8 Description
This command causes the device to immediately enter the Standby mode.
8.44 WRITE BUFFER
8.44.1 Command code
E8h
8.44.2 Feature set
General feature set
−
−
Optional for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.44.3 Protocol
PIO data-out (see 9.6).
Page 222
T13/1321D revision 3
8.44.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
5
4
3
2
1
0
na
na
na
na
na
na
na
na
na
obs
na
obs
DEV
E8h
Device/Head register DEV shall indicate the selected device.
8.44.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
3
na
2
ABRT
1
na
0
na
DRQ
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.44.6 Error outputs
The device shall return command aborted if the command is not supported.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
6
na
5
na
4
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
na
Error register ABRT shall be set to one if this command is not supported. ABRT may be set to one if the device is
not able to complete the action requested by the command.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
Page 223
T13/1321D revision 3
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.44.7 Prerequisites
DRDY set to one.
8.44.8 Description
This command enables the host to write the contents of one sector in the device’s buffer.
The READ BUFFER and WRITE BUFFER commands shall be synchronized within the device such that
sequential WRITE BUFFER and READ BUFFER commands access the same 512 bytes within the buffer.
8.45 WRITE DMA
8.45.1 Command code
CAh
8.45.2 Feature set
General feature set
Mandatory for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.45.3 Protocol
DMA (see 9.7).
8.45.4 Inputs
The Cylinder Low, Cylinder High, Device/Head, and Sector Number specify the starting sector address to be
written. The Sector Count register specifies the number of sectors to be transferred.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
CAh
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High Page 224
T13/1321D revision 3
starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
8.45.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.45.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred. The amount of data transferred is indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
ICRC
obs
BSY
6
WP
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
obs
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
DRQ
na
na
ERR
Error register ICRC shall be set to one if an interface CRC error has occurred during an Ultra DMA data transfer. The
content of this bit is not applicable for Multiword DMA transfers.
WP shall be set to one if the media in a removable media device is write protected.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
Page 225
T13/1321D revision 3
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with address of first unrecoverable error.
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.45.7 Prerequisites
DRDY set to one. The host shall initialize the DMA channel.
8.45.8 Description
The WRITE DMA command allows the host to write data using the DMA data transfer protocol.
8.46 WRITE DMA QUEUED
8.46.1 Command code
CCh
8.46.2 Feature set
Overlapped feature set
−
−
Mandatory for devices implementing the Overlapped feature set but not implementing the PACKET
Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.46.3 Protocol
DMA QUEUED (see 9.9).
8.46.4 Inputs
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
obs
6
LBA
5
4
3
Sector Count
2
1
0
Tag
na
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
CCh
Features number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Page 226
T13/1321D revision 3
Sector count if the device supports command queuing, bits (7:3) contain the Tag for the command being delivered. A
Tag value may be any value between 0 and 31 regardless of the queue depth supported. If queuing is
not supported, this field is not applicable.
Sector number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
8.46.5 Normal outputs
8.46.5.1 Data transmission
Data transfer may occur after receipt of the command or may occur after the receipt of a SERVICE command.
When the device is ready to transfer data requested by a data transfer command, the device sets the following
register content to initiate the data transfer.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
na
na
na
na
CHK
na
Tag
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
SERV
na
DRQ
Interrupt reason register Tag -This field contains the command Tag for the command. A Tag value may be any value between 0
and 31 regardless of the queue depth supported. If the device does not support command queuing
or overlap is disabled, this field is not applicable.
REL - Shall be cleared to zero.
I/O - Shall be cleared to zero indicating the transfer is from the host.
C/D - Shall be cleared to zero indicating the transfer of data.
Device/Head register DEV - Shall indicate the selected device.
Status register BSY - Shall be cleared to zero.
DRDY - Shall be set to one.
DF (Device Fault) - Shall be cleared to zero.
SERV (Service) - Shall be set to one if another command is ready to be serviced.
DRQ - Shall be set to one.
CHK - Shall be cleared to zero.
8.46.5.2 Bus release
If the device performs a bus release before transferring data for this command, the register content upon
performing a bus release shall be as shown below.
Page 227
T13/1321D revision 3
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
ERR
na
Tag
obs
BSY
na
DRDY
obs
DF
na
na
na
DEV
SERV
na
DRQ
na
Sector Count register Tag - If the device supports command queuing, this field shall contain the Tag of the command being
bus released. If the device does not support command queuing, this field shall be zeros.
REL bit shall be set indicating that the device has bus released an overlap command.
I/O shall be cleared to zero.
C/D shall be cleared to zero.
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating bus release.
DRDY shall be set to one.
SERV (Service) shall be cleared to zero if no other queued command is ready for service. SERV shall
be set to one when another queued command is ready for service. This bit shall be set to one
when the device has prepared this command for service.
DF (Device Fault) shall be cleared to zero.
DRQ bit shall be cleared to zero.
ERR bit shall be cleared to zero.
8.46.5.3 Service request
When the device is ready to transfer data or complete a command after the command has performed a bus
release, the device shall set the SERV bit and not change the state of any other register bit (see 6.9). When
the SERVICE command is received, the device shall set outputs as described in data transfer, command
completion, or error outputs depending on the service the device requires.
8.46.5.4 Command completion
When the transfer of all requested data has occurred without error, the register content shall be as shown
below.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
REL
I/O
C/D
na
ERR
00h
Tag
obs
BSY
na
DRDY
obs
DF
na
na
na
DEV
SERV
DRQ
na
na
Sector Count register Tag - If the device supports command queuing, this field shall contain the Tag of the completed
command. If the device does not support command queuing, this field shall be zeros.
REL shall be cleared to zero.
I/O shall be set to one.
C/D shall be set to one.
Page 228
T13/1321D revision 3
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
SERV (Service) shall be cleared to zero when no other queued command is ready for service. SERV
shall be set to one when another queued command is ready for service.
DF (Device Fault) shall be cleared to zero.
DRQ bit shall be cleared to zero.
ERR bit shall be cleared to zero.
8.46.6 Error outputs
The Sector Count register contains the Tag for this command if the device supports command queuing. The
device shall return command aborted if the command is not. The device shall return command aborted if the
device supports command queuing and the Tag is invalid. An unrecoverable error encountered during the
execution of this command results in the termination of the command and the Command Block registers
contain the sector where the first unrecoverable error occurred. If a queue existed, the unrecoverable error shall
cause the queue to abort. The device may remain BSY for some time when responding to these errors.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
ICRC
7
6
WP
obs
BSY
na
DRDY
5
4
3
2
1
0
IDNF
MC
MCR
ABRT NM
na
Tag
REL
I/O
C/D
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
SERV
DRQ
na
na
ERR
Error register ICRC shall be set to one if an interface CRC error has occurred during an Ultra DMA data transfer. The
content of this bit is not applicable for Multiword DMA transfers.
WP shall be set to one if the media in a removable media device is write protected.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Count register Tag - If the device supports command queuing, this field shall contain the Tag of the completed
command. If the device does not support command queuing, this field shall be zeros.
REL shall be cleared to zero.
I/O shall be set to one.
C/D shall be set to one.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Page 229
T13/1321D revision 3
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
SERV (Service) shall be cleared to zero when no other queued command is ready for service. SERV
shall be set to one when another queued command is ready for service.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.46.7 Prerequisites
DRDY set to one. The host shall initialize the DMA channel.
8.46.8 Description
This command executes in a similar manner to a WRITE DMA command. The device may perform a bus
release the bus or may execute the data transfer without performing a bus release if the data is ready to
transfer.
If the device performs a bus release, the host shall reselect the device using the SERVICE command.
Once the data transfer is begun, the device shall not perform a bus release until the entire data transfer has
been completed.
8.47 WRITE MULTIPLE
8.47.1 Command code
C5h
8.47.2 Feature set
General feature set
− Mandatory for devices not implementing the PACKET Command feature set.
− Use prohibited for devices implementing the PACKET Command feature set.
8.47.3 Protocol
PIO data-out (see 9.6).
8.47.4 Inputs
The Cylinder Low, Cylinder High, Device/Head, and Sector Number specify the starting sector address to be
written. The Sector Count register specifies the number of sectors to be transferred.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
Page 230
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
C5h
T13/1321D revision 3
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
8.47.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.47.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred. The amount of data transferred is indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
WP
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
DRQ
na
na
ERR
Error register WP shall be set to one if the media in a removable media device is write protected.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
Page 231
T13/1321D revision 3
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete action requested by the command. ABRT shall be set to one if an address outside
of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.47.7 Prerequisites
DRDY set to one. If bit 8 of IDENTIFY DEVICE word 59 is cleared to zero, a successful SET MULTIPLE MODE
command shall proceed a WRITE MULTIPLE command.
8.47.8 Description
The WRITE MULTIPLE command performs the same as the WRITE SECTOR(S) command except that the
device does not a) set DF, as required, b) set ERR and the bits in the Error register, as required, c) clear DRQ
and BSY, and d) assert INTRQ until data is transferred for all sectors in the block. Data for all of the other
sectors in the block are transferred without the device asserting INTRQ. In addition, the DRQ qualification of
the transfer is required only before the first sector of a block, not before each sector of the block.
The number of sectors per block is defined by a successful SET MULTIPLE command. If no successful SET
MULTIPLE command has been issued, the block is defined by the device’s default value for number of sectors
per block as defined in bits 0-7 in word 47 in the IDENTIFY DEVICE information.
If bit 8 is set to one and bits 0-7 are cleared to zero in word 59 in the IDENTIFY DEVICE information, then the
block count of sectors to be transferred without intervening interrupts shall be programmed by the SET
MULTIPLE MODE command before issuing the first READ MULTIPLE or WRITE MULTIPLE command after a
power-on or hardware reset. If bit 8 in word 1 is set to one and bits 0-7 are not cleared to zero in word 59 in the
IDENTIFY DEVICE information, then the block count of sectors to be transferred without intervening interrupts
may be reprogrammed by the SET MULTIPLE MODE command before issuing the next READ MULTIPLE or
WRITE MULTIPLE command.
When the WRITE MULTIPLE command is issued, the Sector Count register contains the number of sectors
(not the number of blocks) requested.
If the number of requested sectors is not evenly divisible by the block count, as many full blocks as possible
are transferred, followed by a final, partial block transfer. The partial block transfer is for n sectors, where:
n = Remainder (sector count/ block count).
If the WRITE MULTIPLE command is received when WRITE MULTIPLE commands are disabled, the Write
Multiple operation shall be rejected with command aborted.
Page 232
T13/1321D revision 3
Device errors encountered during WRITE MULTIPLE commands are posted after the attempted device write of
the block or partial block transferred. The command ends with the sector in error, even if the error was in the
middle of a block. Subsequent blocks are not transferred in the event of an error.
The contents of the Command Block Registers following the transfer of a data block that had a sector in error
are undefined. The host should retry the transfer as individual requests to obtain valid error information.
Interrupt pending is set when the DRQ bit is set to one at the beginning of each block or partial block.
8.48 WRITE SECTOR(S)
8.48.1 Command code
30h
8.48.2 Feature set
General feature set
−
−
Mandatory for devices not implementing the PACKET Command feature set.
Use prohibited for devices implementing the PACKET Command feature set.
8.48.3 Protocol
PIO data-out (see 9.6).
8.48.4 Inputs
The Cylinder Low, Cylinder High, Device/Head, and Sector Number specify the starting sector address to be
written. The Sector Count register specifies the number of sectors to be transferred.
Register
Features
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Command
7
6
obs
LBA
5
4
3
2
1
0
na
Sector count
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs DEV
Head number or LBA
30h
Sector Count number of sectors to be transferred. A value of 00h indicates that 256 sectors are to be transferred.
Sector Number starting sector number or LBA address bits (7:0).
Cylinder Low starting cylinder number bits (7:0) or LBA address bits (15:8).
Cylinder High starting cylinder number bits (15:8) or LBA address bits (23:16).
Device/Head bit 6 set to one if LBA address, cleared to zero if CHS address.
DEV shall indicate the selected device.
bits (3:0) starting head number or LBA address bits (27:24).
Page 233
T13/1321D revision 3
8.48.5 Normal outputs
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
6
5
4
3
2
1
0
na
DRQ
na
na
na
na
na
ERR
na
na
na
na
na
obs
BSY
na
DRDY
obs
DF
DEV
na
Device/Head register DEV shall indicate the selected device.
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be cleared to zero.
DRQ shall be cleared to zero.
ERR shall be cleared to zero.
8.48.6 Error outputs
An unrecoverable error encountered during the execution of this command results in the termination of the
command. The Command Block registers contain the address of the sector where the first unrecoverable error
occurred. The amount of data transferred is indeterminate.
Register
Error
Sector Count
Sector Number
Cylinder Low
Cylinder High
Device/Head
Status
7
na
obs
BSY
6
WP
na
DRDY
5
MC
4
IDNF
3
MCR
2
ABRT
1
NM
0
na
na
Sector number or LBA
Cylinder low or LBA
Cylinder high or LBA
obs
DEV
Head number or LBA
DF
na
DRQ
na
na
ERR
Error register WP shall be set to one if the media in a removable media device is write protected.
MC shall be set to one if the media in a removable media device changed since the issuance of the last
command. The device shall clear the device internal media change detected state.
IDNF shall be set to one if a user-accessible address could not be found and after an unsuccessful
INITIALIZE DEVICE PARAMETERS command until a valid CHS translation is established (see
8.16.8). IDNF shall be set to one if an address outside of the range of user-accessible
addresses is requested if command aborted is not returned.
MCR shall be set to one if a media change request has been detected by a removable media device.
This bit is only cleared by a GET MEDIA STATUS or a media access command.
ABRT shall be set to one if this command is not supported or if an error, including an ICRC error, has
occurred during an Ultra DMA data transfer. ABRT may be set to one if the device is not able
to complete the action requested by the command. ABRT shall be set to one if an address
outside of the range of user-accessible addresses is requested if IDNF is not set to one.
NM shall be set to one if no media is present in a removable media device.
Sector Number, Cylinder Low, Cylinder High, Device/Head shall be written with the address of first unrecoverable error.
DEV shall indicate the selected device.
Page 234
T13/1321D revision 3
Status register BSY shall be cleared to zero indicating command completion.
DRDY shall be set to one.
DF (Device Fault) shall be set to one if a device fault has occurred.
DRQ shall be cleared to zero.
ERR shall be set to one if an Error register bit is set to one.
8.48.7 Prerequisites
DRDY set to one.
8.48.8 Description
This command writes from 1 to 256 sectors as specified in the Sector Count register. A sector count of 0
requests 256 sectors.
9
Protocol
Commands are grouped into different classes according to the protocol followed for command execution. The
command classes with their associated protocol are defined in state diagrams in this clause, one state diagram
for host actions and a second state diagram for device actions. Figure 12 shows the overall relationship of the
host protocol state diagrams. Figure 13 shows the overall relationship of the device protocol state diagrams.
State diagrams defining these protocols are not normative descriptions of implementations, they are normative
descriptions of externally apparent device or host behavior. Different implementations are allowed. See 3.2.7 for
state diagram conventions.
A device shall not timeout any activity when waiting for a response from the host.
Page 235
T13/1321D revision 3
HHR: Host power on
or Hardware Reset
Power on
HI:HIO: Host bus Idle
RESET asserted
HSR: Host Software
Reset
Reset completed
Command written
Command completed
SRST
Service return
Bus released
HND: Host Non-Data
command
HPIOI: Host PIO dataIn command
HPIOO: Host PIO dataOut command
HDMA: Host DMA
command
HED: Host EXECUTE
DEVICE DIAGNOSTIC command
HDR: Host DEVICE
RESET command
HP: Host PACKET nondata & PIO data command
HPD: Host PACKET
DMA command
Figure 12 − Overall host protocol state sequence
Page 236
HDMAQ: Host DMA
QUEUED command
T13/1321D revision 3
DHR: Device power on
or Hardware Reset
Power on
DI:DIO: Device bus Idle
RESET asserted
Reset completed
D0SR: Device 0 Software Reset
SRST & Device 0
Command written
Command completed
Service return
Bus released
D1SR: Device 1 Software Reset
SRST & Device 1
DND: Device Non-Data command
DPIOO: Device PIO data-Out
command
D0ED: Device 0 EXECUTE
DEVICE DIAGNOSTIC command
DDR: Device DEVICE RESET
command
DPD: Device PACKET DMA
command
DPIOI: Device PIO data-In
command
DDMA: Device DMA
command
D1ED: Device 1 EXECUTE
DEVICE DIAGNOSTIC cmnd
DP: Device PACKET nondata & PIO data command
DDMAQ: Device DMA
QUEUED command
Figure 13 − Overall device protocol state sequence
Page 237
T13/1321D revision 3
9.1 Power-on and hardware reset protocol
This clause describes the protocol for processing of power-on and hardware resets.
If the host asserts RESET-, regardless of the power management mode, the device shall execute the hardware
reset protocol. If the host reasserts RESET- before a device has completed the power-on or hardware reset
protocol, then the device shall restart the protocol from the begining.
The host should not set the SRST bit to one in the Device Control register or issue a DEVICE RESET
command while the BSY bit is set to one in either device Status register as a result of executing the power-on
or hardware reset protocol. If the host sets the SRST bit in the Device Control register to one or issues a
DEVICE RESET command before devices have completed execution of the power-on or hardware reset
protocol, then the devices shall ignore the software reset or DEVICE RESET command.
A host should issue an IDENTIFY DEVICE and/or IDENTIFY PACKET DEVICE command after the power-on or
hardware reset protocol has completed to determine the current status of features implemented by the
device(s).
Figure 14 and the text following the figure decribes the power-on or hardware reset protocol for the host. Figure
15 and the text following the figure decribes the power-on or hardware reset protocol for the devices.
HHR0: Assert_RESETPower on or
hardware reset
required
HHR1: Negate_wait
RESET- asserted
>25 µs (t=0)
HHR0:HHR1
HHR2: Check_status
t > 2 ms
HHR1:HHR2
BSY = 0
HHR2:HI0
Host_idle
BSY = 1
xx:HHR0
HHR2:HHR2
Figure 14 − Host power-on or hardware reset state diagram
HHR0: Assert_RESET- State: This state is entered at power-on or when the host recognizes that a
hardware reset is required.
When in this state, the host asserts RESET-. The host shall remain in this state with RESET- asserted for at
least 25 µs.
Transition HHR0:HHR1: When the host has had RESET- asserted for at least 25 µs, the host shall make a
transition to the HHR1: Negate_wait state.
HHR1: Negate_wait State: This state is entered when RESET- has been asserted for at least 25 µs.
When in this state, the host shall negate RESET-. The host shall remain in this state for at least 2 ms after
negating RESET-. If the host tests CBLID- it shall do so at this time.
Transition HHR1:HHR2: When RESET- has been negated for at least 2 ms, the host shall make a transition to
the HHR2: Check_status state.
HHR2: Check_status State: This state is entered when RESET- has been negated for at least 2 ms.
When in this state the host shall read the Status or Alternate Status register.
Page 238
T13/1321D revision 3
Transition HHR2:HHR2: When BSY is set to one, the host shall make a transition to the HHR2: Check_status
state.
Transition HHR2:HI0: When BSY is cleared to zero, the host shall make a transition to the HI0: Host_idle
state (see Figure 19). If status indicates that an error has occurred, the host shall take appropriate error
recovery action.
DHR0: RESETPDIAG-=X, DASP-=X, BSY=x
DHR1: Release_bus
PDIAG-=R, DASP-=X, BSY=1
Bus released & Device 1
RESET- negated (t=0)
DHR0:DHR1
DHR1:D1HR0
Bus released & Device 0
D0HR0: DASP-_wait
DHR1:D0HR0
PDIAG- =R, DASP- =R, BSY=1
D1HR0: Set_DASPt > 1 ms
PDIAG- =R, DASP- =R, BSY=1
D0HR0:D0HR1
RESET- asserted
xx:DHR0
D0HR1: Sample_DASPPDIAG- =R, DASP- =R, BSY=1
DASP- asserted
D1HR0:D1HR1
Sample DASPD0HR1:D0HR1
DASP- asserted
D0HR1:D0HR2
D1HR1: Set_status
PDIAG- =R, DASP- =A, BSY=1
Status set, passed diagnostic
Device_idle_NS
D1HR1:DI2
BSY=0, PDIAG-=A
450 ms < t ≤ 5 s
D0HR1:D0HR3
Clear bit 7
Status set, failed diagnostic
Device_idle_NS
D1HR1:DI2
BSY=0, PDIAG-=N
D0HR2: Sample_PDIAGPDIAG- =R, DASP- =R, BSY=1
D0HR3: Set_status
PDIAG- =R, DASP- =R, BSY=1
Resample PDIAGD0HR2:D0HR2
PDIAG- asserted
D0HR2a:D0HR3
Clear bit 7
Status set
D0HR3:DI1
BSY=0
Device_idle_S
t > 31 s
D0HR2b:D0HR3
Set bit 7
BSY
v
DRQ
0
REL
0
SERV
0
C/D
0
I/O
0
INTRQ
R
DMARQ
R
PDIAGV
DASPV
Figure 15 − Device power-on or hardware reset state diagram
DHR0: RESET State: This state is entered when a valid assertion of the RESET- signal is recognized.
The device shall not recognize a RESET- assertion shorter than 20 ns as valid. Devices may recognize a
RESET- assertion greater that 20 ns as valid and shall recognize a RESET- assertion equal to or greater than
25 µs as valid.
Transition DHR0:DHR1: When a valid RESET- signal is negated, the device shall make a transition to the
DHR1: Release_Bus state.
Page 239
T13/1321D revision 3
DHR1: Release_bus State: This state is entered when a valid RESET- signal is negated.
When in this state, the device shall release bus signals PDIAG-, INTRQ, IORDY, DMARQ, and DD(15:0) and
shall set BSY to one within 400 ns after entering this state. The device shall determine if the device is Device 0
or Device 1 by checking the jumper, switch, or CSEL.
Transition DHR1:D0HR0: When the device has determined that the device is Device 0, has released the bus
signals, and has set BSY to one, then the device shall make a transition to the D0HR0: DASP-_wait state.
Transition DHR1:D1HR0: When the device has determined that the device is Device 1, has released the bus
signals, and has set BSY to one, then the device shall make a transition to the D1HR0: Set_DASP- state.
D0HR0: DASP-_wait State: This state is entered when the device has released the bus signals, set
BSY to one, and determined that the device is Device 0.
When in this state, the device shall release DASP- within 1 ms of the negation of RESET-.
Transition D0HR0:D0HR1: When at least 1 ms has elapsed since the negation of RESET-, the device shall
make a transition to the D0HR1: Sample_DASP- state.
D0HR1: Sample_DASP- State: This state is entered when at least 1 ms has elapsed since the
negation of RESET-.
When in this state, the device should begin performing the hardware initialization and self-diagnostic testing.
This may revert the device to the default condition (the device’s settings may now be different than they were
before the host asserted RESET-). All Ultra DMA modes shall be disabled.
When in this state, the device shall sample the DASP- signal.
Transition D0HR1:D0HR2: When the sample indicates that DASP- is asserted, the device shall make a
transition to the D0HR2: Sample_PDIAG- state.
Transition D0HR1:D0HR1: When the sample indicates that DASP- is negated and
elapsed since the negation of RESET-, then the device shall make a transition to the
state. When the sample indicates that DASP- is negated and greater than 450 ms
elapsed since the negation of RESET-, then the device may make a transition to the
state.
less than 450 ms have
D0HR1: Sample_DASPbut less than 5 s have
D0HR1: Sample_DASP-
Transition D0HR1:D0HR3: When the sample indicates that DASP- is negated and 5 s have elapsed since the
negation of RESET-, then the device shall clear bit 7 in the Error register and make a transition to the D0HR3:
Set_status state. When the sample indicates that DASP- is negated and greater than 450 ms but less than 5 s
have elapsed since the negation of RESET-, then the device may clear bit 7 in the Error register and make a
transition to the D0HR3: Set_status state.
D0HR2: Sample_PDIAG- State: This state is entered when the device has recognized that DASP- is
asserted.
When in this state, the device shall sample the PDIAG- signal.
Transition D0HR2a:D0HR3: When the sample indicates that PDIAG- is asserted, the device shall clear bit 7 in
the Error register and make a transition to the D0HR3: Set_status state.
Page 240
T13/1321D revision 3
Transition D0HR2b:D0HR3: When the sample indicates that PDIAG- is not asserted and 31 s have elapsed
since the negation of RESET-, then the device shall set bit 7 in the Error register and make a transition to the
D0HR3: Set_state state.
Transition D0HR2:D0HR2: When the sample indicates that PDIAG- is not asserted and less than 31 s have
elapsed since the negation of RESET-, then the device shall make a transition to the D0HR2: Sample_PDIAGstate.
D0HR3: Set_status State: This state is entered when Bit 7 in the Error register has been set or cleared.
When in this state the device shall complete the hardware initialization and self-diagnostic testing begun in the
Sample DASP- state if not already completed.
The diagnostic code shall be placed in bits 6-0 of the Error register (see Table 19). The device shall set the
signature values (see 9.12). The device shall clear the SRST bit to zero in the Device Control register if set set
to one. The content of the Features register is undefined. The device shall set word 93 in the IDENTIFY DEVICE
or IDENTIFY PACKET DEVICE response (see 8.12.52).
If the device does not implement the PACKET command feature set, the device shall clear bits 3, 2, and 0 in
the Status register to zero.
If the device implements the PACKET command feature set, the device shall clear bits 6, 5, 4, 3, 2, and 0 in
the Status register to zero. The device shall return the operating modes to their specified initial conditions.
MODE SELECT conditions shall be restored to their last saved values if saved values have been established.
MODE SELECT conditions for which no values have been saved shall be returned to their default values.
Transition D0HR3:DI1: When hardware initialization and self-diagnostic testing is completed and the status
has been set, the device shall clear BSY to zero and make a transition to the DI1: Device_idle_S state (see
Figure 21).
D1HR0: Set_DASP- State: This state is entered when the device has released the bus, set BSY to one,
and determined that the device is Device 1.
When in this state, the device shall release PDIAG- within 1 ms and assert DASP- within 400 ms of the
negation of RESET-.
When in this state, the device should begin execution of the hardware initialization and self-diagnostic testing.
The device may revert to the default condition (the device’s settings may now be in different conditions than
they were before RESET- was asserted by the host). All Ultra DMA modes shall be disabled.
Transition D1HR0:D1HR1: When DASP- has been asserted, the device shall make a transition to the D1HR1:
Set_status state.
D1HR1: Set_status State: This state is entered when the device has asserted DASP-.
When in this state the device shall complete any hardware initialization and self-diagnostic testing begun in the
Set DASP- state if not already completed. The diagnostic code shall be placed in the Error register (see Table
19). If the device passed self-diagnostics, the device shall assert PDIAG-. The device shall set word 93 in the
IDENTIFY DEVICE or IDENTIFY PACKET DEVICE response (see 8.12.52).
All actions required in this state shall be completed in ≤ 30 s.
The device shall set the signature values (see 9.12). The content of the Features register is undefined. The
device shall clear the SRST bit to zero in the Device Control register if set set to one.
Page 241
T13/1321D revision 3
If the device does not implement the PACKET command feature set, the device shall clear bits 3, 2, and 0 in
the Status register to zero.
If the device implements the PACKET command feature set, the device shall clear bits 6, 5, 4, 3, 2, and 0 in
the Status register to zero. The device shall return the operating modes to their specified initial conditions.
MODE SELECT conditions shall be restored to their last saved values if saved values have been established.
MODE SELECT conditions for which no values have been saved shall be returned to their default values.
Transition D1HR1a:DI2: When hardware initialization and self-diagnostic testing is completed, the device
passed its diagnostics, and the status has been set, the device shall clear BSY to zero, assert PDIAG-, and
make a transition to the DI2: Device_idle_NS state (see Figure 21).
Transition D1HR1b:DI2: When hardware initialization and self-diagnostic testing is completed, the device failed
its diagnostic, and the status has been set, the device shall clear BSY to zero, negate PGIAG-, and make a
transition to the DI2: Device_idle_NS state (see Figure 21).
9.2 Software reset protocol
This clause describes the protocol for processing of software reset when the host sets SRST.
If the host sets SRST in the Device Control register to one regardless of the power management mode, the
device shall execute the software reset protocol. If the host asserts RESET- before a device has completed the
software reset protocol, then the device shall execute the hardware reset protocol from the beginning.
The host should not set the SRST bit to one in the Device Control while the BSY bit is set to one in either
device Status register as a result of executing the software reset protocol. If the host sets the SRST bit in the
Device Control register to one before devices have completed execution of the software reset protocol, then the
devices shall restart execution of the software reset protocol from the beginning. If the host issues a DEVICE
RESET command before devices have completed execution of the software reset protocol, the command shall
be ignored.
A host should issue an IDENTIFY DEVICE and/or IDENTIFY PACKET DEVICE command after the software
reset protocol has completed to determine the current status of features implemented by the device(s).
Figure 16 and the text following the figure decribe the software reset protocol for the host. Figure 17 and the
text following the figure describes the software reset protocol for Device 0. Figure 18 and the text following the
figure describes the software reset protocol for Device 1.
HSR0: Set_SRST
Software reset
required
xx:HSR0
SRST asserted
≥ 5 µs (t = 0)
HSR0:HSR1
HSR2: Check_status
HSR1: Clear_wait
t > 2 ms
HSR1:HSR2
BSY = 0
HSR2:HI0
Host_idle
BSY = 1
HSR2:HSR2
Figure 16 − Host software reset state diagram
HSR0: Set_SRST State: This state is entered when the host initiates a software reset.
When in this state, the host shall set SRST in the Device Control register to one. The SRST bit shall be written
to both devices when the Device Control register is written. The host shall remain in this state with SRST set to
Page 242
T13/1321D revision 3
one for at least 5 µs. The host shall not set SRST to one unless the bit has been cleared to zero for at least 5
µs.
Transition HSR0:HSR1: When the host has had SRST set to one for at least 5 µs, the host shall make a
transition to the HSR1: Clear_wait state.
HSR1: Clear_wait State: This state is entered when SRST has been set to one for at least 5 µs.
When in this state, the host shall clear SRST in the Device Control register to zero. The host shall remain in
this state for at least 2 ms.
Transition HSR1:HSR2: When SRST has been cleared to zero for at least 2 ms, the host shall make a
transition to the HSR2: Check_status state.
HSR2: Check_status State: This state is entered when SRST has been cleared to zero for at least 2
ms.
When in this state the host shall read the Status or Alternate Status register.
Transition HSR2:HSR2: When BSY is set to one, the host shall make a transition to the HSR2: Check_status
state.
Transition HSR2:HI0: When BSY is cleared to zero, the host shall check the ending status in the Error
register and the signature (see 9.12) and make a transition to the HI0: Host_idle state (see Figure 19).
D0SR0: SRST
PDIAG- =R, BSY=1
SRST set to one
xx:D0SR0
D0SR3: Set_status
PDIAG- =R, BSY=1
SRST cleared to zero & no Device 1 (t=0)
D0SR0:D0SR3
Clear bit 7
D0SR1: PDIAG-_wait
SRST cleared to zero & PDIAG- =R, BSY=1
Device 1 exists (t=0)
D0SR0:D0SR1
Status set
D0SR3:DI1
BSY=0
Device_
idle_S
t>1 ms
D0SR1:D0SR2
D0SR2: Sample_PDIAGPDIAG- =R, BSY=1
Resample PDIAG-
PDIAG- asserted
D0SR2a:D0SR3
Clear bit 7
D0SR2:D0SR2
t > 31 s
D0SR2b:D0SR3
Set bit 7
BSY
v
DRQ
0
REL
0
SERV
0
C/D
0
I/O
0
INTRQ
R
DMARQ
R
PDIAGV
DASPR
Figure 17 − Device 0 software reset state diagram
Page 243
T13/1321D revision 3
D0SR0: SRST State: This state is entered by Device 0 when the SRST bit is set to one in the Device
Control register.
When in this state, the device shall release PDIAG-, INTRQ, IORDY, DMARQ, and DD(15:0) within 400 ns after
entering this state. The device shall set BSY to one within 400 ns after entering this state.
If the device does not implement the PACKET command feature set, the device should begin performing the
hardware initialization and self-diagnostic testing. The device may revert to the default condition (the device’s
setting may now be in different conditions than they were before the SRST bit was set to one by the host).
However, an Ultra DMA mode setting (either enabled or disabled) shall not be affected by the host setting SRST
to one.
If the PACKET command feature set is implemented, the device may begin performing the hardware
initialization and self-diagnostic testing and the device is not expected to stop any background device activity
(e.g., immediate command, see MMC or MMC-2) that was started prior to the time that SRST was set to one.
The device shall not revert to the default condition and an Ultra DMA mode setting (either enabled or disabled)
shall not be affected by the host setting SRST to one.
Transition D0SR0:D0SR1: When SRST is cleared to zero and the assertion of DASP- by Device 1 was
detected during the most recent power-on or hardware reset, the device shall make a transition to the D0SR1:
PDIAG-_wait state.
Transition D0SR0:D0SR3: When SRST is cleared to zero and the assertion of DASP- by Device 1 was not
detected during the most recent power-on or hardware reset, the device shall clear bit 7 to zero in the Error
register and make a transition to the D0SR3: Set_status state.
D0SR1: PDIAG-_wait State: This state is entered when SRST has been cleared to zero and Device 1 is
present.
The device shall remain in this state for at least 1 ms.
Transition D0SR1:D0SR2: When at least 1 ms has elapsed since SRST was cleared to zero, the device shall
make a transition to the D0SR2: Sample_PDIAG- state.
D0SR2: Sample_PDIAG- State: This state is entered when SRST has been cleared to zero for at least
1 ms.
When in this state, the device shall sample the PDIAG- signal.
Transition D0SR2:D0SR2: When the sample indicates that PDIAG- is not asserted and less than 31 s have
elapsed since SRST was cleared to zero, then the device shall make a transition to the D0SR2:
Sample_PDIAG- state.
Transition D0SR2a:D0SR3: When the sample indicates that PDIAG- is asserted, the device device shall clear
bit 7 to zero in the Error register and shall make a transition to the D0SR3: Set_status state.
Transition D0SR2b:D0SR3: When the sample indicates that PDIAG- is not asserted and 31 s have elapsed
since SRST was cleared to zero, the device shall set bit 7 to one in the Error register and shall make a
transition to the D0SR3: Set_status state.
D0SR3: Set_status State: This state is entered when Bit 7 in the Error register has been set or cleared.
When in this state the device shall complete any hardware initialization and self-diagnostic testing begun in the
SRST state if not already completed.
Page 244
T13/1321D revision 3
All actions required in this state shall be completed within 31 s.
The diagnostic code shall be placed in bits 6-0 of the Error register (see Table 19). The device shall set the
signature values (see 9.12). The content of the Features register is undefined.
If the device does not implement the PACKET command feature set, the device shall clear bits 3, 2, and 0 in
the Status register to zero.
If the device implements the PACKET command feature set, the device shall clear bits 6, 5, 4, 3, 2, and 0 in
the Status register to zero. The device shall return the operating modes to their specified initial conditions.
MODE SELECT conditions shall be restored to their last saved values if saved values have been established.
MODE SELECT conditions for which no values have been saved shall be returned to their default values.
Transition D0SR3:DI1: When hardware initialization and self-diagnostic testing is completed and the status
has been set, the device shall clear BSY to zero and make a transition to the DI1: Device_idle_S state (see
Figure 21).
D1SR0: SRST
PDIAG-=X, BSY=1
SRST = 1
xx:D1SR0
D1SR1: Release_PDIAGPDIAG- =X, BSY=1
SRST= 0
D1SR2: Set_status
PDIAG- =R, BSY=1
PDIAG- released
D1SR0:D1SR1
D1SR1:D1SR2
Status set,
passed diagnostics
Device_
D1SR2:DI2
idle_NS
BSY=0, PDIAG-=A
Status set,
failed diagnostics
Device_
D1SR2:DI2
idle_NS
BSY=0, PDIAG-=N
BSY
v
DRQ
0
REL
0
SERV
0
C/D
0
I/O
0
INTRQ
R
DMARQ
R
PDIAGV
DASPR
Figure 18 − Device 1 software reset state diagram
D1SR0: SRST State: This state is entered by Device 1 when the SRST bit is set to one in the Device
Control register.
When in this state, the device shall release INTRQ, IORDY, DMARQ, and DD(15:0) within 400 ns after entering
this state. The device shall set BSY to one within 400 ns after entering this state.
If the device does not implement the PACKET command feature set, the device shall begin performing the
hardware initialization and self-diagnostic testing. The device may revert to the default condition (the device’s
setting may now be in different conditions than they were before the SRST bit was set to one by the host).
However, an Ultra DMA mode setting (either enabled or disabled) shall not be affected by the host setting SRST
to one.
If the PACKET command feature set is implemented, the device may begin performing the hardware
initialization and self-diagnostic testing and the device is not expected to stop any background device activity
(e.g., immediate command, see MMC and MMC-2) that was started prior to the time that SRST was set to one.
The device shall not revert to the default condition and an Ultra DMA mode setting (either enabled or disabled)
shall not be affected by the host setting SRST to one.
Transition D1SR0:D1SR1: When SRST is cleared to zero, the device shall make a transition to the D1SR1:
Release_PDIAG- state.
D1SR1: Release_PDIAG- State: This state is entered when SRST is cleared to zero.
Page 245
T13/1321D revision 3
When in this state, the device shall release PDIAG- within 1 ms of entering this state.
Transition D1SR1:D1SR2: When PDIAG- has been released, the device shall make a transition to the D1SR2:
Set_status state.
D1SR2: Set_status State: This state is entered when the device has negated PDIAG-.
When in this state the device shall complete the hardware initialization and self-diagnostic testing begun in the
SRST state if not already completed. The diagnostic code shall be placed in the Error register (see Table 19). If
the device passed the self-diagnostics, the device shall assert PDIAG-.
All actions required in this state shall be completed within 30 s.
The device shall set the signature values (see 9.12). The contents of the Features register is undefined.
If the device does not implement the PACKET command feature set, the device shall clear bits 3, 2, and 0 in
the Status register to zero.
If the device implements the PACKET command feature set, the device shall clear bits 6, 5, 4, 3, 2, and 0 in
the Status register to zero. The device shall return the operating modes to their specified initial conditions.
MODE SELECT conditions shall be restored to their last saved values if saved values have been established.
MODE SELECT conditions for which no values have been saved shall be returned to their default values.
Transition D1SR2:DI2: When hardware initialization and self-diagnostic testing is completed and the status
has been set, the device shall clear BSY to zero, assert PDIAG- if it passed its diagnostics, and make a
transition to the DI2: Device_idle_NS state (see Figure 21).
9.3 Bus idle protocol
When the selected device has BSY cleared to zero and DRQ cleared to zero the bus is idle.
If command overlap is implemented and enabled, the host may be waiting for a service request for a released
command. In this case, the device is preparing for the data transfer for the released command.
If command overlap and command queuing are implemented and enabled, the host may be waiting for a service
request for a number of released commands. In this case, the device is preparing for the data transfer for one of
the released commands.
Figure 19 and the text following the figure describe the host state during bus idle for hosts not implementing
command overlap and queuing. Figure 20 and the text following the figure describes the additional host state
during bus idle required for command overlap and queuing. Figure 21 and the text following the figure describe
the device state during bus idle for devices not implementing command overlap and queuing. Figure 22 and the
text following the figure describe the additional device state during bus idle required for command overlap and
queuing.
Page 246
T13/1321D revision 3
HI0: Host_Idle
Command
completed or
power-on,
hardware, or
software reset
xx:HI0
HI1: Check_Status
HI2: Device_Select
BSY = 0 & DRQ = 0 &
wrong device selected
Command to initiate
HI0:HI1
HI1:HI2
Device selected
HI2:HI1
Command to initiate
HIO2:HI1
BSY = 1 or DRQ = 1
HI3: Write_parameters
BSY = 0 & DRQ = 0 &
correct device selected
HI1:HI3
Parameters written
HI3:HI4
HI1:HI1
HI4: Write_command
Command written
HI4:xx
Command protocol
Figure 19 − Host bus idle state diagram
HI0: Host_Idle State: This state is entered when a device completes a command or when a power-on,
hardware ,or software reset has occurred.
When in this state, the host waits for a command to be issued to a device.
Transition HI0:HI1: When the host has a command to issue to a device, the host shall make a transition to
the HI1: Check_Status state.
HI1: Check_Status State: This state is entered when the host has a command to issue to a device.
When in this state, the host reads the device Status or Alternate Status register.
Transition HI1:HI2: When the status read indicates that both BSY and DRQ are cleared to zero but the wrong
device is selected, then the host shall make a transition to the HI2: Device_Select state.
Transition HI1:HI1: When the status read indicates that either BSY or DRQ is set to one, the host shall make
a transition to the HI1: Check_Status state to recheck the status of the selected device.
Transition HI1:HI3: When the status read indicates that both BSY and DRQ are cleared to zero and the
correct device is selected, then the host shall make a transition to the HI3: Write_Parameters state.
HI2: Device_Select State: This state is entered when the wrong device is selected for issuing a new
command.
When in this state, the host shall write to the Device/Head reagister to select the correct device.
Transition HI2:HI1: When the Device/Head register has been written to select the correct device, then the host
shall make a transition to the HI1: Check_Status state.
HI3: Write_Parameters State: This state is entered when the host has determined that the correct
device is selected and both BSY and DRQ are cleared to zero.
When in this state, the host writes all required command parameters to the device Command Block registers
(see clause 8).
Page 247
T13/1321D revision 3
Transition HI3:HI4: When all required command parameters have been written to the device Command Block
registers, the host shall make a transition to the HI4: Write_Command state.
HI4: Write_Command State: This state is entered when the host has written all required command
parameters to the device Command Block registers.
When in this state, the host writes the command to the device Command register.
Transition HI4:xx: When the host has written the command to the device Command register, the host shall
make a transition to the command protocol for the command written.
Page 248
T13/1321D revision 3
HIO0: INTRQ_wait_A
HIO1: Device_select_A
Command completed
& nIEN = 0 & REL = 0
& SERV = 0
xx:HIO0
Select device
HIO0:HIO1
Bus released,
REL=1 & nIEN = 0
xx:HIO0
New command to initiate
HIO0:HIO2
INTRQ asserted
HIO0:HIO3
Device selected
HIO1:HIO0
Bus released,
REL=1 & nIEN = 1
xx:HIO3
nIEN = 1
HIO2:HI1
Check_
status
HIO4: Device_select_B
HIO3: Check_status_A
Command completed
& nIEN = 1 & REL = 0
Select device
& SERV = 0
HIO3:HIO4
xx:HIO3
HIO2: Disable_INTRQ
Device selected
HIO4:HIO3
SERV =0
HIO3:HIO3
New command to initiate
HIO3:HIO2
SERV = 1
HIO3:HIO5
HIO5: Write_SERVICE
Command
completed &
SERV = 1
xx:HIO5
HIO6: INTRQ_wait_B
Service INTRQ
enabled
HIO5:HIO6
>1 outstanding
command
HIO7: Check_status_B
>1 outstanding command
HIO6:HIO7
1 outstanding command
HIO6:HP0
Service
HIO6:HPD0
return
HIO6:HDMAQ0
Tag checked
HIO7:HP0
HIO7:HPD0
HIO7:HDMAQ0
Service
return
HIO5:HIO7
1 outstanding command
& no service INTRQ
HIO5:HP0
HIO5:HPD0
HIO5:HDMAQ0
Service return
Figure 20 − Additional Host bus Idle state diagram with Overlap or overlap and queuing
HIO0: INTRQ_wait_A State: This state is entered when a command has completed with nIEN cleared to
zero, REL set to one, and SERV cleared to zero. This state is entered when the device has released the bus
with nIEN cleared to zero. This state is entered when the host is waiting for INTRQ to be asserted for bus
released commands.
