Apache Labs ANAN-100D SDR HF/6 Meter Transceiver

Apache Labs ANAN-100D SDR HF/6 Meter Transceiver
TechnicalReview
Product
by Mark
Spencer,
WA8SME
Mark
J. Wilson,
K1RO,
k1ro@arrl.org
Apache Labs ANAN-100D
SDR HF/6 Meter Transceiver
An advanced, fully open-source SDR transceiver
Reviewed by Martin Ewing, AA6E
aa6e@arrl.net
Software defined radio
(SDR) is taking off in
Amateur Radio. For
“mainstream” transceivers (typically covering the 160 – 6 meter
bands with 100 W output), there have been
only a few commercial
offerings, mainly from
FlexRadio Systems.
Now we think this market segment is going
to have even more life, thanks to continuing advances in digital technology.
The Apache Labs ANAN series has its own
approach, based on open-source d­ esign.
In this review, we look at the ANAN100D, which is based on some of the same
technology as the FlexRadio 6000 series
transceivers we recently reviewed.1 The
FlexRadio 6000 series and the Apache
Labs ANAN series represent a generation
of SDRs built around fast samplers that are
able to ingest the entire RF spectrum from
near dc to 54 MHz and beyond. They do
require external intelligence provided by
your PC in order to operate, but the heavy
lifting is done in the radio with special field
programmable gate array (FPGA) devices.
Using a technique called digital down-conversion (DDC), FPGAs can divide up the
bandwidth into a number of subbands that
can be displayed as panadapter spectra or
waterfalls, where you can plunk down receivers to listen to your favorite modes —
from CW and SSB to exotic digital modes.
Introduction
The ANAN-100D HF/6 meter transceiver
gives you a detailed view of up to 1 MHz
in one panadapter display on one band plus
another 350 kHz in a second band, using
free PowerSDR mRX PS software running
on your PC. The transmitter produces up to
100 W peak (30 W maximum sustained average). With the “Pure Signal” feature, the
transmitter can produce remarkably low
intermodulation distortion (IMD), making
its signal one of the cleanest on the bands.
As we mentioned above, the ANAN-100D
and its recommended PowerSDR software
are open-source projects, including both
software and hardware design. All the basic
design information and source code are
freely available, if you want to be part of
the project or if you just want to see how
the magic is made. Both hardware and
software have been developed as part of
the OpenHPSDR project (see sidebar “The
Bottom Line
OpenHPSDR Project”). Apache Labs,
an Indian company,
builds and packages
the hardware, while
the software is developed by a number
of ham developers
spread around the
globe. Apache Labs
provides links to key
software, but the developers publish much of
their work to various other
websites. Finding documentation may be a
challenge, but you can eventually find what
you need, so the open arrangement works
pretty well.
The heart of the ANAN-100D is the “Angelia” SDR transceiver board, a descendant
of OpenHPSDR’s “Hermes” design.2 If you
want to make your own, you can buy the assembled Angelia board separately, and you
can even buy the bare eight-layer PC board
by itself. Apache Labs builds the ’100D
system by adding a 100 W power amplifier and other circuitry to the Angelia in a
compact enclosure ready for your operating
desk (see the lead photo and Figure 4). The
radio is shipped with pre-installed firmware
(the internal control and signal processing
software), but you must download software
and documentation from the Internet to get
up and running.
Hardware
On the receiver side, the ANAN-100D has
Units Tested
The Apache Labs ANAN-100D offers good performance in a full-featured SDR based on the open-source
model. Low transmit IMD with Pure Signal enabled is
particularly noteworthy.
Apache Labs ANAN-100D, serial number 32186
PowerSDR_mRX_PS version 3.2.27
Angelia firmware v. 5.0. (2/18/15 – 6/13/15)
QST ® – Devoted entirely to Amateur Radio
www.arrl.org
October 2015 45
Key Measurements
Summary
Table 1
Apache Labs ANAN-100D, serial number 32186
Manufacturer’s Specifications
Measured in the ARRL Lab
Frequency coverage: Receive, 0.01 – 55 MHz;
transmit, 160 – 6 meter amateur bands.
Receive, 0.100 – 61.440 MHz;
transmit, as specified.
124
Power requirement: Not specified.
At 13.8 V dc: transmit, 15 A (typical);
receive, 2.1 A. Operation confirmed
at 11.7 V dc (≥95 W RF output).
Modes of operation: SSB, CW, AM, Digital,
RTTY, FM.
As specified.
124
Receiver
Receiver Dynamic Testing
Minimum discernible signal (MDS): –138 dBm. Noise floor (MDS), 500 Hz DSP BW:
0.137 MHz –130 dBm
0.475 MHz –130 dBm
1.0 MHz
–135 dBm
3.5 MHz
–135 dBm
14 MHz
–135 dBm
50 MHz
–144 dBm
Noise figure: Not specified.
14 MHz, 12 dB; 50 MHz, 3 dB.
Spectral sensitivity: Not specified.
100 kHz screen bandwidth: panadapter,
–135 dBm; waterfall, –142 dBm.
AM sensitivity: Not specified.
10 dB (S+N)/N, 1-kHz, 30% modulation,
6 kHz DSP BW:
1.0 MHz
1.16 µV
3.88 MHz 1.16 µV
50.4 MHz 0.56 µV
FM sensitivity: Not specified.
29 MHz, 0.47 µV; 52 MHz, 0.20 µV.
RM
117
20 60
140
20 kHz Reciprocal Mixing Dynamic Range
BG
20 70
140
20 kHz Blocking Gain Compression (dB)
96
I3
97
20 50
110
20 kHz 3rd-Order Dynamic Range (dB)
RM
105
60
2
140
2 kHz Reciprocal Mixing Dynamic Range
122
BG
122
70
2
140
2 kHz Blocking Gain Compression (dB)
I3
96
50
2
110
2 kHz 3rd-Order Dynamic Range (dB)
20
Reciprocal mixing dynamic range: Not specified.