When in this state, the host waits for INTRQ to be asserted indicating that a device is ready to resume
execution of a bus released command.
Transition HIO0:HIO1: When the host has one or more commands outstanding to both devices, the host may
make a transition to the HIO1: Device_select_A state to sample INTRQ for the other device.
Page 249
T13/1321D revision 3
Transition HIO0:HIO2: When the host has a new command to issue to a device and that device has no
command released or supports command queuing, then the host shall make a transition to the HIO2:
Disable_INTRQ state.
Transition HIO0:HIO3: When the host detects INTRQ asserted, the host shall make a transition to the HIO3:
Check_status A state.
HIO1: Device_select_A State: This state is entered when the host has outstanding, bus released
commands to both devices and nIEN is cleared to zero.
When in this state, the host shall disable INTRQ by setting nIEN to one, shall write the Device/Head register to
select the other device, and then, shall enable INTRQ by clearing nIEN to zero.
Transition HIO1:HIO0: Having selected the other device, the host shall make a transition to the HIO0:
INTRQ_wait_A state.
HIO2: Disable_INTRQ State: This state is entered when the host has a new command to issue to a
device and that device has no outstanding, bus released command or supports command queuing.
When in this state, the host shall set nIEN to one.
Transition HIO2:HI1: When nIEN has been set to one, the host shall make a transition to the HI1:
Check_status state (see Figure 19).
HIO3: Check_status_A State: This state is entered when a command is completed with nIEN set to
one, REL set to one, and SERV cleared to zero. This state is entered when the device has released the bus
and nIEN is set to one. This state is entered when an interrupt has occured indicating that a device is
requesting service.
When in this state, the host shall read the Status register of the device requesting service.
Transition HIO3:HIO4: If SERV is cleared to zero and the host has released commands outstanding to both
devices, then the host may make a transition to the HIO4: Device_select_B state.
Transition HIO3:HIO2: If SERV is cleared to zero and the host has a new command to issue to a device, then
the host shall make a transition to the HIO2: Disable_INTRQ state.
Transition HIO3:HIO3: If SERV is cleared to zero and the host has no new command to issue, then the host
shall make a transition to the HIO3: Check_status state.
Transition HIO3:HIO5: If SERV is set to one, the host shall make a transition to the HIO5: Write_SERVICE
state.
HIO4: Device_select_B State: This state is entered when the host has outstanding, bus released
commands to both devices and nIEN is set to one.
When in this state, the host shall disable INTRQ by setting nIEN to one, shall write the Device/Head register to
select the other device, and then, shall enable INTRQ by clearing nIEN to zero.
Transition HIO4:HIO3: Having selected the other device, the host shall make a transition to the HIO3:
Check_status_A state.
HIO5: Write_SERVICE State: This state is entered when a device has set SERV to one indicating that
the device requests service. This state is entered when a command has completed with SERV set to one.
Page 250
T13/1321D revision 3
When in this state, the host shall write the SERVICE command to the Command register.
Transition HIO5:HIO6: When the device is one that implements the PACKET command feature set and the
Service interrupt is enabled, then the host shall make a transition to the HIO6: INTRQ_wait_B state.
Transition HIO5:HIO7: When the host has more than one released command outstanding to the device and
the Service interrupt is disabled, the host shall make a transition to the HIO7: Check_status_B state.
Transition HIO5:xx: When the Service interrupt is disabled and the host has only one released command
outstanding to the device, the host shall make a transition to the service return for the protocol for the command
outstanding (see Figure 31, Figure 33, or Figure 35).
HIO6: INTRQ_wait_B State: This state is entered when the SERVICE command has been written to a
device implementing the PACKET command feature set and the Service interrupt is enabled.
NOTE − READ DMA QUEUED and WRITE DMA QUEUED commands do not implement the
Service interrupt.
When in this state, the host waits for the assertion of INTRQ.
Transition HIO6:HIO7: When the host has more than one released command outstanding to the device and
INTRQ is asserted, the host shall make a transition to the HIO7: Check_status_B state.
Transition HIO6:xx: When INTRQ has been asserted and the host has only one released command
outstanding to the device, then the host shall make a transition to the service return for the protocol for the
command outstanding (see Figure 31, Figure 33, or Figure 35).
HIO7: Check_status_B State: This state is entered when the SERVICE command has been written
and the host has more than one released command outstanding to the device.
When in this state the host reads the command tag to determine which outstanding command service is
requested for. If a DMA data transfer is required for the command, the host shall set up the DMA engine.
Transition HIO7:xx: When the command for which service is requested has been determined, the host shall
make a transition to the service return for that command protocol (see Figure 31, Figure 33, or Figure 35).
Page 251
T13/1321D revision 3
DI0: Device_Idle_SI
INTRQ=A
Command completed &
nIEN=0 & interrupt pending
xx:DI0
Device/Head register
written & device selected
DI0:DI0
Command written
DI0:xx
Command protocol
Device/Head register
written & device deselected
DI0:DI2
Device/Head register
written & device selected,
nIEN=0 & interrupt pending
DI2:DI0
Status register read or
nIEN set to one
DI0:DI1
Power-on, hardware, or software
reset if Device 1
xx:DI2
DI1: Device_Idle_S
INTRQ=N
Power-on, hardware, or
software reset if Device 0
xx:DI1
Command completed & (no
interrupt pending or
nIEN=1)
xx:DI1
BSY
0
DRQ
0
REL
0
DI2: Device_Idle_NS
INTRQ=R
Device/Head register
written & device deselected
DI1:DI2
Device/Head register
written & device selected &
(no interrupt pending or nIEN=1)
DI2:DI1
Command written
DI1:xx
Command protocol
Device/Head register
written & device selected
DI1:DI1
SERV
0
C/D
0
I/O
0
INTRQ
V
DMARQ
R
PDIAGR
DASPR
Figure 21 − Device bus Idle state diagram
DI0: Device_Idle_SI State (selected/INTRQ asserted): This state is entered when the device has
completed the execution of a command protocol with interrupt pending and nIEN=0.
When in this state, the device shall have DRQ cleared to zero, INTRQ asserted, and BSY cleared to zero.
Reading any register except the Status register shall have no effect.
Transition DI0:xx: If the Command register is written, the device shall clear the device internal interrupt
pending, shall negate or release INTRQ within 400 ns of the negation of DIOW-, shall release PDIAG- if
asserted, and shall make a transition to the command protocol indicated by the content of the Command
register. The host should not write to the Command register at this time.
Transition DI0:DI1: When the Status register is read, the device shall clear the device internal interrupt
pending, negate or release INTRQ within 400 ns of the negation of DIOR-, and make a transition to the DI1:
Device_Idle_S state. When nIEN is set to one in the Device Control register, the device shall negate INTRQ and
make a transition to the DI1: Device_Idle_S state.
Page 252
T13/1321D revision 3
Transition DI0:DI0: When the Device/Head register is written and the DEV bit selects this device or any other
register except the Command register is written, the device shall make a transition to the DI0: Device_Idle_SI
state.
Transition DI0:DI2: When the Device/Head register is written and the DEV bit selects the other device, then
the device shall release INTRQ within 400 ns of the negation of DIOW-, and make a transition to the DI2:
Device_Idle_NS state.
DI1: Device_Idle_S State (selected/INTRQ negated): This state is entered when the device has
completed the execution of a command protocol with no interrupt pending or nIEN=1, or when a pending
interrupt is cleared. This state is also entered by Device 0 at the completion of a power-on, hardware, or
software reset.
When in this state, the device shall have BSY and DRQ cleared to zero and INTRQ negated or released.
When entering this state from a power-on, hardware, or software reset, if the device does not implement the
PACKET command feature set, the device shall set DRDY to one within 30 s of entering this state. When
entering this state from a power-on, hardware, or software reset, if the device does implement the PACKET
command feature set, the device shall not set DRDY to one.
Transition DI1:xx: When the Command register is written, the device shall exit the interrupt pending state,
release PDIAG- if asserted and make a transition to the command protocol indicated by the content of the
Command register.
Transition DI1:DI1: When the Device/Head register is written and the DEV bit selects this device or any
register is written except the Command register, the device shall make a transition to the DI1: Device_Idle_S
state.
Transition DI1:DI2: When the Device/Head register is written and the DEV bit selects the other device, the
device shall make a transition to the DI2: Device_Idle_NS state.
DI2: Device_Idle_NS State (not selected): This state is entered when the device is deselected. This
state is also entered by Device 1 at the completion of a power-on, hardware, or software reset.
When in this state, the device shall have BSY and DRQ cleared to zero and INTRQ shall be released.
When entering this state from a power-on, hardware, or software reset, if the device does not implement the
PACKET command feature set, the device shall set DRDY to one within 30 s of entering this state. When
entering this state from a power-on, hardware, or software reset, if the device does implement the PACKET
command feature set, the device shall not set DRDY to one.
Transition DI2:DI0: When the Device/Head register is written, the DEV bit selects this device, the device has
an interrupt pending, and nIEN is cleared to zero, then the device shall assert INTRQ within 400 ns of the
negation of DIOW- and make a transition to the DI0: Device_Idle_SI state.
Transition DI2:DI1: When the Device/Head register is written, the DEV bit selects this device, and the device
has no interrupt pending or nIEN is set to one, then the device shall make a transition to the DI1: Device_Idle_S
state.
Page 253
T13/1321D revision 3
DIO0: Device_Idle_SIR
INTRQ=A,REL=1, SERV=0
DIO4: Device_Idle_NS
INTRQ=R, REL=x, SERV=x
Command completed &
Command written
interrupt pending & nIEN=0
Command protocol
DIO0:xx
& REL=0 & SERV=0
Device deselected
xx:DIO0
DIO0:DIO4
Bus released & interrupt
pending & nIEN=0 &
REL=1 & SERV=0
Status register read
xx:DIO0
Service required
DIO0:DIO2
Device selected
DIO4:DIO0
DIO1: Device_Idle_SR
INTRQ=N,REL=1, SERV=0
DIO0:DIO1
Command completed & (no
interrupt pending or
nIEN=1) & REL=0 & SERV=0
xx:DIO1
Bus released & (no interrupt pending
or nIEN=1) & REL=1 & SERV=0
xx:DIO1
Command
written
DIO1:xx
Cmnd
protocol
Device deselected
DIO1:DIO4
Device selected
DIO4:DIO1
Service required & nIEN=0
DIO1:DIO2
Service required & nIEN=1
DIO1:DIO3
DIO3: Device_Idle_SS
INTRQ=N, REL=1, SERV=1
Command completed & (no interrupt
pending or nIEN=1) & REL=0 & SERV=1
xx:DIO3
DIO2: Device_Idle_SIS
INTRQ=A, REL=1, SERV=1
Bus released & (no
interrupt pending or
nIEN=1) & REL=1 &
SERV=1
xx:DIO3
Status register read
DIO2:DIO3
Command completed &
interrupt pending & nIEN=0
& REL=0 & SERV=1
SERVICE written
DIO2:DP0
DIO2:DPD0
DIO2:DDMAQ0
Command written
DIO2:xx
xx:DIO2
Bus released & interrupt
pending & nIEN=0 &
REL=1 & SERV=1
DRQ
0
Device selected
DIO4:DIO3
SERVICE written
DIO3:DP0
DIO3:DPD0
DIO3:DDMAQ0 Service
return
Command written
DIO3:xx
Cmnd
protocol
Service return
Command protocol
Device deselected
DIO2:DIO4
Device selected
DIO4:DIO2
xx:DIO2
BSY
0
Device deselected
DIO3:DIO4
REL
v
SERV
v
C/D
x
I/O
x
INTRQ
V
DMARQ
R
PDIAGR
DASPR
Figure 22 − Additional Device bus Idle state diagram with Overlap or overlap and queuing
Page 254
T13/1321D revision 3
DIO0: Device_Idle_SIR State (selected/INTRQ asserted/RELset to one): This state is entered
when the device has completed the execution of a command protocol with interrupt pending, nIEN=0, REL set
to one, and SERV cleared to zero. This state is entered when the device has released an overlapped command
with interrupt pending, nIEN=0, REL set to one, and SERV cleared to zero.
When in this state, the device is preparing for completion of a released command. The device shall have BSY
and DRQ cleared to zero, and INTRQ asserted.
Transition DIO0:xx: When the Command register is written, the device shall clear the interrupt pending, shall
negate or release INTRQ within 400 ns of the negation of DIOW-, and shall make a transition to the command
protocol indicated by the content of the Command register.
NOTE − Since a queue exists, only commands in the queued command set may be written to
the Command register. If any other command is written to the Command register, the queue is
aborted and command aborted is returned for the command (see 6.9).
Transition DIO0:DIO1: When the Status register is read, the device shall clear the interrupt pending, negate or
release INTRQ within 400 ns of the negation of DIOR-, and make a transition to the DIO1: Device_Idle_SR
state.
Transition DIO0:DIO2: When the Device/Head register is written and the DEV bit selects the other device,
then the device shall release INTRQ within 400 ns of the negation of DIOW- and make a transition to the DIO2:
Device_Idle_NS state.
Transition DIO0:DIO2: When the device is ready to continue the execution of a released command, the device
shall make a transition to the DIO2: Device_idle_SIS state.
DIO1: Device_Idle_SR State (selected/INTRQ negated/REL set to one): This state is entered
when the device has completed the execution of a command protocol with no interrupt pending or nIEN=1, REL
set to one, and SERV cleared to zero. This state is entered when the device has released an overlapped
command with no interrupt pending or nIEN=1, REL set to one, and SERV cleared to zero. This state is
entered when a pending interrupt is cleared, REL is set to one, and SERV is cleared to zero.
When in this state, the device is preparing for completion of a released command. The device shall have BSY
and DRQ cleared to zero, and INTRQ negated or released.
Transition DIO1:xx: When the Command register is written, the device shall make a transition to the
command protocol indicated by the content of the Command register.
NOTE − Since a queue exists, only commands in the queued command set may be written to
the Command register. If any other command is written to the Command register, the queue is
aborted and command aborted is returned for the command (see 6.9).
Transition DIO1:DIO4: When the Device/Head register is written and the DEV bit selects the other device, the
device shall make a transition to the DIO4: Device_Idle_NS state.
Transition DIO1:DIO2: When the device is ready to continue the execution of a released command and
nIEN=0, the device shall make a transition to the DIO2: Device_idle_SIS state.
Transition DIO1:DIO3: When the device is ready to continue the execution of a released command and
nIEN=1, the device shall make a transition to the DIO3: Device_idle_SS state.
DIO2: Device_Idle_SIS State (selected/INTRQ asserted/SERV set to one): This state is
entered when the device has completed the execution of a command protocol with interrupt pending, nIEN=0,
Page 255
T13/1321D revision 3
REL set to one, and SERV set to one. This state is entered when the device has released an overlapped with
interrupt pending, nIEN=0, REL set to one, and SERV set to one.
Transition DIO2:DIO3: When the Status register is read, the device shall clear the interrupt pending, negate or
release INTRQ within 400 ns of the negation of DIOR-, and make a transition to the DIO3: Device_Idle_SS
state.
Transition DIO2: DIO4: When the Device/Head register is written and the DEV bit selects the other device, the
device shall release INTRQ within 400 ns of the negation of DIOW- and make a transition to the DIO4:
Device_Idle_NS state.
Transition DIO2:DP0/DPD0/DDMAQ0: When the SERVICE command is written into the Command register,
the device shall set the Tag for the command to be serviced, negate or release INTRQ within 400 ns of the
negation of DIOW-, and make a transition to the Service return of the command ready for service (see Figure 32
Device PACKET non-data and PIO data command protocol, Figure 34 Device PACKET DMA command
protocol, or Figure 36 Device DMA QUEUED command protocol).
Transition DIO2:xx: When any overlapped command other than SERVICE is written to the Command register,
the device shall negate or release INTRQ within 400 ns of the negation of DIOW- and make a transition to the
protocol for the new command.
DIO3: Device_Idle_SS State (selected/INTRQ negated/SERV set to one): This state is entered
when the device has completed the execution of a command protocol with no interrupt pending or nIEN=1, REL
set to one, and SERV set to one. This state is entered when the device has released an overlapped with no
interrupt pending or nIEN=1, REL set to one, and SERV set to one.
Transition DIO3: DIO4: When the Device/Head register is written and the DEV bit selects the other device, the
device shall make a transition to the DIO4: Device_Idle_NS state.
Transition DIO3:DP0/DPD0/DDMAQ0: When the SERVICE command is written into the Command register,
the device shall set the Tag for the command to be serviced and make a transition to the Service return of the
command ready for service (see Figure 32, Figure 34, or Figure 36).
Transition DIO3:xx: When any overlapped command other than SERVICE is written to the Command register,
the device shall make a transition to the protocol for the new command.
DIO4: Device_Idle_NS State (not selected): This state is entered when the device is deselected with
REL or SERV set to one.
When in this state, the device shall have BSY and DRQ cleared to zero and INTRQ shall be released.
Transition DIO4:DIO0: When the Device/Head register is written, the DEV bit selects this device, the device
has an interrupt pending, nIEN is cleared to zero, REL is set to one, and SERV is cleared to zero, then the
device shall assert INTRQ within 400 ns of the negation of DIOW- and make a transition to the DIO0:
Device_Idle_SIR state.
Transition DIO4:DIO1: When the Device/Head register is written, the DEV bit selects this device, the device
has no interrupt pending or nIEN is set to one, REL is set to one, and SERV is cleared to zero, then the device
shall make a transition to the DIO1: Device_Idle_SIR state.
Transition DIO4:DIO2: When the Device/Head register is written, the DEV bit selects this device, the device
has an interrupt pending, nIEN is cleared to zero, REL is set to one, and SERV is set to one, then the device
shall assert INTRQ within 400 ns of the negation of DIOW- and make a transition to the DIO2: Device_Idle_SIS
state.
Page 256
T13/1321D revision 3
Transition DIO4:DIO3: When the Device/Head register is written, the DEV bit selects this device, the device
has no interrupt pending or nIEN is set to one, REL is set to one, and SERV is set to one, then the device shall
make a transition to the DIO3: Device_Idle_SIR state.
9.4 Non-data command protocol
This class includes:
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
−
CFA ERASE SECTORS
CFA REQUEST EXTENDED ERROR CODE
CHECK POWER MODE
FLUSH CACHE
GET MEDIA STATUS
IDLE
IDLE IMMEDIATE
INITIALIZE DEVICE PARAMETERS
MEDIA EJECT
MEDIA LOCK
MEDIA UNLOCK
NOP
READ NATIVE MAX ADDRESS
READ VERIFY SECTOR(S)
SECURITY ERASE PREPARE
SECURITY FREEZE LOCK
SEEK
SET FEATURES
SET MAX ADDRESS
SET MULTIPLE MODE
SLEEP
SMART DISABLE OPERATION
SMART ENABLE/DISABLE AUTOSAVE
SMART ENABLE OPERATION
SMART EXECUTE OFFLINE IMMEDIATE
SMART RETURN STATUS
STANDBY
STANDBY IMMEDIATE
Execution of these commands involves no data transfer. Figure 23 and the text following the figure describes
the host state. Figure 24 and the text following the figure decribes the device state.
See the NOP command description in 8.20 and the SLEEP command in 8.40 for additional protocol
requirements.
Page 257
T13/1321D revision 3
HND0: INTRQ_wait
Command
written & selected
device & nIEN = 0
HI4:HND0
HND1: Check_Status
INTRQ asserted
BSY = 0
HND1:HI0
HND0:HND1
Host_Idle
BSY =1
HND1:HND1
Command
written & selected
device & nIEN = 1
HI4:HND1
Figure 23 − Host Non-Data state diagram
HND0: INTRQ_Wait_State: This state is entered when the host has written a non-data command to the
device and the nIEN bit in the device has been cleared to zero.
When in this state the host may wait for INTRQ to be asserted by the device.
Transition HND0:HND1: When the device asserts INTRQ, the host shall make a transition to the HND1:
Check_Status state.
HND1: Check_Status State: This state is entered when the host has written a non-data command to
the device and the nIEN bit in the device has been set to one, or when INTRQ has been asserted.
When in this state, the host shall read the device Status register. When entering this state from another state
other than when an interrupt has occurred, the host shall wait 400 ns before reading the Status register.
Transition HND1:HI0: When the status read indicates that BSY is cleared to zero, the host shall make a
transition to the HI0: Host_Idle state (see Figure 19). If status indicates that an error has occured, the host
shall take appropriate error recovery action.
Transition HND1:HND1: When the status read indicates that BSY is set to one, the host shall make a
transition to the HND1: Check_Status state to recheck device status.
DND0: Command_Execution
Execution complete & nIEN = 0
Non-data
command written
DND0:DI0
Device_idle_SI
BSY=0, INTRQ=A
DI0:DND0
DI1:DND0
Execution complete & nIEN = 1
DND0:DI1
BSY=0
BSY
1
DRQ
0
REL
0
SERV
0
C/D
0
Device_idle_S
I/O
0
INTRQ
R
DMARQ
R
PDIAGR
DASPR
Figure 24 − Device Non-Data state diagram
DND0: Command_Execution State: This state is entered when a non-data command has been
written to the device Command register.
Page 258
T13/1321D revision 3
When in this state, the device shall set BSY to one within 400 ns of the writing of the Command register, shall
execute the requested command, and shall set the interrupt pending.
Transition DND0:DI0: When command execution completes and nIEN is cleared to zero, then the device shall
set error bits if appropriate, clear BSY to zero, assert INTRQ, and make a transition to the DI0: Device_Idle_SI
state (see Figure 21).
Transition DND0:DI1: When command execution completes and nIEN is set to one, the device shall set error
bits if appropriate, clear BSY to zero, and make a transition to the DI1: Device_Idle_S state (see Figure 21).
9.5 PIO data-in command protocol
This class includes:
−
−
−
−
−
−
−
−
CFA TRANSLATE SECTOR
IDENTIFY DEVICE
IDENTIFY PACKET DEVICE
READ BUFFER
READ MULTIPLE
READ SECTOR(S)
SMART READ DATA
SMART READ LOG SECTOR
Execution of this class of command includes the transfer of one or more blocks of data from the device to the
host. Figure 25 and the text following the figure describes the host states. Figure 26 and the text following the
figure describes the device states.
Page 259
T13/1321D revision 3
HPIOI0: INTRQ_wait
PIO data in
command written
& selected device
& nIEN = 0
HI4:HPIOI0
HPIOI1: Check_Status
INTRQ asserted
BSY = 0 & DRQ = 0
HPIOI0:HPIOI1
HPIOI1:HI0
PIO data in command
written & selected device
& nIEN = 1
Host_idle
BSY = 1
HPIOI1:HPIOI1
HI4:HPIOI1
HPIOI2: Transfer_Data
Data register read
& DRQ data block
transferred & all
data for command
not transferred &
nIEN = 0
HPIOI2:HPIOI0
BSY = 0 & DRQ = 1
HPIOI1:HPIOI2
Data register read & DRQ data block
transferred & all data for command
not transferred & nIEN = 1
HPIOI2:HPIOI1
Data register read & DRQ data
block transfer not complete
HPIOI2:HPIOI2
Data register read & all data for
command transferred
HPIOI2:HI0
Host_idle
Figure 25 − Host PIO data-In state diagram
HPIOI0: INTRQ_Wait State: This state is entered when the host has written a PIO data-in command to
the device and nIEN is cleared to zero, or at the completion of a DRQ data block transfer if all the data for the
command has not been transferred and nIEN is cleared to zero.
When in this state, the host shall wait for INTRQ to be asserted.
Transition HPIOI0:HPIOI1: When INTRQ is asserted, the host shall make a transition to the HPIOI1:
Check_Status state.
HPIOI1: Check_Status State: This state is entered when the host has written a PIO data-in command
to the device and nIEN is set to one, or when INTRQ is asserted.
When in this state, the host shall read the device Status register. When entering this state from the HI4 state,
the host shall wait 400 ns before reading the Status register. When entering this state from the HPIOI2 state,
the host shall wait one PIO transfer cycle time before reading the Status register. The wait may be
accomplished by reading the Alternate Status register and ignoring the result.
Transition HPIOI1:HI0: When BSY is cleared to zero and DRQ is cleared to zero, then the device has
completed the command with an error. The host shall perform appropriate error recovery and make a transition
to the HI0: Host_Idle state (see Figure 19).
Transition HPIOI1:HPIOI1: When BSY is set to one , the host shall make a transition to the HPIOI1:
Check_Status state.
Page 260
T13/1321D revision 3
Transition HPIOI1:HPIOI2: When BSY is cleared to zero and DRQ is set to one, the host shall make a
transition to the HPIOI2: Transfer_Data state.
HPIOI2: Transfer_Data State: This state is entered when the BSY is cleared to zero, DRQ is set to
one, and the DRQ data block transfer has not completed.
When in this state, the host shall read the device Data register to transfer data.
Transition HPIOI2:HPIOI0: When the host has read the device Data register and the DRQ data block has
been transferred, all blocks for the command have not been transferred, and nIEN is cleared to zero, then the
host shall make a transition to the HPIOI0: INTRQ_Wait state.
Transition HPIOI2:HPIOI1: When the host has read the device Data register and the DRQ data block has
been transferred, all blocks for the command have not been transferred, and nIEN is set to one, then the host
shall make a transition to the HPIOI1: Check_Status state.
Transition HPIOI2:HPIOI2: When the host has read the device status register and the DRQ data block transfer
has not completed, then the host shall make a transition to the HPIOI2: Transfer_Data state.
Transition HPIOI2:HI0: When the host has read the device Data register and all blocks for the command have
been transferred, then the host shall make a transition to the HI0: Host_Idle state (see Figure 19). The host
may read the Status register.
Page 261
T13/1321D revision 3
DPIOI0: Prepare_Data
BSY=1, DRQ=0, INTRQ=N
Command execution aborted with error & nIEN=0
DPIOI0:DI0
BSY=0, INTRQ=A
PIO data in
command written
DI0:DPIOI0
DI1:DPIOI0
Command execution aborted with error & nIEN=1
DPIOI0:DI1
BSY=0
Device_idle_SI
Device_idle_S
DPIOI2: Data_Rdy_INTRQ
BSY=0, DRQ=1, INTRQ=A
Data ready to transfer & nIEN=0
DPIOI0:DPIOI2
Data ready to transfer & nIEN=1
DPIOI0:DPIOI1
DPIOI1: Transfer_Data
BSY=0, DRQ=1, INTRQ=N
Data register read &
DRQ data block
transferred & all data
for command not
transferred
Status register read
DPIOI2:DPIOI1
Data register read & DRQ data
block transfer not complete
DPIOI1:DPIOI1
DPIOI1:DPIOI0
Data register read & all data for command transferred
DPIOI1:DI1
BSY=0
BSY
v
DRQ
v
REL
0
SERV
0
C/D
0
I/O
0
INTRQ
V
DMARQ
R
Device_idle_S
PDIAGR
DASPR
Figure 26 − Device PIO data-In state diagram
DPIOI0: Prepare_Data State: This state is entered when the device has a PIO data-in command written
to the Command register.
When in this state, device shall set BSY to one within 400 ns of the writing of the Command register and
prepare the requested data for transfer to the host.
For IDENTIFY DEVICE and IDENTIFY PACKET DEVICE commands, if the device tests CBLID- it shall do so
and update bit 13 in word 93.
Transition DPIOI0:DI0: When an error is detected that causes the command to abort and nIEN is cleared to
zero, then the device shall set the appropriate error bits, clear BSY to zero, assert INTRQ, and make a
transition to the DI0: Device_Idle_SI state (see Figure 21).
Transition DPIOI0:DI1: When an error is detected that causes the command to abort and nIEN is set to one,
then the device shall set the appropriate error bits, clear BSY to zero, and make a transition to the DI1:
Device_Idle_S state (see Figure 21).
Transition DPIOI0:DPIOI1: When the device has a DRQ data block ready to transfer and nIEN is set to one,
then the device shall make a transition to the DPIOI1: Transfer_Data state.
Page 262
T13/1321D revision 3
Transition DPIOI0:DPIOI2: When the device has a DRQ data block ready to transfer and nIEN is cleared to
zero, then the device shall make a transition to the DPIOI2: Data_Ready_INTRQ state.
DPIOI1: Data_Transfer State: This state is entered when the device is ready to transfer a DRQ data
block and nIEN is set to one, or when the INTRQ indicating that the device is ready to transfer a DRQ data
block has been acknowleged by a read of the Status register.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, and the device has a data
word ready in the Data register for transfer to the host.
Transition DPIOI1:DPIOI1: When the Data register is read and transfer of the DRQ data block has not
completed, then the device shall make a transition to the DPIOI1: Data_Transfer state.
Transition DPIOI1:DPIOI0: When the Data register is read and the transfer of the current DRQ data block has
completed, but all blocks for this request have not been transferred, then the device shall make a transition to
the DPIOI0: Prepare_Data state.
Transition DPIOI1:DI1: When the Data register is read and all blocks for this request have been transferred,
then the device shall clear BSY to zero and make a transition to the DI1: Device_Idle_S state (see Figure 21).
The interrupt pending is not set on this transition.
DPIOI2: Data_Ready_INTRQ State: This state is entered when the device has a DRQ data block
ready to transfer and nIEN is cleared to zero.
When in this state, BSY is cleared to zero, DRQ is set to one, and INTRQ is asserted.
Transition DPIOI2:DPIOI1: When the Status register is read, then the device shall clear the interrupt pending,
negate INTRQ, and make a transition to the DPIOI1: Data_Transfer state.
9.6 PIO data-out command protocol
This class includes:
−
−
−
−
−
−
−
−
−
−
−
CFA WRITE MULTIPLE WITHOUT ERASE
CFA WRITE SECTORS WITHOUT ERASE
DOWNLOAD MICROCODE
SECURITY DISABLE PASSWORD
SECURITY ERASE UNIT
SECURITY SET PASSWORD
SECUITY UNLOCK
SMART WRITE LOG SECTOR
WRITE BUFFER
WRITE MULTIPLE
WRITE SECTOR(S)
Execution of this class of command includes the transfer of one or more blocks of data from the host to the
device. Figure 27 and the text following the figure describes the host states. Figure 28 and the text following the
figure describes the device states.
Page 263
T13/1321D revision 3
HPIOO0: Check_Status
PIO data out command written
HI4:HPIOO0
BSY = 0 & DRQ = 0
HPIOO0:HI0
Host_idle
BSY = 1
HPIOO0:HPIOO0
HPIOO2: INTRQ_wait
INTRQ asserted
HPIOO2:HPIOO0
BSY = 0 & DRQ = 1
HPIOO0:HPIOO1
HPIOO1: Transfer_Data
Data register written &
DRQ data block
transferred & nIEN = 1
HPIOO1:HPIOO0
Data register written & DRQ data
block transferred & nIEN = 0
HPIOO1:HPIOO2
Data register written & DRQ data
block transfer not complete
HPIOO1:HPIOO1
Figure 27 − Host PIO data-Out state diagram
HPIOO0: Check_Status State: This state is entered when the host has written a PIO data-out
command to the device; when a DRQ data block has been written and nIEN is set to one; or when a DRQ data
block has been written, nIEN is cleared zero, and INTRQ has been asserted.
When in this state, the host shall read the device Status register. When entering this state from the HI4 state,
the host shall wait 400 ns before reading the Status register. When entering this state from the HPIOO1 state,
the host shall wait one PIO transfer cycle time before reading the Status register. The wait may be
accomplished by reading the Alternate Status register and ignoring the result.
Transition HPIOO0:HI0: When BSY is cleared to zero and DRQ is cleared to zero, then the device has
completed the command and shall make a transition to the HI0: Host_Idle state (see Figure 19). If an error is
reported, the host shall perform appropriate error recovery.
Transition HPIOO0:HPIOO0: When BSY is set to one and DRQ is cleared to zero, the host shall make a
transition to the HPIOO0: Check_Status state.
Transition HPIOO0:HPIOO1: When BSY is cleared to zero and DRQ is set to one, the host shall make a
transition to the HPIOO1: Transfer_Data state.
HPIOO1: Transfer_Data State: This state is entered when the BSY is cleared to zero, DRQ is set to
one.
When in this state, the host shall write the device Data register to transfer data.
Page 264
T13/1321D revision 3
Transition HPIOO1:HPIOO2: When the host has written the device Data register, the DRQ data block has
been transferred, and nIEN is cleared to zero, then the host shall make a transition to the HPIOO2:
INTRQ_Wait state.
Transition HPIOO1:HPIOO0: When the host has written the device Data register, the DRQ data block has
been transferred, and nIEN is set to one, then the host shall make a transition to the HPIOO0: Check_Status
state.
Transition HPIOO1:HPIOO1: When the host has written the device Data register and the DRQ data block
transfer has not completed, then the host shall make a transition to the HPIOO1: Transfer_Data state.
HPIOO2: INTRQ_Wait State: This state is entered when the host has completed a DRQ data block
transfer and nIEN is cleared to zero.
When in this state, the host shall wait for INTRQ to be asserted.
Transition HPIOO2:HPIOO0: When INTRQ is asserted, the host shall make a transition to the HPIOO0:
Check_Status state.
Page 265
T13/1321D revision 3
DPIOO0: Prepare
BSY=1, DRQ=0, INTRQ=N
PIO data out
command written
DI0:DPIOO0
DI1:DPIOO0
Error or all data for command transferred & nIEN = 0
DPIOO0:DI0
BSY=0, INTRQ=A
Device_idle_SI
Error or all data for command transferred & nIEN = 1
DPIOO0:DI1
BSY=0
Device_idle_S
DPIOO2: Ready_INTRQ
BSY=0, DRQ=1, INTRQ=A
Ready to receive susequent
DRQ data block & nIEN=0
DPIOO0:DPIOO2
Ready to receive susequent
DRQ data block & nIEN=1
DPIOO0b:DPIOO1
Ready to receive first
DRQ data block
DPIOO0a:DPIOO1
DPIOO1: Transfer_Data
BSY=0, DRQ=1, INTRQ=N
Data register written
& DRQ data block
transferred
DPIOO1:DPIOO0
BSY
v
DRQ
v
REL
0
Status register read
DPIOO2:DPIOO1
Data register written & DRQ data
block transfer not complete
DPIOO1:DPIOO1
SERV
0
C/D
0
I/O
0
INTRQ
V
DMARQ
R
PDIAGR
DASPR
Figure 28 − Device PIO data-Out state diagram
DPIOO0: Prepare State: This state is entered when the device has a PIO data-out command written to
the Command register or when a DRQ data block has been transferred.
When in this state, device shall set BSY to one within 400 ns of the writing of the Command register, shall
clear DRQ to zero, and negate INTRQ. The device shall check for errors, determine if the data transfer is
complete, and if not, prepare to receive the next DRQ data block.
Transition DPIOO0a:DPIOO1: When the device is ready to receive the first DRQ data block for a command,
the device shall make a transition to the DPIOO1: Transfer_Data state.
Transition DPIOO0b:DPIOO1: When the device is ready to receive a subsequent DRQ data block for a
command and nIEN is set to one, then the device shall set the interrupt pending and make a transition to the
DPIOO1: Transfer_Data state.
Page 266
T13/1321D revision 3
Transition DPIOO0:DPIOO2: When the device is ready to receive a subsequent DRQ data block for a
command and nIEN is cleared to zero, then the device shall set the interrupt pending and make a transition to
the DPIOO2: Ready_INTRQ state.
Transition DPIOO0:DI0: When all data for the command has been transferred or an error occurs that causes
the command to abort, and nIEN is cleared to zero, then the device shall set the interrupt pending, set
appropriate error bits, clear BSY to zero, assert INTRQ, and make a transition to the DI0: Device_Idle_SI state
(see Figure 21).
Transition DPIOO0:DI1: When all data for the command has been transferred or an error occurs that causes
the command to abort, and nIEN is set to one, then the device shall set the interrupt pending, set appropriate
error bits, clear BSY to zero, and make a transition to the DI1: Device_Idle_S state (see Figure 21).
DPIOO1: Data_Transfer State: This state is entered when the device is ready to receive a DRQ data
block.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, and the device recieves a
data word in the Data register.
Transition DPIOO1:DPIOO1: When the Data register is written and transfer of the DRQ data block has not
completed, then the device shall make a transition to the DPIOO1: Data_Transfer state.
Transition DPIOO1:DPIOO0: When the Data register is written and the transfer of the current DRQ data block
has completed, then the device shall make a transition to the DPIOO0: Prepare state.
DPIOO2: Ready_INTRQ State: This state is entered when the device is ready to receive a DRQ data
block and nIEN is cleared to zero.
When in this state, BSY is cleared to zero, DRQ is set to one, and INTRQ is asserted.
Transition DPIOO2:DPIOO1: When the Status register is read, the device shall clear the interrupt pending,
negate INTRQ, and make a transition to the DPIOO1: Data_Transfer state.
9.7 DMA command protocol
This class includes:
−
−
READ DMA
WRITE DMA
Execution of this class of command includes the transfer of one or more blocks of data from the host to the
device or from the device to the host using DMA transfer. The host shall initialize the DMA channel prior to
transferring data. A single interrupt is issued at the completion of the successful transfer of all data required by
the command or when the transfer is aborted due to an error. Figure 29 and the text following the figure
describes the host states. Figure 30 and the text following the figure describes the device states.
Page 267
T13/1321D revision 3
HDMA0: Check_Status
DMA command written
BSY = 0 & DRQ = 0
HDMA0:HI0
HI4:HDMA0
Host_Idle
BSY = 1 & DMARQ negated
HDMA0:HDMA0
HDMA2: INTRQ_wait
INTRQ asserted
HDMA2:HDMA0
(BSY = 0 & DRQ = 1 & DMARQ asserted) or
(BSY = 1 & DRQ = 0 & DMARQ asserted)
HDMA0:HDMA1
HDMA1: Transfer_Data
All data for command
transferred & nIEN = 1
HDMA1:HDMA0
All data for command
transferred & nIEN = 0
HDMA1:HDMA2
Figure 29 − Host DMA state diagram
HDMA0: Check_Status State: This state is entered when the host has written a DMA command to the
device; when all data for the command has been transferred and nIEN is set to one; or when all data for the
command has been transferred, nIEN is cleared zero, and INTRQ has been asserted.
When in this state, the host shall read the device Status register. When entering this state from the HI4 state,
the host shall wait 400 ns before reading the Status register. When entering this state from the HDMA1 state,
the host shall wait one PIO transfer cycle time before reading the Status register. The wait may be
accomplished by reading the Alternate Status register and ignoring the result.
Transition HDMA0:HI0: When the BSY is cleared to zero and DRQ is cleared to zero, then the device has
completed the command and shall make a transition to the HI0: Host_Idle state (see Figure 19). If an error is
reported, the host shall perform appropriate error recovery.
Transition HDMA0:HDMA0: When BSY is set to one, DRQ is cleared to zero, and DMARQ is neagted, then
the host shall make a transition to the HDMA0: Check_Status state.
Transition HDMA0:HDMA1: When BSY is cleared to zero, DRQ is set to one, and DMARQ is asserted; or if
BSY is set to one, DRQ is cleared to zero, and DMARQ is asserted, then the host shall make a transition to
the HDMA1: Transfer_Data state. The host shall have set up the host DMA engine prior to making this
transition.
HDMA1: Transfer_Data State: This state is entered when BSY is cleared to zero, DRQ is set to one,
and DMARQ is asserted; or BSY is set to one, DRQ is cleared to zero, and DMARQ is asserted. The host
shall have initialized the DMA channel prior to entering this state.
When in this state, the host shall perform the data transfer as described in the Multiword DMA timing or the
Ultra DMA protocol.
Transition HDMA1:HDMA2: When the host has transferred all data for the command and nIEN is cleared to
zero, then the host shall make a transition to the HDMA2: INTRQ_Wait state.
Page 268
T13/1321D revision 3
Transition HDMA1:HDMA0: When the host has transferred all data for the command and nIEN is set to one,
then the host shall make a transition to the HDMA0: Check_Status state.
HDMA2: INTRQ_Wait State: This state is entered when the host has completed the transfer of all data
for the command and nIEN is cleared to zero.
When in this state, the host shall wait for INTRQ to be asserted.
Transition HDMA2:HDMA0: When INTRQ is asserted, the host shall make a transition to the HDMA0:
Check_Status state.
DDMA0: Prepare
BSY=1, DRQ=0, DMARQ=N
Error caused command abort & nIEN=0
DMA command written
Device_idle_SI
DDMA0:DI0
DI0:DDMA0
DI1:DDMA0
BSY=0, INTRQ=A
Ready to transfer data
DDMA0:DDMA1
Error caused command abort & nIEN=1
DDMA0:DI1
Device_idle_S
BSY=0
DDMA1: Transfer_Data
BSY=0, DRQ=1, DMARQ=A,
or BSY=1, DRQ=0, DMARQ=A
Error or all data transferred & nIEN=0
DDMA1:DI0
Device_idle_SI
BSY=0, INTRQ=A
Error or all data transferred & nIEN=1
DDMA1:DI1
Device_idle_S
BSY=0
BSY
v
DRQ
v
REL
0
SERV
0
C/D
0
I/O
0
INTRQ
N
DMARQ
V
PDIAGR
DASPR
Figure 30 − Device DMA state diagram
DDMA0: Prepare State: This state is entered when the device has a DMA command written to the
Command register.
When in this state, device shall set BSY to one, shall clear DRQ to zero, and negate INTRQ. The device shall
check for errors, and prepare to transfer data.
Transition DDMA0:DI0: When an error is detected that causes the command to abort and nIEN is cleared to
zero, the device shall set the appropriate error bits, enter the interrupt pending state, and make a transition to
the DI0: Device_Idle_SI state (see Figure 21).
Transition DDMA0:DI1: When an error is detected that causes the command to abort and nIEN is set to one,
then the device shall set the appropriate error bits, enter the interrupt pending state, and make a transition to
the DI1: Device_Idle_S state (see Figure 21).
Page 269
T13/1321D revision 3
Transition DDMA0:DDMA1: When the device is ready transfer data for the command, the device shall make a
transition to the DDMA1: Transfer_Data state.
DDMA1: Data_Transfer State: This state is entered when the device is ready to transfer data.
When in this state, BSY is cleared to zero, DRQ is set to one, and INTRQ is negated; or BSY is set to one,
DRQ is cleared to zero, and INTRQ is negated. Data is transferred as decribed in Multiword DMA timing or Ultra
DMA protocol.
Transition DDMA1:DI0: When the data transfer has completed or the device choses to abort the command
due to an error and nIEN is cleared to zero, then the device shall set error bits if appropriate, enter the interrupt
pending state, and make a transition to the DI0: Device_Idle_SI state (see Figure 21).
Transition DDMA2:DI1: When the data transfer has completed or the device choses to abort the command
due to an error and nIEN is set to one, then the device shall set error bits if appropriate, enter the interrupt
pending state, and make a transition to the DI1: Device_Idle_S state (see Figure 21).
9.8 PACKET command protocol
This class includes:
−
PACKET
The PACKET command has a set of protocols for non-DMA data transfer commands and a set of protocols for
DMA data transfer commands. Figure 31 and the text following the figure describes the host protocol for the
PACKET command when non-data, PIO data-in, or PIO data-out is requested. Figure 32 and the text following
the figure describes the device protocol for the PACKET command when non-data, PIO data-in, or PIO data-out
is requested. Figure 33 and the text following the figure describes the host protocol for the PACKET command
when DMA data transfer is requested. Figure 34 and the text following the figure describes the device protocol
for the PACKET command when DMA data transfer is requested.
Page 270
T13/1321D revision 3
HP0: Check_Status_A
PACKET command written
HI4:HP0
BSY = 1
HP0:HP0
HP1: Send_Packet
BSY = 0 & DRQ = 1
HP0:HP1
BSY = 0 & DRQ = 0
HP0:HI0
Data register written & command
packet transfer not complete
HP1:HP1
Host_Idle
Command packet transfer
complete, nIEN=1
HP2: Check_Status_B
Command packet transfer
complete, nIEN=0
HP1:HP3
HP1:HP2
Service return
HIOx:HP2
BSY = 1
HP2:HP2
HP3: INTRQ_wait
(Ready to transfer data or
command compete) & nIEN = 0
HP2:HP3
INTRQ asserted
HP3:HP2
BSY = 0 & DRQ = 0 & REL=0 & SERV=0, no queue
HP2:HI0
Host_Idle
BSY = 0 & DRQ = 0 & REL=0 & SERV=0 & nIEN=0, queue
HP2a:HIO0
Command complete
BSY = 0 & DRQ = 0 & REL=0 & SERV=0 & nIEN=1, queue
HP2a:HIO3
Command complete
BSY = 0 & DRQ = 0 & REL=1 & SERV=0 & nIEN=0
HP2b:HIO0
Bus release
BSY = 0 & DRQ = 0 & REL=1 & SERV=0 & nIEN=1
HP2b:HIO3
Bus release
BSY = 0 & DRQ = 0 & SERV=1
Bus release or command complete
HP2:HIO5
BSY = 0 & DRQ = 1
HP2:HP4
HP4: Transfer_Data
Data register written or
read & DRQ data
block transferred
HP4:HP2
Data register written or read & DRQ data
block transfer not complete
HP4:HP4
Figure 31 − Host PACKET non-data and PIO data command state diagram
HP0: Check_Status_A State: This state is entered when the host has written a PACKET command to
the device.
When in this state, the host shall read the device Status register. When entering this state from the HI4 state,
the host shall wait 400 ns before reading the Status register.
Page 271
T13/1321D revision 3
Transition HP0:HP0: When BSY is set to one, the host shall make a transition to the HP0: Check_Status_A
state.
Transition HP0:HP1: When BSY is cleared to zero and DRQ is set to one, then the host shall make a
transition to the HP1: Send_Packet state.
Transition HP0:HI0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, and SERV
is cleared to zero, then the command is completed and the host shall make a transition to the HI0: Host_Idle
state (see Figure 19). If an error is reported, the host shall perform appropriate error recovery.
HP1: Send_Packet State: This state is entered when BSY is cleared to zero, DRQ is set to one.
When in this state, the host shall write a byte of the command packet to the Data register.
Transition HP1:HP1: When the Data register has been written and the writing of the command packet is not
completed, the host shall make a transition to the HP1: Send_Packet state.
Transition HP1:HP2: When the Data register has been written, the writing of the command packet is
completed, and nIEN is set to one, the host shall make a transition to the HP2: Check_Status_B state.
Transition HP1:HP3: When the Data register has been written, the writing of the command packet is
completed, and nIEN is cleared to zero, the host shall make a transition to the HP3: INTRQ wait state.
HP2: Check_Status_B State: This state is entered when the host has written the command packet to
the device, when INTRQ has been asserted, when a DRQ data block has been transferred, or from a service
return.
When in this state, the host shall read the device Status register. When entering this state from the HP1 or
HP4 state, the host shall wait one PIO transfer cycle time before reading the Status register. The wait may be
accomplished by reading the Alternate Status register and ignoring the result.
Transition HP2:HP2: When BSY is set to one, and DRQ is cleared to zero, the host shall make a transition to
the HP2: Check_Status_B state.
Transition HP2:HP3: When the host is ready to transfer data or the command is complete, and nIEN is
cleared to zero, then the host shall make a transition to the HP3: INTRQ_Wait state.
Transition HP2:HP4: When BSY is cleared to zero and DRQ is set to one, then the host shall make a
transition to the HP4: Transfer_Data state.
Transition HP2:HI0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, SERV is
cleared to zero, and the device queue is empty, then the command is completed and the host shall make a
transition to the HI0: Host_Idle state (see Figure 19). If an error is reported, the host shall perform appropriate
error recovery.
Transition HP2a:HIO0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, SERV
is cleared to zero, nIEN is cleared to zero, and the device has a queue of released commands, then the
command is completed and the host shall make a transition to the HIO0: Command completed state (see
Figure 19). If an error is reported, the host shall perform appropriate error recovery.
Transition HP2a:HIO3: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, SERV
is cleared to zero, nIEN is set to one, and the device has a queue of released commands, then the command is
completed and the host shall make a transition to the HIO3: Command completed state (see Figure 19). If an
error is reported, the host shall perform appropriate error recovery.
Page 272
T13/1321D revision 3
Transitions HP2b:HIO0: When BSY is cleared to zero, DRQ is cleared to zero, REL is set to one, SERV is
cleared to zero, and nIEN is cleared to zero, then the host shall make a transition to the HIO0: INTRQ_wait_A
state (see Figure 20). The bus has been released.
Transitions HP2b:HIO3: When BSY is cleared to zero, DRQ is cleared to zero, REL is set to one, SERV is
cleared to zero, and nIEN is set to one, then the host shall make a transition to the HIO3: Check_status_A
state (see Figure 20). The bus has been released.
Transitions HP2:HIO5: When BSY is cleared to zero, DRQ is cleared to zero, and SERV is set to one, then
the host shall make a transition to the HIO5: Write_SERVICE state (see Figure 20). The command is
completed or the bus has been released, and another queued command is ready for service. If an error is
reported, the host shall perform appropriate error recovery.
HP3: INTRQ_Wait State: This state is entered when the command packet has been transmitted, the
host is ready to transfer data or when the command has completed, and nIEN is cleared to zero.
When in this state, the host shall wait for INTRQ to be asserted.
Transition HP3:HP2: When INTRQ is asserted, the host shall make a transition to the HP2: Check_Status_B
state.
HP4: Transfer_Data State: This state is entered when BSY is cleared to zero, DRQ is set to one, and
C/D is cleared to zero.
When in this state, the host shall read the byte count then read or write the device Data register to transfer
data. If the bus has been released, the host shall read the Sector Count register to determine the Tag for the
queued command to be executed.
Transition HP4:HP2: When the host has read or written the device Data register and the DRQ data block has
been transferred, then the host shall make a transition to the HP2: Check_Status_B state.
Transition HP4:HP4: When the host has read or written the device status register and the DRQ data block
transfer has not completed, then the host shall make a transition to the HP4: Transfer_Data state.
Page 273
T13/1321D revision 3
DP0: Prepare_A
BSY=1, DRQ=0, INTRQ=N, C/D=x, I/O=x
PACKET command written
xx:DP0
DP1: Receive_packet
BSY=0, DRQ=1, INTRQ=N, C/D=1, I/O=0
Ready to receive command packet
DP0:DP1
Data register written & command
packet transfer complete
DP1:DP2
Data register written &
command packet transfer
not complete
DP1:DP1
DP2: Prepare_B
BSY=1, DRQ=0, INTRQ=N, C/D=x, I/O=x
Service return
xx:DP2
Command complete & nIEN=1 & no queue
Device_Idle_S
DP2:DI1
BSY=0, REL=0, C/D=1, I/O=1
Command complete & nIEN=0 & no queue
DP2:DI0
Device_Idle_SI
BSY=0, REL=0, C/D=1, I/O=1, INTRQ=A
Command complete & nIEN=1 & queue
DP2a:DIO1
Device_Idle_SR
BSY=0, REL=0, C/D=1, I/O=1
Command complete & nIEN=0 & queue
DP2a:DIO0
Device_Idle_SIR
BSY=0, REL=0, C/D=1, I/O=1, INTRQ=A
Command complete & nIEN = 1 & service required Command complete & nIEN = 0 & service required
Device_Idle_SS
DP2a:DIO3
Device_Idle_SIS
DP2a:DIO2
BSY=0, REL=0, SERV=1, C/D=1, I/O=1 BSY=0, REL=0, SERV=1, C/D=1, I/O=1, INTRQ=A
Device_Idle_SIR
Bus release & nIEN = 0
DP2b:DIO0
BSY=0, REL=1, C/D=1, I/O=1, INTRQ=A
Bus release & nIEN = 1
DP2b:DIO1
BSY=0, REL=1, C/D=1, I/O=1
Device_Idle_SR
Bus release & nIEN = 0 & service required Bus release & nIEN = 1 & service required
DP2b:DIO3
Device_Idle_SS
Device_Idle_SIS
DP2b:DIO2
BSY=0,
REL=1,
SERV=1,
C/D=1,
I/O=1
BSY=0, REL=1, SERV=1, C/D=1, I/O=1, INTRQ=A
DP3: Ready_INTRQ
BSY=0, DRQ=1, INTRQ=A, C/D=0, I/O=x
DP4: Transfer_Data
BSY=0, DRQ=1, INTRQ=N, C/D=0, I/O=x
Data register
read/written & DRQ
data block transfer
not complete
DP4:DP4
Ready to transfer DRQ SERVICE written & service
interrupt enabled
data block & nIEN=1
DP2a:DP3
DP2:DP4
Service status read
Set byte count & Tag
DP3:DP2
Data register read/written &
Ready to transfer DRQ
DRQ data block complete
data block & nIEN=0
DP4:DP2
DP2b:DP3
Set byte count & Tag
Status register read
DP3:DP4
BSY
v
DRQ
v
REL
x
SERV
x
C/D
v
I/O
v
INTRQ
V
DMARQ
R
PDIAGR
DASPR
Figure 32 − Device PACKET non-data and PIO data command state diagram
.
DP0: Prepare_A State: This state is entered when the device has a PACKET written to the Command
register.
Page 274
T13/1321D revision 3
When in this state, device shall set BSY to one, clear DRQ to zero, and negate INTRQ within 400 ns of the
receipt of the command and shall prepare to receive a command packet. If the command is a queued
command, the device shall verify that the Tag is valid.
Transition DP0:DP1: When the device is ready to receive the command packet for a command, the device
shall make a transition to the DP1: Receive_Packet state.
DP1: Receive_Packet State: This state is entered when the device is ready to receive the command
packet.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, C/D is set to one, I/O is
cleared to zero, and REL is cleared to zero. When in this state, the device Data register is written.
Transition DP1:DP1: If the Data register is written and the entire command packet has not been received, then
the device shall make a transition to the DP1: Receive_Packet state.
Transition DP1:DP2: When the Data register is written and the entire command packet has been received,
then the device shall make a transition to the DP2: Prepare_B state.
DP2: Prepare_B State: This state is entered when the command packet has been received or from a
Service return.
When in this state, device shall set BSY to one, clear DRQ to zero, and negate INTRQ. Non-data transfer
commands shall be executed while in this state. For data transfer commands, the device shall check for
errors, determine if the data transfer is complete, and if not, prepare to transfer the next DRQ data block.
If the command is overlapped and the release interrupt is enabled, the device shall bus release as soon as the
command packet has been received.
Transition DP2:DP4: When the device is ready to transfer a DRQ data block for a command and nIEN is set to
one, then the device shall set the command Tag and byte count, set the interrupt pending, and make a
transition to the DP4: Transfer_Data state.
Transition DP2b:DP3: When the device is ready to transfer a DRQ data block for a command and nIEN is
cleared to zero, then the device shall set the command Tag and byte count, set the interrupt pending, and
make a transition to the DP3: Ready_INTRQ state.
Transition DP2a:DP3: When the service interrupt is enabled and the device has SERVICE written to the
Command register, then the device shall set the command Tag and byte count and make a transition to the
DP3: Ready_INTRQ state.
Transition DP2:DI0: When the command has completed or an error occurs that causes the command to abort,
the device has no other command released, and nIEN is cleared to zero, then the device shall set the interrupt
pending, set appropriate error bits, set C/D and I/O to one, clear BSY to zero, and make a transition to the DI0:
Device_Idle_SI state (see Figure 21).
Transition DP2:DI1: When the command has completed or an error occurs that causes the command to abort,
the device has no other command released, and nIEN is set to one, then the device shall set appropriate error
bits, set C/D and I/O to one, clear BSY to zero, and make a transition to the DI1: Device_Idle_S state (see
Figure 21).
Transition DP2a:DIO0: When the command has completed or an error occurs that causes the command to
abort, the device has another command released but not ready for service, and nIEN is cleared to zero, then the
device shall set the interrupt pending, set appropriate error bits, set C/D and I/O to one, clear BSY to zero, and
make a transition to the DIO0: Device_Idle_SIR state (see Figure 22).
Page 275
T13/1321D revision 3
Transition DP2a:DIO1: When the command has completed or an error occurs that causes the command to
abort, the device has another command released but not ready for service, and nIEN is set to one, then the
device shall set appropriate error bits, set C/D and I/O to one, clear BSY to zero, and make a transition to the
DIO1: Device_Idle_SR state (see Figure 22).
Transition DP2a:DIO2: When the command has completed or an error occurs that causes the command to
abort, the device has another command ready for service, and nIEN is cleared to zero, then the device shall set
the interrupt pending, set appropriate error bits, set C/D and I/O to one, set SERV to one, clear BSY to zero,
and make a transition to the DIO2: Device_Idle_SIS state (see Figure 22).
Transition DP2a:DIO3: When the command has completed or an error occurs that causes the command to
abort, the device has another command ready for service, and nIEN is set to one, then the device shall set
appropriate error bits, set C/D and I/O to one, set SERV to one, clear BSY to zero, and make a transition to
the DIO3: Device_Idle_SS state (see Figure 22).
Transition DP2b:DIO0: When the command is released and nIEN is cleared to zero, then the device shall set
the interrupt pending, set appropriate error bits, set C/D and I/O to one, set REL to one, clear BSY to zero, and
make a transition to the DIO0: Device_Idle_SIR state (see Figure 22).
Transition DP2b:DIO1: When the command is released and nIEN is set to one, then the device shall set
appropriate error bits, set C/D and I/O to one, set REL to one, clear BSY to zero, and make a transition to the
DIO1: Device_Idle_SR state (see Figure 22).
Transition DP2b:DIO2: When the command is released, the device has another command ready for service,
and nIEN is cleared to zero, then the device shall set the interrupt pending, set appropriate error bits, set C/D
and I/O to one, set REL to one, set SERV to one, clear BSY to zero, and make a transition to the DIO2:
Device_Idle_SIS state (see Figure 22).
Transition DP2b:DIO3: When the command is released, the device has another command ready for service,
and nIEN is set to one, then the device shall set appropriate error bits, set C/D and I/O to one, set REL to one,
set SERV to one, clear BSY to zero, and make a transition to the DIO3: Device_Idle_SS state (see Figure 22).
DP3: Ready_INTRQ State: This state is entered when the device is ready to transfer a DRQ data block
and nIEN is cleared to zero. This state is entered to interrupt upon receipt of a SERVICE command when
service interrupt is enabled.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is asserted, C/D is cleared to zero, and
I/O is set to one for PIO data-out or cleared to zero for PIO data-in.
Transition DP3:DP2: When the Status register is read to respond to a service interrupt, the device shall make
a transition to the DP2: Prepare_B state.
Transition DP3:DP4: When the Status register is read when the device is ready to transfer data, then the
device shall clear the interrupt pending, negate INTRQ, and make a transition to the DP4: Data_Transfer state.
DP4: Data_Transfer State: This state is entered when the device is ready to transfer a DRQ data block.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, C/D is cleared to zero, I/O is
set to one for PIO data-out or cleared to zero for PIO data-in, and a data word is read/written in the Data
register.
Transition DP4:DP4: When the Data register is read/written and transfer of the DRQ data block has not
completed, then the device shall make a transition to the DP4: Data_Transfer state.
Page 276
T13/1321D revision 3
Transition DP4:DP2: When the Data register is read/written and the transfer of the current DRQ data block has
completed, then the device shall make a transition to the DP2: Prepare_B state.
HPD1: Send_Packet
HPD0: Check_Status_A
PACKET command written
HI4:HPD0
BSY = 1
HPD0:HPD0
BSY = 0 & DRQ = 1
HPD0:HPD1
BSY = 0 & DRQ = 0
HPD0:HI0
HPD2: Check_Status_B
Data register written & command
packet transfer not complete
HPD1:HPD1
Host_Idle
Command packet
transfer complete,
nIEN=1
HPD1:HPD2
Command packet transfer
complete, nIEN=0
HPD1:HPD3
BSY = 0 & DRQ = 0 & REL=0 & SERV=0, no queue
HPD2:HI0
Host_Idle
Service return
HIOx:HPD2
BSY = 0 & DRQ = 0 & REL=0 & SERV=0 & nIEN=0, queue
HPD2a:HIO0
Command complete
BSY = 0 & DRQ = 0 & REL=0 & SERV=0 & nIEN=1, queue
HPD2a:HIO3
Command complete
BSY = 0 & DRQ = 0 & REL=1 & SERV=0 & nIEN=0
HPD2b:HIO0
BSY = 1
HPD2:HPD2
Bus release
BSY = 0 & DRQ = 0 & REL=1 & SERV=0 & nIEN=1
HPD2b:HIO3
Bus release
BSY = 0 & DRQ = 0 & SERV=1
HPD2:HIO5
Bus release or command complete
(Ready to transfer data or
command compete) & nIEN = 0
HPD3: INTRQ_wait
HPD2:HPD3
INTRQ asserted
HPD3:HPD2
BSY = 0 & DRQ = 1& DMARQ asserted
HPD2:HPD4
HPD4: Transfer_Data
All data for command
transferred & nIEN=1
HPD4:HPD2
All data for command transferred & nIEN=0
HPD4:HPD3
Figure 33 − Host PACKET DMA command state diagram
HPD0: Check_Status_A State: This state is entered when the host has written a PACKET command
to the device.
When in this state, the host shall read the device Status register. When entering this state from the HI4 state,
the host shall wait 400 ns before reading the Status register.
Page 277
T13/1321D revision 3
Transition HPD0:HPD0: When BSY is set to one, the host shall make a transition to the HPD0:
Check_Status_A state.
Transition HPD0:HPD1: When BSY is cleared to zero and DRQ is set to one, then the host shall make a
transition to the HPD1: Send_Packet state.
Transition HPD0:HI0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, and
SERV is cleared to zero, then the command is completed and the host shall make a transition to the HI0:
Host_Idle state (see Figure 19). If an error is reported, the host shall perform appropriate error recovery.
HPD1: Send_Packet State: This state is entered when BSY is cleared to zero, DRQ is set to one.
When in this state, the host shall write a byte of the command packet to the Data register.
Transition HPD1:HPD1: When the Data register has been written and the writing of the command packet is
not completed, the host shall make a transition to the HPD1: Send_Packet state.
Transition HPD1:HPD2: When the Data register has been written, the writing of the command packet is
completed, and nIEN is set to one, the host shall make a transition to the HPD2: Check_Status_B state.
Transition HPD1:HPD3: When the Data register has been written, the writing of the command packet is
completed, and nIEN is cleared to zero, the host shall make a transition to the HPD3: INTRQ wait state.
HPD2: Check_Status_B State: This state is entered when the host has written the command packet
to the device, when INTRQ has been asserted, when a DRQ data block has been transferred, or from a service
return.
When in this state, the host shall read the device Status register. When entering this state from the HPD1 or
HPD4 state, the host shall wait one PIO transfer cycle time before reading the Status register. The wait may be
accomplished by reading the Alternate Status register and ignoring the result.
Transition HPD2:HPD2: When BSY is set to one, and DRQ is cleared to zero, the host shall make a transition
to the HPD2: Check_Status_B state.
Transition HPD2:HPD3: When the host is ready to transfer data or the command is complete, and nIEN is
cleared to zero, then the host shall make a transition to the HPD3: INTRQ_Wait state.
Transition HPD2:HPD4: When BSY is cleared to zero, DRQ is set to one, and DMARQ is asserted, then the
host shall make a transition to the HPD4: Transfer_Data state. The host shall have set up the DMA engine
before this transition.
Transition HPD2:HI0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, SERV is
cleared to zero, and the device queue is empty, then the command is completed and the host shall make a
transition to the HI0: Host_Idle state (see Figure 19). If an error is reported, the host shall perform appropriate
error recovery.
Transition HPD2a:HIO0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, SERV
is cleared to zero, nIEN is cleared to zero, and the device has a queue of released commands, then the
command is completed and the host shall make a transition to the HIO0: Command completed state (see
Figure 19). If an error is reported, the host shall perform appropriate error recovery.
Transition HPD2a:HIO3: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero, SERV
is cleared to zero, nIEN is set to one, and the device has a queue of released commands, then the command is
completed and the host shall make a transition to the HIO3: Command completed state (see Figure 19). If an
error is reported, the host shall perform appropriate error recovery.
Page 278
T13/1321D revision 3
Transition HPD2b:HIO0: When BSY is cleared to zero, DRQ is cleared to zero, REL is set to one, SERV is
cleared to zero, and nIEN is cleared to zero, then the host shall make a transition to the HIO0: INTRQ_wait_A
state (see Figure 20). The bus has been released.
Transition HPD2b:HIO3: When BSY is cleared to zero, DRQ is cleared to zero, REL is set to one, SERV is
cleared to zero, and nIEN is set to one, then the host shall make a transition to the HIO3: Check_status_A
state (see Figure 20). The bus has been released.
Transition HPD2:HIO5: When BSY is cleared to zero, DRQ is cleared to zero, and SERV is set to one, then
the host shall make a transition to the HIO5: Write_SERVICE state (see Figure 20). The command is
completed or the bus has been released, and another queued command is ready for service. If an error is
reported, the host shall perform appropriate error recovery.
HPD3: INTRQ_Wait State: This state is entered when the command packet has been transmitted, the
host is ready to transfer data or when the command has completed, and nIEN is cleared to zero.
When in this state, the host shall wait for INTRQ to be asserted.
Transition HPD3:HPD2: When INTRQ is asserted, the host shall make a transition to the HPD2:
Check_Status_B state.
HPD4: Transfer_Data State: This state is entered when BSY is cleared to zero, DRQ is set to one, and
DMARQ is asserted.
When in this state, the host shall read or write the device Data port to transfer data. If the bus has been
released, the host shall read the Sector Count register to determine the Tag for the queued command to be
executed.
Transition HPD4:HPD2: When all data for the request has been transferred and nIEN is set to one, then the
host shall make a transition to the HPD2: Check_Status_B state.
Transition HPD4:HPD3: When all data for the request has been transferred and nIEN is cleared to zero, then
the host shall make a transition to the HPD3: INTRQ_wait state.
Page 279
T13/1321D revision 3
DPD0: Prepare_A
BSY=1, DRQ=0, INTRQ=N,
C/D=x, I/O=x, DMARQ=N
PACKET command written
xx:DPD0
DPD1: Receive_packet
BSY=0, DRQ=1, INTRQ=N,
C/D=1, I/O=0, DMARQ=N
Data register written &
command packet transfer
not complete
Ready to receive command packet
DPD0:DPD1
Data register written & command packet transfer complete
DPD1:DPD0
DPD1:DPD1
DPD2: Prepare_B
BSY=1, DRQ=0, INTRQ=N, C/D=x, I/O=x, DMARQ=N
Service return
xx:DPD2
Command complete & nIEN=1 & no queue
DPD2:DI1
BSY=0, REL=0,C/D=1, I/O=1
Command complete & nIEN=0 & no queue
DPD2:DI0
Device_Idle_SI
BSY=0, REL=0, C/D=1, I/O=1, INTRQ=A
Command complete & nIEN=1 & queue
Command complete & nIEN=0 & queue
DPD2a:DIO0
Device_Idle_SIR
BSY=0, REL=0, C/D=1, I/O=1, INTRQ=A
Device_Idle_S
Device_Idle_SR
DPD2a:DIO1
BSY=0, REL=0, C/D=1, I/O=1
Command complete & nIEN = 1 & service required Command complete & nIEN = 0 & service required
Device_Idle_SS
DPD2a:DIO3
DPD2a:DIO2
Device_Idle_SIS
BSY=0, REL=0, SERV=1, C/D=1, I/O=1 BSY=0, REL=0, SERV=1, C/D=1, I/O=1, INTRQ=A
Bus release & nIEN = 0
DPD2b:DIO0
Device_Idle_SIR
BSY=0, REL=1, C/D=1, I/O=1, INTRQ=A
Bus release & nIEN = 1
DPD2b:DIO1
BSY=0, REL=1, C/D=1, I/O=1
Device_Idle_SR
Bus release & nIEN = 0 & service required Bus release & nIEN = 1 & service required
Device_Idle_SIS
DPD2b:DIO2
DPD2b:DIO3
Device_Idle_SS
BSY=0, REL=1, SERV=1, C/D=1, I/O=1, INTRQ=A BSY=0, REL=1, SERV=1, C/D=1, I/O=1
DPD3: Ready_INTRQ
BSY=0, DRQ=1, INTRQ=A,
C/D=0, I/O=x, DMARQ=N
DPD4: Transfer_Data
BSY=0, DRQ=1, INTRQ=N,
C/D=0, I/O=x, DMARQ=A
Ready to transfer DMA SERVICE written & service
data & nIEN=1
interrupt enabled
DPD2a:DPD3
DPD2:DPD4
Set byte count & Tag
Data transfer completed
or terminated
DPD4:DPD2
Service status read
DPD3:DPD2
Ready to transfer DMA
data & nIEN=0
DPD2b:DPD3
Set byte count & Tag
Status register read
DPD3:DPD4
BSY
v
DRQ
v
REL
x
SERV
x
C/D
v
I/O
v
INTRQ
V
DMARQ
V
PDIAGR
DASPR
Figure 34 − Device PACKET DMA command state diagram
DPD0: Prepare_A State: This state is entered when the device has a PACKET written to the Command
register.
Page 280
T13/1321D revision 3
When in this state, device shall set BSY to one, clear DRQ to zero, and negate INTRQ within 400 ns of the
receipt of the command and shall prepare to receive a command packet. If the command is a queued
command, the device shall verify that the Tag is valid.
Transition DPD0:DPD1: When the device is ready to receive the command packet for a command, the device
shall make a transition to the DPD1: Receive_Packet state.
DPD1: Receive_Packet State: This state is entered when the device is ready to receive the command
packet.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, C/D is set to one, I/O is
cleared to zero, and REL is cleared to zero. When in this state, the device Data register is written.
Transition DPD1:DPD1: If the Data register is written and the entire command packet has not been received,
then the device shall make a transition to the DPD1: Receive_Packet state.
Transition DPD1:DPD2: When the Data register is written and the entire command packet has been received,
then the device shall make a transition to the DPD2: Prepare_B state.
DPD2: Prepare_B State: This state is entered when the command packet has been received or from a
Service return.
When in this state, device shall set BSY to one, clear DRQ to zero, and negate INTRQ. The device shall check
for errors, determine if the data transfer is complete, and if not, prepare to transfer the DMA data.
If the command is overlapped and the release interrupt is enabled, the device shall bus release as soon as the
command packet has been received.
Transition DPD2:DPD4: When the device is ready to transfer DMA data for a command and nIEN is set to
one, then the device shall set the command Tag and byte count, set the interrupt pending, and make a
transition to the DPD4: Transfer_Data state.
Transition DPD2b:DPD3: When the device is ready to transfer DMA data for a command and nIEN is cleared
to zero, then the device shall set the command Tag and byte count, set the interrupt pending, and make a
transition to the DPD3: Ready_INTRQ state.
Transition DPD2a:DPD3: When the service interrupt is enabled and the device has SERVICE written to the
Command register, then the device shall set the command Tag and byte count and make a transition to the
DPD3: Ready_INTRQ state.
Transition DPD2:DI0: When the command has completed or an error occurs that causes the command to
abort, the device has no other command released, and nIEN is cleared to zero, then the device shall set the
interrupt pending, set appropriate error bits, set C/D and I/O to one, clear BSY to zero, and make a transition to
the DI0: Device_Idle_SI state (see Figure 21).
Transition DPD2:DI1: When the command has completed or an error occurs that causes the command to
abort, the device has no other command released, and nIEN is set to one, then the device shall set appropriate
error bits, set C/D and I/O to one, clear BSY to zero, and make a transition to the DI1: Device_Idle_S state
(see Figure 21).
Transition DPD2a:DIO0: When the command has completed or an error occurs that causes the command to
abort, the device has another command released but not ready for service, and nIEN is cleared to zero, then the
device shall set the interrupt pending, set appropriate error bits, set C/D and I/O to one, clear BSY to zero, and
make a transition to the DIO0: Device_Idle_SIR state (see Figure 22).
Page 281
T13/1321D revision 3
Transition DPD2a:DIO1: When the command has completed or an error occurs that causes the command to
abort, the device has another command released but not ready for service, and nIEN is set to one, then the
device shall, set appropriate error bits, set C/D and I/O to one, clear BSY to zero, and make a transition to the
DIO1: Device_Idle_SR state (see Figure 22).
Transition DPD2a:DIO2: When the command has completed or an error occurs that causes the command to
abort, the device has another command ready for service, and nIEN is cleared to zero, then the device shall set
the interrupt pending, set appropriate error bits, set C/D and I/O to one, set SERV to one, clear BSY to zero,
and make a transition to the DIO2: Device_Idle_SIS state (see Figure 22).
Transition DPD2a:DIO3: When the command has completed or an error occurs that causes the command to
abort, the device has another command ready for service, and nIEN is set to one, then the device shall set
appropriate error bits, set C/D and I/O to one, set SERV to one, clear BSY to zero, and make a transition to
the DIO3: Device_Idle_SS state (see Figure 22).
Transition DPD2b:DIO0: When the command is released and nIEN is cleared to zero, then the device shall
set the interrupt pending, set appropriate error bits, set C/D and I/O to one, set REL to one, clear BSY to zero,
and make a transition to the DIO0: Device_Idle_SIR state (see Figure 22).
Transition DPD2b:DIO1: When the command is released and nIEN is set to one, then the device shall, set
appropriate error bits, set C/D and I/O to one, set REL to one, clear BSY to zero, and make a transition to the
DIO1: Device_Idle_SR state (see Figure 22).
Transition DPD2b:DIO2: When the is released, the device has another command ready for service, and nIEN
is cleared to zero, then the device shall set the interrupt pending, set appropriate error bits, set C/D and I/O to
one, set REL to one, set SERV to one, clear BSY to zero, and make a transition to the DIO2: Device_Idle_SIS
state (see Figure 22).
Transition DPD2b:DIO3: When the command is released, the device has another command ready for service,
and nIEN is set to one, then the device shall set appropriate error bits, set C/D and I/O to one, set REL to one,
set SERV to one, clear BSY to zero, and make a transition to the DIO3: Device_Idle_SS state (see Figure 22).
DPD3: Ready_INTRQ State: This state is entered when the device is ready to transfer DMA data and
nIEN is cleared to zero. This state is entered to interrupt upon receipt of a SERVICE command when service
interrupt is enabled.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is asserted, C/D is cleared to zero, and
I/O is set to one for PIO data-out or cleared to zero for PIO data-in.
Transition DPD3:DPD2: When the Status register is read to respond to a service interrupt, the device shall
make a transition to the DPD2: Prepare_B state.
Transition DPD3:DPD4: When the Status register is read and the device is ready to transfer data, then the
device shall clear the interrupt pending, negate INTRQ, and make a transition to the DPD4: Data_Transfer state.
DPD4: Data_Transfer State: This state is entered when the device is ready to transfer DMA data.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, C/D is cleared to zero, I/O is
set to one for data-out or cleared to zero for data-in, DMARQ is asserted, and data is transferred as described
in Multiword DMA timing or Ultra DMA protocol.
Transition DPD4:DPD2: When the data transfer has been completed, then the device shall make a transition
to the DPD2: Prepare_B state.
Page 282
T13/1321D revision 3
9.9 READ/WRITE DMA QUEUED command protocol
This class includes:
−
−
READ DMA QUEUED
WRITE DMA QUEUED
Execution of this class of command includes the transfer of one or more blocks of data from the host to the
device or from the device to the host using DMA transfer. All data for the command may be transferred without
a bus release between the command receipt and the data transfer. This command may bus release before
transferring data. The host shall initialize the DMA channel prior to transferring data. When data transfer is
begun, all data for the request shall be transferred without a bus release. Figure 35 and the text following the
figure describes the host states. Figure 36 and the text following the figure describes the device states.
HDMAQ0: Check_Status
DMA QUEUED command written
HI4:HDMAQ0
BSY = 0 & DRQ = 0 & REL=0 & SERV=0, no queue
HDMAQ0:HI0
Host_Idle
DMA QUEUED service return
HIO5:HDMAQ0
HIO7:HDMAQ0
BSY = 0 & DRQ = 0 & REL=0 & SERV=0 & nIEN=0, queue
HDMAQ0a:HIO0
Command complete
BSY = 0 & DRQ = 0 & REL=0 & SERV=0 & nIEN=1, queue
HDMAQ0a:HIO3
Command complete
BSY = 0 & DRQ = 0 & REL=1 & SERV=0 & nIEN=0
HDMAQ0b:HIO0
Bus release
BSY = 0 & DRQ = 0 & REL=1 & SERV=0 & nIEN=1
HDMAQ0b:HIO3
Bus release
BSY = 0 & DRQ = 0 & SERV=1
HDMAQ0:HIO5
Bus release or command complete
BSY = 1, DMARQ=N
HDMAQ0:HDMAQ0
HDMAQ2: INTRQ_wait
INTRQ asserted
HDMAQ2:HDMAQ0
BSY = 0 & DRQ = 1& DMARQ asserted
HDMAQ0:HDMAQ1
HDMAQ1: Transfer_Data
All data for command
transferred & nIEN=1
HDMAQ1:HDMAQ0
All data for command transferred & nIEN=0
HDMAQ1:HDMAQ2
Figure 35 − Host DMA QUEUED state diagram
HDMAQ0: Check_Status State: This state is entered when the host has written a READ/WRITE DMA
QUEUED command to the device, when all data for the command has been transferred and nIEN is set to one,
or when all data for the command has been transferred, nIEN is cleared to zero, and INTRQ has been asserted.
It is also entered when the SERVICE command has been written to continue execution of a bus released
command.
Page 283
T13/1321D revision 3
When in this state, the host shall read the device Status register. When entering this state from the HI4, HIO5,
or HIO7 state, the host shall wait 400 ns before reading the Status register. When entering this state from the
HDMAQ1 state, the host shall wait one PIO transfer cycle time before reading the Status register. The wait
may be accomplished by reading the Alternate Status register and ignoring the result. When entering this state
from the DMA QUEUED service return, the host shall check the Tag for the command to be serviced before
making a transition to transfer data.
Transition HDMAQ0:HDMAQ0: When BSY is set to one and DMARQ is negated, the host shall make a
transition to the HDMAQ0: Check_Status state.
Transition HDMAQ0:HDMAQ1: When BSY is cleared to zero, DRQ is set to one, and DMARQ is asserted,
then the host shall make a transition to the HDMAQ1: Transfer_Data state. The host shall have set up the DMA
engine before making this transition.
Transition HDMAQ0:HI0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero,
SERV is cleared to zero, and the device queue is empty, then the command is completed and the host shall
make a transition to the HI0: Host_Idle state (see Figure 19). If an error is reported, the host shall perform
appropriate error recovery.
Transition HDMAQ0a:HIO0: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero,
SERV is cleared to zero, nIEN is cleared to zero, and the device has a queue of released commands, then the
command is completed and the host shall make a transition to the HIO0: Command completed state (see
Figure 19). If an error is reported, the host shall perform appropriate error recovery.
Transition HDMAQ0a:HIO3: When BSY is cleared to zero, DRQ is cleared to zero, REL is cleared to zero,
SERV is cleared to zero, nIEN is set to one, and the device has a queue of released commands, then the
command is completed and the host shall make a transition to the HIO3: Command completed state (see
Figure 19). If an error is reported, the host shall perform appropriate error recovery.
Transition HDMAQ0b:HIO0: When BSY is cleared to zero, DRQ is cleared to zero, REL is set to one, SERV
is cleared to zero, and nIEN is cleared to zero, then the host shall make a transition to the HIO0:
INTRQ_wait_A state (see Figure 20). The bus has been released.
Transition HDMAQ0b:HIO3: When BSY is cleared to zero, DRQ is cleared to zero, REL is set to one, SERV
is cleared to zero, and nIEN is set to one, then the host shall make a transition to the HIO3: Check_status_A
state (see Figure 20). The bus has been released.
Transition HDMAQ0:HIO5: When BSY is cleared to zero, DRQ is cleared to zero, and SERV is set to one,
then the host shall make a transition to the HIO5: Write_SERVICE state (see Figure 20). The command is
completed or the bus has been released, and another queued command is ready for service. If an error is
reported, the host shall perform appropriate error recovery.
HDMAQ1: Transfer_Data State: This state is entered when BSY is cleared to zero, DRQ is set to one,
and DMARQ is asserted.
When in this state, the host shall read or write the device Data port to transfer data. If the bus has been
released, the host shall read the Tag in the Sector Count register to determine the queued command to be
executed and initialize the DMA channel.
Transition HDMAQ1:HDMAQ0: When all data for the request has been transferred and nIEN is set to one,
then the host shall make a transition to the HDMAQ0: Check_Status state.
Transition HDMAQ1:HDMAQ2: When all data for the request has been transferred and nIEN is cleared to zero,
then the host shall make a transition to the HDMAQ2: INTRQ_wait state.
Page 284
T13/1321D revision 3
HDMAQ2: INTRQ_Wait State: This state is entered when the command has completed, and nIEN is
cleared to zero.
When in this state, the host shall wait for INTRQ to be asserted.
Transition HDMAQ2:HDMAQ0: When INTRQ is asserted, the host shall make a transition to the HDMAQ0:
Check_Status state.
DDMAQ0: Prepare
BSY=1, DRQ=0, INTRQ=N, DMARQ=N
Service return
DMA QUEUED command written
xx:DDMAQ0
xx:DDMAQ0
Command complete & nIEN=1 & no queue
Device_Idle_S
DDMAQ0:DI1
BSY=0, REL=0
Command complete & nIEN=1 & queue
Device_Idle_SR
DDMAQ0a:DIO1
BSY=0, REL=0
Command complete & nIEN = 1 & service
required
Device_Idle_SS
DDMAQ0a:DIO3
Device_Idle_SIR
Command complete & nIEN=0 & no queue
DDMAQ0:DI0
Device_Idle_SI
BSY=0, REL=0, INTRQ=A
Command complete & nIEN=0 & queue
DDMAQ0a:DIO0
Device_Idle_SIR
BSY=0, REL=0, INTRQ=A
BSY=0, REL=0, SERV=1
Command complete & nIEN = 0 & service required
DDMAQ0a:DIO2
Device_Idle_SIS
BSY=0, REL=0, SERV=1, INTRQ=A
Bus release & nIEN = 0
DDMAQ0b:DIO0
BSY=0, REL=1, INTRQ=A
Bus release & nIEN = 1
DDMAQ0b:DIO1
BSY=0, REL=1
Bus release & nIEN = 0 & service required
Device_Idle_SIS
DDMAQ0b:DIO2
BSY=0, REL=1, SERV=1, INTRQ=A
Device_Idle_SR
Bus release & nIEN = 1 & service required
DDMAQ0b:DIO3
BSY=0, REL=1, SERV=1
Device_Idle_SS
Ready to transfer DMA data
DDMAQ0:DDMAQ1
DDMAQ1: Transfer_Data
BSY=0, DRQ=1, INTRQ=N, DMARQ=A
Data transfer completed
DDMAQ1:DDMAQ0
BSY
v
DRQ
v
REL
x
SERV
x
C/D
x
I/O
x
INTRQ
V
DMARQ
V
PDIAGR
DASPR
Figure 36 − Device DMA QUEUED command state diagram
DDMAQ0: Prepare State: This state is entered when the device has a READ/WRITE DMA QUEUED or
SERVICE command written to the Command register, when the data has been transferred, or when the
command has completed.
When in this state, device shall set BSY to one, clear DRQ to zero, and negate INTRQ. If the command is a
queued command, the device shall verify that the Tag is valid. If commands are queued, the Tag for the
command to be serviced shall be placed into the Sector Count register.
Page 285
T13/1321D revision 3
Transition DDMAQ0:DDMAQ1: When the device is ready to transfer the data for a command, then the device
shall make a transition to the DDMAQ1: Transfer_Data state.
Transition DDMAQ0:DI0: When the command has completed or an error occurs that causes the command to
abort, the device has no other command released, and nIEN is cleared to zero, then the device shall set the
interrupt pending, set appropriate error bits, clear BSY to zero, assert INTRQ, and make a transition to the DI0:
Device_Idle_SI state (see Figure 21).
Transition DDMAQ0:DI1: When the command has completed or an error occurs that causes the command to
abort, the device has no other command released, and nIEN is set to one, then the device shall set appropriate
error bits, clear BSY to zero, assert INTRQ, and make a transition to the DI1: Device_Idle_S state (see Figure
21).
Transition DDMAQ0a:DIO0: When the command has completed or an error occurs that causes the command
to abort, the device has another command released but not ready for service, and nIEN is cleared to zero, then
the device shall set the interrupt pending, set appropriate error bits, clear BSY to zero, assert INTRQ, and
make a transition to the DIO0: Device_Idle_SIR state (see Figure 22).
Transition DDMAQ0a:DIO1: When the command has completed or an error occurs that causes the command
to abort, the device has another command released but not ready for service, and nIEN is set to one, then the
device shall, set appropriate error bits, clear BSY to zero, and make a transition to the DIO1: Device_Idle_SR
state (see Figure 22).