I3
22 +35
20 -40
20 kHz 3rd-Order Intercept (dBm)
I3
22 +30
-40
2
2 kHz 3rd-Order Intercept (dBm)
-29*
I3
TX -20
-49**‡
-35
Transmit 3rd-Order IMD (dB)
-49*
I9
TX -20
< -60**
-70
Transmit 9th-Order IMD (dB)
QS1510-PR098
80 M
Key: ‡ Off Scale
Intercept values were determined
using –97 dBm reference.
Worst case band, predistortion off.
20 M
Worst case band, predistortion on.
46 October 2015
Blocking gain compression dynamic range:
Blocking gain compression dynamic range,
Not specified.
500 Hz DSP BW:
20 kHz offset 5/2 kHz offset
3.5 MHz
124 dB
124/122 dB
14 MHz
124 dB
124/122 dB
50 MHz
116 dB
116/116 dB
14 MHz, 20/5/2 kHz offset: 117/110/105 dB.
ARRL Lab Two-Tone IMD Testing* (500 Hz DSP bandwidth)
Measured
Band
Spacing IMD Level
Input Level
Measured
IMD DR**
Calculated
IP3
3.5 MHz
20 kHz
–135 dBm
–97 dBm
–39 dBm
96 dB
–19 dBm
+9 dBm
+20 dBm
14 MHz
20 kHz
–135 dBm
–97 dBm
–38 dBm
97 dB
–18 dBm
+11 dBm
+22 dBm
14 MHz
5 kHz
–135 dBm
–97 dBm
–39 dBm
96 dB
–18 dBm
+9 dBm
+22 dBm
14 MHz
2 kHz
–135 dBm
–97 dBm
–39 dBm
96 dB
–18 dBm
+9 dBm
+22 dBm
50 MHz
20 kHz
–144 dBm
–97 dBm
–53 dBm
91 dB
–35 dBm
–7 dBm
–4 dBm
Second-order intercept point: Not specified.
14 MHz, +65 dBm; 21 MHz, + 51 dBm;
50 MHz, +83 dBm.
FM two-tone third-order dynamic range:
Not specified.
FM adjacent channel rejection: Not specified
20 kHz spacing: 29 MHz, 58 dB; 52 MHz,
45 dB. 10 MHz spacing: 29 MHz, 82 dB;
52 MHz, 89 dB.¹
29 MHz, 84 dB; 52 MHz, 81 dB.
Squelch sensitivity: Not specified.
29 MHz, 0.18 µV; 52 MHz, 0.09 µV.
DSP noise reduction: Not specified.
10 dB.
Notch filter depth: Not specified.
Auto-notch, >60 dB. Attack time, 188 ms
(single tone), 316 ms (two tones).
ARRL, the national association for Amateur Radio®
www.arrl.org QS1510-ProdRev02
0
–10
Measured in the ARRL Lab
S-meter sensitivity: Not specified.
S-9 signal, 14 MHz, 46.2 µV;
50 MHz, 60.9 µV.
IF/audio response: Not specified.
Range at –6 dB points (bandwidth)†:
CW: 300 – 902 Hz (602 Hz)
Equivalent Rectangular BW: 507 Hz
USB, 194 – 2638 Hz (2444 Hz)
LSB, 196 – 2676 Hz (2480 Hz)
AM, 160 – 4988 Hz (9656 Hz)
Transmitter
Transmitter Dynamic Testing
Power output: 0 – 100 W. As specified, except 0 – 30 W, AM.
Spurious-signal and harmonic suppression:
>50 dB (HF); >60 dB (50 MHz).
HF, typically 58 dB, worst case 52 dB
(24.9 MHz). 50 MHz, 60 dB. Complies
with FCC emission standards.
SSB carrier suppression: >90 dB.
>70 dB.
Undesired sideband suppression: >90 dB.
>70 dB.
Third-order intermodulation distortion (IMD):
Not specified.
100 W PEP, 3rd/5th/7th/9th order:
Pure Signal on (see text):
–52/–54/<–60/<–60 dB (HF, typical)
–52/–50/–51/<–60 dB (50 MHz)
–49/–56/–59/<–60 dB (worst case, 160 m)
Pure Signal off (see text):
–38/–38–44/–52 dB (HF, typical)
–40/–37/–44/–52 dB (50 MHz)
–29/–35/–39/–49 dB (worst case, 15 m)
CW keyer speed range: Not specified.
1 to 45.8 WPM, iambic Mode B.
CW keying characteristics: Not specified.
See Figures 1 and 2.
Response (dB)
–20
Manufacturer’s Specifications
–40
–50
–60
–70
–80
–90
–100
fc-4
fc-2
fc+2
fc
Frequency in kHz
Figure 2 — Spectral display of the ANAN-100D
transmitter during keying sideband testing.
Equivalent keying speed is 60 WPM using
external keying. Spectrum analyzer resolution
bandwidth is 10 Hz, and the sweep time is 30
seconds. The transmitter was being operated
at 100 W PEP output on the 14 MHz band, and
this plot shows the transmitter output ±5 kHz
from the carrier. The reference level is 0 dBc,
and the vertical scale is in dB.
QS1510-ProdRev03
0
50 MHz
14 MHz
–20
SSB, 142 ms; FM, 109 ms (29 MHz),
129 ms (52 MHz).††
Transmitted phase noise: Not specified.
See Figure 3.
Level in dBc/Hz
Receive-transmit turn-around time (tx delay):
Not specified.
–60
–80
–100
–120
–140
QS1510-ProdRev01
–160
Size (height, width, depth, incl protrusions): 3.3 × 10.4 × 11.1 inches. Weight, 10 lbs.
–180
100 Hz
Price: $3489.00.
*ARRL Product Review testing includes Two-Tone IMD results at several signal levels.