Transition DDMAQ0a:DIO2: When the command has completed or an error occurs that causes the command
to abort, the device has another command ready for service, and nIEN is cleared to zero, then the device shall
set the interrupt pending, set appropriate error bits, set SERV to one, clear BSY to zero, assert INTRQ, and
make a transition to the DIO2: Device_Idle_SIS state (see Figure 22).
Transition DDMAQ0a:DIO3: When the command has completed or an error occurs that causes the command
to abort, the device has another command ready for service, and nIEN is set to one, then the device shall set
appropriate error bits, set SERV to one, clear BSY to zero, and make a transition to the DIO3: Device_Idle_SS
state (see Figure 22).
Transition DDMAQ0b:DIO0: When the bus is released and nIEN is cleared to zero, then the device shall set
the interrupt pending, set appropriate error bits, set REL to one, clear BSY to zero, assert INTRQ, and make a
transition to the DIO0: Device_Idle_SIR state (see Figure 22).
Transition DDMAQ0b:DIO1: When the bus is released and nIEN is set to one, then the device shall, set
appropriate error bits, set REL to one, clear BSY to zero, and make a transition to the DIO1: Device_Idle_SR
state (see Figure 22).
Transition DDMAQ0b:DIO2: When the bus is released, the device has another command ready for service,
and nIEN is cleared to zero, then the device shall set the interrupt pending, set appropriate error bits, set REL
to one, set SERV to one, clear BSY to zero, assert INTRQ, and make a transition to the DIO2: Device_Idle_SIS
state (see Figure 22).
Transition DDMAQ0b:DIO3: When the bus is released, the device has another command ready for service,
and nIEN is set to one, then the device shall set appropriate error bits, set REL to one, set SERV to one, clear
BSY to zero, and make a transition to the DIO3: Device_Idle_SS state (see Figure 22).
DDMAQ1: Data_Transfer State: This state is entered when the device is ready to transfer DMA data.
When in this state, BSY is cleared to zero, DRQ is set to one, INTRQ is negated, DMARQ is asserted, and
data is transferred as described in Multiword DMA timing or Ultra DMA protocol.
Page 286
T13/1321D revision 3
Transition DDMAQ1:DDMAQ0: When the data transfer has been completed, then the device shall make a
transition to the DDMAQ0: Prepare state.
9.10 EXECUTE DEVICE DIAGNOSTIC command protocol
This class includes:
−
EXECUTE DEVICE DIAGNOSTIC
If the host asserts RESET- before devices have completed executing their EXECUTE DEVICE DIAGNOSTIC
protocol, then the devices shall start executing the power-on or hardware reset protocol from the beginning.
If the host sets SRST to one in the Device Control register before the devices have completed execution of their
EXECUTE DEVICE DIAGNOSTIC protocol, then the devices shall start executing their software reset protocol
from the beginning.
Figure 37 and the text following the figure describe the EXECUTE DEVICE DIAGNOSTIC protocol for the host.
Figure 38 and the text following the figure describe the EXECUTE DEVICE DIAGNOSTIC protocol for Device 0.
Figure 39 and the text following the figure describe the EXECUTE DEVICE DIAGNOSTIC protocol for Device 1.
HED0: Wait
EXECUTE DEVICE DIAGNOSTIC
command written & nIEN=1
HI4:HED0
HED2: Check_status
2 ms timeout complete
HED0:HED2
BSY = 0
HED2:HI0
Host_idle
BSY = 1
HED2:HED2
HED1: INTRQ_Wait
EXECUTE DEVICE DIAGNOSTIC
command written & nIEN=0
HI4:HED1
INTRQ asserted
HED1:HED2
Figure 37 − Host EXECUTE DEVICE DIAGNOSTIC state diagram
HED0: Wait State: This state is entered when the host has written the EXECUTE DEVICE DIAGNOSTIC
command to the devices and nIEN is set to one.
The host shall remain in this state for at least 2 ms.
Transition HED0:HED1: When at least 2 ms has elapsed since the command was written, the host shall make
a transition to the HED1: Check_status state.
HED1: INTRQ_wait: This state is entered when the host has written the EXECUTE DEVICE DIAGNOSTIC
command to the devices and nIEN is cleared to zero.
When in this state the host shall wait for INTRQ to be asserted.
Transition HED1:HED2: When INTRQ is asserted, the host shall make a transition to the HED2: Check_status
state.
HED2: Check_status State: This state is entered when at least 2 ms since the command was written
or INTRQ has been asserted.
Page 287
T13/1321D revision 3
When in this state, the host shall read the Status or Alternate Status register.
Transition HED2:HED2: When BSY is set to one, the host shall make a transition to the HED1: Check_status
state.
Transition HED2:HI0: When BSY is cleared to zero, the host shall check the results of the command (see
9.16) and make a transition to the HI0: Host_idle state (see Figure 19).
D0ED0: Release_bus
EXECUTE DEVICE DIAGNOSTIC
command written
DI0:D0ED0
DI1:D0ED0
Bus released & no Device 1
D0ED0:D0ED3
Clear bit 7
Bus released & Device 1 exists
D0ED0:D0ED1
D0ED1: PDIAG-_wait
1 ms wait complete
D0ED1:D0ED2
D0ED2: Sample_PDIAG-
Resample PDIAGD0ED2:D0ED2
D0ED3: Set_status
PDIAG- asserted
D0ED2a:D0ED3
Clear bit 7
6 s timeout
D0ED2b:D0ED3
Set bit 7
BSY
1
DRQ
0
REL
0
SERV
0
C/D
0
Status set & nIEN=1
D0ED3:DI1
BSY=0, INTRQ=A
Device_idle_S
Status set & nIEN=0
D0ED3:DI0
BSY=0
Device_idle_SI
I/O
0
INTRQ
N
DMARQ
R
PDIAGR
DASPR
Figure 38 − Device 0 EXECUTE DEVICE DIAGNOSTIC state diagram
D0ED0: Release_bus State: This state is entered when the EXECUTE DEVICE DIAGNOSTIC
command has been written to Device 0.
When in this state, the device shall release PDIAG-, INTRQ, IORDY, DMARQ, and DD(15:0) and shall set BSY
to one and clear DEV to zero within 400 ns after entering this state.
The device should begin performing the self-diagnostic testing.
Transition D0ED0:D0ED1: When the bus has been released, BSY set to one, and the assertion of DASP- by
Device 1 was detected during the most recent power-on or hardware reset, then the device shall make a
transition to the D0ED1: PDIAG-_wait state.
Transition D0ED0:D0ED3: When the bus has been released, BSY set to one, and the assertion of DASP- by
Device 1 was not detected during the most recent power-on or hardware reset, then the device shall clear bit 7
in the Error register and make a transition to the D0ED3: Set_status state.
D0ED1: PDIAG-_wait State: This state is entered when the bus has been released, BSY set to one,
and Device 1 exists.
Page 288
T13/1321D revision 3
The device shall remain in this state until least 1 ms has elapsed since the command was written.
Transition D0ED1:D0ED2: When at least 1 ms has elapsed since the command was written, the device shall
make a transition to the D0ED2: Sample_PDIAG- state.
D0ED2: Sample_PDIAG- State: This state is entered when at least 1 ms has elapsed since the
command was written.
When in this state, the device shall sample the PDIAG- signal.
Transition D0ED2:D0ED3: When the sample indicates that PDIAG- is asserted, the device shall clear bit 7 in
the Error register and make a transition to the D0ED3: Set_status state.
Transition D0ED2:D0ED2: When the sample indicates that PDIAG- is not asserted and less than 6 s have
elapsed since the command was written, then the device shall make a transition to the D0ED2:
Sample_PDIAG- state.
Transition D0ED2:D0ED3: When the sample indicates that DASP- is not asserted and 6 s have elapsed since
the command was written, then the device shall set bit 7 in the Error register and make a transition to the
D0ED3: Set_status state.
D0ED3: Set_status State: This state is entered when Bit 7 in the Error register has been set or cleared.
When in this state the device shall complete the self-diagnostic testing begun in the Release bus state if not
already completed.
Results of the self-diagnostic testing shall be placed in bits 6-0 of the Error register (see Table 19). The device
shall set the signature values (see 9.12). The contents of the Features register is undefined.
If the device does not implement the PACKET command feature set, the device shall clear bits 3, 2, and 0 in
the Status register to zero.
If the device implements the PACKET command feature set, the device shall clear bits 6, 5, 4, 3, 2, and 0 in
the Status register to zero. The device shall return the operating modes to their specified initial conditions.
MODE SELECT conditions shall be restored to their last saved values if saved values have been established.
MODE SELECT conditions for which no values have been saved shall be returned to their default values.
Transition D0ED3:DI1: When hardware initialization and self-diagnostic testing is completed, the status has
been set, and nIEN is set to one, then the device shall clear BSY to zero, and make a transition to the DI1:
Device_idle_S state (see Figure 21).
Transition D0ED3:DI0: When hardware initialization and self-diagnostic testing is completed, the status has
been set, and nIEN is cleared to zero, then the device shall clear BSY to zero, assert INTRQ, and make a
transition to the DI0: Device_idle_SI state (see Figure 21).
Page 289
T13/1321D revision 3
D1ED0: Release_bus
PDIAG=x
EXECUTE DEVICE
DIAGNOSTIC command written
DI0:D1ED0
DI1:D1ED0
D1ED1: Negate_PDIAGPDIAG- =R
Bus released
D1ED0:D1ED1
D1ED2: Set_status
PDIAG- =N
PDIAG- negated
D1ED1:D1ED2
Status set
D1ED2:DI2
BSY=0, PDIAG-=A
BSY
1
DRQ
0
REL
0
SERV
0
C/D
0
I/O
0
Device_idle_NS
INTRQ
N
DMARQ
R
PDIAGV
DASPR
Figure 39 − Device 1 EXECUTE DEVICE DIAGNOSTIC command state diagram
D1ED0: Release_bus State: This state is entered when the EXECUTE DEVICE DIAGNOSTIC
command is written to Device 1.
When in this state, the device shall release INTRQ, IORDY, DMARQ, and DD(15:0) within 400 ns after entering
this state. The device shall set BSY to one and clear DEV to zero within 400 ns after entering this state.
The device should begin performing the self-diagnostic testing.
Transition D1ED0:D1ED1: When the bus has been released and BSY set to one, then the device shall make a
transition to the D1ED1: Negate_PDIAG- state.
D1ED1: Negate_PDIAG- State: This state is entered when the bus has been released and BSY set to
one.
When in this state, the device shall negate PDIAG- within less than 1 ms of the receipt of the EXECUTE
DEVICE DIAGNOSTIC command.
Transition D1ED1:D1ED2: When PDIAG- has been negated, the device shall make a transition to the D1ED2:
Set_status state.
D1ED2: Set_status State: This state is entered when the device has negated PDIAG-.
When in this state the device shall complete the hardware initialization and self-diagnostic testing begun in the
Release bus state if not already completed. Results of the self-diagnostic testing shall be placed in the Error
register (see Table 19). If the device passed the self-diagnostics, the device shall assert PDIAG-.
The device shall set the signature values (see 9.12). The effect on the Features register is undefined.
If the device does not implement the PACKET command feature set, the device shall clear bits 3, 2, and 0 in
the Status register to zero.
If the device implements the PACKET command feature set, the device shall clear bits 6, 5, 4, 3, 2, and 0 in
the Status register to zero. The device shall return the operating modes to their specified initial conditions.
MODE SELECT conditions shall be restored to their last saved values if saved values have been established.
MODE SELECT conditions for which no values have been saved shall be returned to their default values.
Page 290
T13/1321D revision 3
A requirements for this state shall be completed within 5 s or less from the writing of the command.
Transition D1ED2:DI2: When hardware initialization and self-diagnostic testing is completed and the status
has been set, then the device shall clear BSY to zero, assert PDIAG- if diagnostics were passed, and make a
transition to the DI2: Device_idle_NS state (see Figure 21).
9.11 DEVICE RESET command protocol
This class includes:
−
DEVICE RESET
If the host asserts RESET- before the device has completed executing a DEVICE RESET command, then the
device shall start executing the hardware reset protocol from the begining. If the host sets the SRST bit to one
in the Device Control register before the device has completed executing a DEVICE RESET command, the
device shall start executing the software reset protocol from the beginning.
The host should not issue a DEVICE RESET command while a DEVICE RESET command is in progress. If the
host issues a DEVICE RESET command while a DEVICE RESET command is in progress, the results are
indeterminate.
Figure 40 and the text following the figure describe the DEVICE RESET command protocol for the host. Figure
41 and the text following the figure describe the DEVICE RESET command protocol for the device.
HDR0: Wait
DEVICE RESET
command written
HDR1: Check_status
400 ns timeout complete
HDR0:HDR1
HI4:HDR0
BSY = 0
HDR1:HI0
Host_idle
BSY = 1
HDR1:HDR1
Figure 40 − Host DEVICE RESET command state diagram
HDR0: Wait State: This state is entered when the host has written the DEVICE RESET command to the
device.
The host shall remain in this state for at least 400 ns.
Transition HDR0:HDR1: When at least 400 ns has elapsed since the command was written, the host shall
make a transition to the HDR1: Check_status state.
HDR1: Check_status State: This state is entered when at least 400 ns has elapsed since the
command was written.
When in this state the host shall read the Status register.
Transition HDR1:HDR1: When BSY is set to one, the host shall make a transition to the HDR1: Check_status
state.
Page 291
T13/1321D revision 3
Transition HDR1:HI0: When BSY is cleared to zero, the host shall make a transition to the HI0: Host_idle
state (see Figure 19). If status indicates that an error has occurred, the host shall take appropriate action.
DDR0: Release_bus
DEVICE RESET
command written
DDR1: Set_status
Status set
DDR1:DI1
BSY=0
Bus released
DDR1:DDR2
DI0:DDR0
DI1:DDR0
BSY
1
DRQ
0
REL
0
SERV
0
C/D
0
I/O
0
INTRQ
N
DMARQ
R
Device_idle_S
PDIAGR
DASPR
Figure 41 − Device DEVICE RESET command state diagram
DDR0: Release_bus State: This state is entered when the DEVICE RESET command is written.
When in this state, the device shall release INTRQ, IORDY, DMARQ, and DD(15:0) within 400 ns after entering
this state. The device shall set BSY to one within 400 ns after entering this state.
Transition DDR0:DDR1: When the bus has been released and BSY set to one, the device shall make a
transition to the DDR1: Set_status state.
DDR1: Set_status State: This state is entered when the device has released the bus and set BSY to
one.
When in this state the device should stop execution of any uncompleted command. The device should end
background activity (e.g., immediate commands, see MMC and MMC-2).
The device should not revert to the default condition. If the device reverts to the default condition, the device
shall report an exception condition by setting CHK to one in the Status register. MODE SELECT conditions
shall not be altered.
The device shall set the signature values (see 9.12). The content of the Features register is undefined.
The device shall clear bit 7 in the ERROR register to zero. The device shall clear bits 6, 5, 4, 3, 2, and 0 in the
Status register to zero.
Transition DDR1:DI1: When the status has been set, the device shall clear BSY to zero and make a transition
to the DI1: Device_idle_S state (see Figure 21).
9.12 Signature and persistence
A device not implementing the PACKET command feature set shall place the signature in the Command Block
registers listed below for power-on reset, hardware reset, software reset, and the EXECUTE DEVICE
DIAGNOSTIC command.
If the device does not implement the PACKET command feature set, the signature shall be:
Sector Count
01h
Sector Number
01h
Cylinder Low
00h
Cylinder High
00h
Device/Head
00h
Page 292
T13/1321D revision 3
A device implementing the PACKET command feature set shall place the signature in the Command Block
registers listed below for power-on reset, hardware reset, software reset, the EXECUTE DEVICE DIAGNOSTIC
command, and the DEVICE RESET command. The DEVICE RESET command shall not change the value of
the DEV bit when writing the signature into the Device/Head register for a device implementing the PACKET
command feature set. If the device implements the PACKET command feature set, the signature is also written
in the registers for the IDENTIFY DEVICE and READ SECTOR(S) commands.
If the device implements the PACKET command feature set, the signature shall be:
Sector Count
01h
Sector Number
01h
Cylinder Low
14h
Cylinder High
EBh
Device/Head
000x0000b where x equals 0
except when responding to a
DEVICE RESET command. For
the DEVICE RESET command the
value of x is not changed from that
existing when the command is
written to the Command register.
If the PACKET command feature set is implemented by a device, then the signature values written by the
device in the Command Block registers following power-on reset, hardware reset, software reset, or the DEVICE
RESET command shall not be changed by the device until the device receives a command that sets DRDY to
one. Writes by the host to the Command Block registers that contain the signature values shall overwrite the
signature values and invalidate the signature.
9.13 Ultra DMA data-in commands
9.13.1 Initiating an Ultra DMA data-in burst
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and10.2.4.1.
a) The host shall keep DMACK- in the negated state before an Ultra DMA burst is initiated.
b) The device shall assert DMARQ to initiate an Ultra DMA burst when DMACK- is negated. After assertion of
DMARQ the device shall not negate DMARQ until after the first negation of DSTROBE.
c) Steps (c), (d), and (e) may occur in any order or at the same time. The host shall assert STOP.
d) The host shall negate HDMARDY-.
e) The host shall negate CS0-, CS1-, DA2, DA1, and DA0. The host shall keep CS0-, CS1-, DA2, DA1, and
DA0 negated until after negating DMACK- at the end of the burst.
f) Steps (c), (d), and (e) shall have occurred at least tACK before the host asserts DMACK-. The host shall
keep DMACK- asserted until the end of an Ultra DMA burst.
g) The host shall release DD(15:0) within tAZ after asserting DMACK-.
h) The device may assert DSTROBE tZIORDY after the host has asserted DMACK-. Once the device has driven
DSTROBE the device shall not release DSTROBE until after the host has negated DMACK- at the end of
an Ultra DMA burst.
i) The host shall negate STOP and assert HDMARDY- within tENV after asserting DMACK-. After negating
STOP and asserting HDMARDY-, the host shall not change the state of either signal until after receiving
the first negation of DSTROBE from the device (i.e., after the first data word has been received).
j) The device shall drive DD(15:0) no sooner than tZAD after the host has asserted DMACK-, negated STOP,
and asserted HDMARDY-.
k) The device shall drive the first word of the data transfer onto DD(15:0). This step may occur when the
device first drives DD(15:0) in step (j).
l) To transfer the first word of data the device shall negate DSTROBE within tFS after the host has negated
STOP and asserted HDMARDY-. The device shall negate DSTROBE no sooner than tDVS after driving the
first word of data onto DD(15:0).
Page 293
T13/1321D revision 3
9.13.2 The data-in transfer
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.2.
a) The device shall drive a data word onto DD(15:0).
b) The device shall generate a DSTROBE edge to latch the new word no sooner than tDVS after changing the
state of DD(15:0). The device shall generate a DSTROBE edge no more frequently than tCYC for the
selected Ultra DMA mode. The device shall not generate two rising or two falling DSTROBE edges more
frequently than t2cyc for the selected Ultra DMA mode.
c) The device shall not change the state of DD(15:0) until at least tDVH after generating a DSTROBE edge to
latch the data.
d) The device shall repeat steps (a), (b), and (c) until the Ultra DMA burst is paused or terminated by the
device or host.
9.13.3 Pausing an Ultra DMA data-in burst
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.3.
9.13.3.1 Device pausing an Ultra DMA data-in burst
a) The device shall not pause an Ultra DMA burst until at least one data word of an Ultra DMA burst has been
transferred.
b) The device shall pause an Ultra DMA burst by not generating additional DSTROBE edges. If the host is
ready to terminate the Ultra DMA burst, see 9.13.4.2.
c) The device shall resume an Ultra DMA burst by generating a DSTROBE edge.
9.13.3.2 Host pausing an Ultra DMA data-in burst
a) The host shall not pause an Ultra DMA burst until at least one data word of an Ultra DMA burst has been
transferred.
b) The host shall pause an Ultra DMA burst by negating HDMARDY-.
c) The device shall stop generating DSTROBE edges within tRFS of the host negating HDMARDY-.
d) When operating in Ultra DMA modes 2, 1, or 0: If the host negates HDMARDY- within tSR after the device
has generated a DSTROBE edge, then the host shall be prepared to receive zero or one additional data
words; or, if the host negates HDMARDY- greater than tSR after the device has generated a DSTROBE
edge, then the host shall be prepared to receive zero, one or two additional data words. While operating in
Ultra DMA modes 4 or 3 the host shall be prepared to receive zero, one, two or three additional data words
after negating HDMARDY-. The additional data words are a result of cable round trip delay and tRFS timing
for the device.
e) The host shall resume an Ultra DMA burst by asserting HDMARDY-.
9.13.4 Terminating an Ultra DMA data-in burst
9.13.4.1 Device terminating an Ultra DMA data-in burst
Burst termination is completed when the termination protocol has been executed and DMACK- negated.
The device shall terminate an Ultra DMA burst before command completion.
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.4.
a) The device shall initiate termination of an Ultra DMA burst by not generating additional DSTROBE edges.
Page 294
T13/1321D revision 3
b) The device shall negate DMARQ no sooner than tSS after generating the last DSTROBE edge. The device
shall not assert DMARQ again until after DMACK- has been negated.
c) The device shall release DD(15:0) no later than tAZ after negating DMARQ.
d) The host shall assert STOP within tLI after the device has negated DMARQ. The host shall not negate
STOP again until after the Ultra DMA burst is terminated.
e) The host shall negate HDMARDY- within tLI after the device has negated DMARQ. The host shall continue
to negate HDMARDY- until the Ultra DMA burst is terminated. Steps (d) and (e) may occur at the same
time.
f) The host shall drive DD(15:0) no sooner than tZAH after the device has negated DMARQ. For this step, the
host may first drive DD(15:0) with the result of the host CRC calculation (see 9.15);
g) If DSTROBE is negated, the device shall assert DSTROBE within tLI after the host has asserted STOP. No
data shall be transferred during this assertion. The host shall ignore this transition on DSTROBE.
DSTROBE shall remain asserted until the Ultra DMA burst is terminated.
h) If the host has not placed the result of the host CRC calculation on DD(15:0) since first driving DD(15:0)
during (f), the host shall place the result of the host CRC calculation on DD(15:0) (see 9.15).
i) The host shall negate DMACK- no sooner than tMLI after the device has asserted DSTROBE and negated
DMARQ and the host has asserted STOP and negated HDMARDY-, and no sooner than tDVS after the host
places the result of the host CRC calculation on DD(15:0).
j) The device shall latch the host’s CRC data from DD(15:0) on the negating edge of DMACK-.
k) The device shall compare the CRC data received from the host with the results of the device CRC
calculation. If a miscompare error occurs during one or more Ultra DMA bursts for any one command, at
the end of the command the device shall report the first error that occurred (see 9.15).
l) The device shall release DSTROBE within tIORDYZ after the host negates DMACK-.
m) The host shall not negate STOP nor assert HDMARDY- until at least tACK after negating DMACK-.
n) The host shall not assert DIOR-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after negating DMACK.
9.13.4.2 Host terminating an Ultra DMA data-in burst
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.5.
a) The host shall not initiate Ultra DMA burst termination until at least one data word of an Ultra DMA burst
has been transferred.
b) The host shall initiate Ultra DMA burst termination by negating HDMARDY-. The host shall continue to
negate HDMARDY- until the Ultra DMA burst is terminated.
c) The device shall stop generating DSTROBE edges within tRFS of the host negating HDMARDY-.
d) When operating in Ultra DMA modes 2, 1, or 0: If the host negates HDMARDY- within tSR after the device
has generated a DSTROBE edge, then the host shall be prepared to receive zero or one additional data
words; or, if the host negates HDMARDY- greater than tSR after the device has generated a DSTROBE
edge, then the host shall be prepared to receive zero, one or two additional data words. While operating in
Ultra DMA modes 4 or 3 the host shall be prepared to receive zero, one, two or three additional data words
after negating HDMARDY-. The additional data words are a result of cable round trip delay and tRFS timing
for the device.
e) The host shall assert STOP no sooner than tRP after negating HDMARDY-. The host shall not negate
STOP again until after the Ultra DMA burst is terminated.
f) The device shall negate DMARQ within tLI after the host has asserted STOP. The device shall not assert
DMARQ again until after the Ultra DMA burst is terminated.
g) If DSTROBE is negated, the device shall assert DSTROBE within tLI after the host has asserted STOP. No
data shall be transferred during this assertion. The host shall ignore this transition on DSTROBE.
DSTROBE shall remain asserted until the Ultra DMA burst is terminated.
h) The device shall release DD(15:0) no later than tAZ after negating DMARQ.
i) The host shall drive DD(15:0) no sooner than tZAH after the device has negated DMARQ. For this step, the
host may first drive DD(15:0) with the result of the host CRC calculation (see 9.15).
j) If the host has not placed the result of the host CRC calculation on DD(15:0) since first driving DD(15:0)
during (9), the host shall place the result of the host CRC calculation on DD(15:0) (see 9.15).
Page 295
T13/1321D revision 3
k) The host shall negate DMACK- no sooner than tMLI after the device has asserted DSTROBE and negated
DMARQ and the host has asserted STOP and negated HDMARDY-, and no sooner than tDVS after the host
places the result of the host CRC calculation on DD(15:0).
l) The device shall latch the host’s CRC data from DD(15:0) on the negating edge of DMACK-.
m) The device shall compare the CRC data received from the host with the results of the device CRC
calculation. If a miscompare error occurs during one or more Ultra DMA burst for any one command, at the
end of the command, the device shall report the first error that occurred (see 9.15) .
n) The device shall release DSTROBE within tIORDYZ after the host negates DMACK-.
o) The host shall neither negate STOP nor assert HDMARDY- until at least tACK after the host has negated
DMACK-.
p) The host shall not assert DIOR-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after negating DMACK.
9.14 Ultra DMA data-out commands
9.14.1 Initiating an Ultra DMA data-out burst
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.6.
a)
b)
c)
d)
e)
The host shall keep DMACK- in the negated state before an Ultra DMA burst is initiated.
The device shall assert DMARQ to initiate an Ultra DMA burst when DMACK- is negated.
Steps (c), (d), and (e) may occur in any order or at the same time. The host shall assert STOP.
The host shall assert HSTROBE.
The host shall negate CS0-, CS1-, DA2, DA1, and DA0. The host shall keep CS0-, CS1-, DA2, DA1, and
DA0 negated until after negating DMACK- at the end of the burst.
f) Steps (c), (d), and (e) shall have occurred at least tACK before the host asserts DMACK-. The host shall
keep DMACK- asserted until the end of an Ultra DMA burst.
g) The device may negate DDMARDY- tZIORDY after the host has asserted DMACK-. Once the device has
negated DDMARDY-, the device shall not release DDMARDY- until after the host has negated DMACK- at
the end of an Ultra DMA burst.
h) The host shall negate STOP within tENV after asserting DMACK-. The host shall not assert STOP until after
the first negation of HSTROBE.
i) The device shall assert DDMARDY- within tLI after the host has negated STOP.
After asserting DMARQ
and DDMARDY- the device shall not negate either signal until after the first negation of HSTROBE by the
host.
j) The host shall drive the first word of the data transfer onto DD(15:0). This step may occur any time during
Ultra DMA burst initiation.
k) To transfer the first word of data: the host shall negate HSTROBE no sooner than tUI after the device has
asserted DDMARDY-. The host shall negate HSTROBE no sooner than tDVS after the driving the first word
of data onto DD(15:0).
9.14.2 The data-out transfer
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.7.
a) The host shall drive a data word onto DD(15:0).
b) The host shall generate an HSTROBE edge to latch the new word no sooner than tDVS after changing the
state of DD(15:0). The host shall generate an HSTROBE edge no more frequently than tCYC for the selected
Ultra DMA mode. The host shall not generate two rising or falling HSTROBE edges more frequently than
t2cyc for the selected Ultra DMA mode.
c) The host shall not change the state of DD(15:0) until at least tDVH after generating an HSTROBE edge to
latch the data.
d) The host shall repeat steps (a), (b), and (c) until the Ultra DMA burst is paused or terminated by the device
or host.
Page 296
T13/1321D revision 3
9.14.3 Pausing an Ultra DMA data-out burst
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.8.
9.14.3.1 Host pausing an Ultra DMA data-out burst
a) The host shall not pause an Ultra DMA burst until at least one data word of an Ultra DMA burst has been
transferred.
b) The host shall pause an Ultra DMA burst by not generating an HSTROBE edge. If the host is ready to
terminate the Ultra DMA burst, see 9.14.4.1.
c) The host shall resume an Ultra DMA burst by generating an HSTROBE edge.
9.14.3.2 Device pausing an Ultra DMA data-out burst
a) The device shall not pause an Ultra DMA burst until at least one data word of an Ultra DMA burst has been
transferred.
b) The device shall pause an Ultra DMA burst by negating DDMARDY-.
c) The host shall stop generating HSTROBE edges within tRFS of the device negating DDMARDY-.
d) When operating in Ultra DMA modes 2, 1, or 0: If the device negates DDMARDY- within tSR after the host
has generated an HSTROBE edge, then the device shall be prepared to receive zero or one additional data
words; or, if the device negates DDMARDY- greater than tSR after the host has generated an HSTROBE
edge, then the device shall be prepared to receive zero, one or two additional data words. While operating
in Ultra DMA modes 4 or 3 the device shall be prepared to receive zero, one, two or three additional data
words after negating DDMARDY-. The additional data words are a result of cable round trip delay and tRFS
timing for the host.
e) The device shall resume an Ultra DMA burst by asserting DDMARDY-.
9.14.4 Terminating an Ultra DMA data-out burst
9.14.4.1 Host terminating an Ultra DMA data-out burst
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.9.
a) The host shall initiate termination of an Ultra DMA burst by not generating additional HSTROBE edges.
b) The host shall assert STOP no sooner than tSS after the last generated an HSTROBE edge. The host shall
not negate STOP again until after the Ultra DMA burst is terminated.
c) The device shall negate DMARQ within tLI after the host asserts STOP. The device shall not assert
DMARQ again until after the Ultra DMA burst is terminated.
d) The device shall negate DDMARDY- within tLI after the host has negated STOP. The device shall not assert
DDMARDY- again until after the Ultra DMA burst termination is complete.
e) If HSTROBE is negated, the host shall assert HSTROBE within tLI after the device has negated DMARQ.
No data shall be transferred during this assertion. The device shall ignore this transition on HSTROBE.
HSTROBE shall remain asserted until the Ultra DMA burst is terminated.
f) The host shall place the result of the host CRC calculation on DD(15:0) (see 9.15).
g) The host shall negate DMACK- no sooner than tMLI after the host has asserted HSTROBE and STOP and
the device has negated DMARQ and DDMARDY-, and no sooner than tDVS after placing the result of the
host CRC calculation on DD(15:0).
h) The device shall latch the host’s CRC data from DD(15:0) on the negating edge of DMACK-.
i) The device shall compare the CRC data received from the host with the results of the device CRC
calculation. If a miscompare error occurs during one or more Ultra DMA bursts for any one command, at
the end of the command, the device shall report the first error that occurred (see 9.15).
j) The device shall release DDMARDY- within tIORDYZ after the host has negated DMACK-.
k) The host shall neither negate STOP nor negate HSTROBE until at least tACK after negating DMACK-.
l) The host shall not assert DIOW-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after negating DMACK.
Page 297
T13/1321D revision 3
9.14.4.2 Device terminating an Ultra DMA data-out burst
Burst termination is completed when the termination protocol has been executed and DMACK- negated.
The device shall terminate an Ultra DMA burst before command completion.
The following steps shall occur in the order they are listed unless otherwise specified. Timing requirements are
shown in 10.2.4 and 10.2.4.10.
a) The device shall not initiate Ultra DMA burst termination until at least one data word of an Ultra DMA burst
has been transferred.
b) The device shall initiate Ultra DMA burst termination by negating DDMARDY-.
c) The host shall stop generating an HSTROBE edges within tRFS of the device negating DDMARDY-.
d) When operating in Ultra DMA modes 2, 1, or 0: If the device negates DDMARDY- within tSR after the host
has generated an HSTROBE edge, then the device shall be prepared to receive zero or one additional data
words; or, if the device negates DDMARDY- greater than tSR after the host has generated an HSTROBE
edge, then the device shall be prepared to receive zero, one or two additional data words. While operating
in Ultra DMA modes 4 or 3 the device shall be prepared to receive zero, one, two or three additional data
words after negating DDMARDY-. The additional data words are a result of cable round trip delay and tRFS
timing for the host.
e) The device shall negate DMARQ no sooner than tRP after negating DDMARDY-. The device shall not assert
DMARQ again until after DMACK- is negated.
f) The host shall assert STOP within tLI after the device has negated DMARQ. The host shall not negate
STOP again until after the Ultra DMA burst is terminated.
g) If HSTROBE is negated, the host shall assert HSTROBE within tLI after the device has negated DMARQ.
No data shall be transferred during this assertion. The device shall ignore this transition of HSTROBE.
HSTROBE shall remain asserted until the Ultra DMA burst is terminated.
h) The host shall place the result of the host CRC calculation on DD(15:0) (see 9.15).
i) The host shall negate DMACK- no sooner than tMLI after the host has asserted HSTROBE and STOP and
the device has negated DMARQ and DDMARDY-, and no sooner than tDVS after placing the result of the
host CRC calculation on DD(15:0).
j) The device shall latch the host’s CRC data from DD(15:0) on the negating edge of DMACK-.
k) The device shall compare the CRC data received from the host with the results of the device CRC
calculation. If a miscompare error occurs during one or more Ultra DMA bursts for any one command, at
the end of the command, the device shall report the first error that occurred (see 9.15).
l) The device shall release DDMARDY- within tIORDYZ after the host has negated DMACK-.
m) The host shall neither negate STOP nor HSTROBE until at least tACK after negating DMACK-.
n) The host shall not assert DIOW-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after negating DMACK.
9.15 Ultra DMA CRC rules
The following is a list of rules for calculating CRC, determining if a CRC error has occurred during an Ultra DMA
burst, and reporting any error that occurs at the end of a command.
1) Both the host and the device shall have a 16-bit CRC calculation function.
2) Both the host and the device shall calculate a CRC value for each Ultra DMA burst.
3) The CRC function in the host and the device shall be initialized with a seed of 4ABAh at the beginning of an
Ultra DMA burst before any data is transferred.
4) For each STROBE transition used for data transfer, both the host and the device shall calculate a new CRC
value by applying the CRC polynomial to the current value of their individual CRC functions and the word
being transferred. CRC is not calculated for the return of STROBE to the asserted state after the Ultra
DMA burst termination request has been acknowledged.
5) At the end of any Ultra DMA burst the host shall send the results of the host CRC calculation function to
the device on DD(15:0) with the negation of DMACK-.
Page 298
T13/1321D revision 3
6)
7)
8)
9)
10)
11)
The device shall then compare the CRC data from the host with the calculated value in its own CRC
calculation function. If the two values do not match, the device shall save the error. A subsequent Ultra
DMA burst for the same command that does not have a CRC error shall not clear an error saved from a
previous Ultra DMA burst in the same command. If a miscompare error occurs during one or more Ultra
DMA bursts for any one command, the device shall report the first error that occurred. If the device detects
that a CRC error has occurred before data transfer for the command is complete, the device may complete
the transfer and report the error or abort the command and report the error.
For READ DMA, WRITE DMA, READ DMA QUEUED, or WRITE DMA QUEUED commands: When a
CRC error is detected, the error shall be reported by setting both ICRC and ABRT (bit 7 and bit 2 in the
Error register) to one. ICRC is defined as the “Interface CRC Error” bit. The host shall respond to this error
by re-issuing the command.
For a REQUEST SENSE packet command (see SPC T10/955D for definition of the REQUEST SENSE
command): When a CRC error is detected during transmission of sense data the device shall complete the
command and set CHK to one. The device shall report a Sense key of 0Bh (ABORTED COMMAND). The
device shall preserve the original sense data that was being returned when the CRC error occurred. The
device shall not report any additional sense data specific to the CRC error. The host device driver may retry
the REQUEST SENSE command or may consider this an unrecoverable error and retry the command that
caused the Check Condition.
For any packet command except a REQUEST SENSE command: If a CRC error is detected, the device
shall complete the command with CHK set to one. The device shall report a Sense key of 04h
(HARDWARE ERROR). The sense data supplied via a subsequent REQUEST SENSE command shall
report an ASC/ASCQ value of 08h/03h (LOGICAL UNIT COMMUNICATION CRC ERROR). Host drivers
should retry the command that resulted in a HARDWARE ERROR.
A host may send extra data words on the last Ultra DMA burst of a data-out command. If a device
determines that all data has been transferred for a command, the device shall terminate the burst. A device
may have already received more data words than were required for the command. These extra words are
used by both the host and the device to calculate the CRC, but, on an Ultra DMA data-out burst, the extra
words shall be discarded by the device.
The CRC generator polynomial is: G(X) = X16 + X12 + X5 + 1. Table 47 describes the equations for 16-bit
parallel generation of the resulting polynomial (based on a word boundary).
NOTE − Since no bit clock is available, the recommended approach for calculating CRC is to
use a word clock derived from the bus strobe. The combinational logic is then equivalent to
shifting sixteen bits serially through the generator polynomial where DD0 is shifted in first and
DD15 is shifted in last.
NOTE − If excessive CRC errors are encountered while operating in an Ultra mode, the host
should select a slower Ultra mode. Caution: CRC errors are detected and reported only while
operating in an Ultra mode.
Page 299
T13/1321D revision 3
DD(15:0)
CRCOUT (15:0)
CRCIN (15:0:)
Combinational
Logic
Edge
Triggered
Register
Device
f1-f16
Word
Clock
Figure 42 − Example Parallel CRC generator
Table 47 − Equations for parallel generation of a CRC polynomial
CRCIN0 = f16
CRCIN8 = f8 XOR f13
CRCIN1 = f15
CRCIN9 = f7 XOR f12
CRCIN2 = f14
CRCIN10 = f6 XOR f11
CRCIN3 = f13
CRCIN11 = f5 XOR f10
CRCIN4 = f12
CRCIN12 = f4 XOR f9 XOR f16
CRCIN5 = f11 XOR f16
CRCIN13 = f3 XOR f8 XOR f15
CRCIN6 = f10 XOR f15
CRCIN14 = f2 XOR f7 XOR f14
CRCIN7 = f9 XOR f14
CRCIN15 = f1 XOR f6 XOR f13
f1 = DD0 XOR CRCOUT15
f9 = DD8 XOR CRCOUT7 XOR f5
f2 = DD1 XOR CRCOUT14
f10 = DD9 XOR CRCOUT6 XOR f6
f3 = DD2 XOR CRCOUT13
f11 = DD10 XOR CRCOUT5 XOR f7
f4 = DD3 XOR CRCOUT12
f12 = DD11 XOR CRCOUT4 XOR f1 XOR f8
f5 = DD4 XOR CRCOUT11 XOR f1
f13 = DD12 XOR CRCOUT3 XOR f2 XOR f9
f6 = DD5 XOR CRCOUT10 XOR f2
f14 = DD13 XOR CRCOUT2 XOR f3 XOR f10
f7 = DD6 XOR CRCOUT9 XOR f3
f15 = DD14 XOR CRCOUT1 XOR f4 XOR f11
f8 = DD7 XOR CRCOUT8 XOR f4
f16 = DD15 XOR CRCOUT0 XOR f5 XOR f12
NOTES −
1 f = feedback
2 DD = Data to or from the bus
3 CRCOUT = 16-bit edge triggered result (current CRC)
4 CRCOUT(15:0) are sent on matching order bits of DD(15:0)
5 CRCIN = Output of combinatorial logic (next CRC)
9.16
Single device configurations
9.16.1 Device 0 only configurations
In a single device configuration where Device 0 is the only device and the host selects Device 1, Device 0 shall
respond as follows:
1) A write to the Device Control register shall complete as if Device 0 was the selected device;
Page 300
T13/1321D revision 3
2) A write to a Command Block register, other than the Command register, shall complete as if
Device 0 was selected;
3) A write to the Command register shall be ignored, except for EXECUTE DEVICE DIAGNOSTIC;
4) A read of the Control Block or Command Block registers, other than the Status or Alternate Status
registers, shall complete as if Device 0 was selected;
5) A read of the Status or Alternate status register shall return the value 00h.
NOTE − Even though Device 1 is not present, the register content may appear valid for Device
1. Further means may be necessary to determine the existence of Device 1 (e.g., issuing a
command).
9.16.2 Device 1 only configurations
Host support of Device 1 only configurations is host specific.
In a single device configuration where Device 1 is the only device and the host selects Device 0, Device 1 shall
respond to accesses of the Command Block and Control Block registers in the same way would if Device 0
was present. This is because Device 1 cannot determine if Device 0 is, or is not, present.
Host implementation of read and write operations to the Command and Control Block registers of non-existent
Device 0 are host specific.
NOTE − The remainder of this subclause is a recommendation for hosts. The host implementor
should be aware of the following when supporting Device 1 only configurations:
1) Following a hardware reset or software reset, the following steps may be used to reselect
Device 1:
a) Write to the Device/Head register with DEV bit set to one;
b) Using one or more of the Command Block registers that may be both written and
read, such as the Sector Count or Sector Number, write a data pattern other than
00h or FFh to the register(s);
c) Read the register(s) written in step (b). If the data read is the same as the data
written, proceed to step (e);
d) Repeat steps (a) to (c) until the data matches in step (c) or until 31 s has past. After
31 s the host may assume that Device 1 is not functioning properly;
e) Read the Status register and Error registers. Check the Status and Error register
contents for any error conditions that Device 1 may have posted.
2) Following the execution of an EXECUTE DEVICE DIAGNOSTIC command, no interrupt
pending should be set to signal command completion. After writing the EXECUTE DEVICE
DIAGNOSTIC command to the Command register, execute steps (a) to (e) as described in
(1) above;
3) At all other times, do not write zero into the DEV bit of the Device/Head register. All other
commands execute normally.
10
Timing
10.1 Deskewing
For PIO and Multiword DMA modes all timing values shall be measured at the connector of the selected
device. The host shall account for cable skew.
For Ultra DMA modes unless otherwise specified, timing parameters shall be measured at the connector of the
host or device to which the parameter applies.
Page 301
T13/1321D revision 3
10.2 Transfer timing
The minimum cycle time supported by the device in PIO mode 3, 4 and Multiword DMA mode 1, 2 respectively
shall always be greater than or equal to the minimum cycle time defined by the associated mode e.g., a device
supporting PIO mode 4 timing shall not report a value less than 120 ns, the minimum cycle time defined for PIO
mode 4 timings.
See 3.2.8 for timing diagram conventions.
10.2.1 Register transfers
Figure 43 defines the relationships between the interface signals for register transfers. Peripherals reporting
support for PIO mode 3 or 4 shall power-up in a PIO mode 0, 1, or 2.
For PIO modes 3 and above, the minimum value of t0 is specified by word 68 in the IDENTIFY DEVICE
parameter list. Table 48 defines the minimum value that shall be placed in word 68.
Both hosts and devices shall support IORDY when PIO mode 3 or 4 are the currently selected mode of
operation.
Page 302
T13/1321D revision 3
t0
ADDR valid
(See note 1)
t2
t1
t9
t2i
DIOR-/DIOWWRITE
DD(7:0)
(See note 2)
t3
t4
READ
DD(7:0)
(See note 2)
t5
t6
t6z
IORDY
(See note 3,3-1)
IORDY
(See note 3,3-2)
IORDY
(See note 3,3-3)
tA
tC
tRD
tB
tC
NOTES −
1 Device address consists of signals CS0-, CS1- and DA(2:0)
2 Data consists of DD(7:0).
3 The negation of IORDY by the device is used to extend the register transfer cycle. The determination
of whether the cycle is to be extended is made by the host after tA from the assertion of DIOR- or
DIOW-. The assertion and negation of IORDY are described in the following three cases:
3-1 Device never negates IORDY, devices keeps IORDY released: no wait is generated.
3-2 Device negates IORDY before tA, but causes IORDY to be asserted before tA. IORDY is
released prior to negation and may be asserted for no more than 5 ns before release: no
wait generated.
3-3 Device negates IORDY before tA. IORDY is released prior to negation and may be
asserted for no more than 5 ns before release: wait generated. The cycle completes
after IORDY is reasserted. For cycles where a wait is generated and DIOR- is asserted,
the device shall place read data on DD(7:0) for tRD before asserting IORDY.
4 DMACK- shall remain negated during a register transfer.
Figure 43 − Register transfer to/from device
Page 303
T13/1321D revision 3
Table 48 − Register transfer to/from device
Register transfer timing parameters
Mode 0 Mode 1 Mode 2 Mode 3 Mode 4 Note
ns
ns
ns
ns
ns
t0
Cycle time
(min)
600
383
330
180
120
1,4,5
t1
Address valid to DIOR-/DIOW- (min)
70
50
30
30
25
setup
t2
DIOR-/DIOW- pulse width
8-bit
(min)
290
290
290
80
70
1
t2i
DIOR-/DIOW- recovery time
(min)
70
25
1
t3
DIOW- data setup
(min)
60
45
30
30
20
t4
DIOW- data hold
(min)
30
20
15
10
10
t5
DIOR- data setup
(min)
50
35
20
20
20
t6
DIOR- data hold
(min)
5
5
5
5
5
t6Z DIOR- data tristate
(max)
30
30
30
30
30
2
t9
DIOR-/DIOW- to address valid hold
(min)
20
15
10
10
10
tRD Read Data Valid to IORDY active
(min)
0
0
0
0
0
(if IORDY initially low after tA)
tA
IORDY Setup time
35
35
35
35
35
3
tB
IORDY Pulse Width
(max)
1250
1250
1250
1250
1250
tC
IORDY assertion to release
(max)
5
5
5
5
5
NOTES −
1 t0 is the minimum total cycle time, t2 is the minimum DIOR-/DIOW- assertion time, and t2i is the minimum
DIOR-/DIOW- negation time. A host implementation shall lengthen t2 and/or t2i to ensure that t0 is equal to
or greater than the value reported in the devices IDENTIFY DEVICE data. A device implementation shall
support any legal host implementation.
2 This parameter specifies the time from the negation edge of DIOR- to the time that the data bus is released
by the device.
3 The delay from the activation of DIOR- or DIOW- until the state of IORDY is first sampled. If IORDY is inactive
then the host shall wait until IORDY is active before the register transfer cycle is completed. If the device is
not driving IORDY negated at the tA after the activation of DIOR- or DIOW-, then t5 shall be met and tRD is
not applicable. If the device is driving IORDY negated at the time tA after the activation of DIOR- or DIOW-,
then tRD shall be met and t5 is not applicable.
4 ATA/ATAPI standards prior to ATA/ATAPI-5 inadvertently specified an incorrect value for mode 2 time t0 by
utilizing the 16-bit PIO value
5 Mode shall be selected no faster than the highest mode supported by the slowest device.
10.2.2 PIO data transfers
Figure 44 defines the relationships between the interface signals for PIO data transfers. Peripherals reporting
support for PIO mode 3 or 4 shall power-up in a PIO mode 0, 1, or 2.
For PIO modes 3 and above, the minimum value of t0 is specified by word 68 in the IDENTIFY DEVICE
parameter list. Table 49 defines the minimum value that shall be placed in word 68.
IORDY shall be supported when PIO mode 3 or 4 are the current mode of operation.
NOTE − Some devices implementing the PACKET Command feature set prior to ATA/ATAPI-4
power-up in PIO mode 3 and enable IORDY as the default.
Page 304
T13/1321D revision 3
t0
ADDR valid
(See note 1)
t2
t1
t9
t2i
DIOR-/DIOWWRITE
DD(15:0)
DD(7:0)
(See note 2)
t3
READ
DD(15:0)
DD(7:0)
(See note 2)
t4
t5
t6
t6z
IORDY
(See note 3,3-1)
IORDY
(See note 3,3-2)
IORDY
(See note 3,3-3)
tA
tC
tRD
tB
tC
NOTES −
1 Device address consists of signals CS0-, CS1- and DA(2:0)
2 Data consists of DD(15:0) for all devices except devices implementing the CFA feature set when 8-bit
transfers is enabled. In that case, data consists of DD(7:0).
3 The negation of IORDY by the device is used to extend the PIO cycle. The determination of whether
the cycle is to be extended is made by the host after tA from the assertion of DIOR- or DIOW-.
The assertion and negation of IORDY are described in the following three cases:
3-1 Device never negates IORDY, devices keeps IORDY released: no wait is generated.
3-2 Device negates IORDY before tA, but causes IORDY to be asserted before tA. IORDY is
released prior to negation and may be asserted for no more than 5 ns before release: no
wait generated.
3-3 Device negates IORDY before tA. IORDY is released prior to negation and may be
asserted for no more than 5 ns before release: wait generated. The cycle completes
after IORDY is reasserted. For cycles where a wait is generated and DIOR- is asserted,
the device shall place read data on DD(7:0) for tRD before asserting IORDY.
4 DMACK- shall be negated during a PIO data transfer.
Figure 44 − PIO data transfer to/from device
Page 305
T13/1321D revision 3
Table 49 − PIO data transfer to/from device
PIO timing parameters
Mode 0 Mode 1 Mode 2 Mode 3 Mode 4 Note
ns
ns
ns
ns
ns
t0
Cycle time
(min)
600
383
240
180
120
1,4
t1
Address valid to DIOR-/DIOW- (min)
70
50
30
30
25
setup
t2
DIOR-/DIOW(min)
165
125
100
80
70
1
t2i
DIOR-/DIOW- recovery time
(min)
70
25
1
t3
DIOW- data setup
(min)
60
45
30
30
20
t4
DIOW- data hold
(min)
30
20
15
10
10
t5
DIOR- data setup
(min)
50
35
20
20
20
t6
DIOR- data hold
(min)
5
5
5
5
5
t6Z DIOR- data tristate
(max)
30
30
30
30
30
2
t9
DIOR-/DIOW- to address valid hold
(min)
20
15
10
10
10
tRD Read Data Valid to IORDY active
(min)
0
0
0
0
0
(if IORDY initially low after tA)
tA
IORDY Setup time
35
35
35
35
35
3
tB
IORDY Pulse Width
(max)
1250
1250
1250
1250
1250
tC
IORDY assertion to release
(max)
5
5
5
5
5
NOTES −
1 t0 is the minimum total cycle time, t2 is the minimum DIOR-/DIOW- assertion time, and t2i is the minimum
DIOR-/DIOW- negation time. A host implementation shall lengthen t2 and/or t2i to ensure that t0 is equal to
or greater than the value reported in the devices IDENTIFY DEVICE data. A device implementation shall
support any legal host implementation.
2 This parameter specifies the time from the negation edge of DIOR- to the time that the data bus is released
by the device.
3 The delay from the activation of DIOR- or DIOW- until the state of IORDY is first sampled. If IORDY is inactive
then the host shall wait until IORDY is active before the PIO cycle is completed. If the device is not driving
IORDY negated at the tA after the activation of DIOR- or DIOW-, then t5 shall be met and tRD is not
applicable. If the device is driving IORDY negated at the time tA after the activation of DIOR- or DIOW-, then
tRD shall be met and t5 is not applicable.
4 Mode may be selected at the highest mode for the device if CS(1:0) and AD(2:0) do not change between read
or write cycles or selected at the highest mode supported by the slowest device if CS(1:0) or AD(2:0) do
change between read or write cycles.
Page 306
T13/1321D revision 3
10.2.3 Multiword DMA data transfer
Figure 45 through Figure 48 define the timing associated with Multiword DMA transfers.
For Multiword DMA modes 1 and above, the minimum value of t0 is specified by word 65 in the IDENTIFY
DEVICE parameter list. Table 50 defines the minimum value that shall be placed in word 65.
Devices shall power-up with mode 0 as the default Multiword DMA mode.
Table 50 − Multiword DMA data transfer
Multiword DMA timing parameters
Mode 0
Mode 1
Mode 2
Note
ns
ns
ns
t0
Cycle time
(min)
480
150
120
see note
tD
DIOR-/DIOW- asserted pulse width
(min)
215
80
70
see note
tE
DIOR- data access
(max)
150
60
50
tF
DIOR- data hold
(min)
5
5
5
tG
DIOR-/DIOW- data setup
(min)
100
30
20
tH
DIOW- data hold
(min)
20
15
10
tI
DMACK to DIOR-/DIOW- setup
(min)
0
0
0
tJ
DIOR-/DIOW- to DMACK hold
(min)
20
5
5
tKR
DIOR- negated pulse width
(min)
50
50
25
see note
tKW
DIOW- negated pulse width
(min)
215
50
25
see note
tLR
DIOR- to DMARQ delay
(max)
120
40
35
tLW
DIOW- to DMARQ delay
(max)
40
40
35
tM
CS(1:0) valid to DIOR-/DIOW(min)
50
30
25
tN
CS(1:0) hold
(min)
15
10
10
tZ
DMACK- to read data released
(max)
20
25
25
NOTE − t0 is the minimum total cycle time, tD is the minimum DIOR-/DIOW- assertion time, and tK (t KR or tKW,
as appropriate) is the minimum DIOR-/DIOW- negation time. A host shall lengthen tD and/or tK to ensure
that t0 is equal to the value reported in the devices IDENTIFY DEVICE data.
Page 307
T13/1321D revision 3
10.2.3.1 Initiating a Multiword DMA data burst
The values for the timings for each of the Multiword DMA modes are contained in Table 50.
CS0-/CS1tM
See note
DMARQ
See note
DMACKtD
tI
DIOR-/DIOWtE
Read
DD(15:0)
tG
Write
DD(15:0)
tG
tF
tH
NOTE − The host shall not assert DMACK- or negate both CS0 and CS1 until the assertion of
DMARQ is detected. The maximum time from the assertion of DMARQ to the assertion of
DMACK- or the negation of both CS0 and CS1 is not defined.
Figure 45 − Initiating a Multiword DMA data transfer
Page 308
T13/1321D revision 3
10.2.3.2 Sustaining a Multiword DMA data burst
The values for the timings for each of the Multiword DMA modes are contained in Table 50.
CS0-/CS1t0
DMARQ
DMACKtK
tD
DIOR-/DIOW-
tE
tE
Read
DD(15:0)
tG
Write
DD(15:0)
tG
tF
tH
tG
tG
tF
tH
Figure 46 − Sustaining a Multiword DMA data transfer
Page 309
T13/1321D revision 3
10.2.3.3 Device terminating a Multiword DMA data burst
The values for the timings for each of the Multiword DMA modes are contained in Table 50.
CS0-/CS1tN
t0
DMARQ
(See note)
tL
DMACKtK
tD
tJ
DIOR-/DIOWtZ
tE
Read
DD(15:0)
tG
tF
Write
DD(15:0)
tG
tH
NOTE − To terminate the data burst, the Device shall negate DMARQ within the tL of the assertion of
the current DIOR- or DIOW- pulse. The last data word for the burst shall then be transferred by the
negation of the current DIOR- or DIOW- pulse. If all data for the command has not been
transferred, the device shall reassert DMARQ again at any later time to resume the DMA operation
as shown in figure 45.
Figure 47 − Device terminating a Multiword DMA data transfer
Page 310
T13/1321D revision 3
10.2.3.4 Host terminating a Multiword DMA data burst
The values for the timings for each of the Multiword DMA modes are contained in Table 50.
CS0-/CS1tN
t0
DMARQ
(See note 2)
DMACK(See note 1)
tK
tD
tJ
DIOR-/DIOWtE
Read
DD(15:0)
Write
DD(15:0)
tZ
tG
tG
tF
tH
NOTE −
1 To terminate the transmission of a data burst, the host shall negate DMACK- within the specified time
after a DIOR- or DIOW- pulse. No further DIOR- or DIOW- pulses shall be asserted for this burst.
2 If the device is able to continue the transfer of data, the device may leave DMARQ asserted and wait
for the host to reassert DMACK- or may negate DMARQ at any time after detecting that DMACKhas been negated.
Figure 48 − Host terminating a Multiword DMA data transfer
Page 311
T13/1321D revision 3
10.2.4 Ultra DMA data transfer
Figure 49 through Figure 58 define the timings associated with all phases of Ultra DMA bursts.
Table 51 contains the values for the timings for each of the Ultra DMA modes.
Page 312
T13/1321D revision 3
Table 51 − Ultra DMA data burst timing requirements
Name
Mode 0
(in ns)
t2CYCTYP
tCYC
min
240
112
t2CYC
230
tDS
tDH
tDVS
15
5
70
tDVH
6
tFS
0
tLI
tMLI
tUI
tAZ
0
20
0
tZAH
tZAD
tENV
20
0
20
tSR
tRFS
tRP
tIORDYZ
tZIORDY
tACK
160
tSS
50
0
20
max
Mode 1
(in ns)
Mode 2
(in ns)
min max
160
73
min max
120
54
Mode 3
(in ns)
Mode 4
(in ns)
min
90
39
min
60
25
max
Comment
max (see Notes 1 and 2)
Typical sustained average two cycle time
Cycle time allowing for asymmetry and clock
variations (from STROBE edge to STROBE edge)
154
115
86
57
Two cycle time allowing for clock variations (from
rising edge to next rising edge or from falling edge to
next falling edge of STROBE)
10
7
7
5
Data setup time at recipient
5
5
5
5
Data hold time at recipient
48
30
20
6
Data valid setup time at sender (from data valid until
STROBE edge) (see Note 4)
6
6
6
6
Data valid hold time at sender (from STROBE edge
until data may become invalid) (see Note 4)
230
0
200
0
170
0
130
0
120 First STROBE time (for device to first negate
DSTROBE from STOP during a data in burst)
150
0
150
0
150
0
100
0
100 Limited interlock time (see Note 3)
20
20
20
20
Interlock time with minimum (see Note 3)
0
0
0
0
Unlimited interlock time (see Note 3)
10
10
10
10
10 Maximum time allowed for output drivers to release
(from asserted or negated)
20
20
20
20
Minimum delay time required for output
0
0
0
0
drivers to assert or negate (from released)
70
20
70
20
70
20
55
20
55 Envelope time (from DMACK- to STOP and
HDMARDY- during data in burst initiation and from
DMACK to STOP during data out burst initiation)
50
30
20
NA
NA STROBE-to-DMARDY- time (if DMARDY- is negated
before this long after STROBE edge, the recipient
shall receive no more than one additional data word)
75
70
60
60
60 Ready-to-final-STROBE time (no STROBE edges
shall be sent this long after negation of DMARDY-)
125
100
100
100
Minimum time to assert STOP or negate DMARQ.
20
20
20
20
20 Maximum time before releasing IORDY
0
0
0
0
Minimum time before driving STROBE (see note 5)
20
20
20
20
Setup and hold times for DMACK- (before assertion
or negation)
50
50
50
50
Time from STROBE edge to negation of DMARQ or
assertion of STOP (when sender terminates a burst)
NOTES −
1 Timing parameters shall be measured at the connector of the sender or receiver to which the parameter applies. For
example, the sender shall stop generating STROBE edges tRFS after the negation of DMARDY-. Both STROBE and
DMARDY- timing measurements are taken at the connector of the sender.
2 All timing measurement switching points (low to high and high to low) shall be taken at 1.5V.
3 tUI , tMLI , and tLI indicate sender-to-recipient or recipient-to-sender interlocks, i.e., either sender or recipient is waiting
for the other to respond with a signal before proceeding. tUI is an unlimited interlock that has no maximum time
value. tMLI is a limited time-out that has a defined minimum. tLI is a limited time-out that has a defined maximum.
4 The test load for tDVS and tDVH shall be a lumped capacitor load with no cable or receivers. Timing for tDVS and tDVH
shall be met for all capacitive loads from 15 to 40 pf where all signals have the same capacitive load value.
5 tZIORDY may be greater than tENV since the device has a pull up on IORDY- giving it a known state when released.
Page 313
T13/1321D revision 3
10.2.4.1 Initiating an Ultra DMA data-in burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
DMARQ
(device)
tUI
DMACK(host)
tACK
tZAD
STOP
(host)
HDMARDY(host)
tFS
tENV
tACK
tFS
tENV
tZAD
tZIORDY
DSTROBE
(device)
tAZ
tDVS
tDVH
DD(15:0)
tACK
DA0, DA1, DA2,
CS0-, CS1NOTES −
1 See 9.13.1 Initiating an Ultra DMA data-in burst.
2 The definitions for the DIOW-:STOP, DIOR-:HDMARDY-:HSTROBE and
IORDY:DDMARDY-:DSTROBE signal lines are not in effect until DMARQ and DMACK
are asserted.
Figure 49 − Initiating an Ultra DMA data-in burst
Page 314
T13/1321D revision 3
10.2.4.2 Sustained Ultra DMA data-in burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
t2CYC
tCYC
tCYC
t2CYC
DSTROBE
at device
tDVH
tDVS
tDVH
tDVS
tDVH
DD(15:0)
at device
DSTROBE
at host
tDH
tDS
tDH
tDS
tDH
DD(15:0)
at host
NOTES −
1 See 9.13.2 The data-in transfer.
2 DD(15:0) and DSTROBE signals are shown at both the host and the device to emphasize that
cable settling time as well as cable propagation delay shall not allow the data signals to be
considered stable at the host until some time after they are driven by the device.
Figure 50 − Sustained Ultra DMA data-in burst
Page 315
T13/1321D revision 3
10.2.4.3 Host pausing an Ultra DMA data-in burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
DMARQ
(device)
DMACK(host)
tRP
STOP
(host)
tSR
HDMARDY(host)
tRFS
DSTROBE
(device)
DD(15:0)
(device)
NOTES −
1 See 9.13.3.2 Host pausing an Ultra DMA data-in burst.
2 The host may assert STOP to request termination of the Ultra DMA burst no sooner than
tRP after HDMARDY- is negated.
3 If the tSR timing is not satisfied, the host may receive zero, one, or two more data words
from the device.
Figure 51 − Host pausing an Ultra DMA data-in burst
Page 316
T13/1321D revision 3
10.2.4.4 Device terminating an Ultra DMA data-in burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
DMARQ
(device)
tMLI
DMACK(host)
tLI
tACK
tLI
STOP
(host)
tACK
tLI
HDMARDY(host)
tSS
DSTROBE
(device)
DD(15:0)
tIORDYZ
tZAH
tAZ
tDVS
tDVH
CRC
tACK
DA0, DA1, DA2,
CS0-, CS1NOTES −
1 See 9.13.4.1 Device terminating an Ultra DMA data-in burst.
2 The definitions for the DIOW-:STOP, DIOR-:HDMARDY-:HSTROBE and IORDY:DDMARDY:DSTROBE signal lines are no longer in effect after DMARQ and DMACK are negated.
Figure 52 − Device terminating an Ultra DMA data-in burst
Page 317
T13/1321D revision 3
10.2.4.5 Host terminating an Ultra DMA data-in burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
DMARQ
(device)
tLI
tMLI
DMACK(host)
tZAH
tAZ
tRP
tACK
STOP
(host)
tACK
HDMARDY(host)
tRFS
tLI
tMLI
tIORDYZ
DSTROBE
(device)
tDVS
DD(15:0)
tDVH
CRC
tACK
DA0, DA1, DA2,
CS0-, CS1NOTES −
1 See 9.13.4.2 Host pausing an Ultra DMA data-in burst.
2 The definitions for the DIOW-:STOP, DIOR-:HDMARDY-:HSTROBE and IORDY:DDMARDY:DSTROBE signal lines are no longer in effect after DMARQ and DMACK are negated.
Figure 53 − Host terminating an Ultra DMA data-in burst
Page 318
T13/1321D revision 3
10.2.4.6 Initiating an Ultra DMA data-out burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
DMARQ
(device)
tUI
DMACK(host)
tACK
tENV
STOP
(host)
tZIORDY
tLI
tUI
DDMARDY(device)
tACK
HSTROBE
(host)
tDVS
tDVH
DD(15:0)
(host)
tACK
DA0, DA1, DA2,
CS0-, CS1NOTES −
1 See 9.14.1 Initiating an Ultra DMA data-out burst.
2 The definitions for the DIOW-:STOP, IORDY:DDMARDY-:DSTROBE and DIOR:HDMARDY-:HSTROBE signal lines are not in effect until DMARQ and DMACK are
asserted.
Figure 54 − Initiating an Ultra DMA data-out burst
Page 319
T13/1321D revision 3
10.2.4.7 Sustained Ultra DMA data-out burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
t2CYC
tCYC
tCYC
t2CYC
HSTROBE
at host
tDVH
tDVS
tDVH
tDVS
tDVH
DD(15:0)
at host
HSTROBE
at device
tDH
tDS
tDH
tDS
tDH
DD(15:0)
at device
NOTES −
1 See 9.14.2 The data out-transfer.
2 DD(15:0) and HSTROBE signals are shown at both the device and the host to emphasize that
cable settling time as well as cable propagation delay shall not allow the data signals to be
considered stable at the device until some time after they are driven by the host.
Figure 55 − Sustained Ultra DMA data-out burst
Page 320
T13/1321D revision 3
10.2.4.8 Device pausing an Ultra DMA data-out burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
tRP
DMARQ
(device)
DMACK(host)
STOP
(host)
tSR
DDMARDY(device)
tRFS
HSTROBE
(host)
DD(15:0)
(host)
NOTES −
1 See 9.14.3.2 Device pausing an Ultra DMA data-out burst.
2 The device may negate DMARQ to request termination of the Ultra DMA burst no sooner
than t RP after DDMARDY- is negated.
3 If the tSR timing is not statisfied, the device may receive zero, one, or two more data
words from the host.
Figure 56 − Device pausing an Ultra DMA data-out burst
Page 321
T13/1321D revision 3
10.2.4.9 Host terminating an Ultra DMA data-out burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
tLI
DMARQ
(device)
tMLI
DMACK(host)
tLI
tSS
tACK
STOP
(host)
tLI
tIORDYZ
DDMARDY(device)
tACK
HSTROBE
(host)
tDVS
DD(15:0)
(host)
tDVH
CRC
tACK
DA0, DA1, DA2,
CS0-, CS1NOTES −
1 See 9.14.4.1 Host terminating an Ultra DMA data-out burst.
2 The definitions for the DIOW-:STOP, IORDY:DDMARDY-:DSTROBE, and
DIOR-:HDMARDY-:HSTROBE signal lines are no longer in effect after DMARQ and
DMACK are negated.
Figure 57 − Host terminating an Ultra DMA data-out burst
Page 322
T13/1321D revision 3
10.2.4.10 Device terminating an Ultra DMA data-out burst
The values for the timings for each of the Ultra DMA modes are contained in 10.2.4.
DMARQ
(device)
DMACK(host)
tLI
tMLI
tACK
STOP
(host)
tRP
tIORDYZ
DDMARDY(device)
tRFS
tLI
tMLI
tACK
HSTROBE
(host)
tDVS
DD(15:0)
(host)
tDVH
CRC
tACK
DA0, DA1, DA2,
CS0-, CS1NOTES −
1 See 9.14.4.2 Device pausing an Ultra DMA data-out burst.
2 The definitions for the DIOW-:STOP, IORDY:DDMARDY-:DSTROBE, and
DIOR-:HDMARDY-:HSTROBE signal lines are no longer in effect after DMARQ and
DMACK are negated.
Figure 58 − Device terminating an Ultra DMA data-out burst
Page 323
T13/1321D revision 3
Annex A
(normative)
Connectors and cable assemblies
The device shall implement one of the connector options described in this annex.
A.1 40-pin Connector
The I/O connector is a 40-pin connector. The header mounted to a host or device is shown in figure A.1 and the
dimensions are shown in table A.1. The connector mounted to the end of the cable is shown in figure A.2 and
the dimensions are shown in table A.2. Signal assignments on these connectors are shown in table A.3.
The pin locations are governed by the cable plug, not the receptacle. The way in which the receptacle is
mounted on the printed circuit board affects the pin positions, and pin 1 shall remain in the same relative
position. This means the pin numbers of the receptacle may not reflect the conductor number of the plug. The
header receptacle may or may not be polarized, and all the signals are relative to pin 20, which is keyed.
By using the plug positions as primary, a straight cable can connect devices. As shown in figure A.3,
conductor 1 on pin 1 of the plug shall be in the same relative position no matter what the receptacle numbering
looks like. If receptacle numbering was followed, the cable would have to twist 180 degrees between a device
with top-mounted receptacles, and a device with bottom-mounted receptacles.
Page 324
T13/1321D revision 3
A1 ± T1
Pos. 2
A13 max
Pos. 40
A3 min
A2 ± T2
A14 ± T6
A5 max
A6 min
A10 ± T5
Pos. 1
Pin 1
Indicator
Pos. 20
Pin removed
Pos. 39
A8 ± T4
Square or round pin
F
Detail A
Detail A
A9 min
A12 min
A7 max
A11 +T5/-T3
A4 ± T3
A10 ± T5
F
Section F-F
Figure A.1 − Host or device 40-pin I/O header
Table A.1 − Host or device 40-pin I/O header
Dimension
Millimeters
Inches
A1
58.17
2.290
A2
48.26
1.900
A3
56.01
2.205
A4
5.84
0.230
A5
9.55
0.376
A6
6.22
0.245
A7
10.16
0.400
A8
0.64
0.025
A9
4.06
0.160
A 10
2.54
0.100
A 11
6.35
0.250
A 12
6.48
0.255
A 13
0.33
0.013
A 14
0.58
0.023
T1
0.51
0.020
T2
0.13
0.005
T3
0.25
0.010
T4
0.03
0.001
T5
0.08
0.003
T6
0.18
0.007
Page 325
T13/1321D revision 3
A5 min
A8 min
A7 max
Pos. 1
Indicator
See NOTE
Mating pins 0.025 sq. or round
A1 ± T3
Pos. 39
A2 ± T1
Pos. 1
A4 ± T2
A8 ± T3
Pos. 2
Pos. 20
Blocked
A3 ± T1
A4 ± T2
Pos. 40
NOTE − The optional polarizing feature is recommended. Some host
receptacles do not provide a polarizing slot.
Figure A.2 − 40-pin I/O cable connector
Table A.2 − 40-pin I/O cable connector
Dimension
Millimeters
Inches
A1
55.37
2.180
A2
48.26
1.900
A3
6.10
0.240
A4
2.54
0.100
A5
6.48
0.255
A6
4.57
0.180
A7
3.81
0.150
A8
1.27
0.050
T1
0.13
0.005
T2
0.08
0.003
T3
0.25
0.010
Page 326
T13/1321D revision 3
Signal name
Table A.3 − 40-pin I/O connector interface signals
Connector
Conductor
Connector
Signal name
contact
contact
1
1
2
2
Ground
3
3
4
4
DD8
5
5
6
6
DD9
7
7
8
8
DD10
9
9
10
10
DD11
11
11
12
12
DD12
13
13
14
14
DD13
15
15
16
16
DD14
17
17
18
18
DD15
19
19
20
20
(keypin)
21
21
22
22
Ground
23
23
24
24
Ground
25
25
26
26
Ground
RESETDD7
DD6
DD5
DD4
DD3
DD2
DD1
DD0
Ground
DMARQ
DIOW-:STOP
DIOR-:HDMARDY:HSTROBE
IORDY:DDMARDY27
27
28
28
:DSTROBE
DMACK29
29
30
30
INTRQ
31
31
32
32
DA1
33
33
34
34
DA0
35
35
36
36
CS037
37
38
38
DASP39
39
40
40
NOTE − Pin 32 was defined as IOCS16 in ATA-2, ANSI X3.279-1996.
40
20
Circuit board
CSEL
Ground
Obsolete (see note)
PDIAG-:CBLIDDA2
CS1Ground
1
2
Circuit board
40
20
1
2
Figure A.3 − 40-pin I/O header mounting
A.1.1 40-conductor cable
The 40-conductor cable assemby is shown in figure A.4 with dimensions in table A.4. Cable capacitance shall
not exceed 35 pf.
Page 327
T13/1321D revision 3
A1
A2
Host Connector
Polarity stripe on
#1 conductor
A3
Device 0
Connector
Device 1
Connector
Figure A.4 − 40-conductor cable configuration
Table A.4 − 40-conductor cable configuration
Dimension
Millimeters
Inches
A1
254.00 min
10.00 min
457.20 max
18.00 max
A2
127.00 min
5.00 min
304.80 max
12.00 max
A3
127.00 min
5.00 min
152.40 max
6.00 max
A.1.2 80-conductor cable assembly using the 40-pin connector
To provide better signal integrity, the optional 80-conductor cable assembly is specified for use with 40-pin
connectors. Use of this assembly is mandatory for systems operating at Ultra DMA modes greater than 2. The
mating half of the connector is as described in A.1. Every other conductor in the 80-conductor cable is
connected to the ground pins in each connector.
The electrical requirements of the 80-conductor ribbon cable are shown in table A.5 and the physical
requirements are described in figure A.5 and table A.6.
Figure A.6 and table A.7 describe the physical dimensions of the cable assembly. The connector in the center
of the cable assembly labeled Device 1 Connector is optional. The System Board connector shall have a blue
base and a black or blue retainer. The Device 0 Connector shall have a black base and a black retainer. The
Device 1 Connector shall have a gray base and a black or gray retainer. The cable assembly may be printed
with connector identifiers.
There are alternative cable conductor to connector pin assignments depending on whether the connector
attaches all even or odd conductors to ground. Table A.8 shows the signal assignments for connectors that
ground the even numbered conductors. Table A.9 shows the signal assignments for connectors that ground the
odd numbered conductors. Only one connector type, even or odd, shall be used in a given cable assembly.
Connectors shall be labeled as grounding the even or odd conductors as shown in figure A.7. Cable assemblies
conforming to table A.8 are interchangable with cable assemblies conforming to table A.9.
All connectors shall have position 20 blocked to provide keying. Pin 28 in Device 1 Connector shall not be
attached to any cable conductor, the connector contact may be removed to meet this requirement (see
5.2.13.2). Pin 34 in the Host Connector shall not be attached to any cable conductor and shall be attached to
Ground within the connector (see 6.7).
Page 328
T13/1321D revision 3
Table A.5 − 80-conductor cable electrical
Conductor
Ground-signal-ground
Single ended impedance
(Ω)
Capacitance
(pF/ft)
(pF/m)
Inductance
(µH)
Propagation delay
(nsec/ft)
(nsec/m)
requirements
30 AWG
70-90
13-22
42-72
0.08-0.16
1.35-1.65
4.43-5.41
A1 ± T1
Polarity stripe
on #1 conductor
A2 ± T1
A3 ± T2
A4 ± T3
A5 ± T4
Figure A.5 − 80-conductor ribbon cable
Table A.6 − 80-conductor ribbon cable
Dimension
Millimeters
Inches
A1
50.800
2.000
A2
50.165
1.975
A3
0.635
0.025
A4
0.6858
0.027
A5
0.3175
0.0125
T1
0.127
0.005
T2
0.0406
0.0016
T3
0.0508
0.002
T4
0.102
0.004
Page 329
T13/1321D revision 3
A1
Polarity stripe on
#1 conductor
Device 1
Connector
1
Figure A.6 − 80-conductor cable configuration
Table A.7 − 80-conductor cable configuration
Dimension
Millimeters
Inches
A1
457.20 max
18.00 max
A2
127.00 min
5.00 min
A3
152.40 max
6.00 max
A2 min shall be greater than or equal to A3.
Page 330
Device 0
tsoH
Host Connector
A3
Devcie
A2
Device 0
Connector
T13/1321D revision 3
Table A.8 − Signal assignments for connectors grounding even conductors
Signal name
Connector
Conductor
Signal name
contact
RESET1
1
2
Ground
Ground
2
3
4
Ground
DD7
3
5
6
Ground
DD8
4
7
8
Ground
DD6
5
9
10
Ground
DD9
6
11
12
Ground
DD5
7
13
14
Ground
DD10
8
15
16
Ground
DD4
9
17
18
Ground
DD11
10
19
20
Ground
DD3
11
21
22
Ground
DD12
12
23
24
Ground
DD2
13
25
26
Ground
DD13
14
27
28
Ground
DD1
15
29
30
Ground
DD14
16
31
32
Ground
DD0
17
33
34
Ground
DD15
18
35
36
Ground
Ground
19
37
38
Ground
(keypin)
20
39
40
Ground
DMARQ
21
41
42
Ground
Ground
22
43
44
Ground
DIOW23
45
46
Ground
Ground
24
47
48
Ground
DIOR25
49
50
Ground
Ground
26
51
52
Ground
IORDY
27
53
54
Ground
CSEL
28
55
56
Ground
DMACK29
57
58
Ground
Ground
30
59
60
Ground
INTRQ
31
61
62
Ground
Reserved
32
63
64
Ground
DA1
33
65
66
Ground
PDIAG34(see note)
67
68
Ground
DA0
35
69
70
Ground
DA2
36
71
72
Ground
CS037
73
74
Ground
CS138
75
76
Ground
DASP39
77
78
Ground
Ground
40
79
80
Ground
NOTE − Pin 34 in the Host Connector shall not be attached to any cable
conductor and shall be attached to Ground within the connector (see 6.7).
Page 331
T13/1321D revision 3
Table A.9 − Signal assignments for connectors grounding odd conductors
Signal name
Conductor
Connector
Signal name
contact
Ground
1
2
1
RESETGround
3
4
2
Ground
Ground
5
6
3
DD7
Ground
7
8
4
DD8
Ground
9
10
5
DD6
Ground
11
12
6
DD9
Ground
13
14
7
DD5
Ground
15
16
8
DD10
Ground
17
18
9
DD4
Ground
19
20
10
DD11
Ground
21
22
11
DD3
Ground
23
24
12
DD12
Ground
25
26
13
DD2
Ground
27
28
14
DD13
Ground
29
30
15
DD1
Ground
31
32
16
DD14
Ground
33
34
17
DD0
Ground
35
36
18
DD15
Ground
37
38
19
Ground
Ground
39
40
20
(keypin)
Ground
41
42
21
DMARQ
Ground
43
44
22
Ground
Ground
45
46
23
DIOWGround
47
48
24
Ground
Ground
49
50
25
DIORGround
51
52
26
Ground
Ground
53
54
27
IORDY
Ground
55
56
28
CSEL
Ground
57
58
29
DMACKGround
59
60
30
Ground
Ground
61
62
31
INTRQ
Ground
63
64
32
Reserved
Ground
65
66
33
DA1
Ground
67
68
34 (see note)
PDIAGGround
69
70
35
DA0
Ground
71
72
36
DA2
Ground
73
74
37
CS0Ground
75
76
38
CS1Ground
77
78
39
DASPGround
79
80
40
Ground
NOTE − Pin 34 in the Host Connector shall not be attached to any cable
conductor and shall be attached to Ground within the connector (see 6.7).
Page 332
T13/1321D revision 3
Pin 1
Termination block
Characters located
approx. where shown
XXX = EVN or ODD
Figure A.7 − Connector labeling for even or odd conductor grounding
Page 333
T13/1321D revision 3
A.2 4-pin power connector
The power connector is a 4-pin connector. The header mounted to a device is shown in figure A.8 and the
dimensions are shown in table A.10. The connector mounted to the end of the cable is shown in figure A.9 and
the dimensions are shown in table A.11. Pin assignments for these connectors are shown in table A.12.
A7 ± T2 X 45°
(2X)
A6 +T3/-T2
-C-A-
A16 min*
A8 min
A12
A10
R A14 Ref
A13 (4X)
Spherical R
∅ A2
A5 ± T2
A9 ± T2
∅ A1 ± T1 (4X)
⊕ ∅ A2 A B C
A3
-BA4
A11
A15
Pin 1
Unless otherwise specified, all tolerances are ± T3
* The tolerance build up of A6 and A7 shall not exceed A16
Figure A.8 − Device 4-pin power header
Table A.10 − Device 4-pin power header
Dimension
Millimeters
Inches
A1
2.10
0.083
A2
3.50
0.138
A3
5.08
0.200
A4
15.24
0.600
A5
6.60
0.260
A6
21.32
0.839
A7
1.65
0.065
A8
7.50
0.295
A9
6.00
0.236
A 10
4.95
0.195
A 11
1.00
0.039
A 12
11.18
0.440
A 13
3.80
0.150
A 14
3.00
0.118
A 15
5.10
0.201
A 16
17.80
0.701
T1
0.04
0.0016
T2
0.15
0.006
T3
0.25
0.010
Page 334
T13/1321D revision 3
A11
A10 (4X)
20 ° Ref
A12
A9 min
A7 ± T2
-A-
A13 min
A5 +T2/-T3
-C-
A8
Pin 1
A6 ± T1 X 45° Ref (2x)
A4 ± T1
∅ A14 (4X)
A2
-B-
A3
∅ A1 Ref
⊕ ∅ T4 A B C
Unless otherwise specified, all tolerances are ± T3.
Figure A.9 − 4-pin power cable connector
Table A.11 − 4-pin power cable connector
Dimension
Millimeters
Inches
A1
2.03
0.080
A2
5.08
0.200
A3
15.24
0.600
A4
6.35
0.250
A5
21.00
0.827
A6
1.78
0.070
A7
7.87
0.310
A8
5.51
0.217
A9
1.19
0.047
A 10
5.08
0.200
A 11
11.18
0.440
A 12
1.19
0.047
A 13
2.00
0.079
A 14
4.06
0.160
T1
0.10
0.004
T2
0.15
0.006
T3
0.25
0.010
T4
0.60
0.024
Table A.12 − 4-pin power connector pin assignments
Power line
Pin
+12 volts
1
+12 volt return
2
+5 volt return
3
+5 volts
4
Page 335
T13/1321D revision 3
A.2.1 Mating performance
Mating force should be 3.85 lbs (1.75 kg) maximum per contact.
Unmating force should be 0.25 lbs (113.5 g) minimum per contact.
A.3 Unitized connectors
The 40-pin I/O signal header and the 4-pin power connector may be implemented in one of two unitized
connectors that provide additional pins for configuration jumpers. The dimensioning of the 40-pin I/O signal area
shall be as defined in figure A.1 and the dimensioning of the 4-pin power connector area shall be as defined in
figure A.8 for both unitized connectors.
The first of the unitized connectors is shown in figure A.10 with dimensions as shown in table A.13. The jumper
pins, A through I, have been assigned as follows:
−
−
−
−
−
−
E-F - CSEL
G-H - Master
G-H and E-F - Master with slave present
No jumper - Slave
A through D - Vendor specific
I - Reserved
The second of the unitized connectors is shown in figure A.11 with dimensions as shown in table A.14. The
jumper pins, A through J, have been assigned as follows:
−
−
−
−
A-B - CSEL
C-D - Slave
E-F - Master
G through J - Vendor specific
Dimensioning of
the 4-pin power
connector area
per figure A.7
Dimensioning of the 40- pin
signal connector area per
figure A.1
A6
Pin H
A7
Pin 40
A3
Pin B
Pin A
A1 ± T1
Pin 1
A4
A2
A5
A8
A9
Tolerances ±0.010 inches unless otherwise specified
Figure A.10 − Unitized connector
Page 336
T13/1321D revision 3
Table A.13 − Unitized connector
Dimension
Millimeters
Inches
A1
2.54
0.100
A2
4.06
0.160
A3
8.40
0.331
A4
5.26
0.207
A5
63.50
2.500
A6
13.54
0.533
A7
2.54
0.100
A8
70.825
2.788
A9
95.50
3.760
T1
0.15
0.006
Dimensioning of
the 4-pin power
connector area
per figure A.7
Dimensioning of the 40-pin
signal connector area per
figure A.1
Pin B
A5
Pin J
A6
Pin 40
A1
Pin 1
A7
Pin A
A2
A3
A4
A8
A9
Tolerances ±0.010 inches unless otherwise specified
Figure A.11 − Unitized connector
Table A.14 − Unitized connector
Dimension
Millimeters
Inches
A1
8.51
0.335
A2
5.51
0.217
A3
57.15
2.250
A4
10.16
0.400
A5
17.88
0.704
A6
8.94
0.352
A7
2.54
0.100
A8
75.29
2.964
A9
100.33
3.950
A.4 50-pin connector
An alternative connector is often used for 2 1/2 inch or smaller devices. This connector is shown in figure A.12
with dimensions shown in table A.15. Signal assignments are shown in table A.16. Although there are 50 pins
in the plug, a 44-pin mating receptacle may be used.
Pins E, F, and 20 are keys and are removed.
Some devices may utilize pins A, B, C, and D for option selection via physical jumpers. If a device uses pins A,
B, C, and D for device selection, when no jumper is present the device should be designated as Device 0.
Page 337
T13/1321D revision 3
When a jumper is present between pins B and D, the device should respond to the CSEL signal to determine
the device number.
A
Pos. 43
A2 ± T1
Pos. 1
A1
Pos. C
Pos. A
Pos. B
Pos. D
A
Pos. 44
Pos. 20
Pin removed
Pos. 2
A2 ± T1
A3 ± T2
A1
Section A-A
Figure A.12 − 50-pin connector
Dimension
A1
A2
A3
T1
T2
Page 338
Table A.15 − 50-pin connector
Millimeters
Inches
2.00
0.079
0.50
0.020
3.86
0.152
0.05
0.002
0.20
0.008
Pos. E and F
Pins removed
T13/1321D revision 3
Table A.16 − Signal assignments for 50-pin connector
Signal name
Connector
Conductor
Connector
Signal name
contact
contact
Option selection pins
A
B
Option selection pins
Option selection pins
C
D
Option selection pins
(keypin)
E
F
(keypin)
RESET1
1
2
2
Ground
DD7
3
3
4
4
DD8
DD6
5
5
6
6
DD9
DD5
7
7
8
8
DD10
DD4
9
9
10
10
DD11
DD3
11
11
12
12
DD12
DD2
13
13
14
14
DD13
DD1
15
15
16
16
DD14
DD0
17
17
18
18
DD15
Ground
19
19
20
20
(keypin)
DMARQ
21
21
22
22
Ground
DIOW-:STOP
23
23
24
24
Ground
DIOR-:HDMARDY25
25
26
26
Ground
:HSTROBE
IORDY:DDMARDY27
27
28
28
CSEL
:DSTROBE
DMACK29
29
30
30
Ground
INTRQ
31
31
32
32
Obsolete (see note)
DA1
33
33
34
34
PDIAGDA0
35
35
36
36
DA2
CS037
37
38
38
CS1DASP39
39
40
40
Ground
+5 V (logic)
41
41
42
42
+5 V (motor)
Ground(return)
43
43
44
44
Reserved - no connection
NOTE − Pin 32 was defined as IOCS16 in ATA-2, ANSI X3.279-1996.