Two-tone, Third-order Dynamic Range figures comparable to previous reviews are shown
on the first line in each group. The “IP3” column is the calculated Third-order Intercept Point.
Second-order intercept points were determined using –97 dBm reference.
**Third order two-tone dynamic range values shown are best case. IMD DR depends on band activity
and signal strengths. See text and February 2010 QST, page 52, for an explanation.
†Default values, CW 500 Hz BW, SSB 2.4 kHz BW, AM 10 kHz BW; bandwidth is
adjustable via DSP.
††Default values. Turnaround times and CW delay are adjustable in PowerSDR.
0.28 0.32
10 kHz
100 kHz
Frequency Offset
1 MHz
0.04 0.08 0.12 0.16 0.20 0.24
Time (s)
0.28 0.32
(A)
QS1510-ProdRev01
Figure 1 — CW keying waveform for
the ANAN-100D showing the first two
dits in full-break-in (QSK) mode using
external keying. Equivalent keying
speed is 60 WPM (A) and 48 WPM
(B). The upper trace is the actual key
­closure; the lower trace is the RF envelope. (Note that the first key closure
starts at the left edge of the figure.)
Horizontal divisions are 10 ms. The
transceiver was being operated at
100 W output on the 14 MHz band.
1 kHz
Figure 3 — Spectral display of the ANAN-100D
transmitter output during phase noise testing.
Power output is 100 W on the 14 MHz band (blue
trace) and on the 50 MHz band (red trace). The
carrier, off the left edge of the plot, is not shown.
This plot shows composite transmitted noise
100 Hz to 1 MHz from the carrier. The reference
level is 0 dBc, and the vertical scale is in dBc/Hz.
0
0.04 0.08 0.12 0.16 0.20 0.24
Time (s)
fc+4
–40
Transmit-receive turn-around time (PTT release S-9 signal, AGC fast, 240 ms.††
to 50% audio output): Not specified.
0
–30
0
(A)
QST ® – Devoted entirely to Amateur Radio
0.01 0.02 0.03 0.04 0.05 0.06
Time (s)
0.07 0.08
(B)
www.arrl.org
October 2015 47
two analog-to-digital converter (ADC)
channels sampling the input RF (10 kHz –
55 MHz) at a 122.88 MHz rate with 16-bit
resolution. Firmware defines up to seven
independent receivers operating at bandwidths up to ~350 kHz that can be placed
within the 55 MHz range. Output of each
of these receivers can be displayed as a
panadapter, waterfall, and in other ways,
supporting one or two audio detectors.
As a transmitter, the ’100D supports a
single modulation channel with digital
upconversion to the operating frequency
followed by a 100 W power amplifier.
Traditional modes include SSB, AM, FM,
and CW. RTTY and digital modes are supported as SSB audio via additional digimode software.
Software
The ANAN-100D is entirely controlled
through the user’s PC, excepting only the
on/off switch. The recommended software
for the Windows PC is the OpenHPSDR
“mRX PS” extension of PowerSDR. PowerSDR is open-source software familiar
to many amateurs as the basis for several
SDR products, including earlier FlexRadio
transceivers. For the OpenHPSDR project,
PowerSDR has been enhanced by Doug
Wigley, W5WC, Warren C. Pratt, NRØV,
and perhaps others. The “beta” User Guide
has been prepared by Ken Hopper, N9VV,
and Bill Diaz, KC9XG.3
PowerSDR supports up to four receivers as
two spectral displays (see Figure 5). It will
allow you to “stitch” together three receivers to appear as one panadapter display covering up to about 1.05 MHz. Normally, you
have a single audio channel, but in “multiRX” mode you can have two. If you’re
operating split frequency, for example, you
can listen to DX with one ear, and the pileup with the other. Separately, PowerSDR
will support another independent receiver
called “RX2,” which is normally connected
to its own ADC and its own RF input jack.
This can be useful for diversity reception or
for monitoring a second band.
your operation, but keeping up to date is a
good idea.
The PowerSDR mRX PS software for
Windows appears to be the only fully operational transceiver software available for
the ANAN-100D. However, with the open
nature of the project, it is not surprising that
there are other choices. I looked at two,
which currently work with the ’100D for
receiving, but not transmitting.
CuSDR was developed for Windows by
Hermann von Hasseln, DL3HVH. It provides views of the entire HF spectrum
along with detailed windows for specific
receiving bands of interest.5 The visual interface is modern and attractive. CuSDR is
available in beta test, but it is not officially
released.
Ghpsdr3-Qt, developed by John Melton,
GØORX/N6LYT, is another software system developed for OpenHPSDR radios.
It features an interesting client/server
structure that may make it easier to share
work among different computers. Source
or binary versions are available for Linux,
MacOS, or Windows.6
These alternative projects show that it is
possible for experienced programmers to
develop their own back-end systems for
special needs or just for fun.
Computers and Networks
The User Guide recommends a computer
with a 2.8 GHz Intel i3 processor or better,
6 GB RAM, and a 1280 × 1024 display
­running Windows 7. I think PowerSDR
mRX PS software should run well on most
recent Windows PCs. It worked okay for me
on a range of computers, from multicore
Figure 4 — The ANAN-100D’s rear panel connections include three antennas,
an external transverter, external 10 MHz oscillator, and amplifier PTT.
Angelia’s internal computations are performed in an Altera Cyclone IV FPGA.
Open source firmware for this device has
been developed by Joe Martin, K5SO.4
If your radio does not have the current
firmware version, you can download and
install the latest one. A new version may
bring new features, bug fixes, or reliability
improvements that may or may not affect
48 October 2015
Figure 5 — PowerSDR offers spectral and waterfall displays, along with access to all transceiver functions.