A.5 68-pin PCMCIA connector
This clause defines the pinouts used for the 68-pin alternative connector for the AT Attachment Interface. This
connector is defined in the PCMCIA PC Card Standard. This clause defines a pinout alternative that allows a
device to function as an AT Attachment Interface compliant device, while also allowing the device to be
compliant with PC Card ATA mode defined by PCMCIA. The signal protocol allows the device to identify the
host interface as being 68-pin as defined in this standard or PC Card ATA.
To simplify the implementation of dual-interface devices, the 68-pin AT Attachment Interface maintains
commonality with as many PC Card ATA signals as possible, while supporting full command and signal
compliance with this standard.
The 68-pin pinout shall not cause damage or loss of data if a PCMCIA card is accidentally plugged into a host
slot supporting this interface. The inversion of the RESET signal between this standard and PCMCIA interfaces
prevents loss of data if the device is unable to reconfigure itself to the appropriate host interface.
A.5.1 Signals
This specification relies upon the electrical and mechanical characteristics of PCMCIA and unless otherwise
noted, all signals and registers with the same names as PCMCIA signals and registers have the same meaning
as defined in PCMCIA.
Page 339
T13/1321D revision 3
The PC Card-ATA specification is used as a reference to identify the signal protocol used to identify the host
interface protocol.
A.5.2 Signal descriptions
Any signals not defined below shall be as described in this standard, PCMCIA, or the PC Card ATA
documents.
Table A.15 shows the signals and relationships such as direction, as well as providing the signal name of the
PCMCIA equivalent.
Page 340
T13/1321D revision 3
Pin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Signal
Ground
DD3
DD4
DD5
DD6
DD7
CS0-
Hst
x
x
x
x
x
x
x
SELATA-
x
CS1-
x
INTRQ
x
Vcc
x
Table A.15 − Signal assignments for 68-pin connector
Dir Dev PCMCIA
Pin
Signal
Hst
Dir
→
x
Ground
35
Ground
x
→
↔
x
D3
36
CD1x
←
↔
x
D4
37
DD11
x
↔
↔
x
D5
38
DD12
x
↔
↔
x
D6
39
DD13
x
↔
↔
x
D7
40
DD14
x
↔
→
x
CE141
DD15
x
↔
→
i
A10
42
CS1x
→
→
x
OE43
←
44
DIORx
→
→
x(1)
A9
45
DIOWx
→
→
i
A8
46
47
48
→
i
WE49
←
x
READY/
50
IREQ→
x
Vcc
51
Vcc
x
→
52
53
54
55
M/Sx
→
→
i
A7
56
CSEL
x
→
→
i
A6
57
←
→
i
A5
58
RESETx
→
→
i
A4
59
IORDY
o
←
→
i
A3
60
DMARQ
o
←
→
x
A2
61
DMACKo
→
→
x
A1
62
DASPx
↔
Dev
x
x
x
x
x
x
x
x(1)
i
x
x
PCMCIA
Ground
CD1D11
D12
D13
D14
D15
CE2VS1IORDIOWR-
x
Vcc
17
18
19
20
21
22
23
24
25
26
27
28
DA2
DA1
x
x
29
DA0
x
→
x
A0
63
PDIAG-
x
↔
x
30
31
32
33
DD0
DD1
DD2
x
x
x
x
↔
↔
↔
←
x
x
x
x
D0
D1
D2
WP/
IOIS16
Ground
64
65
66
67
DD8
DD9
DD10
CD2-
x
x
x
x
↔
↔
↔
←
x
x
x
x
VS2RESET
WAITINPACKREGBVD2/
SPKRBVD1/
STSCHG
D8
D9
D10
CD2-
34
Ground
x
→
x
68
Ground
x
→
Key:
Dir = the direction of the signal between host and device.
x in the Hst column = this signal shall be supported by the Host.
x in the Dev column = this signal shall be supported by the device.
i in the Dev column = this signal shall be ignored by the device while in 68-pin mode.
o = this signal is Optional.
Nothing in Dev column = no connection should be made to that pin.
x
Ground
x(2)
x(2)
i
x
x(3)
x(3)
o
x
NOTES −
1 The device shall support only one CS1- signal pin.
2 The device shall support either M/S- or CSEL but not both.
3 The device shall hold this signal negated if it does not support the function.
Page 341
T13/1321D revision 3
A.5.2.1 CD1- (Card Detect 1)
This signal shall be grounded by the device. CD1- and CD2- are used by the host to detect the presence of the
device.
A.5.2.2 CD2- (Card Detect 2)
This signal shall be grounded by the device. CD1- and CD2- are used by the host to detect the presence of the
device.
A.5.2.3 CS1- (Device chip select 1)
Hosts shall provide CS1- on both the pins identified in table A.15.
Devices shall recognize only one of the two pins as CS1-.
A.5.2.4 DMACK- (DMA acknowledge)
This signal is optional for hosts and devices.
If this signal is supported by the host or the device, the function of DMARQ shall also be supported.
A.5.2.5 DMARQ (DMA request)
This signal is optional for hosts.
If this signal is supported by the host or the device, the function of DMACK- shall also be supported.
A.5.2.6 IORDY (I/O channel ready)
This signal is optional for hosts.
A.5.2.7 M/S- (Master/slave)
This signal is the inverted form of CSEL. Hosts shall support both M/S- and CSEL though devices need only
support one or the other.
Hosts shall assert CSEL and M/S- prior to applying VCC to the connector.
A.5.2.8 SELATA- (Select 68-pin ATA)
This pin is used by the host to select which mode to use, PC Card-ATA mode or the 68-pin mode defined in
this standard. To select 68-pin ATA mode, the host shall assert SELATA- prior to applying power to the
connector, and shall hold SELATA- asserted.
The device shall not re-sample SELATA- as a result of either a hardware or software reset. The device shall
ignore all interface signals for 19 ms after the host supplies Vcc within the device's voltage tolerance. If
SELATA- is negated following this time, the device shall either configure itself for PC Card-ATA mode or not
respond to further inputs from the host.
A.5.3 Removability considerations
This specification supports the removability of devices that use the protocol. As removability is a new
consideration for devices, several issues need to be considered with regard to the insertion or removal of
devices.
Page 342
T13/1321D revision 3
A.5.3.1 Device recommendations
The following are recommendations to device implementors:
−
−
−
CS0-, CS1-, RESET-, and SELATA- signals be negated on the device to prevent false selection
during hot insertion.
Ignore all interface signals except SELATA- until 19 ms after the host supplies VCC within the
device's voltage tolerance. This time is necessary to de-bounce the device's power-on reset
sequence. Once in the 68-pin mode as defined in this standard, if SELATA- is ever negated
following the 19 ms de-bounce delay time, the device disables itself until VCC is removed.
Provide a method to prevent unexpected removal of the device or media.
A.5.3.2 Host recommendations
The following are recommendations to host implementors:
−
−
−
−
−
Connector pin sequencing to protect the device by making contact to ground before any other
signal in the system.
SELATA- to be asserted at all times.
All devices reset and reconfigured to the same base address each time a device at that address is
inserted or removed.
The removal or insertion of a device at the same address to be detected so as to prevent the
corruption of a command.
Provide a method to prevent unexpected removal of the device or media.
A.6 CompactFlash connector
Device compliant with the CompactFlash Association Specification utilize the connector defined in that
specification.
Page 343
T13/1321D revision 3
Annex B
(normative)
Device determination of cable type
B.1 Overview
This standard requires that, for systems using a cable assembly, an 80-conductor cable assembly shall be
installed before a system may operate with Ultra DMA modes greater than 2. However, some hosts have not
implemented circuitry to determine the installed cable type by detecting whether PDIAG-:CBLID- is connected
to ground as mandated by this standard. The following describes an alternate method for using IDENTIFY
DEVICE or IDENTIFY PACKET DEVICE data from the device to determine the cable type. It is not
recommended that a host use the method described in this annex.
If a host uses IDENTIFY DEVICE or IDENTIFY PACKET DEVICE data from the device to determine the cable
type, then a 0.047 µf capacitor shall be installed from CBLID- to ground at the host connector. The tolerance on
this capacitor is +/- 20% or less. After receiving an IDENTIFY DEVICE or IDENTIFY PACKET DEVICE
command the device detects the presence or absence of the capacitor by asserting PDIAG-:CBLID- to
discharge the capacitor, releasing PDIAG-, and sampling PDIAG-:CBLID- before the installed capacitor could
recharge through the 10 kΩ pull-up resistor(s) on PDIAG-:CBLID- at the device(s).
If the host system has a capacitor on PDIAG-:CBLID- and a 40-conductor cable is installed, the rise time of the
signal will be slow enough that the device will sample PDIAG-:CBLID- while the signal is still below VIL.
Otherwise, if PDIAG-:CBLID- is not connected from the host connector to the devices in an 80-conductor cable
assembly, the device will detect that the signal is pulled above VIH through the resistor(s) on the device(s). The
capacitor test results will then be reported to the host in the IDENTIFY DEVICE or IDENTIFY PACKET DEVICE
data. The host will use the data to determine the maximum transfer rate of which the system is capable and
use this information when setting the transfer rate using the SET FEATURES command.
B.2 Sequence for device detection of installed capacitor
The following is the sequence for a host using IDENTIFY DEVICE or IDENTIFY PACKET DEVICE data from the
device to determine the cable type:
1) the host issues an IDENTIFY DEVICE or IDENTIFY PACKET DEVICE command (according to device
type) first to Device 1 and then to Device 0 after every power-on or hardware reset sequence (the
command is issued to Device 1 first to ensure that Device 1 releases PDIAG-:CBLID- before Device 0 is
selected. Device 0 will be unable to distinguish a discharged capacitor if Device 1 is driving the line to
its electrically low state. Issuing the command to Device 1 forces it to release PDIAG-:CBLID-);
2) the selected device asserts PDIAG-:CBLID- for at least 30 µs after receipt of the IDENTIFY DEVICE or
IDENTIFY PACKET DEVICE command but before transferring data for the command:
3) the device releases PDIAG-:CBLID- and samples it between two and thirteen µs after release;
4) if the device detects that PDIAG-:CBLID- is below VIL, then the device returns a value of zero in bit 13
of word 93 in its IDENTIFY DEVICE or IDENTIFY PACKET DEVICE data (if the host system has a
capacitor on that signal and a 40-conductor cable is installed, the rise time of the signal will be slow
enough that it will be sampled by the device while it is still below VIL);
5) if the device detects that the signal is above VIH, then the device returns a value of one in bit 13 of word
93 in its IDENTIFY DEVICE or IDENTIFY PACKET DEVICE data (this signal is not connected between
the host and the devices in an 80-conductor cable assembly, thus, the sampling device will see this
signal pulled above VIH through the 10 kΩ resistor(s) installed on the device(s);
6) the host then uses its knowledge of its own capabilities and the content of words 88 and 93 to
determine the Ultra DMA modes of which the system is capable;
7) the host then uses the SET FEATURES command to set the transfer mode.
Page 344
T13/1321D revision 3
Host PCB
Host
connector
Device 0
connector
Device 1
connector
PDIAG-:CBLIDconductor
Device 0
PCB
Device 1
PCB
Figure B.1 – Example configuration of a system where the device detects a 40-conductor cable
Table B.1 – Device detection of installed capacitor
Cable assembly
Device 1
Value reported in Device-determined Determination
type
releases PDIAG- ID data by device
cable type
correct?
40-conductor
Yes
0
40-conductor
Yes
80-conductor
Yes
1
80-conductor
Yes
40-conductor
No
0
40-conductor
Yes
80-conductor
No
0
40-conductor
No (see note)
NOTE − Ultra DMA modes 3 or 4 will not be set even though the system supports them.
Table B.2 – Results of device based cable detection if the host does not have the capacitor installed
Cable assembly
Device 1
Value reported in Device-determined Determination
type
releases PDIAG- ID data by device
cable type
correct?
40-conductor
Yes
1
80-conductor
No (see note 1)
80-conductor
Yes
1
80-conductor
Yes
40-conductor
No
0
40-conductor
Yes
80-conductor
No
0
40-conductor
No (see note 2)
NOTES –
1 Ultra DMA mode 3 or 4 may be set incorrectly resulting in ICRC errors.
2 Ultra DMA modes 3 or 4 will not be set even though the system supports them.
Page 345
T13/1321D revision 3
B.3 Using the combination of methods for detecting cable type
Determining the cable assembly type may be done either by the host sensing the condition of the PDIAG:CBLID- signal, by relying on information from the device, or a combination of both methods. Table B.3
describes the results of using both host and device cable detection methods.
Table B.3 – Results of using both host and device cable detection methods
Cable assembly
Device 1
Electrical state of Value reported in Determined Determination
type
Releases PDIAG- CBLID- at host ID data by device cable type
correct?
40-conductor
Yes
1
0
40
Yes
80-conductor
Yes
0
1
80
Yes
40-conductor
No
0
0
40
Yes (see note)
80-conductor
No
0
0
40
No (see note)
NOTE – The “0,0” result is independent of cable type and indicates that Device 1 is incorrectly asserting
PDIAG-. When the host determines this result, it shall not operate with Ultra DMA modes greater than
2 and it may respond in several ways:
a) report that Device 1 is incompatible with Ultra DMA modes higher than 2 and should be used on a
different port in order to use those modes on the port being detected;
b) report that Device 1 is not allowing the cable type to be properly detected;
c) do not notify the user of any problem but detect the cable as a 40-conductor.
The Table B.4 below illustrates intermediate results for all combinations of cable, device, and host, for hosts
that support Ultra DMA modes greater than 2.
Table B.4 – Results for all combinations of device and host cable detection methods
Design options
Intermediate actions and results
Results
80-conDevice
Host
Host uses
Host
Device Capa- ID word
Host
Host
ductor
supports senses
ID data,
capacitor tests for citor
93 Bit
checks
may set
cable
UDMA
PDIAG-: capacitor connected capa- detec13
ID word
UDMA
installed modes >2 CBLIDinstalled to device
citor
ted
value
93 bit 13
mode
>2
No
No
Yes
No
No
No
No
0
No
No
No
Yes
Yes
No
No
Yes
No
1
No
No
Yes
No
Yes
No
No
No
No
0
No
No
Yes
Yes
Yes
No
No
Yes
No
1
No
Yes
No
No
No
Yes
Yes
No
No
0
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
0
Yes
No
Yes
No
No
Yes
No
No
No
0
Yes
No
Yes
Yes
No
Yes
No
Yes
No
1
Yes
Yes
Page 346
T13/1321D revision 3
Annex C
(informative)
Identify device data for devices with more than 1024 logical cylinders
C.1 Definitions and background information
The original IBM PC BIOS (Basic Input/Output System) imposed several restrictions on the support of devices,
and these have been incorporated into many higher level software products. One such restriction limits the
capacity of a device. BIOS software cannot support a device with more than 1,024 cylinders, 16 heads, and 63
sectors per track without translating an input logical geometry to a different output logical geometry. The
maximum addressable capacity of a device that does not require BIOS translation is 1,032,192 sectors.
These rules allow BIOSes using bit shifting translation to access 15,481,935 (16,383∗15∗63) sectors, and
BIOSes using LBA assisted translation to access 16,450,560 (1024∗255∗63) sectors. Extended BIOS
functionality is defined in the ANSI Technical Report Enhanced BIOS Services for Disk Drives, which describes
new services provided by BIOS firmware to support ATA hard disks up to 16 mega-terra-sectors.
C.2 Cylinder, head, and sector addressing
BIOSs and other software that operate a device in CHS translation employ a combination of IDENTIFY DEVICE
data words 1, 3, 6, words 53-58, and words 60-61 to ascertain the appropriate translation to use and determine
the capacity of a device.
Maximum compatibility is facilitated if the following guidelines are used. These guidelines limit the values
placed into words 1, 3, 6, 53-58, and 60-61. Accessing beyond 15,481,935 sectors should be performed using
LBA.
C.2.1 Word 1
For devices with a capacity less than or equal to 1,032,192 sectors, if IDENTIFY DEVICE data word 1 (Default
Cylinders) does not specify a value greater than 1,024, then no guideline is necessary.
If a device is greater than 1,032,192 sectors, but less than or equal to 16,514,064 sectors, the maximum value
that can be placed into this word is determined by the value in word 3 as shown in table C.1.
If a device is greater than 15,481,935 sectors and supports CHS, this word should contain 16,383 (3FFFh).
The INITIALIZE DEVICE PARAMETERS command does not change this value.
The value in this word is changed by the SET MAX ADDRESS command.
Page 347
T13/1321D revision 3
Table C.1 − Word 1 value for devices between 1,032,192 and 16,514,064 sectors
Value in word 3
Maximum value in word 1
1 1h
65,535 FFFFh
2 2h
65,535 FFFFh
3 3h
65,535 FFFFh
4 4h
65,535 FFFFh
5 5h
32,767 7FFFh
6 6h
32,767 7FFFh
7 7h
32,767 7FFFh
8 8h
32,767 7FFFh
9 9h
16,383 3FFFh
10 Ah
16,383 3FFFh
11 Bh
16,383 3FFFh
12 Ch
16,383 3FFFh
13 Dh
16,383 3FFFh
14 Eh
16,383 3FFFh
15 Fh
16,383 3FFFh
16 10h
16,383 3FFFh
C.2.2 Word 3
IDENTIFY DEVICE data word 3 (Default Heads) does not specify a value greater than 16. If the device has less
than or equal to 8,257,536 sectors, then set word 3 to 16 heads. If the device has more than 16,514,064
sectors, then set word 3 to 15 heads. If this value is set to 16 when the device has more than 16,514,064
sectors, some systems will not boot some operating systems.
The INITIALIZE DEVICE PARAMETERS command does not change this value.
C.2.3 Word 6
If the device is above 1,032,192 sectors then the value should be 63. This value does not exceed 63 (3Fh).
The INITIALIZE DEVICE PARAMETERS command does not change this value.
C.2.4 Use of words 53 through 58
Devices with a capacity over 1,032,192 sectors implement words 53-58. Devices with a capacity less than or
equal to 1,032,192 sectors may also implement these words. These words define the address range for all
sectors accessible in CHS mode under 16,514,064. The product of word 54, word 55, and word 56 must not
exceed 16,514,064.
C.2.5 Words 60-61
IDENTIFY DEVICE data words 60-61 contain a 32-bit value that is equal to the total number of sectors that can
be accessed using LBA. If the device is less than or equal to 15,481,935 sectors, this value should be the
product of words 1, 3, and 6. Setting the total number of LBA sectors in this manner reduces the probability of
conflicting device capacities being calculated by different operating systems.
Page 348
T13/1321D revision 3
C.3 Orphan sectors
The sectors, if any, between the last sector addressable in CHS mode and the last sector addressable in LBA
mode are known as "orphan" sectors. A device may or may not allow access to these sectors in CHS
addressing mode.
The values in words 1, 3, and 6 are selected such that the number of orphan sectors is minimized. Normally,
the number of orphan sectors should not exceed ( [word55] x [word56] - 1 ). However, the host system may
create conditions where there are a larger number of orphans sectors by issuing the INITIALIZE DEVICE
PARAMETERS command with values other than the values in words 3 and 6. If the recommendation in C.2.5 is
followed, there will be no orphan sectors and problems associated with new operating systems calculating a
different device size from older operating systems will be minimized.
Page 349
T13/1321D revision 3
Annex D
(informative)
Signal integrity and UDMA implementation guide
D.1 Introduction
The ATA bus (also called the IDE bus) is a storage interface originally designed for the ISA Bus of the IBM
PC/AT . With the advent of faster host systems and devices, the definition of the bus has been expanded to
include new operating modes. Each of the PIO modes, numbered zero through four, is faster than the one
before (higher numbers translate to faster transfer rates). PIO modes 0, 1, and 2 correspond to transfer rates
for the interface as was originally defined with maximum transfer rates of 3.3, 5.2, and 8.3 megabytes per
second (MB/s), respectively. PIO mode 3 defines a maximum transfer rate of 11.1 MB/s, and PIO mode 4
defines a maximum rate of 16.7 MB/s. Additionally, Multiword DMA and Ultra DMA modes have been defined.
Multiword DMA mode 0, 1, and 2 have maximum transfer rates of 4.2, 13.3, and 16.7 MB/s, respectively. Ultra
DMA modes 0, 1, 2, 3 and 4 have maximum transfer rates of 16.7, 25, 33.3, 44.4, and 66.6 MB/s, respectively.
Ultra DMA features such as increased frequencies, double-edge clocking, and non-interlocked signaling require
improved signal integrity on the bus relative to that required by PIO and Multiword DMA modes. For Ultra DMA
modes 0, 1 and 2 this is achieved by the use of partial series termination and controlled slew rates. For modes
3 and 4 an 80-conductor cable assembly is required in addition to partial series termination and controlled slew
rates. This cable assembly has ground lines interspersed between all signal lines on the bus in order to control
impedance and reduce crosstalk, eliminating many of the signal integrity problems inherent to the 40-conductor
cable assembly. However, many of the design considerations and measurement techniques required for the
80-conductor cable assembly are different from those used for the 40-conductor assembly. Hosts and devices
intended to be used with 40- or 80-conductor cables may be designed to meet all requirements for operation
with both types. Unless otherwise stated, 40- and 80-conductor cables are assumed to be 18" long, the
maximum allowed by this standard. Timing and signal integrity issues as discussed apply to this length cable.
D.2 The issues
The following issues and design challenges are discussed along with suggestions for implementation: timing,
crosstalk between signals, ground bounce, and ringing.
D.2.1 Timing
Two of the features Ultra DMA introduced to the bus are double-edge clocking and non-interlocked (also known
as source-synchronous) signaling. Double-edge clocking allows a word of data to be transferred on each edge
of STROBE (this is HSTROBE for an Ultra DMA data out transfer and DSTROBE for a data in transfer),
resulting in doubling the data rate without increasing the fundamental frequency of signaling on the bus. Noninterlocked signaling means that DATA and STROBE are both generated by the sender during a data transfer.
In addition to signal integrity issues such as clocking the same data twice due to ringing on the STROBE
signal and delay-limited interlock timings on the bus, non-interlocked signaling makes settling time and skew
between different signals on the bus critical for proper Ultra DMA operation.
D.2.1.1 Cabling
The 80-conductor cable assembly adds 40 ground lines to the cable interspersed between the 40 signal lines
defined for the 40-conductor cable assembly. These added ground lines are connected inside each connector
on the cable assembly to the seven ground pins defined for the 40-conductor cable assembly. These additional
ground lines allow the return current for each signal line to follow a much lower impedance path to the outgoing
current than was allowed by the grounding scheme in the 40-conductor cable assembly. This results in greatly
reduced crosstalk on the data bus. The controlled impedance and reduced crosstalk of the 80-conductor cable
assembly results in much improved behavior of electrical signals on the bus and reduces the data settling time
to effectively zero regardless of switching conditions. Thus, the signal at the recipient is monotonic, such that
Page 350
T13/1321D revision 3
the first crossing of the input threshold is considered final. Reducing the time allowed for data settling from
greater than 25 ns in Ultra DMA mode 2, to 0 ns with the 80-conductor cable assembly allows nominal cycle
time to be reduced from 60 ns for mode 2, to 30 ns for mode 4.
D.2.1.2 Skew
Skew is the difference in total propagation delay between two signals as they transit the bus. Propagation
delay is the amount of time required for a single input signal at one part of the system to cause a disturbance
to be observed at another part of the system in a system containing continuously distributed capacitance and
inductance. Propagation delay is determined by the velocity of light within the dielectric materials containing
the electric fields in the system. For systems with uniform properties along their length, propagation delay is
often specified as seconds per foot or seconds per meter.
Skew will be positive or negative depending on which signal is chosen as the reference. All skews in the Ultra
DMA timing derivations are defined as STROBE delay minus data delay. A positive skew is a STROBE that is
delayed more than the data.
Skew corresponds to the reduction in setup and hold times that occurs between the sender and the recipient.
In order to insure that data is clocked correctly, the maximum allowable skew in each direction in a system is
less than the difference between the setup or hold time produced by the sender and required by the recipient.
Skew between signals will increase as they transit the bus based on differences in the electrical characteristics
of the paths followed by each signal. An understanding of the origins of skew and its importance to Ultra DMA
requires an explanation of the nature of signal propagation on a ground-signal-ground cable.
D.2.1.3 Source-terminated bus
The bus operates as a “source-terminated” bus, meaning that the only low-impedance connection to ground is
via the source impedance of the drivers in the sender.
R_source
100 Ω
R_rec
1 MΩ
+
−
Cable
Z0=100 Ω TD=4 ns
V_source
Figure D.1 − A transmission line with perfect source termination
On a source-terminated transmission line, the initial voltage level produced at the source propagates through
the system until it reaches the receiving end that, by definition, is an open circuit or at least has high
impedance relative to the characteristic impedance of the transmission line. This open circuit produces a
reflection of the original step with the same polarity and amplitude as the original step but travelling in the
opposite direction. The reflected step adds to the first step to raise the voltage throughout the system to two
times the original step voltage. In a perfectly terminated system (see figure D.1), R_source matches the cable
impedance resulting in an initial step voltage on the transmission line equal to fifty percent of V_source, and the
entire system has reached a steady state at V_source once the reflection returns to the source.
The waveforms that are measured on the bus as a result of this behavior depend on the ratio of the signal rise
time to the propagation delay of the system. If the rise time is shorter than the one-way propagation delay, the
initial voltage step will be visible at the sender. At the recipient the incoming voltage step is instantaneously
doubled as it reflects back to the sender and no step is observed (see figure D.2).
Page 351
T13/1321D revision 3
8v
4v
transmitter
receiver
0v
-4 v
0 ns
20 ns
40 ns
60 ns
80 ns
time
Figure D.2 – Waveforms on a source-terminated bus with rise time less than Tprop
If the rise time is longer than the propagation delay, the sender waveform changes, but the same behavior still
occurs: the reflected step adds to the initial step at the sender while a delayed doubling of the initial step is
observed at the recipient. Because the rising edges of the two steps overlap when measured at the sender,
there is a temporary increase in slew rate instead of a step seen at the sender while the rising edge of the
reflection adds to the edge still being generated by the sender (see figure D.3).
6v
4v
transmitter
receiver
0v
-1 v
0 ns
20 ns
40 ns
time
60 ns
80 ns
Figure D.3 – Waveforms on a source-terminated bus with rise time greater than Tprop
In figure D.4, the source impedance is perfectly matched to the cable impedance with the result that, after the
first reflection returns to the source, there are no further reflections, and the system is at a steady state. In a
system that is not perfectly terminated, there are two possibilities. The first possibility is when the source
Page 352
T13/1321D revision 3
impedance is less than the characteristic impedance of the transmission line, the initial step is greater than fifty
percent of VoH, and the system is at a voltage higher than VoH when the first reflection returns to the recipient
(see figure D.4). In this case another reflection occurs at the source to reduce the system to a voltage below
VoH but closer to VoH than the initial peak. Reflections continue but are further reduced in amplitude each time
they reflect from the termination at the source.
8v
4v
transmitter
receiver
0v
-4 v
0 ns
20 ns
40 ns
60 ns
80 ns
time
Figure D.4 – Waveforms on a source-terminated bus with R_source less than cable Z0
The second possibility is when the source impedance is higher than the characteristic impedance, the initial
step is less than fifty percent of VoH, and multiple reflections back and forth on the bus will be required to bring
the whole system up to a steady state at VoH (see figure D.5).
5v
transmitter
receiver
0v
0 ns
20 ns
40 ns
60 ns
80 ns
time
Figure D.5 – Waveforms on a source-terminated bus with R_source greater than cable Z0
Page 353
T13/1321D revision 3
Note that falling edges exhibit the same transmission line behavior as rising edges. The only difference
between the edges is that VoH and VoL are reversed. In actual systems output impedance and slew rate of the
drivers are often different between rising and falling edges, resulting in different step voltages and waveform
shapes.
For typical implementations using 33 Ω series termination, the effective driving impedance of a sender’s
component I/O viewed from the cable connector ranges from 50 to 90 Ω. The component I/O is the combined
input and/or output circuitry, bond wire, and pin on an IC that is responsible for receiving and/or sending data on
a particular conductor within the bus. The initial voltage step produced when an edge is driven onto the cable
will be equal to the driver’s open-circuit VoH divided by the effective output impedance and the input impedance
of the cable (typically 82 Ω), or a 50 to 60 Ω printed circuit board trace in the case of hosts. This step voltage
will fall in the range from 50 to 70 percent of VoH. For example, for a theoretical source with zero output
impedance using 33 Ω termination driving an 82 Ω cable the resulting step voltage is not greater than 100 ∗ (
82 ÷ (33 + 82 ) ) = 71.3 percent of VoH. Because the thresholds of an input are not centered with respect to the
high and low voltages, the initial voltage step produced by a driver will often cross the recipient’s input threshold
on a rising edge but not on a falling edge. However, since the signal received at the end of the bus is a doubled
version of the initial output from the sender, it will cross the switching thresholds for any reasonably low output
impedance. Because of this the main voltage step only affects skew and delay for signals received at devices
that are not at the end of the cable. The greater the distance a device is from the device end of the cable (i.e.,
closer to the host), the longer the duration of the step observed (see figures D.6 and D.7).
transmitter
receiver
-20 ns
#Avg
16
30 ns
10.00 ns/div
80 ns
repetitive
Figure D.6 – Typical step voltage seen in ATA systems using an 80-conductor cable (measured at
drive and host connectors during read)
Page 354
T13/1321D revision 3
transmitter
receiver
-20 ns
# Avg
16
30 ns
10.0 ns/div
80 ns
repetitive
Figure D.7 – Typical step voltage seen in ATA systems using an 80-conductor cable (measured at host
and drive connectors during write)
In addition to the step produced by the initial voltage driven onto the bus and the subsequent reflection, smaller
steps are produced each time the propagating signal encounters a change in the bus impedance. The major
impedance changes that occur in a system are: 1) at the connections between the cable and the printed
circuit boards (PCBs) of the hosts and devices, 2) along the traces of the PCBs as the result of changing
layers, and 3) at the connection between a motherboard and a backplane.
The transmission line behavior of the 80-conductor cable assembly adds skew to the received signal in two
ways: First, impedance differences along one line versus another will result in different amounts of delay and
attenuation on each line due to reflections on the bus. This produces a time difference between the two
signals’ threshold crossings at the recipient. Secondly, signals received at the device that is not at the end of
the cable may cross the threshold during the initial voltage step or after the reflection from the end of the cable
is received, depending on the supply voltage, series termination, output impedance, VoH, and PCB trace
characteristics of the host.
Factors other than cable characteristics also contribute to skew. Differences in the capacitive loading between
the STROBE and DATA lines on devices attached to the bus will delay propagating signals by differing
amounts. Differences in slew rate or output impedance between drivers when driving the 82 Ω load will result in
skew being generated as the signal is sent at the sender. Differences between the input RC delays on
STROBE and DATA lines will add skew at the recipient.
The fundamental requirement for minimizing skew in the entire system is to make the STROBE and DATA lines
as uniform as possible throughout the system. Methods of achieving this are described in D.1.10.
D.2.1.4 Timing measurements for the 80-conductor cable assembly
The reflections that are present in a system make it difficult to measure skew and delays accurately. For the
received signal at a device, the propagation delay from the device connector to the device integrated circuit (IC)
connector pin is about 300 ps for typical PCBs and trace lengths. The IC is the entire component (die and
package) that contains the ATA bus interface circuitry.
This delay introduces an error of plus or minus 300 ps in timing measurements made at the device connector
since rising edges and falling edges will be measured before and after the step respectively. When comparing
two signals, this results in an error in measured skew of plus or minus 600 ps due to the measurement
Page 355
T13/1321D revision 3
position. This error is small enough relative to the total timing margin of an Ultra DMA system that it may be
ignored in most cases.
Since the trace length on host PCBs are often much longer than those on devices, the propagation time for a
signal from the host connector to the host IC may be as high as 2 ns. This results in a plus or minus 2 ns
accuracy in the measurement of a single signal and a plus or minus 4 ns accuracy for skew between two
signals. These errors are not removed by adding or subtracting an allowance for PCB propagation delay
depending on rising or falling edges because characteristics of the PCB and termination will affect the step
levels and skew that occur at the component I/Os. As a result of this, accurate measurements of skew in
signals received at the host are made either at pins of the host IC, or at points on the PCB traces as close to
the IC pins as possible. Test pads, headers, or unconnected vias in PCB layouts may be designed allowing
connection to DATA, STROBE, and ground for this purpose.
It is important to note that the timing specifications for Ultra DMA in the standard are based on measuring
signals at the interface connector.
D.2.1.5 Simulations for the 80-conductor cable assembly
The difficult nature of measuring skew in actual systems makes simulations a more important tool in
determining the effect on skew of design decisions regarding component I/Os, PCB layout, cable lengths, and
other aspects of system design. Because of the well-controlled impedance of the 80-conductor cable
assembly, single line transmission line models provide accurate predictions of the delay through the bus based
on a given design choice for a given set of conditions on the bus. To be certain of the system-wide
consequences of particular design choices, a large number of simulations encompassing many different
combinations of parameters were used to determine the timing specifications for Ultra DMA mode 4. Results of
these simulations are also the basis of the guidelines that follow.
Output skew is measured at the connector of the sender into capacitive loads to ground of 15 pf and 40 pf. An
alternate loading arrangement is to measure the signal produced at the end of an 18-inch 80-conductor cable
assembly into typical device and host loads of 20 pf or 25 pf that are held uniform across STROBE and DATA
lines. Skew is measured at the crossing of the 1.5 volt threshold. All combinations of rising and falling edges
on the signals involved are used when skew is measured.
Minimizing output skew is the best assurance of reliable signaling across the full range of cable loading and
recipient termination conditions that will occur in systems.
D.2.2 Crosstalk
Although the ground-signal-ground configuration of the 80-conductor cable assembly greatly reduces coupling
between wires on the cable, the host and device connectors generate a large amount of crosstalk because they
still use the original ground configuration with no ground lines separating the 16 signals of the data bus. In
addition, crosstalk between traces on the PCB may reach high levels in systems with long traces or with tight
spacing between traces. Cumulative crosstalk plus ground bounce measured at the connector of the recipient
in typical systems using the 80-conductor cable ranges from 400 mV to 1 V peak, in short pulses with a
frequency content equivalent to the frequency content of the edge rates of the drivers being used. Although this
level of total crosstalk may seem like a hazard to reliable signaling, crosstalk exceeding 800 mV detected at
the recipient does not affect the setup or hold times when it occurs during the interval when other signals are
switching (see figure D.8). This figure was generated using the first falling STROBE edge for a trigger and
showing a middle data signal staying low while all other lines switch high to low. With infinite persistence, the
pattern was then changed to all lines switching low to high for the same STROBE edge. The crosstalk that
occurs on the line staying low while all others switch high to low is in excess of 800 mV but has more hold and
setup time margin than data lines that are switching and therefore it does not reduce setup or hold time margin.
Page 356
T13/1321D revision 3
X1
Y1
Y2
X2
108 ns
118 ns
2.0 ns/div
Y2
Y1
delta Y
800 mV
1.5 V
-700 mV
128 ns
realtime
X2
X1
delta X
1/delta X
119 ns
117 ns
1.8 ns
555.556 MHz
Figure D.8 – Positive crosstalk pulse during a falling edge (does not affect data setup or hold time)
A larger signal integrity hazard exists when crosstalk extends into the middle of the cycle when data could be
clocked. This may result from a high level of reverse crosstalk detected at the recipient as the reflected signal
propagates from the recipient input back to the sender output in the switching lines.
X1
Y2
Y1
X2
-21 ns
29 ns
10 ns/div
Y2
Y1
delta Y
552.125 mV
504.8 mV
47.3250 mV
X2
X1
delta X
1/delta X
79 ns
realtime
42.2 ns
21.2 ns
21.0 ns
47.6190 MHz
Figure D.9 – Reverse crosstalk waveform from reflected edge
(seen at the receiver in the middle of a cycle – marker X1)
Reducing a system’s creation of and susceptibility to forward and reverse crosstalk requires an understanding
of how crosstalk is generated and propagates through the system. Crosstalk results from coupling between
signals in the form of either a capacitance from one signal conductor to another or inductors in the path of each
Page 357
T13/1321D revision 3
signal with overlapping magnetic fields. The capacitive and inductive coupling are easiest to understand if
treated as separate effects.
D.2.2.1 Capacitive coupling
Capacitive coupling in its simplest form consists of a capacitor connecting together two transmission lines
somewhere along their length. When a change in voltage occurs on one line (called the “aggressor” line), a
pulse on the non-switching signal (called the “victim” line) is produced with a peak amplitude proportional to the
rate of change of voltage (dV/dt) on the aggressor line. The pulse on the victim line propagates both forward
and backward from the point of coupling and has the same sign in both directions. Forward and backward are
defined relative to the direction that the aggressor signal was propagating. Forward means that the propagation
is in the same direction that the aggressor signal was propagating. Backward means that propagation is
opposite the direction that the aggressor signal was propagating. Figure D.10 is a schematic of a model for
capacitive coupling. Figure D.11 shows waveforms resulting from capacitive coupling at the sender and recipient
component I/Os of the aggressor and victim lines.
Aggressor_src
100 Ω
T1
Z0=100 Ω TD=5 ns
T2
Z0=100 Ω TD=5 ns
Aggressor_rec
100 Ω
+
−
C1
6 pf
V_source
Victim_src
100 Ω
T3
Z0=100 Ω TD=5 ns
T4
Z0=100 Ω TD=5 ns
Figure D.10 – Model of capacitive coupling
Page 358
Victim_rec
100 Ω
T13/1321D revision 3
1.5 V
-0.5 V
Aggressor at receiver
200 mV
-200 mV
Victim at source
200 mV
-200 mV
0 ns
10 ns
Victim at receiver
20 ns
30 ns
40 ns
50 ns
60 ns
Time
Figure D.11 – Waveforms resulting from capacitive coupling (at transmitter and receiver of aggressor
and victim lines)
D.2.2.2 Inductive coupling
In the following inductive coupling is modeled as an inductor in series with each signal, with some coupling
factor K representing the extent to which the inductors’ magnetic fields overlap. In effect these two inductors
constitute a transformer, creating a stepped-down version of the aggressor signal on the victim line. The
amplitude of the signal produced on the victim line is proportional to the rate of change in current (di/dt) on the
aggressor line. Since the impedance of a transmission line is resistive, for points in the middle of a
transmission line di/dt will be proportional to dV/dt. Because the crosstalk signal produced across the
inductance in the victim line is in series with the transmission line, it has a different sign at each end of the
inductor. Because the current in an inductor always opposes the magnetic field that produced it, the polarity of
the crosstalk signal is reversed from the polarity of the di/dt on the aggressor line that produced it. As a result
of these two facts, inductive crosstalk creates a pulse of forward crosstalk with polarity opposite to the edge on
the aggressor, and a pulse of reverse crosstalk with the same polarity as the aggressor edge. Figure D.12 is a
schematic of a model for inductive coupling. Figure D.13 shows waveforms resulting from inductive coupling at
the sender and recipient component I/Os of the aggressor and victim lines.
Page 359
T13/1321D revision 3
Aggressor_src
100 Ω
T2
Z0=100 Ω TD=5 ns
T1
Z0=100 Ω TD=5 ns
Aggressor_rec
100 Ω
L1 50 nH
+
−
MUTUAL COUPLING=1
V_source
Victim_src
100 Ω
T4
Z0=100 Ω TD=5 ns
T3
Z0=100 Ω TD=5 ns
Victim_rec
100 Ω
L2 50 nH
Figure D.12 – Model of inductive coupling
Note that the box in this figure, figure D.14, and figure D.18 between L1, L2 and K2 is a PSPICE element
representing the inductive coupling between L1 and L2 having the coupling value listed in the figure.
1.5 V
-0.5 V
100 mV
Aggressor at receiver
-100 mV
100 mV
Victim at receiver
-100 mV
10 ns
20 ns
Victim at source
30 ns
40 ns
50 ns
60 ns
70 ns
Time
Figure D.13 – Waveforms resulting from inductive coupling (at transmitter and receiver of aggressor
and victim lines)
Page 360
T13/1321D revision 3
D.2.2.3 Mixed capacitive and inductive coupling
Most occurrences of electromagnetic coupling involve both capacitive and inductive coupling. In this case the
forward and reverse crosstalk contributions of the capacitance and inductance add together. Because the
forward inductive crosstalk and the forward capacitive crosstalk have opposite signs, they tend to cancel, while
the reverse crosstalk from both effects have the same sign and add together. Depending on the ratio of
inductive to capacitive coupling, the forward crosstalk may sum to zero when both effects are added together.
Figure D.14 is a schematic of a model for mixed capacitive and inductive coupling. Figure D.15 shows
waveforms resulting from mixed capacitive and inductive coupling at the sender and recipient component I/Os of
the aggressor and victim lines.
T2
Z0=100 Ω TD=5 ns
T1
Z0=100 Ω TD=5 ns
Aggressor_src
100 Ω
Aggressor_rec
100 Ω
L1 50 nH
+
−
MUTUAL COUPLING=1
V_source
C12
6 pf
Victim_src
100 Ω
T4
Z0=100 Ω TD=5 ns
T3
Z0=100 Ω TD=5 ns
Victim_rec
100 Ω
L2 50 nH
Figure D.14 – Model of capacitive and inductive coupling
Page 361
T13/1321D revision 3
1.5 V
-0.5 V
Aggressor at receiver
200 mV
-200 mV
200 mV
Victim at receiver
-200 mV
10 ns
20 ns
30 ns
Victim at source
40 ns
50 ns
60 ns
70 ns
Time
Figure D.15 – Waveforms resulting from mixed capacitive and inductive coupling (at transmitter and
receiver of aggressor and victim lines)
D.2.2.4 Crosstalk from distributed coupling
When transmission lines are placed parallel with and in close proximity to each other, as is the case for PCB
traces, wires in a ribbon cable, etc., the coupling that occurs is continuous along the length of the transmission
lines. To find the crosstalk waveforms at the source and recipient, divide the transmission lines into segments
and treat each segment as an instance of capacitive and inductive coupling. Each segment produces forward
and reverse crosstalk as the aggressor edge goes by. Sum the contributions from each of these segments,
delaying their arrival at the ends according to the segment’s position along the transmission line. Doing this
shows that the forward crosstalk contributions all add together and arrive simultaneously with the aggressor
edge, while the reverse crosstalk is spread out along the length of the transmission line and produces a long
flat pulse travelling back toward the source. Figure D.16 shows a schematic model for a transmission line with
three coupled conductors, connected as two signal wires and a ground return. The waveform at the source end
of the victim line in figure D.17 shows that the reverse crosstalk pulse begins when the edge is driven onto the
aggressor line and continues to be observed at the source until one system delay after the end of the edge is
terminated at the recipient on the aggressor line. The waveform at the victim recipient’s component I/O shows
that the forward crosstalk arrives simultaneously with the edge on the aggressor line, or even slightly before,
due to the fact that the energy in the crosstalk pulse has been subtracted from the edge on the aggressor,
reducing its rise time at the recipient.
Page 362
T13/1321D revision 3
Aggressor_rec
100 Ω
Aggressor_src
100 Ω
+
−
V12
Victim_src
100 Ω
in 1
out 1
in 2
out 2
In 3
out 3
Victim_rec
100 Ω
T3 coupled
L=500 nH
C=50 pf
LM= 100 nH
CM=1- pf
Figure D.16 – Model of distributed coupling
1.5 V
-0.5 V
Aggressor at receiver
100 mV
-100 mV
Victim at receiver
100 mV
-100 mV
0 ns
10 ns
Victim at source
20 ns
30 ns
40 ns
50 ns
60 ns
Time
Figure D.17 – Waveforms resulting from distributed coupling (at transmitter and receiver of aggressor
and victim lines)
The above simulation results of figures D.11, D.13, D.15, and D.17 are simplified by the assumption that all
transmission lines are perfectly terminated at both ends. In actual systems only the sender end of the bus has
a low-impedance termination to ground, and this termination is seldom perfect. The consequences of this help
to explain some characteristics of crosstalk in a system:
Page 363
T13/1321D revision 3
1) Crosstalk is produced by both the initial and reflected edges on the aggressor lines. Forward crosstalk
produced by the initial edge as it propagates from the sender to the recipient arrives at the same time
as the edge that produced it. The edge on the aggressor signals reflects from the high impedance at
the recipient input (or at the end of the cable) and returns back to the sender. Reverse crosstalk
produced as this reflected edge propagates back to the sender is observed on the victim line at the
recipient.
2) If reverse crosstalk from the initial edge is not perfectly terminated at the sender’s component I/O it will
be reflected (with reduced amplitude) back towards the recipient. The quality of the sender’s
component I/O termination depends on the instantaneous output impedance of drivers as they are
switching, as well as the “on” resistance of the drivers in the high or low state once they have
completed switching. Since the source impedance is made up of the driver output impedance in series
with the termination resistors, the most accurate source termination is achieved by using drivers with
low output impedance combined with high value series resistors, creating a total output impedance
near 75 Ω.
3) Crosstalk is observed with doubled amplitude at the high-impedance endpoint of the system (at the
host input during READ operations and at the device input at the device end of the cable during WRITE
operations) due to the reflection. Since crosstalk occurs as a pulse rather than a step, the initial and
reflected portions of the pulse only sum at the endpoint while the pulse is reflecting, and not at other
points along the bus.
4) Series termination resistors at the receiving end of the bus serve to attenuate the amplitude of
crosstalk observed at the receiving component I/Os. Because the component I/O impedance is
predominantly capacitive, its impedance decreases at high frequencies. At the frequency where the
impedance of the component I/O equals the impedance of the series termination resistor, the crosstalk
pulse amplitude observed at the IC input will be about half of the amplitude measured at the connector.
The formula for determining this frequency is F = 1 ÷ (2 ∗ π ∗ R ∗ C) where F is the frequency, R is the
value of the series termination resistor, and C is the input capacitance of the recipient’s component I/O.
So when crosstalk levels are high enough to be a serious concern, the best place to make
measurements of the crosstalk is at the component I/O or on the IC side of the termination resistor. In
design of systems, this filtering effect is used to reduce a system’s susceptibility to crosstalk by
increasing the value of series termination resistors and placing them close to the connector to
maximize the amount of capacitance on the IC side of the resistor.
In systems using the 80-conductor cable the largest contributors to crosstalk are the connector at the sender,
and the PCB traces in systems with long traces or a large amount of coupling between traces. The connector
at the receiving end of the system generates less crosstalk than the one at the sending end because the net
current flow through the aggressor lines is less at the receiving end. This is because the load on the IC side of
the recipient’s connector is the PCB trace and a small capacitance inside the component I/O; only enough
current flows through the connector to charge this total capacitance. At the sending end of the system, the
instantaneous value of current through the connector is determined by the input impedance of the cable, and
this amount of current flows for a length of time sufficient to charge the entire system including the cable and all
attached devices up to the sender’s VoH .
Crosstalk in the connectors is almost entirely inductive. It is produced in both directions from the connector but
not necessarily in equal amplitudes. The highest amplitude crosstalk is generated by many switching lines
coupling into a small number of victim lines, that lowers the effective source impedance of the crosstalk,
making it approximate a voltage source. This voltage source is in series with the transmission line impedance
on each side of the connector on the victim line. As a result, the crosstalk voltage is divided between the two
directions proportional to the impedance seen in each direction. Figure D.18 shows the schematic of a model
that demonstrates this. The PCB and cable on the victim line have been replaced with resistors to simplify the
resulting waveforms. Figure D.19 shows the current through the inductor on the aggressor line and the crosstalk
voltage produced on the victim line into the resistors representing the PCB and cable impedance. The
waveforms indicate that the crosstalk voltage divides in the expected ratio. In this example the PCB receives
(50 ÷ (82 + 50)) ∗ 100 % = 37.9 % of the total voltage across the inductor, while the cable receives the
remaining 62.1%. In an actual system, the crosstalk at the source is terminated by the driver impedance. The
Page 364
T13/1321D revision 3
crosstalk measured at the recipient’s component I/O on the victim line is double the value of the crosstalk pulse
initially produced into the cable impedance.
Aggressor_src
100 Ω
T2
Z0=100 Ω TD=5 ns
T1
Z0=100 Ω TD=5 ns
Aggressor_rec
100 Ω
L1 50 nH
+
−
V_source
MUTUAL COUPLING=1
Cable
82 Ω
PCB_trace
50 Ω
L2 50 nH
Figure D.18 – Model of voltage divider for connector crosstalk formed by PCB and cable
15 ma
-5 ma
at L1
200 mV
-200 mV
At PCB trace
200 mV
-200 mV
10 ns
0 ns
at cable
20 ns
30 ns
40 ns
50 ns
60 ns
Time
Figure D.19 – Waveforms showing connector crosstalk dividing between PCB and cable
Page 365
T13/1321D revision 3
For each edge on the bus four crosstalk pulses are created on non-switching victim lines due to the combined
crosstalk in the PCB, connector, and cable:
1) Forward crosstalk from the initial edge has the same sign as the edge and is seen at the recipient as a
pulse that arrives with the edge. The amplitude of the pulse is doubled at the recipient’s component
I/O, however because it occurs during the interval when the data is changing it may decrease the
signal’s setup or hold time but it presents a minor risk to data integrity overall.
2) Reverse crosstalk from the initial edge travels back towards the driver as a flat pulse with a width equal
to the transition time of the driver. Based on the degree of mismatch between the driver’s output
impedance and the cable impedance, this pulse may be reflected back towards the recipient with
reduced amplitude. Because it continues to arrive at the recipient well after the driver has completed
switching, it creates a risk of incorrect data at the recipient in the middle of the cycle. However, it is
unlikely to ever create a high enough amplitude at the recipient to cause a problem.
3) Forward crosstalk from the reflected edge arrives back at the driver simultaneously with the reflected
edge on the aggressor lines. Depending on the impedance mismatch at the source, the edge will be
reflected back towards the recipient with reduced amplitude and arrives in the middle of the cycle,
however this edge will seldom create a high enough amplitude at the recipient to cause problems.
4) Reverse crosstalk from the reflected edge on the aggressor lines will be created travelling back toward
the recipient and arrives there in the middle of the cycle. In host systems where the termination
resistors are not placed next to the connector a larger portion of the crosstalk created in the connector
will be reverse crosstalk on the cable side because of the divider formed by the 50 to 60 Ω PCB and
the 82 Ω cable impedance. The pulse will be seen with doubled amplitude by the device at the end of
the cable and presents a serious hazard to data integrity if its amplitude at the recipient’s component
I/O exceeds 800 mV.
D.2.2.5 Measuring crosstalk in a system
To measure the total crosstalk in a system set up a data pattern in which one line in the middle of the data bus
is held low while all other lines are asserted simultaneously. Measure the low line at the recipient connector or
component I/O. This measurement includes ground bounce at the sender IC discussed in D.2.3 as well as the
contributions to crosstalk of the PCBs, connectors, and cables. Determining the exact sources of the different
features of the crosstalk measured by this technique is difficult. An effective method to isolate the crosstalk
produced into a victim line in a given portion of the system is to sever the line before and after the feature being
tested. Terminate the isolated segment to ground at the breaks with resistors equivalent to the transmission
line impedance that is normally seen at those points. Measuring the crosstalk voltage across the termination
resistors will indicate the raw quantity of crosstalk into the victim line produced by that portion of the system,
independent of reflections due to impedance mismatches and attenuation due to capacitance along the bus.
Adjusting for impedance mismatches and delays will allow the crosstalk from that portion to be identified in the
total crosstalk of the system, and adjusting the impedance changes through the system may allow the impact
of that crosstalk to be minimized.
D.2.2.6 System design considerations to minimize crosstalk
Because all crosstalk throughout the system is proportional to edge rate, a major factor in controlling crosstalk
is controlling the output slew rate of the drivers. Another major factor is the impedance match of sources to the
cable including the value and placement of termination resistors. Source impedance matching is important to
prevent reverse crosstalk from reflecting off the source and out to the recipient. Drivers, PCB layout, and
termination resistors are selected to provide a good source termination for crosstalk and the reflected signal
edge. Ideal termination at each connector is when the impedance seen looking back toward the source
matches the cable impedance in the forward direction. For devices, this means that the sum of driver output
impedance and termination resistance match the cable impedance (typically 80 to 85 Ω), minus five to ten
percent to allow for attenuation due to the capacitive loading of other devices on the cable. Because the PCB
traces on a device are short, they have little effect on the device’s output impedance
Due to other design constraints, many hosts PCB traces are so long that, for high-frequency crosstalk, the
impedance at the host connector is determined by the PCB trace impedance and termination resistors (if they
Page 366
T13/1321D revision 3
are located at the connector), rather than by the driver’s output impedance. Because of this, there are two
options for hosts with longer traces to ensure an ideal source termination:
1) Place the termination resistors near the sender’s component I/O and use a PCB trace impedance that
matches the source impedance of the sender’s component I/O plus termination resistor. This ideal
impedance is slightly less than the cable impedance. In this case, trace impedance of 70 to 75 Ω with
a large enough trace spacing to keep crosstalk (especially reverse crosstalk) between PCB traces to a
minimum is ideal.
2) Place the termination resistors near the connector and select PCB trace impedance and termination
resistance to sum to the cable impedance or slightly less. In this case, matching the sender’s
component I/O source impedance to the PCB trace impedance rather than the cable impedance is
ideal, since that is the load that it is immediately driving.
Option 2 is desirable for backward compatibility with older systems using the 40-conductor cable because
placing the resistor near the connector helps to damp the ringing that occurs with that cable. In addition, 50 to
60 Ω traces are easier to implement and produce less crosstalk than higher impedance traces making the
second option a better choice in most cases.
In either case, matching the total output impedance to the cable impedance under all conditions of steady-state
or switching, or experiencing over or undershoot conditions is the best solution.
D.2.3 Ground/Power Bounce
Supply bounce is a form of crosstalk that results from changes in current through power and ground pins of IC
packages. For single-ended drivers, the return current for all signals flows through the power and ground leads,
with the result that any voltage drop across these pins is imposed on all signals equally. Voltage drops across
these pins occur due to both resistance and inductance whenever there is a net current flow into or out of the
signal pins of the IC, though inductance has the greatest effect. In terms of the voltage seen at the recipient’s
component I/O, crosstalk due to supply bounce is indistinguishable from inductive crosstalk, with a sign
opposite the polarity of the edge on the aggressor signal(s). See figure D.20 for a model of ground bounce in an
IC package. See figure D.21 for waveforms resulting from ground bounce at the sender’s and recipient’s
component I/O of the aggressor and victim signals.
In order to measure supply bounce in a functioning system, it is necessary to remove all other sources of
crosstalk (especially reverse crosstalk from points later in the system). To remove the other sources of
crosstalk, disconnect the component I/O pin on which the measurement is being taken from the PCB and
measure the voltage at the component I/O while all other lines are switching. The initial and the reflected edges
on the switching lines will produce supply bounce. Measurements with the victim line in a high state show
power bounce and with the victim line in a low state show ground bounce. The ground inside the IC will
“bounce” and produce crosstalk on a low victim line when many lines are switching from high to low and sinking
current through the ground pins. The power inside the IC will “bounce” and produce crosstalk on a high victim
line when many lines are switching from low to high, and drawing current through the power pins.
Page 367
T13/1321D revision 3
T1
Z0=100 Ω TD=5 ns
Aggressor_src
100 Ω
Aggressor_rec
100 Ω
+
−
V12
L1 50 nH
Victim_rec
100 Ω
Victim_src
100 Ω
T2
Z0=100 Ω TD=5 ns
Figure D.20 – Model of ground bounce in IC package
1.5 V
Aggressor at source
Aggressor at receiver
-0.5 V
400 mV
-400 mV
Victim at receiver
400 mV
-400 mV
0 ns
10 ns
Victim at source
20 ns
30 ns
40 ns
50 ns
60 ns
Time
Figure D.21 – Waveforms resulting from ground bounce (at transmitter and receiver of aggressor and
victim lines)
Page 368
T13/1321D revision 3
D.2.4 Ringing and data settling time (DST) for the 40-conductor cable assembly
High amplitude ringing may occur for some data patterns in systems using the 40-conductor cable assembly.
The sixteen data lines (DD 15:0) in a 40-conductor cable assembly are adjacent to each other and have only
one ground on each side of the data lines. There are only seven ground lines present in the entire cable
assembly. This lack of ground return paths has three negative effects on data signal integrity:
1) Crosstalk between data lines is very high due to inductive coupling.
2) Conductors in the center of the set of data lines (e.g., DD 11) exhibit very high inductance because the
distance from these signal lines to the current return path is large and the ground return path is shared
with many other signal lines.
3) Conductors in the center of the set of data lines are shielded from ground by the other data lines
around them. When these lines are switching in the same direction there is no potential difference and
therefore no effective capacitance between lines.
This combination of factors results in the impedance of the conductors in the center of the set of data lines
rising from 110 to 150 Ω (measured when a single line is asserted or negated) to an almost purely inductive 300
to 600 Ω when all lines are asserted or negated simultaneously in the same direction. Measured impedance
varies with data pattern, edge rate, cable length, loading, and distance from chassis ground.
Unlike the 40-conductor cable, the 80-conductor cable has the additional 40 ground lines making all signals
ground-signal-ground. This makes the 80-conductor cable impedance relatively constant with respect to
pattern. Matching impedance and controlling PCB trace geometry as discussed in D.2.2 will result in well
damped ringing and crosstalk in victim lines that remains below 800 mV.
In the following simplified model of the 40-conductor cable assembly with all data lines switching, a conductor
in the center of the set of data lines is described as a pure inductor, forming a series RLC resonant circuit with
the capacitance of the component I/O and PCB traces, and the combined resistance of the driver source
impedance and source series termination resistor (see figure D.22). The voltage across C will ring sinusoidally
in response to an input pulse at V_source, exponentially decaying over time towards a steady state value. The
formula for determining the frequency of this ringing is F = 1 ÷ (2π ∗ SQRT(LC)) where F is the frequency, R is
the value of the series termination resistor, and C is the input capacitance of the recipient’s component I/O.
The rate of decay is proportional to R/L. Figure D.23 shows the output of a simple RLC model with the
waveforms as seen at the connectors of the sender and recipient.
R
40 Ω
+
−
V_source
L
0.8 µH
C
25 pf
Figure D.22 – Simple RLC model of 40-conductor cable with all data lines switching
Page 369
T13/1321D revision 3
10 v
source end
5v
receiving end
0v
-5 v
0 ns
50 ns
100 ns
150 ns
Time
Figure D.23 – Output of Simple RLC model: waveforms at source and receiving connectors
Data settling time (DST) is defined as the portion of cycle time required for ringing to decrease in amplitude until
a signal reaches the threshold of 2.0 volts (V iH) or 800 mV (V iL ). The worst-case situation for most systems
occurs when all data lines are switching except for one line near the middle of the bus that is being held low
(see figure D.24).
In this situation crosstalk creates a pulse on the signal line being held low that rings with a frequency and
damping determined by the effective RLC parameters of the system. The DST value is the duration of time
between the nominal beginning of the cycle (i.e., when the switching lines cross the 1.5 volt threshold) and the
time when the ringing on the line drops below ViL for the last time as measured at the recipient’s component
I/O.
Page 370
T13/1321D revision 3
X1
driven line
Y1
Y2
victim line
X2
-22 ns
28 ns
10 ns/div
Y2
Y1
delta Y
800 mV
1.5 V
-700 mV
78 ns
repetitive
X2
X1
delta X
1/delta X
17.6 ns
-1.0 ns
18.6 ns
53.7634 MHz
Figure D.24 – DST measurement for a line held low while all others are switching high (ch1 on DD3 at
rec., ch2 on DD11 at rec.)
The same situation also occurs with reversed signal polarity (e.g., one line staying high while others are
switching). Another case arises when all lines are switching simultaneously and the voltage on conductors in
the center of the set of data lines rings back across the switching threshold (see figure D.25). This is normally
only a problem in the high state as low side ringing is greatly reduced by the substrate diode clamp to ground
that is inherent in CMOS logic.
X1
receiving end
source end
Y1
Y2
X2
-22 ns
#Avg
28 ns
10.0 ns/div
Y2
Y1
delta Y
2.0 V
1.5 V
500 mV
X2
X1
delta X
1/delta X
78 ns
repetitive
31.8 ns
5.0 ns
26.8 ns
37.3134 MHz
Figure D.25 – DST measurement for all lines switching (ch1 at source, ch2 at rec.)
Page 371
T13/1321D revision 3
As seen in figure D.25, use of 3.3 volt signaling removes the high side voltage margin provided by the
asymmetric threshold of the recipient input. Consequently it is important to use slew rate controlled drivers to
control ringing.
D.2.4.1 Controlling ringing on a 40-conductor cable assembly
An improved RLC model allows comparison between different termination schemes (see figures D.26 and D.27).
These models include separate capacitors to represent trace and component I/O capacitance at the recipient’s
component I/O, as well as a clamping diode, representing the substrate diode in CMOS logic. Because this
single-line simplified model does not include crosstalk between lines in the data bus, it is not used to predict
DST for a particular design and combination of parameters. However, it does indicate the direction of changes
in ringing frequency and damping in response to changes in system parameters.
R_source
7Ω
R_series_src
33 Ω
R_series_rec
33 Ω
C_trace
15 pf
+
−
Cable
1 µH
C_ICpin
10 pf
D1
D1N914
V_source
Figure D.26 – Improved model of 40-conductor cable ringing with termination at IC
R_source
7Ω
Cable
1 µH
R_series_rec
33 Ω
C_trace
15 pf
+
−
R_series_src
33 Ω
C_ICpin
10 pf
D1
D1N914
V_source
Figure D.27 – Improved model of 40-conductor cable ringing with termination at connector
Comparing the results (figure D.28) given by these models for recipient termination resistors located at the IC
versus the connector shows that greater damping is provided when termination is near the connector.
Page 372
T13/1321D revision 3
10 v
waveform at receiver IC
waveform at receiver connector
5v
0v
waveform at source
connector
-5 v
0 ns
100 ns
50 ns
150 ns
time
Figure D.28 – Results of improved 40-conductor model with termination at IC vs. connector
These simple models are used in a similar way to determine the effects of changing slew rate, termination
resistor value, output impedance, PCB trace length, or the length of the cable.
10 v
1 ns risetime
5 ns risetime
10 ns risetime
5v
0v
-5 v
0 ns
100 ns
50 ns
150 ns
time
Figure D.29 – Results of improved 40-conductor model with source rise time of 1,5,and 10ns
As the results in figure D.29 show, increasing the rise time to above 5 ns results in a large decrease in the
amplitude of the ringing. Drivers with control over the shape of rising and falling edges are used to reduce
ringing even more.
Page 373
T13/1321D revision 3
Figures D.28 and D.29 show that, although the diode clamps the voltage at the recipient at one diode drop
below ground, a ringback pulse appears at around 100 ns. This pulse occurs because the combined series
resistance of the termination resistor and diode is much lower than the impedance of the LC circuit that is
ringing. In addition the diode only clamps the voltage across part of the capacitance involved in the ringing. A
higher-resistance clamping diode would be more effective at dissipating energy from the resonant circuit but
would be less effective at clamping the input voltage.
D.2.4.2 STROBE lines on the 40-conductor cable
Although the data bus on the 40-conductor cable has such a high level of crosstalk that transmission line
effects are barely perceptible, the STROBE lines on the 40-conductor cable have a more controlled impedance
of about 115 Ω because they are in a ground-signal-ground configuration. Although the STROBE lines are well
shielded against crosstalk from each other and from the data bus, some devices using drivers with fast edge
rates and no source termination resistors have experienced problems with overshoot and ringback on the
STROBE lines. Ringing will occur when a large impedance mismatch exists between the driver output
impedance and the 115 Ω transmission line. If the ringback on a falling edge exceeds 800 mV, STROBE may
cross the threshold multiple times and cause extra words to be clocked at the recipient. After these problems
were experienced almost all device and host manufacturers began using series termination resistors on the
STROBE lines at both the sender and the recipient.
With current component I/O technology and the requirement for series termination resistors, ringing on the
STROBE lines is seldom a problem for current systems. However, it is important to keep in mind that these
are high speed edge triggered signals, and the possibility of double crossing of input thresholds due to noise,
ringing, or transmission line reflections still exists. Because of this it is important that all hosts and devices
implement some amount of hysteresis on STROBE inputs in addition to glitch filtering by digital logic after the
inputs.
D.3 System Guidelines for Ultra DMA
This is a summary of recommendations for device, system, and chipset designers. These guidelines are not
strict mandates, but are intended as tools for developing compatible, reliable, high-performance systems.
D.3.1 System capacitance
All hosts and devices are required in the body of the standard to meet maximum values of capacitance as
measured at the connector. These values are specified to be 25 pf at the host and 20 pf at the device. With
typical interface IC and PCB manufacturing technology, this limits host trace length to four to six inches. It is
recommended that capacitance be measured at 20 MHz as this is representative of typical ringing frequencies
on an 18-inch 40-conductor cable assembly.
PCB traces up to 12 inches long may be used if the following conditions are met:
1.
2.
3.
4.
5.
The host chipset uses 3.3 volt signaling,
The host chipset allows timing margin for the additional propagation delay in all delay-limited interlocks,
Termination resistors are chosen to minimize input and output skew and are placed near the
connector,
Total capacitance of traces, additional components, and host component I/Os is held to the minimum
possible, and
An 80-conductor cable is installed for operation at Ultra DMA modes 2 and higher.
In this case capacitance at the connector will exceed the maximum value specified. As a result of this,
systems may not operate reliably with a 40-conductor cable assembly in any Ultra DMA mode above mode 1
(22.2 megabytes per second). Under these conditions it is advisable that a host not set mode 2 or above
without insuring that an 80-conductor cable assembly is installed in the system.
Page 374
T13/1321D revision 3
D.3.2 Pull up and pull down resistors
Pull up and pull down resistors having a value below the specified minimum will increase skew. Use of a higher
resistor value on IORDY (such as 2.0 kΩ or 3.3 kΩ will reduce skew and increase noise margin when IORDY is
negated).
Placement of pull up and pull down resistors on the source side of the series termination minimizes loss of DC
margin due to pull up/pull down current through the series termination resistors.
D.3.3 Cables and connectors
Do not exceed the required 18 inch maximum cable length.
Space device connectors six inches apart on 40- and 80-conductor cable assemblies from twelve to eighteen
inches in length. For cable assemblies shorter than twelve inches, place the connector on the cable for the
device that is not at the end at the center of the cable.
Exceeding a spacing of six inches between device connectors on an 80-conductor cable will cause increased
skew when signaling to or from the device not at the end of the cable. As spacing between the devices
decreases on a 40-conductor cable, the capacitance of the two devices (or the host and the device not at the
end of the cable) act in parallel, resulting in decreased ringing frequency and increased DST. Device connector
spacing closer than six inches with an 80-conductor cable may be done.
In systems using a 40-conductor cable assembly, provide a continuous electrical connection from ground on
the device chassis through the system chassis to the ground plane on the host PCB. Routing the cable in
close contact with the chassis will reduce data settling time, as long as it is done without significantly
increasing the cable length.
D.3.4 PCB and IC design
As has been stated, matching the total output impedance of hosts and devices to the cable impedance is ideal
to minimize reflections and reverse crosstalk due to the impedance mismatch between the PCB and cable.
The impedance of the 80-conductor cable is specified to fall within the range of 70 to 90 Ω and is between 80
and 85 Ω for typical cables with solid wire and PVC insulation.
Keeping the ratio of PCB trace spacing to height above ground plane high helps to control crosstalk between
traces.
Controlling PCB trace characteristics to minimize differences in propagation delay between STROBE and all
DATA lines limits the skew. Factors that affect the delay are:
1) Trace length;
2) Additional capacitance due to stubs, routing on inner layers, pads, and external components such as
pull up resistors and clamping diodes; and
3) Additional inductance due to vias, series components such as termination resistors, and routing across
a break in the ground plane, over areas with no ground plane, or at a larger height above the ground
plane.
In systems using an 80-conductor cable to support Ultra DMA modes 3 and higher, design drivers and series
termination together to provide a stable output impedance matching the cable impedance across switching
conditions and process and temperature variation.
Place series termination resistors as close as possible to the cable header or connector.
Page 375
T13/1321D revision 3
Choose series termination values to equalize input RC delays for the STROBE and DATA lines. For typical
host IC implementations the same type of component I/O is used on all signals and therefore all termination
resistors at both STROBE and DATA may have the same value.
Use sufficient ground and power pins on interface ICs to control supply bounce when many lines are switching
at the same time.
D.3.5 Sender and recipient component I/Os
The 80-conductor cable assembly impedance is less than half that of the typical 40-conductor cable assembly
impedance when multiple lines are switching at the same time. For some types of drivers this will result in
more than double the current draw during switching and as a consequence the amplitude of ground bounce will
also double.
As is required in this standard, design drivers to have a slew rate of 1.25 V/ns or less across the full range of
loading conditions, process, and temperature.
Design component I/Os to produce output setup and hold times at the connector as specified in this standard
across the full range of loading conditions, process, and temperature. Provide margin to allow for skew
introduced between the IC and the connector. Design device PCB traces and component I/Os to present
similar loading between STROBE and DATA at the connector to minimize additional skew added to signaling
between other devices on the bus.
Use hysteresis on both data and STROBE inputs. Initial voltage steps on the bus are at undefined levels and
may be near the thresholds, causing slow slew rates through the threshold that result in high sensitivity to
noise if hysteresis is not used.
Test drivers as well as host and device output characteristics at the connector with the following loading
conditions:
1)
2)
3)
4)
5)
0 pf to ground (open circuit, minimize test fixture capacitance)
15 pf to ground
40 pf to ground
470 Ω to ground, switching low to high
470 Ω to Vcc, switching high to low
All tests (except open circuit) are conducted with the intended series termination resistance in place. Output
skew and slew rates are measured between the series termination and the load.
D.4 Ultra DMA timing assumptions
D.4.1 System delays and skews
−
Source Termination Resistor Delays:
Min rising source transition delay = 0.34 ns
Min falling source transition delay = 0.23 ns
Max falling source transition delay = 2.61 ns
−
Recipient Termination Resistor Delays
Max rising recipient transition delay = 0.12 ns
Max falling recipient transition delay = 0.12 ns
−
transmission skews and delays
Page 376
T13/1321D revision 3
All skews are the STROBE delay minus the data delay. Maximum negative skew is the minimum
STROBE delay minus the maximum data delay for a worst-case system configuration. Maximum positive
skew is the maximum STROBE delay minus the minimum data delay for a worst-case system
configuration. The worst case system configurations were determined through simulation and include all
possible system configurations that meet the requirements of the standard. Included in these values is
skew due to PCB trace length variation, PCB trace impedance variation, recipient component I/O
capacitance variation, sender and recipient series termination variation, pattern variation, and common
mode capacitance variation. Unless otherwise noted, timings are at 1.5 V.
Sender’s component I/O to recipient’s component I/O actual thresholds max negative skew = −5.99 ns
Sender’s component I/O to recipient’s connector max negative skew = −3.98 ns
Sender’s component I/O to recipient’s component I/O actual thresholds max positive skew = 5.38 ns
Sender’s component I/O to recipient’s connector max positive skew = 3.42 ns
Sender’s component I/O to recipient’s component I/O maximum delay = 6.2 ns
For Ultra DMA modes 0, 1, and 2 using a 40-conductor cable, an additional −22 ns is included in the two
max negative skews listed above to account for long data settle time present on that cable due to crosstalk
and ringing. The max positive skew values are not affected by this since the crosstalk and ringing in the
STROBE is not sufficient to increase its settle time.
D.4.2 IC and PCB timings, delays, and skews
It is recommended that the timings shown in this clause be met but they are only an example of timings that
result in a system that meets all requirements for Ultra DMA specified in the body of the standard. A system
that does not meet one or more of the timings below may be able to meet all timing requirements by producing
other timings tighter than shown below. It is advisable that a designer take all the timings that are achieved for
that design and re-derive the worst case timings to determine if all are still met.
−
Possible clocks for bus timing and their characteristics
All frequencies are assumed to have 60 / 40 % asymmetry (worst case)
25 MHz (supports modes 0 and 1)
Typical Period = 40 ns
Clock variation = 1 %
33 or 30 MHz clocks (supports modes 0, 1, and 2)
Typical Period = 30 or 33.3 ns
Clock variation = 1 %
50 MHz (supports modes 0, 1, 2, and 3)
Typical Period = 20 ns
Clock variation = 3.5 %
66 MHz (supports modes 0, 1, 2, 3, and 4)
Typical Period = 15 ns
Clock variation = 3.5 %
−
PCB Traces
Max PCB trace skew = 0.1 ns
Max PCB trace delay = 2.1 ns
Page 377
T13/1321D revision 3
−
IC inputs
Input delay includes bond wire, buffer, routing, and logic component delays between the component I/O and
the flip-flop that first latches the data. Input delay is measured from 1.5 V and includes the delay between
1.5 V and the input's threshold.
Input skew is either positive or negative depending on the directions of the STROBE and data transitions. It
is the difference in STROBE signal delay from the input switching threshold to the internal flip-flop that first
latches data and data delay from the input switching threshold to the same flip-flop. The routing component
of skew that accounts for about 30 % of the value listed here is systematic (i.e., always the same polarity
in a system implementation) and could be either positive or negative.
Min input slew rate for testing = 0.4 V/ns
Max input delay = 5.5 ns
Max input skew = 2.45 ns
Max input skew from 1.5 V to actual thresholds = 1.75 ns
−
IC outputs
Output delay is from the internal active clock edge that generates an output transition until the time that the
transition crosses 1.5 V the associated component I/O of the IC.
Max output disable delay is from the internal enable negation of an I/O output until the time that the signal
is released at the component I/O.
Single component I/O output skew is the difference in delay of rising and falling edges on a single output.
This single component I/O skew does include skew due to noise that may be present on the signal in a
functional system. It may be positive or negative depending on the direction of the STROBE and data
transitions.
Output skew is the difference in the output delay of the active STROBE and the output delay of any data
transition that occurs within cycle time before or after the STROBE transition. This timing is met under all
expected loading conditions and starting voltages. This timing is the combination of:
−
−
−
−
−
single component I/O output skew,
skew due to output routing differences between all data and STROBE signals,
skew due to process, temperature, and voltage variation between all data and STROBE signals at the
moments when transitions are generated,
skew due to clock routing to all data and STROBE logic that generates output transitions, and
skew due to supply bounce differences that may occur between the transitions being compared.
As with the single component I/O output skew, this skew may be positive or negative depending on the
direction of the STROBE and data transitions. Some of the components of this skew (e.g., differences in
routing) may be systematic and could be either positive or negative.
Max output delay = 14 ns
Max output disable delay = 10 ns
Max single component I/O output skew = 2.5 ns
Max output skew = 5.4 ns
Max output skew to support modes 0 and 1 with a 25 MHz clock = 5.0 ns
Max output skew to support modes 0, 1, and 2 with a 50 MHz clock = 5.2 ns
Max output skew to support mode 4 with a 30 or 33 MHz clock = 2.8 ns
Up to 3 ns of additional output delay may be needed for data compared to STROBE in cases that use 30
and 33 MHz clocks to support Ultra DMA modes 0, 1, and 2. With these clocks, the data is held by a half
Page 378
T13/1321D revision 3
cycle, and a minimum half cycle is not sufficient to meet the output hold time given the output skews listed
above. An additional delay on data would insure that the required hold time is met even with a short half
clock cycle. Alternatively, improvements in output skew beyond those listed above could also allow the
output hold time to be met with a short half clock cycle.
−
IC flip-flops
The setup and hold times listed here are intended to represent only the flip-flop inside an IC that latches
data. Timing is assumed from the inputs of the flip-flop.
Min flip-flop setup time = 0.5 ns
Min flip-flop hold time = 0.5 ns
D.5 Ultra DMA timing parameters
System timings for Ultra DMA are measured at the connector of the sender or receiver to which the parameter
applies . Internally the IC accounts for input and output delays and skews associated with all signals getting
from the connector to the internal flip-flop of the IC and from the flip-flop of the IC to the connector.
Timings as listed in the body of the specification were derived using the formulas listed below and the timing
assumptions give above in D.4. All applicable clocks were evaluated for each timing parameter and the worst
case value was used in the body of the standard. It is recommended that the system designer re-derive all
timings based on the specific characteristics of the internal clock, IC, and PCB that are to be used to confirm
that timing requirements are met by that implementation.
D.5.1 Typical average two-cycle time (t2CYCTYP)
This is the typical sustained average time of STROBE for the given transfer rate from rising edge to rising edge
or falling edge to falling edge measured at the recipient’s connector.
D.5.2 Cycle time (tCYC )
This is the time allowed for STROBE from rising edge to falling edge or falling edge to rising edge measured at
the recipient’s connector. This timing accounts for STROBE and internal clock variation. The formula for the
minimum value is:
+ (Number of clock cycles to meet minimum typical cycle time with a minimum cycle time due to
clock variation) ∗ (clock cycle time)
– Max single component I/O output skew
D.5.3 Two-cycle time (t2CYC )
This is the time for STROBE for the given transfer rate from rising edge to rising edge or falling edge to falling
edge measured at the recipient’s connector. Since this timing is measured from falling edge to falling edge or
rising edge to rising edge of STROBE, asymmetry in rise and fall time has no affect on the timing. Clock
variation is the only significant contributor to t2CYC variation. The formula for the minimum values is:
+ (2 ∗ (Number of clock cycles to meet minimum typical cycle time with a minimum cycle time due to
clock variation percent) ∗ (clock cycle time)
D.5.4 Data setup time (tDS)
This is the data setup time at the recipient. Since timings are measured at the connector and not at the
component I/O, consider the effect of the termination resistors and traces when generating this number.
Depending on the direction of the data signal and STROBE transitions, the skew between the two changes in
Page 379
T13/1321D revision 3
both the positive and negative directions. A longer data signal delay will reduce the setup time, and a longer
STROBE delay will increase the setup time.
In order to meet the input skews given above in D.4.2, minimize the number of buffers or amount of logic
between the incoming signals and the input latch or flip-flop. This may require the data input buffers to be
routed directly to the input latch with no delay elements and the STROBE signal to be routed directly from its
input buffer to the input latch clock with no delay elements.
The internal latch or flip-flop has a non-zero setup and hold time. tDS is sufficient to insure that the setup time of
the flip-flop is met. The minimum setup required at the threshold of the component I/O is:
+ Max input skew
+ Min flip-flop setup time
The formula for the value at the recipient’s component I/O based on the timings given in D.4 is:
+ (Number of clock cycles to meet typical cycle time with a minimum cycle time due to clock variation)
∗ (clock cycle time)
− (Number of clock cycles used to hold data with a minimum cycle time due to clock variation or with a
minimum cycle symmetry if a half cycle is used) ∗ (clock cycle time)
− Max output skew
+ Sender’s component I/O to recipient’s component I/O actual thresholds max negative skew
In order to meet both setup and hold times over process, temperature, and voltage, clock edges rather than
gate delays are used to generate the hold time. The assumption is made that one 50 or 66.7 MHz clock cycle
or half of a 33 MHz or slower clock cycle has been used to hold data within the sender IC.
After it is shown that the sender is producing setup time that meets the requirement of the recipient, the
specification for setup time at the recipient connector produced by the sender is determined as follows. The tDS
values in the specification were based on the results of the following formula using all possible clocks for the
modes they support.
+ (Number of clock cycles to meet typical cycle time with a minimum cycle time due to clock variation)
∗ (clock cycle time)
− (Number of clock cycles used to hold data with a minimum cycle time due to clock variation or with a
minimum cycle symmetry if a half cycle is used) ∗ (clock cycle time)
− Max output skew
+ Sender’s component I/O to recipient connector max negative skew
D.5.5 Data hold time (tDH)
This is the data hold time at the recipient. This time is sufficient to insure that the hold time of the internal flipflop is met. The longest STROBE delay and shortest data delay is the worst case for hold time. The analysis
is similar to the one for tDS above. The minimum hold required at the component I/O at its threshold is:
+ Maximum input skew
+ Minimum flip-flop hold time
The formula for the value at the recipient’s component I/O based on the timings given in 0 is:
+ (Number of clock cycles used to hold data with a minimum cycle time due to clock variation or with a
minimum cycle symmetry if a half cycle is used) ∗ (clock cycle time)
− Max output skew
− Sender’s component I/O to recipient’s component I/O actual thresholds max positive skew
Page 380
T13/1321D revision 3
After it is shown that the sender is producing hold time that meets the requirement of the recipient, the
specification for hold time at the recipient connector produced by the sender is determined as follows. The tDH
values in the specification were based on the results of the following formula using all possible clocks for the
modes they support.
− (Number of clock cycles used to hold data with a minimum cycle time due to clock variation or with a
minimum cycle symmetry if a half cycle is used) ∗ (clock cycle time)
− Max output skew
− Sender’s component I/O to recipient connector max positive skew
D.5.6 Data valid setup time (tDVS)
This is the data valid setup time measured at the sender’s connector. This timing is measured using a test
load with no cable or recipient. This is the timing that, if met by the sender, will insure that the data setup time
is met at the recipient. It is important that this timing be met using capacitive loads from 15 to 40 pf to insure
reliable operation for any system configuration that meets specification.
In the case of Ultra DMA modes 0, 1, and 2, the data settle time may be long due to crosstalk in the cable and
on the PCB, and the ringing frequency of the system. For modes above 2, there is little or no margin for ringing
on the cable. For these modes, the 80-conductor cable assembly that reduces the crosstalk between signals
is required so that crosstalk and ringing are reduced to a level that does not cross the input switching
thresholds during data setup or hold times. Modes 3 and 4 timing requirements were derived so that they are
met with the same input and output timing characteristics as a system supporting Ultra DMA mode 2. Since it
may be shown using the formulas presented in D.5.4 that sufficient setup time is produced with the system
timings given in D.4, using those same timings in the formula below will produce tDVS values that also represent
sufficient timing for the system. An achievable value for tDVS is calculated as follows:
+ (Number of clock cycles to meet minimum typical cycle time at the minimum cycle time due to
clock variation) ∗ (clock cycle time)
− (Number of clock cycles used to hold data at the minimum cycle time due to clock variation or at the
minimum cycle symmetry if a half cycle is used) ∗ (clock cycle time)
− Max output skew
− Max PCB trace skew
− Max falling source transition delay
+ Min rising source transition delay
D.5.7 Data hold time (tDVH )
This is the data valid hold time measured at the sender’s connector. This timing is measured using a test load
with no cable or recipient. This is the timing that, if met by the sender, will insure that data hold time at the
recipient is met. It is important that this timing be met using capacitive loads from 15 to 40 pf to insure reliable
operation for any system configuration that meets specification.
Since it may be shown using the formulas presented in D.5.5 that sufficient hold time is produced with the
system timings given in D.4, using those same timings in the formula below will produce tDVS values that also
represent sufficient timing for the system. An achievable value for tDVH is calculated as follows:
+ (Number of clock cycles used to hold data at the minimum cycle time due to clock variation or at the
minimum cycle symmetry if a half cycle is used) ∗ (clock cycle time)
− Max output skew
− Max PCB trace skew
− Max falling source transition delay
+ Min rising source transition delay
Page 381
T13/1321D revision 3
D.5.8 First DSTROBE time (tFS)
This is the time for the device to first negate DSTROBE to clock the first word of data after the device has
detected that the host has negated STOP and asserted HDMARDY− at the beginning of a data in burst. This
parameter is measured from the time that STOP and HDMARDY− are in the appropriate states at the device
connector until the first falling DSTROBE edge at the device connector.
This timing is used only for the beginning of a read command from the STOP negation and/or HDMARDY–
assertion to first DSTROBE (all falling edges). The device detects that these two control signals from the host
have changed. Timing is started from the point that both signals have changed to the appropriate state.
Synchronization may be done with two flip-flops. After synchronization is achieved, data is driven on to the bus
and internal clock cycles counted off to meet the minimum setup time before generating the first STROBE
edge. In order for an IC based on a 25, 30, or 33 MHz clock to meet tFS, data needs to be driven onto the bus
no later than 2.5 clock cycles after the control signal transitions. This could be done by synchronizing with
both edges of the system clock or by using only one edge to synchronize and then driving data onto the bus on
the next inactive edge of the clock after the signals are detected at the output of the second synchronization
flip-flop. With a 50 MHz clock, the first word of data needs to be driven out no later than three cycles after the
control transitions and with a 66 MHz clock, it may be four cycles. The formula for the maximum tFS timing is
as follows:
+
+
+
+
+
Max falling recipient transition delay
Max PCB trace delay
Max input delay
Min flip-flop setup time
The time for two, three, or four clock cycles at the maximum period due to frequency variation to
synchronize the control signals and start the data transfer cycle. For 25, 30 and 33 MHz based
systems, the data would be driven out one half cycle after the incoming signal is synchronized
since data is held one half cycle when using these clock frequencies and therefore sent on a half
cycle.