ARRL, the national association for Amateur Radio®
www.arrl.org The OpenHPSDR Project
The ANAN-100D’s Angelia board was developed by Apache Labs as part of the
Open Source High Performance Software Defined Radio (OpenHPSDR) project,
which has supported a number of limited production products for experimenters
beginning in 2006.*
There have been several generations of products, and the OpenHPSDR documentation is full of project names that are historically important. In particular, the Angelia
board descends from an earlier single-board SDR called Hermes. Much of the original OpenHPSDR hardware/firmware development, including the Hermes transceiver, was done by Phil Harman, VK6PH. Hermes in turn incorporates technology
from earlier projects such as Metis, Alex, Mercury, Penelope, and Pennywhistle.
Some of this terminology is carried forward into PowerSDR’s options settings and the
ANAN-100D User Guide.
*See OpenHPSDR.org and S. Cowling, WA2DFI, “The High Performance Software Defined Radio
Project,” QEX, May/Jun 2014, pp 3 – 13.
laptops on up using Windows 7. I was able
to run some experiments with my Intel Core
i7 920 system with various BIOS settings.
With only a single core running, PowerSDR
had problems, but with two or more cores it
seemed to be adequate. Your mileage may
vary!
PowerSDR needs a display with at least
1024 × 768 resolution. A larger display (or
multiple displays) will provide screen space
for logging programs, digimode software,
and other operating niceties. The software
has an interesting “collapsed” mode that
rearranges the display into a more compact
format.
The ANAN-100D with PowerSDR does
place a heavy load on your Ethernet as
well as your PC. The User Guide recommends gigabit Ethernet, although current
software versions can only support a 100
Mb/s rate. Some combinations of operating modes are prohibited because they will
require more than 100 Mb/s. For example,
you can’t stitch radios or use 384 Kb/s
sampling while using the Pure Signal feature. More of these “corner cases” should
be supported when 1000 Mb/s (gigabit)
Ethernet is implemented. You can connect
your computer and the ’100D to your home
network router, but if needed you can use
an isolated Ethernet connection with no
router at all.
Documentation
The first thing to do after you unpack
the ANAN-100D is to download the
User Guide, which is well-written but
rather uneven in its coverage.7 To use the
Guide ­effectively, it helps to know your
way around PowerSDR and some of the
OpenHPSDR lingo. If PowerSDR is new
to you, I would recommend reading the
manual (software parts) for the FlexRadio
Systems FLEX-5000, which has a similar
user interface.8 (The Flex document is a fine
example of a comprehensive user manual.)
Thorough documentation is hard work, and
it can be a difficult thing to accomplish in
distributed open source projects.
Most of the information you might want
about PowerSDR and the ANAN-100D is
available on the Internet, but some of it is
fragmented across various websites, and
those websites may not be quite in sync.
You may need to ask the users and gurus on
the Yahoo Apache-Labs group.9 There is
also a notable wiki site that compiles useful
links and documents.10
Many options are available through the
setup menus, but unfortunately the User
Guide does not explain all of them. Some
major items, such as the ALEX button are
not explained anywhere. (It enables automatic receiver filter switching.) I would
prefer that the more exotic and experimental options be hidden unless you enter
an “expert” mode; otherwise a newbie is
tempted to twiddle settings and possibly set
bad operating conditions or get lost in the
forest of options.
Some settings seem to apply to unfinished
development work that should probably not
appear in “production” software. (Sometimes open source can be a little too open!)
It’s not too much to expect that all controls should be documented, with a clear
explanation of whether special capabilities like CESSB or Pure Signal should be
considered the default choice for normal
operation.
Operating Modes
For CW operation, PowerSDR includes an
iambic keyer function claiming to cover 1
to 60 WPM, so you only need to attach your
paddle or straight key. Unfortunately, we
found that the ANAN-100D (with current
software) was not able to reliably track either the internal iambic keyer or an external
keying signal above about 48 WPM. Figure 1 shows the problem when we try to run
our standard 60 WPM keying test.
CW with the internal keyer was good in
the 20 – 30 WPM range that I normally
send with a paddle. The audio sidetone is in
sync with the RF output without perceptible
delay. Sidetone pitch can be varied from
200 to 2250 Hz. Semi-break-in operation
is provided, with a variable transmit hold
time. Typical of SDR transceivers, the long
transmit/receive switching time prevents
full break-in (QSK) operation.
In CW mode, PowerSDR provides an adjustable audio peaking filter (APF) that can
tailor the audio response to your liking. You
can manually zero beat with another station
by visually tuning to the IF band center, and
if reception is clear enough, you can use
the 0 BEAT button to automatically center
the signal.
PowerSDR also sports a CWX feature that
allows you to send CW from preset buffers
or “live” from your keyboard. This should
help for contesting or for any Morse keyboarding.
For voice operation, the ANAN-100D
provides a front panel 1⁄8-inch jack for
microphone and PTT (push to talk). The
User Guide helpfully explains how to
connect various microphones. You can
select optional “phantom” bias and change
the pin-out by setting on-board jumpers.
(Unfortunately, changing Angelia’s jumpers requires a tricky disassembly of the
’100D.) Alternatively, you can connect
line level audio and PTT through the back
panel DB-25 connector. It took me a while
to figure out how to use a low-output microphone. It required not only selecting
+20 dB boost, but also increasing the MAX
GAIN setting on the transmit setup dialog.
SSB and AM modes offer a selectable
gain compression to increase average
“talk power.” There is also a “downward
expansion” option that acts as a noise gate
— quickly attenuating the audio when the
input level falls below a certain level, helping control background noise. In addition to
QST ® – Devoted entirely to Amateur Radio
www.arrl.org
October 2015 49
gain compression, a special “CESSB Overshoot Control” option is available to minimize overshoot peaks that would otherwise
limit the maximum power.11
An audio equalizer provides either 3- or
10-band adjustment of frequency response
separately for transmit and receive. Some
of the audio characteristics can be stored
in “transmit profiles,” which are labeled
storage areas that allow you to switch your
configuration between, say, normal and DX
modes. (DX might need more compression
and less bass response to give more punch,
for example.)
AM is supported on receive with filters up
to ±10 kHz bandwidth, providing good results on broadcast stations, especially in the
SAM (synchronous AM) mode. Transmit
filters can also be set to pass modulation up
to 10 kHz. If you ask for more than 3 kHz
(all that is needed for normal voice communications), you will get a mild scolding from
PowerSDR, but you can use all 10 kHz.
The ’100D with PowerSDR supports FM
voice operation with either 2.5 or 5.0 kHz
deviation. A convenient control panel allows you to select CTCSS tones and repeater offsets, including normal, reverse,
and simplex. A MEMORY tab brings up
a window that lets you remember all the
relevant settings for an operating channel,
such as a repeater. (This feature is handy
for storing commonly used frequencies and
settings for any mode.)
An adjustable squelch control is available
on any mode, but will be particularly important for channelized FM work. There is
no tuning indicator to show carrier offset,
but you can see the offset on the panadapter
during a pause in modulation.
For digital QSOs, you can use the
ANAN-100D with Fldigi, WSJT-x, and
similar software. The convenient way to operate with these sound card modes is to internally connect the digimode software with
PowerSDR using Virtual Audio Cables.12
VAC operation requires no sound cards, but
you will need to obtain some additional software. The PowerSDR-Fldigi writeup also
gives some clues about the ’100D CAT interface, which is not otherwise documented,
apart from the FLEX-5000 guide.
tions. The output level is fairly low, about
–10 dBm. PowerSDR provides a flexible
antenna switching matrix and transverter
setup dialog that lets you specify the frequency and gain relationships for up to 13
different transverters. You can also make
use of programmable band select outputs
to control transverters, antenna relays, or
other gear.
Diversity Reception
With the VFO SYNC button, you can force
VFOs A and B to the same frequency. The
two receiver channels (RX1 and RX2) can
then be used with separate antennas in two
different diversity modes. You can receive
RX1 in one ear and RX2 in the other (using
the PAN adjustments). Signals then seem
to come from different directions, and that
may help you dig the desired signal out of
the noise and QRM. When the signal fades
out in one channel, it may still be audible in
the other, depending on how independent
your antennas are in terms of spatial separation, polarization, and so on.
The ANAN-100D also supports a diversity
combining mode. This adds the RX1 and
RX2 signals together with adjustable phase
and amplitude. You can use this simply to
Taming IMD with Predistortion Linearization (Pure Signal)
Intermodulation distortion (IMD) is a problem for transmitters because it causes
audible distortion of your voice or multi-tone digital signal, and it causes your signal
to broaden and interfere with other stations near your frequency. For years, engineers
have used the technique called “predistortion linearization” to reduce IMD in commercial
RF power amplifiers.† We present a quick overview here.
To our knowledge, the “Pure Signal” feature of PowerSDR, as implemented by ­Warren
C. Pratt, NRØV, is the first application of predistortion linearization to a commercial
Amateur Radio HF transceiver.†† Pratt received the 2014 ARRL Technical Innovation
Award for this work.
Part A of Figure A offers a graphical view. Your “linear” amplifier is never quite linear
— that’s what produces IMD. It might have some power compression at high power as
shown by the “Amp” curve in the figure. If we introduce predistortion before the amplifier with a matching power expansion characteristic, the overall result should be much
improved linearity. The amplifier’s power transfer curve will vary depending on operating
conditions, so a fixed predistortion won’t do. With good software, your radio can continuously compare the RF output with the desired waveform and adjust the predistortion as
needed.
Part B of Figure A is a basic block diagram. The receiver operates during transmission allowing control software to adapt to changing conditions. Everything you need is
provided in the ANAN-100D with PowerSDR Pure Signal feature. You can also use it to
linearize an external power amplifier, but you will have to feed back a small sample of
the amplifier’s output to the receiver with some extra hardware.
Predistortion linearization is a fascinating development that can work particularly well
in SDR radios like the ’100D, helping you transmit a cleaner signal, improve your audio
quality, and reduce QRM for everyone.
†www.highfrequencyelectronics.com/Sep04/HFE0904_Nezami.pdf
††www.wavenode.com/PureSignal.pdf
Figure A — Predistortion linearization
method graphical
concept (A) and simplified block diagram with
adaptive control (B).
For 2 meters and higher frequencies, you
can set up the ANAN-100D to work with
external transverters. The ’100D has dedicated transverter input and output connec50 October 2015
ARRL, the national association for Amateur Radio®
www.arrl.org There is an auto notch filter (ANF) that will
try to eliminate steady interfering tones —
“heterodynes” as we knew them of yore.
The binaural (BIN) option separates the
audio reception into two components for
left and right headphones. It produces a
pleasing stereo effect. I’m not sure if it improves copy, but it won’t hurt! The Power­
SDR setup options give you the chance to
experiment with many different settings for
the NR and ANF functions.
There are two noise blanking functions —
NB or NB2. They generally work by snipping out a short piece of audio when there
is a brief noise peak. Depending on your
situation, you may benefit from one option
or the other.
PowerSDR has an RA (radio astronomy)
option. As a radio astronomer in a former life, I found that intriguing. With this
feature, you can measure what we would
call “total power” — the sum of the power
across frequency channels. This is averaged, written to a disk file, and graphed
against time. Astronomically, you might
record the response as the Sun or Moon
drifts through the beam of your VHF
Yagi, helping you check the pointing of
your antenna and the sensitivity of your
system. Moving back to Earth, you might
monitor your interference level throughout
the day.
Another feature of the ANAN-100D is the
setup option for your operating region. Regions are predefined for many countries.
The one you choose determines the allowed operating bands, modes, and license
classes. There is also an “extended” region
that allows transmission on any frequency,
Making use of the extended region, we did
some quick tests transmitting on the proposed 630 meter and 2200 meter bands. We
got around 100 W output at 475 kHz, and
about 40 W at 137 kHz. The transmitter has
high harmonic output, however, because it
lacks the correct low-pass filters for those
bands. You would have to supply your own
filters when the bands open up to Amateur
Radio.
For contest operators who need to reduce
the complexity of working with computer
logging systems, PowerSDR implements
a FocusMaster feature that returns mouse/
keyboard focus to a supplementary program such as the N1MM+ logger after a
radio operation is completed.
Receiver IMD
The ANAN-100D, like some other SDR
receivers, shows a two-tone third-order
dynamic range that depends on the level of
activity and signal strengths in the RF pass
band of the receiver, as explained in the
QS1510-ProdRev06
0
–10
–20
Response (dB)
Other Features
PowerSDR provides two interesting noise
reduction options — NR and NR2. They
both offer the ability to help signals stand
out from noise. They sound rather different,
so you need to try both of them to find what
works for you. I found that NR2 was particularly good in some conditions. It has a somewhat unnatural sound, but it was effective.
which should only be used with care. You
are always responsible to be sure that on-air
operation is confined to bands and modes
that are legal for your license.
–30
–40
–50
–60
–70
–80
–90
–100
fc-4
fc-2
fc+2
fc
Frequency in kHz
fc+4
(A)
0
–10
–20
Response (dB)
gain a bit of signal to noise with two antennas, or you can use it to null out an interfering signal. With some tricky settings you
can create a deep null in a particular direction that can help you eliminate a specific
source of interference. It can be hard to find
the null if propagation is varying too much,
but I achieved a 35 dB null on a local broadcast AM station.
–30
–40
–50
–60
–70
–80
–90
–100
fc-4
fc-2
fc+2
fc
Frequency in kHz
fc+4
(B)
Figure 6 — ANAN-100D during normal operation (A) and with Pure Signal on (B).
February 2010 issue of QST, page 52. The
values in the data table are the best observed
dynamic ranges. Typically, on a quiet band,
the user will have a third order IMD dynamic range of about 65 dB.
Transmitter Linearity
The ANAN-100D with PowerSDR offers a
very interesting transmitter IMD-reduction
feature called “Pure Signal” (PS), implemented by Warren C. Pratt, NRØV, and
based on the principle of predistortion
linearization. (See the sidebar, “Taming
IMD with Predistortion Linearization (Pure
Signal.”)
IMD improvements measured in the ARRL
Lab were dramatic. A typical IMD measurement before and after applying Pure
Signal is shown in Figure 6. The typical 3rd
order IMD without PS was a ­respectable
–38 dB (Figure 6A), but with PS IMD it
was an exceptional –50 dB (Figure 6B) —
the best we’ve ever seen in the Lab.
Pure Signal is provided on an “opt in” basis.
You need to set it up and enable it before
using. At some point, we would hope that
a feature like this will be well enough established that it will be the default choice
for modes requiring good linearity — AM,
SSB, and some digital modes.
Frequency Reference
The internal frequency reference is a
±0.1 ppm TCXO, which is excellent for
most amateur work. In a normal working environment, you can expect stability
somewhat better than this. PowerSDR offers an automatic frequency calibration
against an external signal such as WWV,
but you can do better with Fldigi software.
On our unit, I measured an offset of about
0.04 ppm against WWV at 20 MHz.
The ANAN-100D offers the ability to work
with an external 10 MHz oscillator as a reference, which may be helpful if you want to
run a VHF+ station from a common oscillator or if you have a GPS disciplined oscillator. To switch from internal to external
reference requires changing jumpers on the
Angelia board, which (again!) means disassembling the radio. (Apache Labs notes that
the ANAN-200D model does not require
opening the case and moving jumpers for
the 10 MHz reference or mic/bias adjustments.) Apache Labs sells a small add-on
daughter card for the Angelia that provides
automatic switching between internal and
external references, if you need it.
QST ® – Devoted entirely to Amateur Radio
www.arrl.org
October 2015 51
Thermal Design
Apache Labs rates the ANAN-100D for
a maximum continuous output of 30 W,
which could affect long transmissions
on RTTY or FM for example. This is apparently due to the limited air circulation
and use of the radio’s case as a heat sink.
Running at 30 W continuously, the case
gets quite hot to the touch. If you go to the
Yahoo support group, you will find ideas
to improve cooling and increase operating
power, such as drilling holes or adding fins.
But the standard cooling fan is quiet and
fine for lower duty cycle voice, CW, or data
modes running up to 100 W peak.
Conclusions
The ANAN-100D is an advanced SDR
transceiver that developed from the
OpenHPSDR project. It has a strong heritage as an “open” system, where you have
access to all the design information and
where, more importantly, you have access
to advanced users and designers. You are
joining a project almost as much as you are
purchasing a radio. That’s good, because it
means there are a lot of new features being
hatched, and help is easy to find. Some
amateurs, on the other hand, might prefer
the stability and thorough documentation
that you get with more traditional products
and companies.
The ANAN-100D’s operating “look and
feel” is very similar to the last-generation
FlexRadio 3000/5000 SDR radios because
of their shared PowerSDR interface. It is
a mature and time-tested software system
that is very usable.
If you hanker for a high performance SDR
radio and you lean toward open source solutions, the ANAN-100D could be the radio
for you.
Manufacturer: Apache Labs Pvt Ltd, 1023
Tower B4, Spaze I-Tech Park, Sector 49,
Sohna Road, Gurgaon 122001, Haryana,
India; e-mail support@apache-labs.com;
apache-labs.com. Available from several
US dealers.
Notes
1M. Ewing, AA6E, “FlexRadio Systems FLEX6300 Transceiver, FLEX-6700 Transceiver, and
SmartSDR for Windows Software,” Product
­Review, QST, Apr 2014, pp 47 – 59.
2apache-labs.com/productPdf/1031_Angelia.
pdf
3apache-labs.com/download_file.
php?downloads_id=1017
4www.k5so.com/HPSDR_downloads.html
5plus.google.com/107168125384405552048/
about
6openhpsdr.org/wiki/index.php?title=
Ghpsdr3-Qt
7apache-labs.com/instant-downloads.html
8goo.gl/JbC5g5 (pdf)
9groups.yahoo.com/neo/groups/apache-labs/
info
10anan-100d.wikidot.com
11D. Hershberger, W9GR, “Controlled Envelope
Single Sideband,” QEX, Nov/Dec 2014,
pp 3 – 13.
12groups.yahoo.com/group/apache-labs/
files/PowerSDR-Fldigi_Guide_V2.0.pdf
See the Digital
Edition of QST for a
video overview of
the Apache Labs
ANAN-100D SDR
HF/6 Meter
Transceiver and
PowerSDR_mRX_PS.
MFJ-2910 80/160 Meter Matching
Network for 43-Foot Verticals
Reviewed by Phil Salas, AD5X
ad5x@arrl.net
When fed with a 1:4 unun (unbalancedto-unbalanced transformer), the 43-foot
vertical antenna has a reasonable SWR
on 60 through 10 meters, thereby keeping
SWR-related losses low on those bands. On
160 and 80 meters, however, this antenna
provides such a severe mismatch that you
will throw away most of your power in your
coax and unun on these bands because of
the extreme SWR-related losses.
I have previously written articles describing
a base matching unit for the 43-foot vertical that significantly reduces SWR-related
52 October 2015
coax and unun mismatch losses on 160 and
80 meters, while preserving the 60 – 10
meter compromise SWR of the original
antenna.13, 14 For those who do not wish
to undertake that construction project, the
MFJ-2910, shown at the base of the author’s 43-foot vertical in Figure 7, is an option to consider.
Bottom Line
The MFJ-2910 provides a worthwhile addition to 43-foot vertical for
those who wish to use this antenna
on 160 and 80 meters.
ARRL, the national association for Amateur Radio®
First, a Discussion on RF Voltages
The MFJ-2910 is rated for 1.5 kW PEP or
750 W continuous RF power on all amateur
bands from 160 – 6 meters. The power limitations are due to the extremely high voltages and high currents that can be generated
because of the low radiation resistance
and high reactance of this antenna on 160
meters, and the finite Q of the matching inductor. As it is important to understand the
high voltages and currents that can occur,
a discussion of the calculations is in order.
You will be matching your amplifier power
into the sum of the radiation resistance
(Rr), inductive losses (Rind), and ground
www.arrl.org Table 2
160 Meter Inductor Performance
160 meter inductor loss and peak voltage vs ground loss at 1500 W
Assumptions: radiation resistance, 3 Ω; antenna reactance, –j 600 Ω; inductor loss 2 W
Ground Loss RF Current Inductor loss Peak Voltage at Antenna Base
(Ω)(ARMS )(W)
(VPK)
0
10
15
17.32
10.0
8.66
600
200
150
14,700
8,500
7,400
Figure 8 — Inside the MFJ-2910. The unun, relays, and coils are visible. The black material on
some of the relays is a silicone sealant type material that helps support the inductor.
Figure 7 — MFJ-2910 connected to the base
of the reviewer’s 43-foot vertical.
losses (Rg) — Rtotal = Rr + Rind + Rg. Since
Power = I2R, I = P / R total and so inductor
loss = I2Rind. Finally, the peak voltage is
=
Z
R total 2 + X 2 .
VPK = 2 × Z × I where
From EZNEC modeling software, a 43-foot
vertical antenna has an impedance of about
3 – j 600 Ω on 160 meters. And when inputting the coil dimensions into HamCalc
(300 mm length × 45.7 mm diameter, wire
diameter = 1.65 mm, N = 94), the inductor
has a calculated resistive loss of about 2 Ω
on 160 meters.15 Table 2 shows the approximate inductor power loss and antenna base
peak voltage versus ground losses. As you
can see, the inductor loss can get quite high
as ground loss is reduced. However, average
inductor loss will typically be 20 – 30% of
that shown for SSB or CW operation.
MFJ-2910 Discussion
Figure 8 shows a network of 13 relays to
select taps on a large air-wound inductor to
provide 43-foot vertical base-matching on
160 and 80 meters. On 60 – 10 meters the
relay network bypasses the inductor, and
so the input reverts to the original unun-fed
antenna. The relays have contact-to-contact
breakdown voltage ratings of 1000 VRMS,
and contact-to-coil breakdown voltage
ratings of 5000 VRMS. Six sets of relay contacts in series protect against arcing on 160
meters (up to 8500 VPK). And to improve
the contact-to-coil breakdown voltage, the
affected relay coils are disconnected from
RF and dc ground on 160 meters. Further,
a 1⁄4-inch arc-gap built into the MFJ-2910
printed-circuit board may provide some
additional protection, but it will probably
not arc before the relays are damaged. Note
that the relay voltage break-down ratings
will be exceeded at 1500 W if your ground
loss is lower than about 10 Ω and you have
negligible tuner and feed-line losses. In all
fairness, however, achieving 10 Ω ground
loss can be quite a challenge.16
On 160 meters all relays are de-energized,
thus selecting the maximum loading coil
inductance and the correct shunt tap for
lowest SWR. On 80 meters, –12 V energizes relays that bypass much of the
160 meter loading coil and change the shunt
tap for best SWR. For 60 to 10 meters,
+12 V energizes other relays to bypass
the 160/80 meter matching components
and revert the antenna to its normal 60 to
10 meter configuration.
All dc power is provided through the interconnecting RF cable via an included
three position bias-T (MFJ-4118). An isolated 12 V dc power source, such as the
MFJ-1316 or 12 V dc wall transformer,
is required as both positive and negative
voltages must be provided for proper operation. About 60 mA is required for –12 V
(80 meter operation) and +12 V (60 – 10
meter operation). No current is required for
the default (0 V) 160 meter operation.
MFJ-2910 Setup and Adjustments
Because of the very high voltages developed across the antenna base-to-ground
on 160 meters, a new fiberglass-insulated
antenna base is provided as part of the
MFJ-2910 package. Depending on your
antenna base insulator, you may or may not
elect to install this new base.
The MFJ-2910 must be adjusted for proper
operation as each antenna installation will
be a little different. To make the adjustments, remove six externally accessible
screws and two nuts, slide the MFJ-2910
from its housing, and mount it to the antenna base. An antenna analyzer is necessary to determine the proper 160 and
80 meter resonance frequencies and best
SWR. A #2 Phillips screwdriver is used to
change the inductor tap positions. You will
need to take the MFJ-4118 bias-T and a battery pack to the antenna base to enable the
different bands while you make the necessary adjustments.
Note: Incorrectly connecting the bias-T can
damage your antenna analyzer by applying
QST ® – Devoted entirely to Amateur Radio
www.arrl.org
October 2015 53
performance I achieved (again, based on
my ground loss of about 13 Ω). With an
SWR of 3:1 or less across the band, SWRrelated losses are negligible.
Some Final Observations
The maximum power I have available is
1200 W, and no arcing occurred at this
power level. It was nice to not even need
an antenna tuner on my selected 160 and
80 meter CW frequencies! Of course, an
in-shack antenna tuner will be needed
with a solid-state amplifier for extendedfrequency operation on 160 and 80 meters,
and also for 60 – 10 meters with the 43-foot
vertical’s standard unun matching network.
(A)
Finally, because of the design of the
MFJ-2910 you won’t have to worry about
moisture collecting inside the unit. Any
moisture will easily drain out through the
bottom/connector side of the unit. My only
concern is that the holes in the bottom plate
may be large enough for some insects to
move into the unit over time. Ants are a big
problem in the Dallas, Texas, area where
I live.
(B)
Figure 9 — MFJ-2910 160 meter SWR performance is shown at (A), and the 80 meter SWR
performance is shown at (B).
±12 V dc to the analyzer’s RF port. You
must connect the MFJ-4118 RF/DC OUT/IN
connector to the MFJ-2910 RF input. The
antenna analyzer connects to the MFJ-4118
RF IN/OUT connector.
I found that the cover raised the 160 meter
center frequency by 50 kHz, and the
80 meter center frequency by 25 kHz. Keep
this in mind when selecting your center frequencies during the tuning process.
There are four inductor taps that must be
set for proper operation. One tap sets the
160 meter center frequency, one tap sets
the 80 meter center frequency, one tap sets
the 160 meter SWR, and one tap sets the 80
meter SWR.
Start the tuning process on 80 meters by
setting the bias-T to –12 V. Determine the
lowest SWR frequency and move the top
tap on the short coil to move the resonant
frequency. Moving the tap toward the
ground end of the coil raises the resonant
frequency. Next select 160 meters by setting the bias-T off, or just unplug the dc
source. Determine the lowest SWR frequency and adjust the tap on the long coil.
Again, moving the tap toward ground raises
the resonant frequency.
I have done a lot of experimenting with my
own 160/80 meter matching unit, and I’ve
found that the input SWR tap points are
quite consistent from unit-to-unit, antennato-antenna, and installation-to-installation.
With the MFJ-2910, connect the 80 meter
input tap (the closest tap to ground) to the
8th wire from the ground end of the coil
(the count should include the bottom wire).
Then connect the 160 meter input tap point
(the next tap point above the 80 meter tap)
to the 13th wire from the ground end of the
coil — five wires above the 80 meter tap. I
believe that if you start with these input tap
settings, they will require little or no further
adjustment.
The resonant frequencies will change a bit
when the MFJ-2910 cover is re-installed.
54 October 2015
Figure 9A shows the measured SWR performance of the MFJ-2910 on 160 meters.
As I operate mostly CW, I centered the unit
between 1800 – 1850 kHz. As you can see,
the SWR stays below 2.5:1 over this range,
resulting in negligible SWR-related coax
and unun losses when using RG-213 or better quality coaxial cable. Incidentally, my
measured ground loss is 13 Ω — a lower
ground loss will result in a narrower SWR
bandwidth.
Figure 9B shows the 80 meter SWR
ARRL, the national association for Amateur Radio®
Conclusion
For those who want to operate 160 and
80 meters with a 43-foot vertical, the
MFJ-2910 is one solution worth considering. It eliminates the severe SWR-related
coax and unun losses normally associated
with this antenna on those bands. However,
while the MFJ-2910 will handle the extremely high RF voltages that will occur
with this electrically short antenna when
mounted over a typical ground system, very
well-designed ground systems that achieve
less than 10 Ω ground loss may result in
voltage breakdown within the MFJ-2910
when running full legal limit.
Manufacturer: MFJ Enterprises, 300
Industrial Park Rd, Starkville, MS 39759;
tel 662-323-5869; www.mfjenterprises.
com.
Notes
13P. Salas, AD5X, “160 and 80 Meter Matching
Network for Your 43 Foot Vertical,” QST, part 1,
Dec 2009, pp 30 – 32; part 2, Jan 2010, pp 34
– 35.
14P. Salas, AD5X, “160- and 80-Meter Matching
Network for Your 43-foot Vertical – UPDATED,”
available from www.ad5x.com (look in the
Articles section).
15hamwaves.com/antennas/inductance.html
16R. Severns, N6LF, “Vertical Antenna Ground
Systems at HF,” available from www.kkn.net/
dayton2004/HF_vertical_ground_
system_design_N6LF_Dayton.pdf
www.arrl.org 
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