+ The time for as many cycles as required to meet the tDVS minimum timing for the first word of data at
the maximum period due to frequency variation.
+ Max output buffer delay
+ Max PCB trace delay
+ Max falling source transition delay
D.5.9 Limited interlock time (tLI)
The time is for limited interlock from sender to recipient or recipient to sender. This is the interlock time in this
protocol that has a specified maximum.
The value of tLI needs to be large enough to give a recipient of the signal enough time to respond to an input
signal from the sender of the signal. The derivation of tLI is similar to that of tFS since both involve the recipient
of the signal responding to the control signal of the sender of the signal. As with tFS, the number of internal
clock cycles that an IC may take to respond is dependent on the frequency of the clock being used. For a 25,
30 MHz clock, the maximum time to respond is three cycles, for 33 MHz clock it is four, for a 50 MHz clock it
is five, and for a 66 MHz clock it is seven cycles maximum for modes 0 through 2. Modes 3 and 4 require a
faster response time. For a 30 or 33MHz clock it is two cycles, for a 50 MHz clock it is three cycles and for a
66 MHz clock it is four clock cycles maximum. The formula for the values of tLI is as follows:
+
+
+
+
+
Max falling recipient transition delay or max rising recipient transition delay
Max PCB trace delay
Max input delay
Min flip-flop setup time
The time for two, three, four, five, or seven clock periods (depending on clock used and modes
supported) at the maximum period due to frequency variation to synchronize the signals to the
internal clock and respond appropriately.
Page 382
T13/1321D revision 3
+ Max output buffer delay
+ Max PCB trace delay
+ Max falling source transition delay
D.5.10 Limited interlock time with minimum (tMLI)
The time is for the minimum limited interlock from sender to recipient.
This timing insures that some control signals are in their proper state before DMACK– is negated. It is
important that STROBE and the control lines are in their proper states because all signals revert to their nonUltra DMA definitions at the negation of DMACK–. If the signals are not in their proper state, the selected
device or another device may see a false read or write STROBE or data request. All control signals need to be
in their proper state and detectable at the device before DMACK– is negated so tMLI has to overcome the
following:
+ Sender’s component I/O to recipient’s component I/O maximum delay
+ Max input delay
+ Min flip-flop setup time
The value calculated by the formula above for tMLI for all modes is under 14 ns. The specified value for this
timing allows for additional margin.
D.5.11 Unlimited interlock time (tUI)
This interlock timing is measured from an action of a device to a reaction by the host. In order to allow the host
to indefinitely delay the start of a read or write transfer, this value has no maximum. The reason for this
parameter is to ensure that one event occurs before another, for this reason the minimum is set to 0. In
practice the host will take some non-zero positive time to respond to the incoming signal from the device.
D.5.12 Maximum driver release time (tAZ)
This is the maximum time that an output driver has to make the transition from being asserted or negated to
being released. During data bus direction turn around, the driver of the bus is required to release the data. For
the beginning of a read burst, the host releases the data bus before or on the same internal clock cycle that it
asserts DMACK–. For the end of a read burst, the device releases the data bus before or on the same clock
cycle that it negates DMARQ. If the same clock is used, the maximum delay is calculated using the following
formula:
+ Max output skew
− Min falling source transition delay
The value calculated by the formula above for tAZ for all modes is under 6 ns. The specified value for this timing
allows for additional margin.
D.5.13 Minimum delay time (tZAH )
This is the minimum time that the host waits after the negation of DMARQ at the termination of a data in
transfer to begin driving data onto the bus for purpose of transferring the CRC word to the device. In this case
the device is allowed to continue driving the bus for a maximum of tAZ after the DMARQ negation. The host is
required to wait tZAH after the DMARQ negation to drive the data. Skew on the cable is the major factor to
consider here and a longer data delay than DMARQ delay (i.e., max negative skew) is the worst case. For
modes using a 40-conductor cable, the component of maximum negative skew associated with data settle time
as listed in D.4.1 should not be included since the bus is being released for this timing. To avoid bus
contention, this value is calculated using the following formula:
+ Max specified tAZ
Page 383
T13/1321D revision 3
– Sender’s component I/O to recipient’s component I/O actual thresholds max negative skew
The value calculated by the formula above for tZAH is under 17 ns in all cases. The specified value for this timing
allows for additional margin.
D.5.14 Minimum driver assert/negate time (tZAD )
This is the minimum time after STOP is negated or HDMARDY− is asserted (whichever comes later) that a
device drives the data bus at the initiation of a read operation. This is when the data bus is changed from host
driving or released to device driving.
The use of STOP negated and HDMARDY− asserted guarantees that a system failure has not occurred leaving
the host in a Multiword DMA mode and the device in an Ultra DMA mode. STOP is the same signal line as
DIOW–, and HDMARDY– is the same signal line as DIOR–. Multiword DMA mode never asserts both DIOW–
and DIOR– at the same time. The negation of STOP and assertion of HDMARDY– is equivalent to both DIOW–
and DIOR– being asserted. Since the device requires both signals to be in this state before driving the bus, it
insures that the host is in Ultra DMA mode, not Multiword DMA, and has released the data bus.
The STOP negation and HDMARDY− assertion are required by the standard to meet tENV timing. The tENV timing
is a minimum of 20 ns from the point where the host releases the bus, no additional delay is necessary based
on the tZAH evaluation that is applicable to the conditions of this timing also.
Even though tZAD is 0 ns minimum for all modes, in practice, most devices will take two flip-flop delays to
synchronize the incoming STOP and HDMARDY− transitions making tZAD time dependant on the clock
frequency used by the device. Since the data is driven long enough before the first STROBE to meet the setup
time requirement, this synchronization time has been taken into account in the tFS derivation of D.5.8.
D.5.15 Envelope time (tENV)
This time is from which the host asserts DMACK– until it negates STOP and asserts HDMARDY– at the
beginning of a data in burst, and the time from which the host asserts DMACK– until it negates STOP at the
beginning of a data out burst. Since tENV only applies to outputs from the host, the timings are synchronous with
the host clock. Based on an argument similar to the one for tMLI in D.5.10, the minimum for tENV is 20 ns. This
insures that all control signals at all the devices are in their proper (non-Ultra DMA mode) states before
DMACK– is asserted and are sensed as changing only after DMACK– has been asserted. The 20 ns accounts
for cable and gate skew between DMACK– and the control signals on device inputs. Since tENV involves
synchronous events only and an increase in tENV reduces the performance of the specification, a maximum is
specified.
Enough internal clock cycles are used between the assertion of DMACK– and the other control signals to
insure tENV minimum is met. For a 25, 30, or 33 MHz clock this is a single cycle, for 50 or 66 MHz clocks this
is two cycles. The following formula is used to verify that the minimum tENV value of 20 ns is met by any
particular system implementation:
+ (One or two host clock cycles (depending on frequency used) at the minimum period due to
frequency variation to delay control signals inside the IC) ∗ (clock cycle time)
– Max output skew
– PCB trace skew
– Max falling source transition delay
+ Min falling source transition delay
Using the number of clock cycles specified above for each possible frequency, the minimum is met. Based on
the number of clock cycles needed to meet the minimum, reasonable maximums for tENV are determined.
Rather than limiting the possible cycles to generate tENV, the following assumption was made: for a 25 or 30
MHz clock a single cycle is used; for a 33 or 50 MHz clock a maximum of two cycles is used; and with a 66
Page 384
T13/1321D revision 3
MHz clock a maximum of three clock cycles is used. Using these numbers of cycles, the formula to determine
the maximum tENV is as follows:
+ (One, two, or three cycles (depending on frequency used) at the maximum period due to frequency
variation to delay control signals inside the IC) ∗ (clock cycle time)
+ Max output skew
+ PCB trace skew
+ Max falling source transition delay
− Min falling source transition delay
It may be possible that fewer or more clock cycles are used with some frequencies given reduced output skew.
If the timing characteristics of D.4 are just met, the following number of clock cycles for the internal IC delay to
meet tENV minimum and maximum values are used.
1)
2)
3)
4)
5)
with 25 MHz, delay is one cycle
with 30 MHz, delay is one cycle
with 33 MHz, delay is one or two cycles
with 50 MHz, delay is two cycles
with 66 MHz, delay is two or three cycles
D.5.16 STROBE to DMARDY– time (tSR )
If DMARDY– is negated before this maximum time after a STROBE edge, then the recipient will not receive
more than one additional STROBE (i.e., one more word of valid data). This timing is applicable only to modes
0, 1, and 2 because the transfer rate of modes 3 and 4 is too high to insure that only one additional STROBE
will be sent after DMARDY– is negated.
D.5.17 DMARDY– to final STROBE time (tRFS)
This is the maximum time after DMARDY– is negated after which the sender will not transmit any more
STROBE edges (i.e., no additional valid data words). This timing gives the sender time to detect the negation
of DMARDY– and respond by not sending any more STROBES. The tRFS time may affect the number of words
transferred.
Since tRFS involves a response to a request for a pause, the sender needs to stop sending data as soon as
practical. An example of an input synchronization method is to use two flip-flops where the first is clocked on
the active edge of the internal clock and the second on the unused (inactive) edge of the clock. The action to
stop the STROBE signal would be taken on the next active clock edge (i.e., if there had been a STROBE
scheduled for that edge it would not be sent). In this example a half cycle of the clock gives adequate time to
avoid metastability while synchronizing the signal. The following timing diagram shows one possible case:
tRFS range
Next STROBE would have
been here
Clock
STROBE
DMARDYFF1
FF2
Figure D.30 – DMARDY- to final STROBE tRFS syncronization
Page 385
T13/1321D revision 3
The diagram above shows the range of possible STROBE to DMARDY– transition relationships and the
possible synchronization flip-flop responses. When a 66 MHz or higher clock frequency is used, two clock
periods may be used to synchronize the data as long as no STROBE edge is sent on the subsequent clock
edges until the transfer is resumed.
The tRFS time may be the longest when the DMARDY– transition occurs before an internal clock cycle, but, due
to skews and missed setup time, the transition is not clocked into the first flip-flop until the next clock (the
dotted line transition on FF1 and later on FF2). When this happens one clock cycle before a STROBE
transition is generated (as shown by the left tRFS range marker near the middle of the DMARDY– transition
range in the diagram above), the next STROBE transition will occur (as shown in dotted lines). For all other
cases, the tRFS time will be shorter. The maximum tRFS is calculated using the following formula:
+
+
+
+
+
Max rising recipient transition delay
Max PCB trace delay
Max input delay
Min flip-flop setup time
(One or two clock cycles at the maximum system clock period due to frequency variation for
synchronization) ∗ (clock cycle time)
+ Max output delay
+ Max PCB trace delay
+ Max falling source transition delay
D.5.18 DMARDY– to pause time (tRP)
This is the minimum time after DMARDY– is negated after which the recipient may assert STOP or negate
DMARQ–. After this time the recipient will not receive any more STROBE edges (i.e., no additional valid data
words). STROBE edges may arrive at the recipient until this time. Since this time parameter applies to the
recipient only (as the recipient waits for STROBEs), the parameter is measured at the recipient connector.
Because of this, the output delay of DMARDY– from inside the IC to the connector and the input delay of a
STROBE edge from the connector to the associated internal IC flip-flop are considered.
There are two ways to determine the tRP minimum. One method is to consider how long it will take from the
negation of DMARDY− at the recipient for the sender to see the negation and become paused. This would
involve synchronizing DMARDY– as it is done for tRFS, and then taking one more system clock cycle to change
the state of the state machine to a paused state. Using this method, the minimum time is calculated using the
following formula:
+
+
+
+
Sender’s component I/O to recipient’s component I/O maximum delay
Max input delay
Min flip-flop setup time
(Two or three clock cycles (depending on clock used) at the maximum period due to clock frequency
variation) ∗ (clock cycle time)
A second method to calculate this value is to consider how long it might be for the last STROBE to be detected
after negating DMARDY–, and make sure tRP is long enough so that the internal assertion of STOP occurs after
the last STROBE has latched the last word of data. This method is applied in the following formula:
+
+
+
+
+
Sender’s component I/O to recipient’s component I/O maximum delay
Maximum tRFS for mode
Sender’s component I/O to recipient’s component I/O maximum delay
Max input delay
Min flip-flop setup time
Using both of the above, it may be shown that tRP is met given the tRFS requirement and is sufficient to receive
the last STROBE for all modes with all clock frequencies. All of the numbers are measured at the connector,
Page 386
T13/1321D revision 3
and the time to wait internal to the IC will be longer than the value of tRP. For higher frequency clocks, the
internal delay may need to be more than one clock cycle longer than the value of tRP in order to account for total
output and input delays.
D.5.19 Maximum IORDY release time (tIORDYZ)
This is the maximum time allowed for the device to release IORDY:DDMARDY–:DSTROBE at the end of a
burst. The tIORDYZ time allows IORDY to be asserted immediately after DMACK– is asserted. DMACK– being
asserted may be used to enable the IORDY output. As soon as the DMACK– is negated, the component I/O
cell will be released. For this implementation, the following formula determines the maximum tIORDYZ:
+
+
+
+
+
Max falling recipient transition delay
Max PCB trace delay
Max in delay (in this case to enable IORDY)
Max output disable delay
Max trace delay
D.5.20 Minimum IORDY assert time (tZIORDY)
This is the minimum time allowed for the device to assert IORDY:DDMARDY–:DSTROBE when the host
asserts DMACK– at the beginning of a burst.
When STOP is negated and HDMARDY– is asserted, it is important that the IORDY:DDMARDY–:DSTROBE
signal be electrically high (DSTROBE asserted or DDMARDY– negated). This could be achieved by the
IORDY:DDMARDY–:DSTROBE signal being driven by the device but it also occurs when this signal is released
by the device because of the pull up at the host required by the standard. Since the correct state of
IORDY:DDMARDY–:DSTROBE occurs when it is released, no maximum tZIORDY is required. As with some
other timings having no maximum defined, the signal will eventually change state as governed by other timing
parameters.
For Ultra DMA, DDMARDY–:DSTROBE is only driven during a data burst. At the initiation of a data in burst,
the device may wait until the time to generate the first DSTROBE and enable DSTROBE in a negated state.
The device may wait tZIORDY then assert DSTROBE and, for the first data transfer, the device would negate
DSTROBE. In both cases the host sees a negation for the first DSTROBE. The first STROBE of a burst is
never a low-to-high transition. At the initiation of a data out burst, the device waits until ready before asserting
DDMARDY–. If the device does not use this implementation, it waits tZIORDY then negates DDMARDY– (i.e.,
drive it electrically high). Then, to signal that the device is ready to receive data, the device may negate
DDMARDY–. Both implementations are equivalent since the negated state of this signal will appear the same
to the host as the released state.
Since this timing was defined for the sole purpose of requiring DMARDY– to be asserted before IORDY is
driven, the minimum value for this timing in all modes is 0 ns.
Page 387
T13/1321D revision 3
D.5.21 Setup and hold before DMACK– time (tACK)
The tACK value is defined for the setup and hold times before assertion and after negation of DMACK–. It is
applied to all control signals generated by the host related to an Ultra DMA burst. These signals are STOP,
HDMARDY–, HSTROBE and the address lines. The burst begins with the assertion of DMACK– and ends with
the negation of DMACK–. For this burst period, all control signals start, remain, and end in specific states as
defined by the protocol. Since there may be some skew between signals from the host to the device due to
transmission and component I/O circuitry affects, the host is required to set up all the control signals before
asserting DMACK–. This insures that by the time all the signals reach the device, they will all be in the proper
state when DMACK– is asserted. Using tACK as the hold time for the signals after the negation of DMACK–
insures that at the termination of the burst, the control signals as seen by the device are in the proper states.
This avoids any device state machine confusion. Based on timing analysis (the same as the analysis used for
tMLI in D.5.10), the minimum for tACK is 20 ns.
D.5.22 STROBE to DMARQ/STOP time (tSS)
This is the minimum time after a STROBE edge before a device as a sender negates DMARQ or a host as a
sender asserts STOP to terminate a transfer. This time is to allow at least one recipient clock cycle between
the last STROBE and the termination signal to avoid the possibility of a race condition between the two events
and ensure the last word is seen as valid by the recipient. The formula used to determine tSS minimum is:
+ Sender’s component I/O to recipient’s component I/O actual thresholds max positive skew
+ Max input skew
+ (One recipient clock cycle at the maximum period due to frequency variation) ∗ (clock cycle time)
For modes 0 and 1, a 25 MHz recipient clock is assumed and for all other modes a 30 MHz recipient clock is
assumed. While the value specified could have been lower for modes using 30 MHz or higher clock
frequencies, tSS is specified to be the same value for all modes for extra margin.
D.6 Ultra DMA Protocol Considerations
D.6.1 Recipient pauses
Ultra DMA protocol allows the recipient to pause a burst at any point in the transfer. The clauses below discuss
some of the issues and design considerations associated with the Ultra DMA recipient pausing protocol.
D.6.1.1 DMARDY– minimum negation time
An Ultra DMA recipient pause is initiated through the recipient's negation of DMARDY–. Once DMARDY– is
negated, the protocol allows for additional words to be transferred. Pausing is typically done for two reasons.
One is that the recipient’s input FIFO or buffer is almost full and would overflow if the burst continued. The
second is that the recipient is preparing to terminate the burst. Normally the case of pausing to free space in
the FIFO or buffer would lead to DMARDY− being negated for at least a few transfer cycles. However, there is
no minimum time for the negation of DMARDY−. The recipient does not have to wait for possible additional
words or for any minimum time from when the recipient negates DMARDY− until it re-asserts DMARDY−. If,
after negating DMARDY−, the recipient becomes ready, it may immediately reassert DMARDY−. Based on the
implementation of the sender, a negation and immediate re-assertion of DMARDY− may cause a subsequent
STROBE to be delayed. It is recommended that some hysteresis be used in the FIFO trigger points for
assertion and negation of DMARDY− to avoid DMARDY− being negated after every word or two.
D.6.1.2 Number of additional words from sender
An Ultra DMA burst may be paused with zero, one, or two additional data transfers as seen at the recipient
connector for modes 0, 1 and 2, and up to three additional transfers for modes 3 and 4. This does not imply
that the sender is allowed to send up to two or three more STROBES after it detects the negation of DMARDY–
. In most cases it would be a violation of tRFS to do so. Rather than counting words after detecting the negation
Page 388
T13/1321D revision 3
of DMARDY−, under all conditions the sender stops generating STROBE edges within tRFS of the recipient
negating DMARDY–. Even in cases where tRFS is met and less than the maximum number of words are sent, it
is still possible for the recipient to see the maximum number of STROBE edges after it negates DMARDY–.
This is due to the delay of the signals through the cable. An example of this is explained below and shown in
figure D.31.
In mode 2 when the STROBE time is 60 ns and signal delays add up to 6 ns, both STROBE from sender to
recipient and DMARDY– from recipient to sender experience a cable delay of 6 ns. While the recipient negates
DMARDY– after the sender toggles STROBE, it does not receive the STROBE transition until after the
DMARDY– negation. This would account for the first word received. By the time the sender detects the
DMARDY– negation, there are only 49 ns until the next STROBE. This STROBE is within tRFS so the sender
may send STROBE without violating the protocol. To the recipient, this would be the second transfer after it
has negated DMARDY–, but to the sender it would be the first and only allowable STROBE transition after
detecting the DMARDY– negation.
60 ns
STROBE @ sender
49 ns
DMARDY- @ sender
6 ns
STROBE @ recipient
6 ns
DMARDY- @ recipient
5 ns
Figure D.31 – STROBE and DMARDY- at sender and recipient
D.6.1.3 Sender output data handling during
In most cases of a recipient pause, a sender stops toggling STROBE in less than one transfer cycle time after
DMARDY− negates at its input in order to meet tRFS. Since the incoming DMARDY− negation is asynchronous
with the sender's internal clock, following synchronous logic design rules the incoming DMARDY− signal should
be synchronized with the internal clock. In this condition data may be gated or latched to the bus but never
strobed.
If an output register is used when data is transferred from memory for presentation on the bus, no assumptions
are made that that data has been or will be transferred. If a pointer in memory is incremented or the data is
cleared from memory when it is sent to the output register, data may be lost unless some recovery mechanism
is present to decrement the pointer or restore the data if it is never strobed due to a burst termination after a
pause. During a pause, other bus activity (like a Status register read) might occur. A design using an output
register would have data in that register overwritten during this other activity. Other designs may involve similar
considerations. It is most important to remember that data on the bus is not sent and is not to be treated as
sent until there is a valid STROBE edge.
D.6.1.4 Additional words at recipient
After DMARDY- is negated, the recipient may receive additional words. There will be some output delay of
DMARDY– from the logic that first generates it inside the IC to the connector, and there will be input delay of
STROBE from the connector to inside the IC. In addition to this, data may be pipelined before the FIFO and
there may be logic delays between triggering a “near full” condition in the FIFO and generating the DMARDY−
Page 389
T13/1321D revision 3
negation. The depth of the recipient's input FIFO where it triggers a condition to negate DMARDY− to avoid an
overflow is therefore dependent on the particular design approach. Consider all FIFO near full trigger threshold to
DMARDY− negation delays, the cable delay, tRFS time, input delays, input data pipelining, and the minimum
cycle time for the mode supported when determining the FIFO trigger point.
The recipient may receive STROBE edges until tRP after it negates DMARDY–. The receipt of two or three
words by the recipient after a pause has been initiated is not an indication that the sender has paused. The
recipient waits until tRP after the pause was initiated before taking any other action (e.g., terminating the burst).
Waiting tRP allows for cable delays between the recipient and sender and allows the sender time to complete its
process of transitioning to a paused state. The process of switching to a paused state may take additional
system clocks after the sender has sent it’s last STROBE transition.
Since the recipient’s and sender’s clocks are asynchronous with respect to each other, there is not a single
fixed number of words that the recipient will receive after negating DMARDY–. Every time a recipient begins a
pause, a sender may send from zero to the maximum number of words allowed by the protocol . The Ultra
DMA protocol does not give the recipient any means of pausing or stopping on an exact, predetermined
boundary.
D.6.2 CRC calculation and comparison
For each STROBE transition used for data transfer, both the host and device calculate a new CRC value. Only
words successfully transferred in the transfer phase of the burst are used to calculate CRC. This includes
words transferred after a pause has been requested. Words put on the bus but never strobed are not to be
used for CRC calculation. In addition, if STROBE is negated at the end of a pause and then the burst is
terminated, the protocol requires STROBE to be re-asserted after DMARQ is negated or STOP is asserted. No
data is transferred on this STROBE edge and any data on the bus that was not strobed during the transfer
phase of the burst is not used in the CRC calculation on this re-assertion of STROBE.
It is not advisable to use STROBE to clock the CRC generator. Noise on the STROBE signal could cause the
recipient's CRC generator to see a glitch and double clock the generator on a single edge. At the same time,
the glitch seen by the CRC generator may not affect the data input portion of the logic. This type of
implementation has lead to CRC errors on systems where data is properly received but the wrong CRC value is
determined. Using different versions of STROBE to clock the CRC generator and to clock data into the FIFO or
buffer may also lead to a fatal error. Noise, lack of setup or hold time, race conditions in the logic, or other
problems could result in the wrong data being clocked into the FIFO or buffer. At the same time the correct
data may be clocked into the CRC generator since it is using a different instance of STROBE. In this case, the
resultant CRC value is correct when the data in the recipient is not. This fatal error has been seen on an
implementation of the Ultra DMA protocol.
Most designs will internally generate a delayed version of STROBE that is synchronous with the recipient
clock. This synchronized version of the STROBE is then used to place data into the FIFO or buffer. It is
advisable for the recipient to use the same clock that places data into its FIFO or buffer to clock data into its
CRC generator. Following this design approach will maximize the probability of clocking the same data into
both the CRC generator and FIFO or buffer and clocking both the same number of times.
This standard includes the equations that define the XOR manipulations to make on each bit and the structure
required to perform this calculation using a clock generated from STROBE. Through the given equations, the
correct CRC is calculated by using a small number of XOR gates, a single 16-bit latch, and a word clock (one
clock per STROBE edge). The equations define the value and order of each bit, and the order of each bit is
mapped to the same order lines of the bus. The CRC register is pre-set to 4ABAh. This requires pre-setting
the latch (CRCOUT) to 4ABAh before the first word clock occurs. After that, CRCIN15 to the latch is tied
through to CRCOUT15. When the burst is terminated CRCOUT15 is the final CRC bit 15 that is sent or
received on DD15. This direct matching of bit order is true for all CRC bits. The proper use of the data sent on
the bus bits DD0 through DD15 during the burst transfer is defined in the equations. The DD15 on the bus has
the same value as bit DD15 in the equations to calculate CRC. This direct mapping is true for all bits strobed
on the bus during a burst.
Page 390
T13/1321D revision 3
Once the burst is terminated and the host sends the CRC data to the device (the host always sends the CRC
independent of whether the burst was a data in or data out transfer), the device compares this to the CRC it has
calculated. While other CRC validation implementations may be possible, a CRC input register may be used
on the device in combination with a digital comparitor to verify that the CRC value in the input register matches
the value in its own CRC calculation register.
D.6.3 The IDENTIFY DEVICE command
A device communicates its Ultra DMA capabilities and current settings to the host in the data returned by the
device as a result of an IDENTIFY DEVICE command.
For the PIO and Multiword DMA protocols, only the host generates data STROBES so the minimum cycle
times reported for those protocols in the IDENTIFY DEVICE data are used by the host for both data in and data
out transfers to insure that the device’s capabilities are not exceeded. For the Ultra DMA protocol, both the
host and device strobe data depending on the direction of the transfer. The host determines a mode setting
based on both the device’s capabilities and its own. The sender may send data (toggle STROBE) at a
minimum period of tCYC . A recipient receives data at the minimum tCYC for the currently active mode. If the
device indicates that it is capable of an Ultra DMA mode, it receives at the minimum time for that mode, no
additional cycle time information is required.
D.6.4 STROBE minimums and maximums
The Ultra DMA protocol does not define a maximum STROBE time. The sender may strobe as slowly as it
chooses independent of the mode that has been set, though it has to meet the specified setup and hold times
for the mode that has been set. The sender is also not required to maintain a consistent cycle time throughout
the burst. It would not be a violation of protocol for the cycle time to change on every cycle so long as all
cycles are longer than or equal to the minimum cycle time for the mode that is set. An upper timing bound or
PLL is not used by the recipient to qualify the STROBE signal. Regardless of the frequency of the STROBE,
the recipient has to be able to meet the setup and hold times of the received signal specified for the mode that
has been set. The limit on the maximum STROBE time is determined by the Ultra DMA device driver or BIOS
time-out. This time out is typically on the order of a few seconds. If a device begins to strobe once every ten
seconds during a data in burst, this would not be in violation of the protocol. However, this could cause a
software driver to determine that the device is not responding and perform a recovery mechanism. The recovery
will often be a hardware reset to the device.
Unlike a recipient pause where the recipient has to wait tR P after negating DMARDY− before the pause is
complete. The sender may consider the burst paused as soon as it meets the data hold time tDVH . The
implication of this is that data to the recipient may stop on any word. After each word, the recipient waits (with
exception of pauses or stops) but never requires an additional word before allowing the burst to be terminated.
D.6.5 Typical STROBE cycle timing
Neither minimum nor typical cycle times are required to be used by the sender. Other cycle times may be
used by systems that do not have internal clocks that provide a frequency to generate signals at those cycle
times. The typical mode 1 cycle time of 80 ns will not be met using a common system clock rate of 66.7 MHz.
Instead a STROBE cycle time of 90 ns for mode 1 is used and is not a violation of the specification. A typical
cycle time of 90 ns reflects 22.2 megabytes per second.
Page 391
T13/1321D revision 3
D.6.6 Holding data to meet setup and hold times
Following are three examples of holding data in an attempt to meet the setup and hold times. The first method
is to use the same clock edge to change data and the STROBE but delay the data through some gates. The
second method is to use one edge of the clock to change the STROBE and then the next opposite edge to
change data (half cycle). The third method is to use one active edge of the clock to change STROBE and the
next to change data.
Using gate delays to hold data may lead to large variations in hold time over process, temperature, and supply
voltage. Meeting Ultra DMA mode 4 timings with gate delays to hold data is not advisable and could lead to
timing violations under some conditions. Mode 4 hold time may be met by a single 66.6MHz clock cycle with
all timings being met. With a slower 25, 30, or 33 MHz clock, a half cycle rather than full cycle hold would be
required in order to still meet the setup time requirements for the higher modes. If the data transitions are not at
the middle of a mode 4 cycle, either the setup or hold time margin will be reduced.
D.6.7 Opportunities for the host to delay the start of a burst
After a device has asserted DMARQ, the host has one opportunity to delay the start of the burst indefinitely for
a data in burst and two opportunities for a data out burst. For both a data in and a data out burst, the first
opportunity that the host has to delay the burst is by delaying the assertion of DMACK–. This delay has no
specified maximum limit. This is necessary for cases of overlap in system bus access that may cause a delay
in the time it takes for the host to become ready to receive data from a device after sending a data in command.
For a data out burst, the host may delay the first STROBE signal. The difference in overhead between delaying
and not delaying may seem small but may still be used to optimize for a faster overall system data transfer
rate. The device does not delay its STROBE indefinitely since the device controls the signal that starts the
transfer process (DMARQ).
Note that it is a violation of the protocol to terminate the burst unless at least one word has been transferred.
After asserting DMACK– the host sends or receives at least one word of data before terminating a burst.
D.6.8 Maximums on all control signals from the device
The timings for all signals from the device used to perform burst initiation, pause, and burst termination have
maximum values. This is to bound the time it takes to perform burst initiation, pause, and termination so the
host always knows in advance how long tasks performed by the device may take. Rather than waiting a few
seconds for a command or burst to time-out, the host determines that a problem exists if activity is not
detected within the specified maximums and sets time-outs for functions performed by the device. For
instance, the longest the initiation of a data in burst may take from the host assertion of DMACK– to the first
STROBE is tENV max plus tRS max. Also, the host may require a burst to terminate in a timely manner in order
to service some other device on the bus or the system depending on the chip set design.
Page 392
T13/1321D revision 3
Annex E
(informative)
Bibliography
AT Attachment Interface with Extensions (ATA-2), ANSI X3.279-1996
AT Attachment - 3 (ATA-3), ANSI X3.298-1997
AT Attachment with Packet Interface Extension (ATA/ATAPI-4), ANSI NCITS.317-1998
BIOS Enhanced Disk Drive Specification (EDD), NCITS TR-21
Address Offset Reserved Area Boot, T13/1407DT
ATA Packet Interface (ATAPI) for Streaming Tape, QIC-1571
Suite of 2.5” Form Factor Specifications, SFF-8200, SFF-82012
Suite of 3.5” Form Factor Specifications, SFF-8300, SFF-8301, SFF-83022
1) QIC documents are published by:
Quarter-Inch Cartridge Drive Standards, Inc.
311 East Carrillo Street
Santa Barbara, CA 93101
Tel: 805-963-3853
Fax: 805-962-1541
2) SFF documents are published by:
SFF
14426 Black Walnut Court, Saratoga, California 95070
FaxAccess: 408 741-1600
SFF documents may be obtained from:
Global Engineering
15 Inverness Way East
Englewood, CO 80112-5704
Tel: 303-792-2181 or 800-854-7179
Fax: 303-792-2192
Page 393
T13/1321D revision 3
Annex F
(informative)
Command set summary
The following four tables are provided to facilitate the understanding of the command set. Table F.1 provides
information on which command codes are currently defined. Table F.2 provides a list of all of the commands in
order of command code. Table F.3 provides a summary of all commands with the protocol, required use,
command code, and registers used for each. Table F.4 shows the status and error bits used by each
command.
Table F.1 − Command matrix
0x
1x
2x
3x
4x
5x
6x
7x
8x
9x
Ax
Bx
Cx
Dx
Ex
Fx
x0
C
O
C
C
C
O
R
C
V
C
C
C
F
R
C
V
x1
R
E
O*
O*
O*
R
R
E
V
C
C
R
V
R
C
C
x2
R
E
O
O
R
R
R
E
V
C
C
R
V
R
C
C
x3
C
E
O
O
R
R
R
E
V
R
R
R
V
R
C
C
x4
R
E
R
R
R
R
R
E
V
E
R
R
C
R
C
C
x5
R
E
R
R
R
R
R
E
V
E
R
R
C
R
C
C
x6
R
E
R
R
R
R
R
E
V
E
R
R
C
R
C
C
Key:
C = a defined command.
R = Reserved, undefined in current specifications.
V = Vendor specific commands.
O = Obsolete.
E=a retired command.
Page 394
x7
R
E
R
R
R
R
R
E
F
E
R
R
C
R
C
V
x8
C
E
R
C
R
R
R
E
V
E
R
A
C
R
C
C
x9
R
E
R
R
R
R
R
E
V
E
R
A
O*
R
E
C
xA
R
E
R
R
R
R
R
E
V
V
R
A
C
C
R
V
xB
R
E
R
R
R
R
R
E
V
R
R
A
O*
E
R
V
xC
R
E
R
O
R
R
R
E
V
R
R
A
C
E
C
V
xD
R
E
R
R
R
R
R
E
V
R
R
A
C
E
C
V
xE
R
E
R
R
R
R
R
E
V
R
R
A
R
C
O
V
xF
R
E
R
R
R
R
R
E
V
R
R
A
R
C
C
V
F=If the device does not implement the CFA feature
set, this command code is Vendor specific.
A=Reserved for assignment by the CompactFlash
Association
* indicates that the entry in this table has changed
from ATA/ATAPI-4, NCITS 317-1998.
T13/1321D revision 3
Table F.2 − Commands sorted by command value
Command name
Command code
NOP
00h
CFA REQUEST EXTENDED ERROR CODE
03h
DEVICE RESET
08h
READ SECTOR(S)
20h
WRITE SECTOR(S)
30h
CFA WRITE SECTORS WITHOUT ERASE
38h
READ VERIFY SECTOR(S)
40h
SEEK
70h
CFA TRANSLATE SECTOR
87h
EXECUTE DEVICE DIAGNOSTIC
90h
INITIALIZE DEVICE PARAMETERS
91h
DOWNLOAD MICROCODE
92h
PACKET
A0h
IDENTIFY PACKET DEVICE
A1h
SERVICE
A2h
SMART
B0h
CFA ERASE SECTORS
C0h
READ MULTIPLE
C4h
WRITE MULTIPLE
C5h
SET MULTIPLE MODE
C6h
READ DMA QUEUED
C7h
READ DMA
C8h
WRITE DMA
CAh
WRITE DMA QUEUED
CCh
CFA WRITE MULTIPLE WITHOUT ERASE
CDh
GET MEDIA STATUS
DAh
MEDIA LOCK
DEh
MEDIA UNLOCK
DFh
STANDBY IMMEDIATE
E0h
IDLE IMMEDIATE
E1h
STANDBY
E2h
IDLE
E3h
READ BUFFER
E4h
CHECK POWER MODE
E5h
SLEEP
E6h
FLUSH CACHE
E7h
WRITE BUFFER
E8h
IDENTIFY DEVICE
ECh
MEDIA EJECT
EDh
SET FEATURES
EFh
SECURITY SET PASSWORD
F1h
SECURITY UNLOCK
F2h
SECURITY ERASE PREPARE
F3h
SECURITY ERASE UNIT
F4h
SECURITY FREEZE LOCK
F5h
SECURITY DISABLE PASSWORD
F6h
READ NATIVE MAX ADDRESS
F8h
SET MAX ADDRESS
F9h
Page 395
T13/1321D revision 3
proto
ND
ND
PI
PO
PO
ND
DR
PO
DD
ND
ND
PI
PI
ND
ND
ND
ND
ND
ND
ND
P
PI
DM
DMO
PI
ND
PI
ND
PO
ND
PO
ND
PO
PO
ND
P
ND
ND
ND
ND
ND
ND
ND
ND
PI
PI
ND
PO
Table F.3 − Command codes and parameters
Command
typ PKT Command
fea
code
CFA ERASE SECTORS
O
N
C0h
CFA REQUEST EXTENDED ERROR
O
N
03h
CFA TRANSLATE SECTOR
O
N
87h
CFA WRITE MULTIPLE W/OUT ERASE
O
N
CDh
CFA WRITE SECTORS W/OUT ERASE
O
N
38h
CHECK POWER MODE
M
M
E5h
DEVICE RESET
O
M
08h
DOWNLOAD MICROCODE
O
N
92h
EXECUTE DEVICE DIAGNOSTIC
M
M
90h
FLUSH CACHE
M
M
E7h
GET MEDIA STATUS
O
O
DAh
IDENTIFY DEVICE
M
N
ECh
IDENTIFY PACKET DEVICE
N
M
A1h
IDLE
M
O
E3h
IDLE IMMEDIATE
M
M
E1h
INITIALIZE DEVICE PARAMETERS
M
N
91h
MEDIA EJECT
O
O
EDh
MEDIA LOCK
O
O
DEh
MEDIA UNLOCK
O
O
DFh
NOP
O
M
00h
PACKET
N
M
A0h
READ BUFFER
O
N
E4h
READ DMA
M
N
C8h
READ DMA QUEUED
O
N
C7h
READ MULTIPLE
M
N
C4h
READ NATIVE MAX ADDRESS
O
N
F8h
READ SECTOR(S)
M
N
20h
READ VERIFY SECTOR(S)
M
N
40h
SECURITY DISABLE PASSWORD
O
O
F6h
SECURITY ERASE PREPARE
O
O
F3h
SECURITY ERASE UNIT
O
O
F4h
SECURITY FREEZE
O
O
F5h
SECURITY SET PASSWORD
O
O
F1h
SECURITY UNLOCK
O
O
F2h
SEEK
M
N
70h
SERVICE
N
O
A2h
SET FEATURES
M
M
EFh
SET MAX ADDRESS
O
N
F9h
SET MULTIPLE MODE
M
N
C6h
SLEEP
M
M
E6h
SMART DISABLE OPERATIONS
O
O
B0h
SMART ENABLE/DISABLE AUTOSAVE
O
O
B0h
SMART ENABLE OPERATIONS
O
O
B0h
SMART EXECUTE OFF_LINE
O
O
B0h
SMART READ DATA
O
O
B0h
SMART READ LOG SECTOR
O
O
B0h
SMART RETURN STATUS
O
O
B0h
SMART WRITE LOG SECTOR
O
O
B0h
Page 396
FR
y
Parameters used
SC SN CY DH
y
y
y
y
D
y
y
y
y
y
y
y
y
y
y
y
y
D
D
y
y
y
D
D*
y
y
y
y
D
D
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
D
D
y
D
D
D
D
y
D
D
y
y
y
y
y
y
D
y
y
y
y
D
D
D
D
D
D
y
y
y
D
D
y
y
D
D
y
D
y
D
y
D
y
D
y
D
y
D
y
D
y
D
(continued)
T13/1321D revision 3
proto
ND
ND
PO
DM
DMO
PO
PO
VS
-
-
-
Table F.3 − Command codes and parameters (concluded)
Command
typ PKT Command
Parameters
fea
code
F
SC SN
R
STANDBY
M
O
E2h
y
STANDBY IMMEDIATE
M
M
E0h
WRITE BUFFER
O
N
E8h
WRITE DMA
M
N
CAh
y
y
WRITE DMA QUEUED
O
N
CCh
y
y
y
WRITE MULTIPLE
M
N
C5h
y
y
WRITE SECTOR(S)
M
N
30h
y
y
Vendor specific
V
V
9Ah,C0hC3h,8xh,
F0h,F7h,
FAh-FFh
Retired
E
E
11h-1Fh,
71h-7Fh,
94h-99h,
DBh-DDh,
E9h
Obsolete
B
B
10h,
21h-23h,
31h-33h,
3Ch, 41h,
50h, C9h,
CBh, EEh
Reserved: all remaining codes
R
R
proto = command protocol
Key:
DM = DMA command
PO = PIO data-out command
P=PACKET command
DMO = Overlapped/queued DMA
ND = Non-data command
VS = Vendor specific command
DR = DEVICE RESET command
typ=Command type
M = Mandatory
V = Vendor specific implementation
CY = Cylinder registers
SN = Sector Number register
R = Reserved
E = Retired
SC = Sector Count register
FR = Features register (see
command descriptions for use)
D = only the device parameter is
valid and not the head parameter
d = the device parameter is valid,
the usage of the head parameter
vendor specific.
used
CY
DH
y
y
y
y
D
D
D
y
y
y
y
PI = PIO data-in command
O = Optional
DD = EXECUTE DEVICE DIAGNOSTIC
PKT fea=Command type when PACKET
Command feature set implemented
N=Not to be used
B = Obsolete
DH = Device/Head register
y = the register contains a valid
parameter for this command. For the
Device/Head register, y means both the
device and head parameters are used.
D* = Addressed to device 0 but both
devices execute the command.
Page 397
T13/1321D revision 3
Table F.4 − Register functions and selection addresses except PACKET and SERVICE commands
Addresses
Functions
CS0CS1DA2
DA1
DA0
Read (DIOR-)
Write (DIOW-)
N
N
x
x
x
Released
Not used
Control block registers
N
A
N
x
x
Released
Not used
N
A
A
N
x
Released
Not used
N
A
A
A
N
Alternate Status
Device Control
N
A
A
A
A
Obsolete(see note)
Not used
Command block registers
A
N
N
N
N
Data
Data
A
N
N
N
A
Error
Features
A
N
N
A
N
Sector Count
Sector Count
A
N
N
A
A
Sector Number
Sector Number
A
N
A
N
N
Cylinder Low
Cylinder Low
A
N
A
N
A
Cylinder High
Cylinder High
A
N
A
A
N
Device/Head
Device/Head
A
N
A
A
A
Status
Command
A
A
x
x
x
Released
Not used
Key:
A = signal asserted
N = signal negated
x = don’t care
NOTE − This register is obsolete. It is recommended that a device not respond to a read of this address.
Table F.5 − Register functions and selection addresses for PACKET and SERVICE commands
Addresses
Functions
CS0CS1DA2
DA1
DA0
Read (DIOR-)
Write (DIOW-)
N
N
x
x
x
Released
Not used
Control block registers
N
A
N
x
x
Released
Not used
N
A
A
N
x
Released
Not used
N
A
A
A
N
Alternate Status
Device Control
N
A
A
A
A
Obsolete(see note)
Not used
Command block registers
A
N
N
N
N
Data
Data
A
N
N
N
A
Error
Features
A
N
N
A
N
Interrupt reason
A
N
N
A
A
A
N
A
N
N
Byte count low
Byte count low
A
N
A
N
A
Byte count high
Byte count high
A
N
A
A
N
Device select
Device select
A
N
A
A
A
Status
Command
A
A
x
x
x
Released
Not used
Key:
A = signal asserted
N = signal negated
x = don’t care
NOTE − This register is obsolete. A device should not respond to a read of this address.
Page 398
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising