NEWS BRIEFS IN-DEPTH ARTICLES NEW PRODUCTS 2

NEWS BRIEFS IN-DEPTH ARTICLES NEW PRODUCTS 2
Volume Thirty-Nine
NEWS BRIEFS
IN-DEPTH ARTICLES
NEW PRODUCTS
2
Implementing a trimless IF VCO (Part 1)
3
Managing noise in cell-phone handsets
8
Data Converters
• 10-bit ADC with track/hold converts at 20Msps
• 10-bit, multichannel, parallel-I/O ADCs fit in QSOP packages
and more
(MAX1425)
(MAX1090–93)
11
11
(MAX4076/77/78)
(MAX4291/92/94)
12
12
(MAX4624–28)
(MAX4534/35)
14
14
(MAX1790)
(MAX1698)
14
15
(MAX1665)
16
(MAX6351–60)
(MAX6361/63/64)
17
17
(MAX3188E/89E)
(MAX3316E–19E)
18
18
(MAX1608/09)
19
(MAX1740/41)
19
Amplifiers and Comparators
• 40µA rail-to-rail op amps sustain 95dB AVOL with 10kΩ load
• 1.8V op amps deliver 120dB AVOL and drive 2kΩ loads
and more
Analog Switches/Multiplexers
• Analog switches guarantee 0.5Ω at +5V (world’s lowest RON)
• Single 4-to-1/dual 2-to-1 muxltiplexers feature rail-to-rail
fault protection
and more
Power-Management ICs
• Low-noise step-up converter is 90% efficient
• High-efficiency current source drives white-LED backlighting
and more
Battery Management
• Li+ battery-pack protector blocks voltage and charge faults
Supervisors
• µP supervisors monitor three supply voltages
• Backup-battery reset ICs offered in SOT packages
and more
Interface ICs
• ±15kV ESD-protected RS-232 transmitters in SOT packages
• 2.5V/1mA RS-232 devices protect against ESD to ±15kV
and more
I/O Expanders
• Octal serial-parallel I/O expander suits SMBus systems
Level Translators
• World’s smallest SIM/smart-card interfaces operate below 1.5V
Sensor-Signal Conditioner
• Signal conditioner enables highly accurate sensors and 4–20mA output (MAX1459)
Temperature Sensor ICs
• Resistor-programmable temperature switches combine flexibility
with simplicity
(MAX6509/10)
20
(MAX6101–05)
21
(MAX2360/62/64)
(MAX2683/84)
21
22
(MAX3665)
22
(MAX3770/71)
23
Voltage References
• 0.5%, SOT23 series references at shunt-reference prices
Wireless ICs
• Highly integrated transmitter ICs serve dual-band cellular phones
• 3.5GHz, SiGe active mixers deliver +11dBm IP3 performance
and more
20
Fiber-Optic ICs
• 3V, 622Mbps, low-power transimpedance amplifier provides
high sensitivity
• 2.125Gbps, 3.3V Fibre Channel repeaters tolerate 0.7UI jitter
and more
News Briefs
MAXIM REPORTS RECORD REVENUES AND EARNINGS FOR THE THIRD
QUARTER OF FISCAL 2000
Maxim Integrated Products, Inc., (MXIM) reported record net revenues of $226.5 million for its fiscal
third quarter ending March 25, 2000, a 53.9% increase over the $147.2 million reported for the same quarter a year
ago. Net income increased to a record $74.7 million in the third quarter, compared to $47.7 million last year, a
56.7% increase. Diluted earnings per share were $0.23 for the third quarter, a 53.3% increase over the $0.15
reported for the same period a year ago.
During the quarter, the Company increased cash and short-term investments by $97.7 million after paying
$33.9 million for capital equipment and repurchasing 100,000 shares of its common stock for $5.0 million. The
Company generated higher than normal cash balances principally due to lower than normal stock repurchases.
Accounts receivable increased by $21.7 million in the third quarter to $125.1 million, and inventories increased
$7.0 million to $51.4 million during the quarter.
Gross margin for the third quarter increased slightly to 69.9%, compared to 69.8% in the second quarter.
Research and development expense was $36.4 million or 16.1% of net revenues in the third quarter, compared to
$32.3 million or 16.0% of net revenues in the second quarter. During the quarter, the Company recorded a writedown
of equipment of $8.8 million to cost of goods sold and recorded a charge to selling, general and administrative
expenses of $4.5 million related to technology licensing. Due to recent accounting changes, the Company recorded
an expense in the third quarter of approximately $5.7 million in medicare taxes on employees’ realized stock option
gains. The tax payments were previously recorded within Stockholders’ Equity as an offset against the proceeds
received from the exercise of stock options. Included in interest income and other, net, was a $4.5 million gain from
the cash sale of the Company’s 50% interest in its high-frequency packaging and assembly business. This business
was jointly owned with Tektronix and was set up to facilitate the 1994 acquisition of the Tektronix integrated
circuit operation.
Bookings in the quarter were approximately $304 million, an 8% increase over the previous quarter’s level
of $283 million, and a 78% increase over the third quarter of last year. Turns orders received in the quarter were
$95 million, a 3% increase over the $93 million received in the prior quarter (turns orders are customer orders that are
for delivery within the same quarter and may result in revenue within the same quarter if the Company has available
inventory that matches those orders). The bookings increase is primarily attributable to strength in the U.S. distribution
channels and Europe. There was significant growth in bookings for the Company’s products targeted for the highfrequency end markets and products with broad-based industrial applications. Bookings on Maxim by U.S. distributors
were $78.8 million during the quarter and exceeded customer bookings on those distributors by $17.6 million.
Bookings on distributors by their customers were up by $14.3 million, a 30% increase over the previous quarter.
Third quarter ending backlog shippable within the next 12 months was approximately $345 million, including
approximately $271 million requested for shipment in the fourth quarter of fiscal 2000. The Company’s second
quarter ending backlog shippable within the next 12 months was approximately $285 million, including approximately $231 million that was requested for shipment in the third quarter. All of these backlog numbers have been
adjusted to be net of cancellations and estimated future U.S. distribution ship and debit pricing adjustments.
Jack Gifford, Chairman, President, and Chief Executive Officer, commented on the quarter: “This was
another strong quarter for Maxim, with record bookings, revenues, and earnings. Relative to many younger technology companies, Maxim continues to offer investors a finite price to earnings ratio. Our annualized bookings rate
was over $1 billion. Turn orders as a percentage of total bookings were in line with historical levels, at approximately 31%. We continue to believe that bookings and turns growth will moderate to reflect our short-term forecasted end-market consumption levels. During the past 12 months, we have increased our manufacturing capacity
to support current and projected consumption trends.”
Gifford continued: “I have never been more encouraged about the broad acceptance of Maxim’s products
in the global marketplace. Clearly our products, and the engineering value they represent, are helping to fuel the
worldwide growth in digital communications and microprocessor-based electronics. It is my opinion that worldwide
economies depend on state-of-the-art technologies, including Maxim’s, that are dominated by U.S. corporations. I
believe this to be a long-term growth engine for Maxim and other U.S. corporations offering these enabling technologies.”
Gifford concluded: “During the quarter, our professional recruiting efforts were highly successful, indicating Maxim’s continuing appeal to world-class technologists in all disciplines.”
the design of a trimless IF VCO, seeks to develop an
appreciation for the magnitude of the task.
Implementing a
trimless IF
VCO (Part 1)
VCO topology
While several oscillator topologies are viable for
construction of a practical RF VCO, the one that has
proven successful in many commercial VCO modules
and countless discrete VCO circuits is the Colpitts
common-collector topology (Figure 1). This topology is
useful for a wide range of operating frequencies, from IF
to RF.
Part 1 of this two-part article explores the design fundamentals needed to implement a trimless, fixed-frequency,
IF voltage-controlled oscillator (VCO) and points out
the challenges in guaranteeing proper circuit operation.
VCOs are essential components in the architecture of
most wireless systems. In dual-conversion approaches, a
fixed-frequency IF VCO is required to control the
frequency translation from IF to baseband and/or
baseband to IF.
A flexible, low-cost, and reasonably high-performance
VCO may be constructed with an inductor-capacitor
(LC) tank circuit consisting of a low-cost surface-mount
inductor and varactor diode. The oscillator tank is a
parallel-resonant circuit controlling the oscillation
frequency; any change in the inductor or capacitor
changes the oscillation frequency. The inductor and
varactor can implement the variable resonance as a
parallel- or series-mode network.
Dual-conversion systems require two oscillators.
Typically, the first (RF VCO) tunes over the full range
of input channel frequencies, and the second (IF VCO)
operates at a single frequency established by the
frequency plan. The RF VCO is available as a module,
IC, or discrete-component circuit, with modules and ICs
being more common. For IF VCOs, small, cost-effective
modules are nearly absent from the market. Probable
reasons include the need for many arbitrary IF frequencies and the need for large-valued inductors that cannot
be laser-trimmed (adjusted) in production. As a result,
the IF VCO is usually implemented as a discrete circuit
or as part of an IC.
The parallel-mode network may be used at lower
frequencies where large-value varactors are impractical
and the inductor value can be made larger. The parallelmode configuration also permits a straightfoward
analysis of the oscillator. For the remainder of this
article, the trimless IF VCO will be illustrated with a
Colpitts-style oscillator, using a parallel-mode LC tank
(Figure 2).
The Colpitts oscillator is discussed in several textbooks
(Clarke and Hess 1978, Hayward 1994, Rohde 1998),
and various equations have been derived to predict the
behavior of oscillators in general and the Colpitts
topology in particular. The oscillator is generalized with
a feedback-amplifier model of the circuit. Expressions for
the exact oscillation frequency may be derived by equating
Maxim has pioneered a new VCO IC intended for use in
wireless systems whose other board-level RF/IF ICs lack
that function. Part 2 of this article will introduce the IC,
discuss its development, and detail the simple and costeffective applications it makes possible.
A discrete-component VCO offers sufficient degrees of
freedom to meet the performance objectives of most
systems (tuning range, output power, phase noise, current
consumption, cost, etc.). For higher volume, cost-sensitive
modern products, however, production-line adjustment of
the oscillation frequency is not acceptable. The RF
engineer is therefore compelled to devise a VCO that
requires no adjustments during assembly, i.e., a trimless
VCO. The design is not trivial. In addition to an understanding of VCO design fundamentals, it requires
substantial RF engineering effort to ensure that the design
is properly centered and that the oscillator tunes to the
desired frequency over all allowed variations in
component values, temperature, and supply voltage. The
following discussion, while explaining pertinent issues in
TANK
Figure 1. The basic Colpitts oscillator
3
impedances in that model, but those expressions are
cumbersome and provide little insight into the design
process.
VCC
CC
RB
VBIAS
Alternatively, the Colpitts oscillator can be analyzed in a
simpler but less accurate manner, which provides a set of
design equations that are clearer, more insightful, and
useful for first-order oscillator design. First, the Colpitts
oscillator may be redrawn as an LC amplifier with positive
feedback (Figure 3). This view is useful in calculating the
loop gain, oscillation amplitude, and phase noise. To
predict startup behavior and oscillation frequency, the
original circuit can also be redrawn as a negative
impedance plus resonator structure (Figure 4). Equations
from these two views are combined as a set of governing
equations for the Colpitts oscillator (Meyer 1998).
CO
C1
TO OUTPUT
BUFFER
L
IQ
C2
CVAR
Figure 2. Use of the Colpitts topology in a VCO
Basic design equations for the Colpitts
oscillator
VCC
Ignoring parasitic elements, the basic equations for this
analysis assume that CC >> C1 and C2, and C1 > Cπ (Cπ
is the base-emitter capacitance). Calculate the oscillation
frequency (fO) as follows:
fO =
1
, C T = C V + C12 ,
2π L ✕ CT
CV
L
REQ
C2
(1)
C
✕ CO
C ✕ C2
C V = VAR
, C12 = 1
C VAR + C O
C1 + C 2
C1
IQ
Calculate the quality factor of the resonant tank circuit
(QT) as follows:
QV =
1
2 π ✕ C V ✕ R S ✕ fO
QT =
R EQ
2 π ✕ L ✕ fO
, R QC = Q V 2 ✕ R S ,
Figure 3. LC amplifier model
(2)
, R EQ = R QL R QC
Estimate the oscillation amplitude as follows:
VO ≅ 2 ✕ I Q ✕ R EQ ✕
J1 (β)
,
J 0(β)
VO ≅ I Q ✕ R EQ ✕ 1.4
jXL
jXC
(3)
RL
OSCILLATOR
DEVICE
Calculate the loop gain and startup criteria as follows:
C + C2
1
Loop gain = g m ✕ R EQ ✕ , where n = 1
n
C2
(4)
REQUIRES RL < RN
Startup criteria :
R EQ
gm
>>
,
(2 π ✕ fO ✕ C1 ) (2 π ✕ fO ✕ C 2 )
QT2
∫
(5)
Figure 4. Reflection amplifier model
minimum 2 : 1 ratio
4
-RN
A wider tuning range causes greater oscillator phase noise
through two well-understood phenomena: a reduction in
the tank-circuit Q and noise on the tuning line. To achieve
a wider tuning range, the varactor must be coupled more
heavily into the tank circuit. This coupling reduces the Q of
CV (the effective variable capacitance) as shown in Equation 2. Lower Q for CV reduces the net Q of the tank and,
consequently, increases the phase noise, per Equation 6.
Calculate the Colpitts oscillator phase noise (PN) at an
offset frequency (fm) from the carrier as follows:
PN = i n
2
1  fO 
✕
✕

VO 2  2Q O 
2
✕
R EQ 2
fm
(6)
Trimless VCO approach
The second factor in reducing phase noise is thermal
noise on the tuning input, which creates FM-sideband
noise. This noise increases with tuning range, and it can
exceed the oscillator’s inherent phase noise. The phase
noise induced by thermal noise is given by:
Developing a trimless VCO is relatively simple in concept.
Oscillator-frequency adjustments can be eliminated if the
oscillator has sufficient extra tuning range to overcome all
the error sources (e.g., component tolerances) that produce
shifts in frequency. At first glance, it may seem intuitive
and simple enough just to provide plenty of oscillator
tuning range and tune out all the error sources. For a given
tuning-voltage range, however, finite variable capacitance
imposes a fundamental limit on the frequency-tuning range,
and the VCO’s electrical-performance requirements often
constrain the tuning range before that limit is reached.
 2 ✕ K V ✕ Vn 
PN = 20 log 
 , where K V =
2 ✕ fm


Hz
, Vn = noise density
V
V
at VTUNE input at fm in
Hz
VCO gain in
Unfortunately, several negative consequences attend an
oscillator with excessive tuning range. Very wide ranges
require heavy capacitive coupling of the varactor to the tank,
which substantially reduces the tank-circuit Q. The result is
greater phase noise (reduced tank amplitude vs. transistor
noise), greater sensitivity to tuning-line noise (which translates directly into frequency modulation), the possibility of
too much voltage swing across the varactor, potential startup
problems, and greater challenge in designing the loop filter.
These factors lead to the conclusion that excessive tuning
range is undesirable. Indeed, it should be no greater than the
minimum necessary to absorb all error sources.
(7)
It is evident in both cases that phase noise degrades with
increasing tuning range. To preserve low phase noise in
a trimless VCO, therefore, it is critically important to
provide just enough tuning range to meet the guaranteed
bandwidth and accommodate the expected error sources.
As the varactor is coupled-in more heavily, more tankvoltage swing appears across the varactor, and the varactor
voltage swing must be limited to avoid forward-biasing the
varactor. This sets a limit on signal power in the tank and,
consequently, on the oscillator’s phase noise. Finally,
startup problems may occur if the tank-circuit equivalent
series resistance (ESR) becomes too high (refer to the
equations). A VCO with very wide frequency-tuning range
may not start up properly, especially over the extremes of
temperature. With the goal of providing just enough tuning
range, the question is—how much?
Glossary
CO = varactor coupling capacitance
CT = total tank capacitance
CVAR = varactor capacitance
fm = offset frequency of PN in Hz
fO = frequency of oscillation
Error sources in the oscillation frequency
gm = bipolar transistor (oscillator) transconductance
The trimless VCO’s frequency tuning range is increased
to accommodate error sources in the oscillation frequency.
These error sources fall in two categories: error in the
component values and error from design centering. The
LC components that set the oscillation frequency are not
ideal, of course. They contribute the following:
in = collector shot noise
IQ = oscillator transistor bias current
QL = inductor Q
QT = tank Q
QV = effective varactor Q
REQ = equivalent tank parallel resistance
• Part-to-part variations (tolerance)
RS = varactor series resistance
• Non-ideal performance (limited frequency response
due to inductance, capacitance, and series resistance in
the leads)
VO = RMS tank voltage
5
The quickest way to compute net frequency skew due to
the various errors is to utilize a spreadsheet program that
contains the detailed formula for oscillation frequency,
based on L and C values in the circuit.
• Error induced by parasitic capacitance and inductance
in the circuit layout
On the other hand, design-centering errors result from
uncertainty in centering the VCO tuning range during
the design process.
Frequency shifts and tuning range
Component-tolerance error
The frequency-tuning range, obtained by varying the
tuning voltage from V TUNE(LOW) to V TUNE(HIGH), has
high- and low-frequency endpoints (fHIGH and fLOW)
with a “center” frequency (f CENTER ) defined as the
midpoint between fHIGH and fLOW (Figure 5). Ideally,
the tuning range should be positioned with fCENTER at the
desired oscillation frequency (Figure 5a). However,
component errors and design-centering errors can shift
the frequency-tuning limits.
Each capacitive and inductive component affecting the
oscillation frequency in an LC oscillator has only limited
part-to-part accuracy, and this tolerance error contributes
to error in the oscillation frequency. Table 1 lists the
typical tolerances for the frequency-setting components
in the oscillator.
Table 1. Oscillator frequency-setting
component tolerances
Component
The desired oscillation frequency cannot be reached if
the system provides inadequate tuning voltage over the
worst-case conditions, which results in insufficient
frequency range (Figure 5b). Clearly, a careful determination of the required tuning range is necessary. That is
accomplished by calculating the frequency skew caused
by all error sources, and verifying that fLOW < fOSC and
fHIGH > fOSC under the worst-case conditions (Figure 5c).
Tolerance
±15% at VTUNE = 0.4V,
±10% at VTUNE = 2.4V
±5%
Varactor
Inductor
Capacitors
±5%
Parasitic Capacitance
±10%
Parasitic Inductance
±6%
Verification of the design
Oscillator-Device
Impedance
±15%
Once circuit-board layout and component value selection
are complete, the design requires verification and measurements (even more than most RF circuits). Nominally, you
must check the tuning range, startup behavior, phase noise,
etc., for compliance with design targets. In addition,
measurements must be made over a statistically significant
number of manufacturing runs to determine the tuning
range and the mean center frequency, and its location with
respect to the desired oscillation frequency.
Design-centering error
Design centering is often overlooked as a source of error
in establishing the oscillation frequency. To maximize
use of the available frequency-tuning range, the tuning
limits must be symmetric with respect to the desired
oscillation frequency. Any error in establishing this
center point, caused by inaccuracies in modeling the
components’ initial or mean values, reduces the tuning
range available to absorb error sources. To guarantee the
oscillation frequency over all conditions of temperature,
supply voltage, component tolerances, etc., the tuning
range must be wide enough to accommodate this error.
fOSC
a)
fLOW
b)
1
, CT = C V + C12 ,
2 π L x CT
fCENTER
c)
fHIGH
LOW SKEW
fHI
You can calculate total frequency error using the
frequency-of-oscillation formula, by multiplying each
element by a variation scaling factor:
fO =
NOMINAL
HIGH SKEW
fLOW
fHIGH
LOW SKEW
HIGH SKEW
fLOW
FREQUENCY
a) NOMINAL FREQUENCY RANGE, PERFECTLY CENTERED
b) INSUFFICIENT TUNING RANGE, FREQUENCY SKEW LOW AND HIGH
c) SUFFICIENT TUNING RANGE, FREQUENCY SKEW LOW AND HIGH
(8)
C
✕ CO
C ✕C
, C12 = 1 2
C V = VAR
C VAR + C O
C1 + C 2
Figure 5. Tuning range and frequency shifts
6
All this work is necessary to produce a robust, reproducable design with the desired electrical performance.
Because the tasks usually require several iterations, you
can easily take months to achieve a discrete-component
design that is acceptable and production worthy.
Development of a trimless IF VCO requires a detailed
circuit design, inclusion of all error sources, verification
on the circuit board, and monitoring over production to
ensure a viable result. Maxim has met this challenge
with a new IC (to be described in Part 2), which solves
the VCO design problems while dramatically reducing
the time necessary to implement a trimless IF VCO.
References
Clarke, Kenneth, and Donald Hess. 1978.
Communications Circuits: Analysis and Design.
Chap. 6.
Hayward, Wes. 1994. Radio Frequency Design.
Chap. 7.
Meyer, Dr. Robert. 1998. Internal communication.
Rohde, Ulrich. 1998. Microwave and Wireless
Synthesizers. Chap. 4.
Part 2 of this article will introduce the IC, discuss its
development, and present a detailed description and
performance summary (Engineering Journal Vol. 40). An
application that illustrates the simplicity, small size, and
cost effectiveness of the device will also be included.
7
Noise-propagation mechanisms
Managing noise
in cell-phone
handsets
Noise propagates by conduction and radiation. Conduction
channels noise through a wire, printed-circuit trace, metal
chassis, or electrical component. Radiation transfers noise
energy through the air or other dielectric such as circuitboard material. Conducted noise can be filtered with traditional circuit techniques. Radiated noise, if not reduced at
the source, requires shielding. Conducted noise that finds
an efficient antenna becomes radiated noise. Although
radiated noise is often controlled with shielding, conductive coatings, and gaskets, these measures may be unnecessary if noise is confined to the conductive mode by
proper PC layout and filtering. It is best, if possible, to
keep noise conductive and not let it radiate.
Modern cell phones must operate in the face of many
unwanted signals. In a typical phone, the signal amplitude
may be only 0.35µV—over 100dB below the amplitude of
nearby noise. To amplify this signal to a level suitable for
demodulation, cell phones often use intermediate
frequency (IF) sections with over 80dB of gain.
To meet the required bit-error rate (BER), noise must be
managed. Shielding and filtering are effective but add
extra weight, size, and cost while shortening battery life.
A better approach is to design for low noise from the start
so that known noise spectra don’t interfere with radio
performance. Managing noise in this way requires an
understanding of the following:
The power amplifier
A PA generates noise by drawing large currents. A 3.6V,
50% efficient PA whose signal faces a 3dB loss before
reaching the antenna can draw up to 800mA from a
lithium ion (Li+) cell. This current flows through resistance in the battery connectors, PCB traces, and ground
paths, producing noise on power-supply lines. The
problem is compounded in phones that use burst transmission as specified by the GSM and IS-136 TDMA
standards. Short PA bursts impose transients on the power
supply and the distribution subsystem.
• Noise-propagation mechanisms
• Points of greatest noise sensitivity
• Noise-generating circuits.
Cell-phone handsets
A digital cell phone is a marvel of packaging, human
interface, and power conservation. The RF sections
consist of filters, low-noise amplifiers, mixers, a power
amplifier (PA), and a frequency synthesizer. A mixedmode ASIC connects transmit and receive sections to IF
signals. Working in concert with a digital ASIC
containing DSP and system-control processors, the
mixed-mode ASIC contains data converters for modulating and demodulating the IF signals. The systemcontrol processor also manages the human interface and
power management.
A popular method for powering the burst-mode PAs is to
boost the supply voltage to reduce peak current, minimize
noise, and allow less expensive PA technology. Still, the
Li+
PROTECTION
BOOST
SMPS
P-CH
S
ENABLE
BUCK
SMPS
2.0V
DIGITAL ASIC
LDO1
FREQUENCY
SYNTHESIZER
LDO2
RECEIVE
RF
LDO3
TRANSMIT
LDO4
STANDBY
POWER
ENABLE
A power-distribution subsystem manages the battery
pack and distributes operating voltages within the
handset (Figure 1). Cell phones may also include a
switch-mode power supply to boost the cell voltage to a
level appropriate for the PA. New low-voltage ASICs
can receive power from a small, step-down switch-mode
supply, and remaining RF and analog circuits can be
powered from linear low-dropout (LDO) regulators. The
various regulators turn on under processor control,
selecting operations demanded by the wireless protocol
(GSM or IS-95, for example).
ENABLE
PA
S
ENABLE
Figure 1. Switch-mode and linear regulators distribute power cleanly
and efficiently.
8
need to supply a current peak often results in an overspecified boost converter. A better solution is to store the
boosted energy on a capacitor; then the boost converter need
only recharge the capacitor between transmitter bursts. A
remaining problem occurs with a typical DC-DC converter
when a capacitor voltage drop is sensed; it attempts to
replenish the charge as quickly as possible, drawing a current
surge from the Li+ cell that creates new noise. A solution
incorporated into new ICs (MAX1687/MAX1688) limits the
PA reservoir capacitor charge rate either with a user-set peak
battery current or an automatically set adaptive current limit
(Figure 2). As a result, the capacitor and power converter
work in concert to maintain efficient power conversion while
minimizing the system disruptions that can accompany PA
current surges. To further control noise, these chips can be
disabled during transmit bursts.
sidebands at the RF output. The MAX881 also senses the
negative bias voltage to ensure that drain current is
controlled when the PA’s main supply is applied. This
safety interlock prevents PA damage (Figure 3).
PA bias
The VCO gain is 25MHz/2V or 12.5MHz/V. High gain
makes the VCO sensitive to noise on the control line. If the
phase detector and VCO are separated in a high-gain PLL,
the VCO often picks up radiated noise, requiring a shielded
cable to preserve the VCO noise spectrum. The following
are other disturbances that can also modulate the VCO:
PLL frequency synthesizer
In many cell phones, the first local oscillator (LO) is
generated by a phase-lock loop (PLL) frequency synthesizer. In AMPS phones, the voltage-controlled oscillator
(VCO) tunes in 30kHz steps over a ±12.5MHz range near
880MHz. (Actual VCOs generate frequencies offset by
the first IF.) If you assume the PLL operates from 3V, the
entire 25MHz tuning range should be covered with a 2V
tuning voltage (control voltage). This provides a margin
that ensures the PLL won’t saturate in response to transients or temperature drift.
The bias voltage on a GaAs-FET PA controls the bias
current, which sets PA gain and output impedance. Since
the bias pin is an amplitude-modulated input, any bias
noise appears at the RF output and is radiated from the
antenna along with the desired signal. GaAs PAs use
depletion-mode MOSFETs, which conduct maximum
drain current without gate bias. To control drain current,
the gate must be negative (below ground). To produce a
bias voltage that’s stable, quiet, and well defined, it’s
common practice to follow an inverting charge pump with
an op amp regulator. Though flexible, this approach
doesn’t yield the physically smallest circuit.
1) Power-supply noise, injected into the PLL phase
detector
2) Power-supply noise injected into the VCO
3) Power-supply noise passed to the output of an active
integrator or loop filter (watch op amp PSRR to
minimize this)
The smallest available circuit for generating PA bias is the
MAX881, which combines an inverting charge pump and
negative regulator in a tiny 10-pin µMAX package. All
bias needs for a GaAs PA are addressed in this IC. For
normal operating conditions, output noise and ripple
(~1mVp-p) are low enough to prevent unwanted noise
LI+
SHDN
C1+
C1
C1-
P-CH
IN
CHARGE PUMP
C3
NEGOUT
C4
C2
IN LX1
FB
LX2
5V
2A PEAK
OUT
R1
MAX1687
RF
IN
FB
ON
GND
RESERVOIR
CAPACITORS
S
PA
0.5V
S
PA
GaAs
MAX881R
R2
ON
RF
TO
ANTENNA
BIAS
Li+
OFF
µC
OUT
L1
450µA (MAX)
PA POWER
SWITCH
POK
0.4625V
CHG
GND
Figure 2. For burst systems such as IS-136 and GSM, large transients
on the battery are minimized by reservoir capacitors and a
boost converter.
Figure 3. An interlock feature of the MAX881R protects the GaAs PA
from damage.
9
4) Noise on the crystal oscillator (TCXO/VCTCXO). The
oscillator signal in high-Q circuits should be clean and
noise free, but excess power-supply noise can raise the
oscillator’s noise floor. Because the PLL multiplies
noise within the loop bandwidth by the PLL division
ratio (~30,000 for an AMPS handset), the frequency
synthesizer is very sensitive to noise from the TCXO.
Improved efficiency
The latest switch-mode power supplies (SMPS) designed
for cellular phones offer small size, high efficiency, lowdropout voltage, small external components, and noisecontrol features. For example, the MAX1692 step-down
power converter uses pulse-width modulation (PWM) and
synchronous rectification to achieve over 90% efficiency
with a low, predictable noise spectrum. Operating from a
single Li+ cell ranging from 3V to 4.2V, it generates
supply voltages down to 1.25V for large ASICs.
5) Noise caused by VCO output load impedance variations that reflect back into the VCO and pull its
operating frequency
For systems where loop bandwidth shapes the noise
spectrum to fall between DC and 500kHz, items 1 to 4
can be improved with passive filtering. The frequency
synthesizer needs a separate LDO to avoid noise
conducted from the power supply. Nevertheless, for
digital phone systems, residual phase noise caused by
modulation by the power supply is too great. An LDO
provides a clean regulated supply voltage for the
frequency synthesizer, but it also can produce noise.
To control interference in high-gain RF sections such as
the IF, the MAX1692 can be synchronized (at frequencies
between 500kHz and 1MHz) with an external crystalcontrolled clock such as that generated by the TCXO.
High-frequency SMPS operation is crucial for the use of
small external components and noise-spectrum planning.
Switch-mode supplies produce a noise spectrum in which
the lowest frequency is the SMPS’ fundamental
switching frequency. The spacing between harmonics is
equal to this fundamental, but other aspects of the
spectrum are difficult to predict. Noise power distributed
among the harmonics is a function of wave shape (vs.
time), current level, inductor value, capacitor values, and
PCB layout.
Broadband noise source
An LDO regulator’s voltage reference and error amplifier
can have significant noise content. A low-noise device
like the MAX8877 combats this by bringing the reference
voltage out to a pin that allows bypassing the noise to
ground with a capacitor. A 0.0µF capacitor, for instance,
lowers output noise to 30µVRMS over a 10Hz to 100kHz
bandwidth (Figure 4). This improvement lowers PLL
noise at 900MHz by up to 20dB. LDOs also isolate
handset sections from each other. Within the LDO
bandwidth, the MAX8877 suppresses power-supply noise
at 10kHz by 60dB. In terms of PCB area, this suppression
is a bargain (the IC comes in a SOT23 package). Passive
components that provide such filtering, especially at low
frequencies, would be much larger.
IN
P- CH
OUT
Li+
SHUTDOWN
Switching noise can be conducted on the input, output,
and ground lines, or radiated by the PCB traces. Always
minimize the ripple and noise conducted from an SMPS,
but also realize that adding filter networks to reduce
conducted noise may actually increase radiated noise.
Such noise radiates from the layout and then propagates
efficiently throughout the system, appearing to come from
everywhere.
To best handle the problem of cell-phone noise, understand the phone’s noise-coupling mechanisms, noisesensitive circuit nodes, and noise-generating circuits. A
boost power converter and a large capacitor can minimize
the conducted noise from PA transients in a GSM/TDMA
system. The radiated noise from an SMPS depends
heavily on the PC layout, and a realistic schematic representation can guide the layout for first-time success. Small
linear regulators provide active noise filtering and, with
reference bypassing, can yield the very low noise levels
required by frequency synthesizers. Finally, placing an IF
in the quiet zone between noise harmonics of the power
supply can eliminate the signal contamination that spoils
bit-error rates in a digital cell phone. To allow the most
effective trade-offs, these noise-planning steps should be
considered early in the design.
REGULATED
OUTPUT
VOLTAGE
CONTROL
VREF
1.25V
CB
BYPASS CAPACITOR
REDUCES NOISE
FROM REFERENCE
Figure 4. Output noise from an LDO regulator is reduced by adding
a bypass capacitor (CB) to the voltage reference.
10
NEW
10-bit ADC with
track/hold converts
at 20Msps
The MAX1425, a 10-bit, analog-to-digital
converter (ADC) with 20Msps digitizing
rate, targets imaging, high-speed communication, and instrumentation applications
that require wide bandwidth, good
linearity, and excellent dynamic performance. Unlike other high-speed 10-bit
ADCs, this monolithic device achieves a
full 61dB signal-to-noise plus distortion
ratio (SINAD) and a 72dB spurious-free
dynamic range (SFDR) at 2MHz input
frequency. It achieves this performance
over a ±2V input range, while operating
from a single 5V supply, through use of a
fully differential pipeline architecture and
advanced 0.6µm CMOS process.
10-bit,
multichannel,
parallel I/O ADCs
fit in QSOP
packages
The MAX1090–MAX1093 10-bit,
400ksps, parallel-interface ADCs feature
an 8-bit interface, 3V or 5V single-supply
operation, an internal reference, and four
or eight input channels. The devices’ small
footprint and low supply current are well
suited for battery-powered applications.
S
A fully differential track/hold amplifier
(T/H) ensures wideband dynamic performance with a low 5pF input capacitance
and 150MHz full-power bandwidth. The
MAX1425 includes a precision, 2.5V
bandgap reference that also generates additional reference voltages. This capability
provides input-range options and automatically ensures the correct DC-bias level for
AC-coupled applications.
A 5V analog supply powers the
MAX1425, with a separate digital supply
to support an output interface of 3V to 5V.
Output data is presented in the two's
complement format. The device is
packaged in a space-saving 28-pin SSOP
specified for the industrial temperature
range (-40°C to +85°C). Prices start at
$3.95 (1000-up, FOB USA). An evaluation kit (MAX1425EVKIT) is available to
save design time.
MAX1093) or 5V (MAX1090/MAX1092).
A V LOGIC pin allows a direct interface
with digital supplies in the 1.8V to 5V
range, without additional circuitry.
Quiescent current is only 1.5mA at a
sample rate of 100ksps, and software
power-down modes further reduce the
supply current to less than 10µA at lower
sampling rates. The MAX1090–MAX1093
are available in 24-pin and 28-pin QSOP
packages, with prices starting at $3.92
(1000-up, FOB USA).
The analog inputs are software-configurable for unipolar/bipolar and singleended/differential operation. The fullscale analog-input range is determined by
the internal +2.5V reference or by an
externally applied reference in the 1V to
VDD range. Each ADC is powered by a
single analog supply of 3V (MAX1091/
11
Lowest power,
3V/18mW, 40MHz
I/Q DACs deliver
70dB SFDR
The MAX5180 family of monolithic
CMOS digital-to-analog converters (DACs)
is capable of 40MHz update rates while
operating from supply voltages in the 2.7V
to 3.3V range, consuming only 18mW at
3.0V. With the addition of four new dual
DACs (two 8-bit and two 10-bit), the family
now has 12 devices: 8- and 10-bit, dual and
single, voltage- and current-output, each
with a 50ppm/°C, low-noise reference.
Guaranteed monotonic, the DACs of this
family deliver ±0.5LSB INL and DNL.
The dual versions provide ±1% FSR gain
and 0.15° phase matching in I/Q reconstruction (transmit) applications. When
operating on 3V, their power consumption
is four-times less than that of comparable
devices. Two selectable idle modes lower
the supply current to 1µA (max) when the
application is inactive.
The four new dual devices include the
current-output MAX5188 (8-bit) and
MAX5182 (10-bit), and the voltageoutput MAX5191 (8-bit) and MAX5185
(10-bit). Intended for applications that
require less precise DAC timing, the new
devices update their outputs alternately
rather than simultaneously. Specifications
and packages for the alternate and simultaneous parts are otherwise identical.
Single 8- and 10-bit DACs are also
available. The 10-bit MAX5181 (current
output) and MAX5184 (voltage output),
for example, deliver the same dynamic
performance as the corresponding dual
versions at just 14mW power dissipation.
Specified for the extended-industrial
temperature range (-40°C to +80°C), the
singles come in 24-pin QSOP packages
and the duals in 28-pin QSOPs. Prices
range from $2.73 for the single 8-bit
MAX5187/MAX5190 to $4.41 for the
dual 10-bit MAX5180/MAX5183 and
MAX5182/MAX5185 (1000-up, FOB
USA). Evaluation kits are available
for $49.50.
Dual 12-bit VOUT
DACs have serial
interface
The MAX5104 is a dual 12-bit DAC with
serial input and voltage output. It operates
on a single 5V supply, draws only 500µA
of supply current, and features Rail-toRail output swings. The output amplifiers
maximize dynamic range with an internal
gain of 2V/V. Settling time is 12µs.
The 3-wire serial interface is SPI™/
QSPI™/MICROWIRE™ compatible.
Each double-buffered DAC input consists
of an input register followed by a DAC
register, which allows the DAC to be
updated with 16-bit serial words independently or simultaneously.
Low-cost 14/16-bit
DACs have serial
inputs and voltage
outputs
The 16-bit MAX5541 and 14-bit MAX5544
are serial-input, voltage-output DACs that
operate on a single 5V supply and provide
full-resolution performance without adjustments. The serial-data inputs are SPI/
QSPI/MICROWIRE compatible, and the
unbuffered voltage outputs are capable of
driving 60kΩ loads. Ranging from 0V to the
applied VREF, the unbuffered outputs also
allow low offset error (1LSB) and low
supply current (0.3mA). Settling time is 1µs.
The double-buffered inputs consist of an
input register followed by a DAC register
and accept 16-bit digital words. Each device
includes a power-on reset circuit that clears
the DAC output to zero in the unipolar
mode. For applications requiring galvanic
isolation, the Schmitt-trigger inputs allow a
direct interface to optocouplers.
The MAX5541/MAX5544 come in 8-pin
SO packages, with prices starting at $4.20
(1000-up, FOB USA).
Other features include programmable
power-down (2µA), hardware power-down
lockout (PDL), a separate referencevoltage input (AC or DC) for each DAC,
and an active-low clear input (CL) that
resets all registers and DACs to zero. Each
device includes an input for adjusting the
digital-logic thresholds and a serial-data
output pin for daisy-chaining.
The MAX5104 comes in a 16-pin QSOP
package. Prices start at $3.75 (1000-up,
FOB USA).
Rail-to-Rail is a registered trademark of Nippon
Motorola, Ltd.
SPI and QSPI are trademarks of Motorola, Inc.
MICROWIRE is a trademark of National
Semiconductor Corp.
40µA rail-to-rail
op amps sustain
95dB AVOL with
10kΩ load
The MAX4076/MAX4077/MAX4078
single/dual/quad rail-to-rail op amps are
open-loop versions of Maxim’s low-power
GainAmp amplifiers. Unlike other rail-torail op amps, these have a patented output
architecture capable of driving a 10kΩ load
while maintaining 95dB open-loop gain.
They feature 230kHz gain-bandwidth
products and draw 40µA supply currents.
Operating from a single supply voltage
between +2.5V and +5.5V, these op amps
suit micropower applications that require
wide output swings while maintaining
excellent distortion and gain accuracy.
MAX4076/MAX4077/MAX4078 devices
are unity-gain stable for load capacitance
up to 100pF. They achieve 84dB total
harmonic distortion at 1kHz. The dual and
quad versions exhibit 90dB of crosstalk at
100kHz.
The MAX4076 comes in space-saving 5pin SOT23 and 8-pin SO packages, the
MAX4077 in 8-pin µ MAX and SO
packages, and the MAX4078 in 14-pin SO
and 16-pin QSOP packages. Prices start at
$0.60 (1000-up, FOB USA).
GainAmp is a trademark of Maxim Integrated
Products.
12
1.8V op amps
deliver 120dB AVOL
and drive 2kΩ loads
The MAX4291/MAX4292/MAX4294
single/dual/quad op amps have rail-to-rail
inputs and outputs. Operating from a single
+1.8V to +5.5V supply, these op amps are
ideal for 2-cell, low-power portable applications. A 100dB power-supply rejection ratio
allows these devices to be powered directly
from a single Li+ cell or from two to three
NiCd, NiMH, or alkaline cells, without producing excessive output error as the cell
voltage decays. Their robust three-stage
design makes no compromise between
specifications, size, and power.
The MAX4291/MAX4292/MAX4294
draw only 100µA of supply current per
amplifier while achieving an open-loop
gain of 120dB, even with a 2kΩ load.
They exhibit a 500kHz gain-bandwidth
product and are unity-gain stable for
capacitive loads up to 100pF. Superior
open-loop gain, excellent load-driving
capability, and 400µV input offset voltage
make these amplifiers well suited for
buffering references.
The MAX4291 is offered in the tiny
5-pin SC70 and space-saving SOT23 packages. The MAX4292 is offered in spacesaving 8-pin µMAX and SO packages. The
MAX4294 is offered in miniature 14-pin
TSSOP and SO packages. Prices start at
$0.23 per amplifier (quad, 50,000-up,
factory direct, FOB USA).
Rail-to-rail, 800kHz,
SC70 op amp has
1µA shutdown
The MAX4400/MAX4401/MAX4402 railto-rail op amps draw only 320µ A per
amplifier while achieving an 800kHz gainbandwidth product. The MAX4401
includes a low-power shutdown mode that
reduces supply current to 1µA (max) and
places its output in a high-impedance state.
Ideal for portable and battery-powered
applications, these op amps operate from a
single +2.5V to +5.5V supply.
Applications that require wide output
swings with good gain accuracy and low
distortion benefit from the patented output
architecture of the MAX4400/MAX4401/
MAX4402. These op amps are capable of
driving a 2kΩ load to within 55mV of each
rail while maintaining a 110dB open-loop
gain. They achieve 0.009% total harmonic
distortion (THD) and are unity-gain-stable
with capacitive loads up to 400pF.
The MAX4400 single op amp is available
in 5-pin SC70 and SOT23 packages. The
MAX4401 single op amp with shutdown is
available in a 6-pin SC70 package. The
MAX4402 dual op amp is available in 8pin SOT23 and SOIC packages. All are
specified over the automotive temperature
range (-40°C to +125°C). Prices start at
$0.33 (1000-up, FOB USA).
200MHz singlesupply op amps in
ultra-small, 5-pin
SC70 packages
The MAX4450/MAX4451 are single/
dual, low-power, 200MHz single-supply
op amps in ultra-small SC70-5 and
SOT23-8 packages, respectively. Combining
single-supply operation, rail-to-rail
outputs, wide bandwidth, and tiny footprints, they are ideal for a variety of
wideband consumer applications, including set-top boxes, surveillance video
systems, digital video cameras, and DVDs.
±2.25V to ±5.5V supplies. They achieve
-3dB bandwidths of 200MHz and 485V/µs
slew rates while drawing only 6.5mA of
quiescent supply current per amplifier. In
addition, they offer 0.1dB gain flatness to
55MHz, low differential gain/phase of
0.02%/0.08°, and an SFDR of -65dBc at
5MHz. The common-mode range includes
ground, and their outputs swing rail-to-rail,
making them highly suitable for lowvoltage, single-supply applications.
The single MAX4450 is offered in 5-pin
SC70 and SOT23 packages, and the dual
MAX4451 is offered in 8-pin SOT23 and
SO packages. Prices start at $0.41 per
amplifier (dual, 50,000-up, FOB USA).
The MAX4450/MAX4451 operate from a
single +4.5V to +11V supply or from dual
10V/µs, rail-to-rail
I/O op amps in
ultra-small
packages
MAX4490/MAX4491/MAX4492 op
amps—in miniature 5-pin SC70, 8-pin
SOT23, and 14-pin TSSOP packages,
respectively—are ideal for applications
demanding the smallest possible board
area. For single-supply applications
(+2.7V to +5.5V), their rail-to-rail
input/output (I/O) provides flexibility and
dynamic range while simplifying circuit
design. The combination of high slew rate
(10V/µs), miniature packaging, and lowvoltage operation makes these op amps
Tiny SOT23
package includes
comparators and
6ppm/°C reference
The MAX9040/MAX9050 ICs include
micropower comparators (with rail-to-rail
inputs and outputs) and a precision
reference in a 5-pin SOT23 package. Ideal
for precision battery-voltage monitoring,
these devices maximize battery life by
eliminating the premature end-of-life
readings common in low-accuracy voltage
monitors. The MAX9040/MAX9050 also
offer 0.4% initial accuracy (“A” grade
only), 6ppm/°C temperature coefficients,
and a fast (400ns) propagation delay while
13
ideal for portable applications such as RF
power-amp control and audio amplification.
These op amps achieve 10MHz gain
bandwidth products while drawing only
800µ A of supply current per amplifier.
Other features include low input bias
current (50pA) and a 2kΩ output-drive
capability. Sample/hold amplifiers and
ADC predriver circuits benefit from the op
amps’ 300pF capacitive-load capability.
The single MAX4490 is available in ultrasmall 5-pin SC70 and SOT23 packages.
The dual MAX4491 is available in 8-pin
SOT23 and SO packages, and the quad
MAX4492 is available in the space-saving
14-pin TSSOP package. Prices start at
$0.22/amplifier (100,000-up, FOB USA).
drawing only 40µA of supply current. The
reference can drive 500µA and is stable
with any capacitive load up to 4.7nF.
The MAX9040 operates from a single
+2.5V to +5.5V supply and provides a
+2.048V reference voltage, which
produces a 500µV LSB in 12-bit systems.
The MAX9050 operates from a single
+2.7V to +5.5V supply and provides a
+2.500V reference voltage. Both
comparators feature internal hysteresis
and crowbar current limiting, which
lowers the supply current for highfrequency switching. Prices start at $0.82
for the “B” grade (50,000-up, FOB USA).
Analog switches
guarantee 0.5Ω
at +5V (world’s
lowest RON)
The MAX4624–MAX4628 low-voltage
analog switches have the world’s lowest
on-resistance (RON). Maximum RON for
the single-pole/single-throw (SPST)
MAX4626/MAX4627/MAX4628 is only
0.5Ω at 5V. The single-pole/double-throw
(SPDT) MAX4624/MAX4625 exhibit 1Ω
max RON at +5V. RON flatness is 0.10Ω
for the SPST devices and 0.15Ω for the
SPDT parts. Because they operate from a
single +1.8V to +5.5V supply, all are
suitable for use in portable equipment.
These switches are also ideal as lowvoltage relay replacements. Maximum
continuous current through the SPST
switches is ±400mA, with a peak of
±800mA (pulsed at 1ms, with a 10%
maximum duty cycle). Overcurrent protection prevents damage from short circuits
and excessive loads.
The MAX4626 is normally open (NO), the
MAX4627 is normally closed (NC), and
the MAX4628 has one NO and one NC
switch. The MAX4624 is break-beforemake, and the MAX4625 is make-beforebreak. For 5V operation, all feature
TTL/CMOS-logic compatibility and fast
switching (50ns tON and 50ns tOFF).
The MAX4624/MAX4625/MAX4628 are
available in 6-pin SOT23 packages, and
the MAX4626/MAX4627 are available in
5-pin SOT23 packages. Prices start at
$0.95 for the MAX4624/MAX4625/
MAX4628 and $0.90 for the MAX4626/
MAX4627 (2500-up, FOB USA).
Single 4-to-1/
dual 2-to-1 muxes
feature rail-to-rail
fault protection
The MAX4534 (single 4-to-1) and
MAX4535 (dual 2-to-1) fault-protected
multiplexers operate with dual ±4.5V to
±20V supplies or a single 9V to 36V
supply. All analog inputs are protected to
±40V with power off and to ±25V with
±15V supply voltages applied. The typical
fault-response time is 20ns.
All on-resistances are 400Ω (max), matched
(within a device) to within 10Ω (max).
Each channel has rail-to-rail signal-handling
capability, with overvoltage clamping (to
the appropriate supply rail) at 150mV
beyond the rails. All channels are open (off)
when power is off, and all digital inputs
guarantee TTL/CMOS compatibility for
+12V or ±15V supplies.
Single-supply
SPDT analog
switch in tiny
SC70 package
The MAX4599 is a single-pole/doublethrow (SPDT) analog switch that operates
on a single +2.0V to +5.5V supply. It offers
60Ω (max) RON at 5V and fast switching
(tON = 30ns max, tOFF = 25ns max).
R ON is flat to within 4Ω (max) and
matches to within 1Ω (max) for channels
within a device. Other specifications
include 5pC maximum charge injection, a
-3dB bandwidth of 200MHz, ±5nA
leakage at +25°C, break-before-make
switching, -76dB off-isolation at 1MHz,
and 0.12% THD.
Package options include the tiny 6-pin
SC70 and SOT23. Prices start at $0.56
(2500-up, FOB USA).
Available packages include the 14-pin
TSSOP, narrow SO, and plastic DIP.
MAX4534/MAX4535 prices start at $1.70
(1000-up, FOB USA).
Low-noise step-up
converter is 90%
efficient
The MAX1790 DC-DC boost converter is a
90% efficient regulator with fast response.
It includes fixed-frequency, current-mode,
PWM circuitry, and a 0.21Ω, 1.6A/14V
N-channel MOSFET.
High switching frequency (pin-selectable
as 640kHz or 1.2MHz) allows easy
filtering and faster loop performance. This
feature, combined with an external
compensation pin that provides flexibility
in determining loop dynamics, enables the
use of small, low-ESR ceramic output
capacitors. The MAX1790 derives output
voltages as high as 12V from inputs as low
as 2.6V.
Soft-start capability is programmed with
an external capacitor that sets the inputcurrent ramp rate, and shutdown mode
lowers supply current to 0.1µ A. The
MAX1790 comes in a space-saving 8-pin
µMAX package. The ultra-small package
and high switching frequency allow the
14
total solution to be less than 1.1mm high.
Prices start at $2.79 (1000-up, FOB USA).
VIN
2.6V TO 5V
IN
ON/OFF
VOUT
LX
SHDN
MAX1790
FREQ
GND
SS
FB
COMP
High-efficiency
current source
drives white-LED
backlighting
The MAX1698 is a switch-mode step-up
controller that regulates LED current at
power levels up to 5W. It’s the most
efficient driver available for the chains of
white or colored LEDs used in backlit
displays for PDAs, digital cameras, and
laptop computers. When combined with
white LEDs, the MAX1698 offers greater
Dynamically
adjustable
synchronous stepdown controller
powers Intel’s
SpeedStep
notebook CPUs
The MAX1717 buck controller features a
dynamically adjustable output, ultra-fast
transient response, ±1% DC accuracy, and
the high efficiency needed to power
leading-edge CPU cores. Maxim’s proprietary, constant-on-time PWM control
(Quick-PWM) handles wide ratios of
input/output voltage with ease and provides
a 100ns instant-on response to load transients while maintaining a near-constant
switching frequency. The MAX1717 has
been optimized for ICs with two or more
operational modes, e.g., CPUs with
SpeedStep or Gemini technology.
simplicity, lower cost, higher efficiency,
longer bulb life, and greater reliability than
is available in displays using fluorescent
(CCFL) and electroluminescent (EL) lamps.
More than 90% efficient, the MAX1698
minimizes loss with a low, 300mV currentsense threshold. A sense resistor (15Ω typ)
sets the current through a primary chain of
LEDs. An equivalent resistor lets you add
chains of matching LEDs that closely match
the primary chain. Control and dimming of
the LED current are accomplished with the
adjust pin rather than high-value currentlimiting resistors, which dissipate power
Output voltage can be dynamically
adjusted from 0.9V to 2V through the 5-bit
DAC inputs, and the output slew-rate
control minimizes battery and inductor
surge currents. This slew-rate control can
be tailored to a given application, providing
a just-in-time arrival at the new DAC
setting. An internal multiplexer uses only
five digital input pins to accept two 5-bit
DAC settings.
The MAX1717 comes in a 24-pin QSOP
package specified for the extended temperature range (-40°C to +85°C). Prices start
at $3.89 (1000-up, FOB USA). A
preassembled evaluation kit with recommended external components (MAX1717EVKIT) is available to reduce design time.
Quick-PWM is a trademark of Maxim Integrated
Products.
SpeedStep is a trademark of Intel Corp.
Gemini is a trademark of Advanced Micro
Devices.
and lower efficiency. The MAX1698 also
includes a soft-start circuit that eliminates
input-current surges at turn-on.
The MAX1698EUB comes in a small 10pin µ MAX package only 1.09mm high,
which occupies half the area of an 8-pin
SO, specified for the extended-industrial
temperature range (-40°C to +85°C). Prices
start at $1.40 (1000-up, FOB USA). A fully
assembled evaluation kit (MAX1698EVKIT) is available to help speed designs.
Ultra-highefficiency dual buck
controller powers
next-generation
notebook CPUs
The MAX1715 dual PWM step-down
controller provides the high efficiency,
excellent transient response, and high DCoutput accuracy needed in stepping down
from high battery voltage to the lower
chipset and RAM voltages required in
notebook computers.
The chip includes a free-running,
constant-on-time, on-demand PWM
controller with input feed-forward. Its
control algorithm provides ultra-fast
transient response, a wide differential
input-output range, low supply current,
and tight load regulation. The resulting IC
is simple, easy to compensate, and not
subject to the noise sensitivity of conventional fixed-frequency current-mode
PWMs.
The MAX1715 was designed to generate
the low supply voltages required for CPU
core and I/O, or 1.8V chipsets and 2.5V
RAM. It comes in a small 28-pin QSOP
package specified for the extended temperature range (-40°C to +85°C). Prices start at
$3.60 (1000-up, FOB USA). A preassembled evaluation kit (MAX1715EVKIT)
with recommended external components is
available to reduce design time.
15
Smallest stepdown DC-DC
for 1.8V logic
achieves 87%
peak efficiency
with no inductor
The MAX1730 regulated charge-pump
step-down converter generates a fixed 1.8V
or 1.9V output voltage from a +2.7V to
+5.5V input. Typical applications generate
a 1.8V core logic supply from a 3.6V Li+
battery. The MAX1730 needs no inductor.
Its fractional-conversion technique
performs DC-DC conversion with efficiency much greater than that of a linear
regulator, yet with minimal increase in
package size. Maximum efficiency for an
ideal linear regulator in this application can
be less than 50%, but the MAX1730
extends battery life with an efficiency
exceeding 80%.
Switching frequencies up to 2MHz not
only enable the use of small flying capacitors (0.22µF), they allow the entire circuit
to fit in just 0.05in2 (32mm2). Quiescent
supply current is a low 75µ A. In
shutdown, the supply current drops to less
than 1µA, and the output disconnects from
the input. A proprietary soft-start circuit
prevents excessive inrush current from the
input supply during startup, making the
MAX1730 compatible with sources of
higher output impedance, such as alkaline
and Li+ batteries.
The MAX1730EUB comes in a small 10pin µ MAX package only 1.09mm high
and half the area of an 8-pin SO, specified
for the extended-industrial temperature
range (-40°C to +85°C). Prices start at
$1.45 (1000-up, FOB USA).
Li+ battery-pack
protector blocks
voltage and charge
faults
The MAX1665 provides complete protection for Li+ battery packs (with 2, 3, or
4 cells in series) against overvoltage,
undervoltage, cell shorts, overcharging,
and overdischarge current. By controlling
external N-channel MOSFETs, the internal
power-MOSFET drivers reliably disconnect
the cells from the pack terminals when a
fault condition occurs, simplifying the
design of battery-pack safety circuitry; you
just select the power MOSFETs and
connect the Li+ cells.
The MAX1665 connects directly to battery
voltages as high as 20V without an external
linear regulator and features a ±1%
accurate, factory-trimmed, 4.3V overvoltage threshold limit. It is also a “micropower” device, drawing less than 1µA in
shutdown and only 15µ A in operation.
MAX1665 applications do not require an
external current-sense resistor. Instead,
overcurrent detection is accomplished by
monitoring the voltage drop across both
N-channel protection MOSFETs.
The MAX1665S protects two series cells,
the MAX1665V three series cells, and the
MAX1665X four series cells. All are
available in 8-pin narrow SO packages
specified for the extended industrial
temperature range (-40°C to +85°C).
Prices for the 4-cell protector start at $1.55
(1000-up, FOB USA).
3.3V, 0.05in2 buck /
boost charge pump
delivers 100mA
The MAX1759 regulated charge-pump
converter generates an adjustable or fixed
(3.3V) output voltage from a +1.6V to
+5.5V input. This capability is useful for
Li+ battery inputs, which can be at 4.2V
when fully charged and below 2.9V when
nearly discharged. The MAX1759’s
ability to maintain its regulated output
voltage whether the input is above or
below that level is unique among chargepump converters. It guarantees 100mA
output currents without an inductor, with
only three small ceramic capacitors. The
entire circuit fits in 0.05in2 (32mm2).
Quiescent supply current is a low 50µA
and drops to 1µ A in shutdown. Unlike
more typical step-up converters, the load
is completely disconnected from the input
during shutdown. The MAX1759 operates
up to 1.5MHz, allowing high output
current with very small external capacitors. An open-drain power-OK output
signals when the output is out of regulation.
Foldback short-circuit protection limits
output current when the output is shorted
16
to ground. The chip also includes softstart circuitry that limits the inrush current
during power-up. The MAX1759EUB is
specified for the extended-industrial
temperature range (-40°C to +85°C).
Prices start at $2.45 (1000-up, FOB USA).
VTH5
RST
VCC3
RST3
RESET
GENERATOR
VTH3
RST5
These ICs offer ±2.5% reset threshold
accuracy over temperature. When supply
voltage falls below the preselected
threshold, the IC asserts and maintains a
reset signal for at least 140ms after VCC
returns above the threshold. The available
reset outputs (guaranteed valid to 1.0V)
are open-drain active-low (MAX803),
push-pull active-low (MAX809), and
push-pull active-high (MAX810).
RSTIN
1.22V
WDI
WATCHDOG
TIMER
GND
...IN THIS!
The MAX803/MAX809/MAX810 come in
a 3-pin SC70 package, with prices starting
at $0.94 (2500 or 10,000 minimum, FOB
USA).
UAL SIZ
CT
SOT23-6
mm
0
x 3.0
maintained for at least 150ms after VCC
returns above the reset threshold.
and an auxiliary user-adjustable reset
input (MAX6364). Supply current for all
three devices is only 10µA.
Each device is offered with a push-pull
active-low, push-pull active-high, or opendrain active-low reset output guaranteed
valid to 1.0V. Other available features
include a manual reset input (MAX6361),
a battery-on output indicator (MAX6363),
UNREGULATED
DC VOLTAGE
MAX6361/MAX6363/MAX6364 devices
are available in a 6-pin SOT package, with
prices starting at $1.50 (2500 minimum,
FOB USA).
2.4V TO 5.5V
VCC
R1
1µF
VCC
RESET µP
RESET
RESET IN
GND
R2
MAX6364
BUS
The MAX6361/MAX6363/MAX6364 ICs
feature ±2.5% reset threshold accuracy
over temperature. They automatically
switch RAM memory to the backup power
when VCC falls below the reset threshold
and the backup battery voltage. When the
supply voltage declines below the reset
threshold, a reset signal is asserted and
GLITCH
FILTER
E
MAX6361/MAX6363/MAX6364 µP-supervisory circuits are designed to monitor +2.5V
to +5.0V supply rails in digital and µ P
systems. Besides reducing cost and eliminating external components and adjustments,
these devices are the first ICs to combine
reset and backup-battery switchover in SOT
packages, which are ≈70% smaller than
currently available alternatives.
MR
VCC5
0
Backup-battery
reset ICs offered
in SOT packages
MAX803/MAX809/MAX810 µP-supervisory circuits are designed to monitor
+2.5V to +5.0V supply rails in digital and
µ P systems. Besides reducing cost and
eliminating external components and
adjustments, these devices are the first
fully integrated power-on reset ICs
available in a 3-pin SC70 package, which
is half the size of the SOT package. Supply
current (12µA) is 30% lower than that of
existing devices.
YOU GET ALL THIS...
3.0
The MAX6358/MAX6359/MAX6360
include a watchdog timer for monitoring
software execution. They provide a 46.4s
timeout at startup and 2.9s timeouts thereafter. The MAX6355/MAX6356/MAX6357
provide a third user-adjustable voltage
monitor in place of the watchdog timer.
The available reset outputs include 3V and
All devices operate over the extendedindustrial temperature range (-40°C to
+85°C). Prices start at $1.38 (2500-up,
FOB USA).
m
Each IC asserts its reset outputs (3V logic,
5V logic, or both) when any monitored
voltage falls below its reset threshold.
(The user selects from a variety of
factory-set, precision threshold voltages
associated with 2.5V, 3V, 3.3V, or 5V
power supplies.) Available in small 5- or
6-pin SOT23 packages, all devices include
an active-low, TTL/CMOS-compatible,
debounced manual-reset input.
Industry-standard
power-on reset ICs
now in SC70
packages
m
The MAX6351–MAX6360 microprocessor (µP) supervisors each monitor two
to three supply voltages simultaneously.
These IC supervisors provide substantial
improvements in accuracy and system
reliability over comparable discretecomponent circuits. They draw supply
currents of only 20µA and maintain valid
outputs for as long as any monitored
voltage remains above 1V.
5V active-low push-pull and active-low
open-drain. The MAX6351 provides two
reset outputs.
A
µP supervisors
monitor three
supply voltages
OUT
3.6V Li+
BATTERY
BATT
VCC
GND
1µF
SRAM
GND
17
±15kV ESDprotected RS-232
transmitters
in SOT packages
The MAX3188E/MAX3189E single RS232 transmitters are designed for space and
cost-constrained applications requiring
minimal RS-232 communications. To
ensure component compliance with strict
European ESD standards, the transmitter
input is protected to ±15kV using IEC
1000-4-2 Air-Gap Discharge, to ±8kV
using IEC 1000-4-2 Contact Discharge, and
to ±15kV using the Human Body Model.
Power consumption and heat dissipation are
minimized by the devices’ low supply
current (only 1µA from a ±4.5V to ±6.0V
supply). They guarantee RS-232 compliant
performance for data rates up to 250kbps
for the MAX3188E, and up to 1Mbps for
the MAX3189E. The MAX3188E/
MAX3189E have three-state transmitter
outputs and can be run off the charge pump
of more than 50 RS-232 devices from
Maxim. To save space and increase design
flexibility, the transmitters can be combined
with a SOT receiver from Maxim’s
MAX3180E line, forming a complete SOTpackage RS-232 transceiver.
The MAX3188E/MAX3189E are available in a 6-pin SOT23 package specified
for the extended temperature range (-40°C
to +85°C). Prices start at $0.55 (1000-up,
FOB USA).
True RS-232 Operation
at 1Mbps
2.5V/1µA
RS-232 devices
protect against
ESD to ±15kV
The MAX3316E–MAX3319E RS-232
transceivers are designed for low-voltage
systems, delivering RS-232 performance
from a single +2.5V supply. To ensure
compliance with strict European ESD
standards, the RS-232 inputs and outputs
are protected to ±15kV using IEC 1000-42 Air-Gap Discharge, to ±8kV using IEC
1000-4-2 Contact Discharge, and to
±15kV using the Human Body Model.
To minimize power and heat dissipation,
the devices employ Maxim’s revolutionary AutoShutdown Plus™ to draw
only 1µA of supply current from a +2.25V
to +3.0V source. They guarantee RS-232
performance for data rates up to 460kbps.
The MAX3318E/MAX3319E feature a
logic-level output (READY) that asserts
when the charge pump is regulating and
the device is ready to begin transmitting.
The MAX3316E family is available in
TSSOP and SSOP packages specified for
the commercial or extended temperature
ranges. Prices start at $2.24 (100-up, FOB
USA).
AutoShutdown Plus is a trademark of Maxim
Integrated Products.
World’s first
5Tx/3Rx RS-232
transceivers with
±15KV ESD protection on all I/O pins
MAX3237E/MAX3238E/MAX3248E
RS-232 transceivers are complementary
serial ports with five transmitters and
three receivers, which makes them ideal
for cell-phone data cables, modems, settop boxes, and other devices requiring RS232 compliant communications. All input
and output pins (RS-232 and logic pins
included) have integrated ESD structures
that protect the device to ±15kV using the
Human Body Model, to ±15KV using the
IEC 1000-4-2 Contact Discharge method,
and to ±4kV using the IEC 1000-4-2 AirGap Discharge method.
The devices’ AutoShutdown Plus and low
supply current (only 10nA from a 3.0V to
5.5V supply) minimize power and heat
dissipation. The MAX3237E guarantees
RS-232 compliant performance for data
rates up to 1.0Mbps, and the MAX3238E/
MAX3248E transmit RS-232 compliant
data up to 250kbps. The MAX3248E
operates with logic levels as low as 1.8V,
which eliminates the need for level
shifters in low-voltage logic systems.
MAX3237E receivers can be active or
inactive in shutdown, but the MAX3238E/
MAX3248E transceivers have a continuously active extra receiver for monitoring
the ring indicator. All three devices come
in an SSOP package, specified for the
18
commercial (0°C to +70°C) or extended
(-40°C to +85°C) temperature range.
Prices start at $3.87 (1000-up, FOB USA).
+3.3V
1µF
26
VCC
28 C1+
0.1µF
0.1µF*
25 C11 C2+
0.1µF
MAX3238E
MAX3248E
V- 4
0.1µF
3 C224 T1IN
T1
23 T2IN
LOGIC
INPUTS
V+ 27
T2
22 T3IN
T3
19 T4IN
T4
17 T5IN
T5
T1OUT 5
T2OUT 6
T3OUT 7
RS-232
OUTPUTS
T4OUT 10
T5OUT 12
14 FORCEOFF
AutoShutdown Plus
INVALID 15
13 FORCEON
16 R1OUTB
21 R1OUT
R1IN 8
R1
5k
LOGIC
OUTPUTS
20 R2OUT
R2IN 9
R2
5k
18 R3OUT
R3IN 11
R3
5k
GND
2
*C3 MAY BE RETURNED TO EITHER VCC OR GND.
RS-232
INPUTS
Octal serial-parallel
I/O expander suits
SMBus systems
The MAX1608/MAX1609 I/O expanders
are designed to provide remote I/O
expansion through an SMBus serial
interface. Each has eight high-voltage,
open-drain outputs that also provide bidirectional capability by serving as TTLcompatible logic inputs. Typical applications include high-side MOSFET loadswitch drivers in power-management
systems, pushbutton switch monitors, and
general-purpose I/Os.
high-impedance state at power-up, are
intended for driving P-channel MOSFETs.
These output conditions enable powerplane sequencing by ensuring that the
MOSFETs are off at power-up.
(MAX1608EVKIT) is available to speed
the design cycle. The MAX1608/MAX1609
come in a space-saving 16-pin QSOP
package, priced from $2.61 (1000-up,
FOB USA).
Other features include thermal-overload
and output-overcurrent protection, a wide
supply voltage range (2.7V to 5.5V), and
ultra-low supply current. An evaluation kit
SMBus is a trademark of Intel Corp.
+2.7V TO +5.5V
100k
V+
SMBUS
TO/
FROM
HOST
ALERT
IO1
SMBDATA
IO2
SMBCLK
IO3
ADD1
The MAX1608 outputs, low at power-up,
are intended for driving N-channel
MOSFETs. MAX1609 outputs, in the
The MAX1740/MAX1741 SIM/smartcard level translators provide level
shifting and ESD protection for smartcard and subscriber identification module
(SIM) ports. The devices integrate two
unidirectional level shifters for the reset
P-CH
SMBSUS
ADD
World’s smallest
SIM/smart-card
interfaces operate
below 1.5V
100k
0.1µF
MAX1609
The eight I/Os are continuously monitored
and can be used as inputs. Two internal 8bit latches allow the outputs to be toggled
using the SMBus suspend input, without
the inherent latency associated with reprogramming inputs and outputs through the
serial bus.
100k
LOAD1
GND
LOAD2
LOAD3
IO7
and clock signals, a bidirectional level
shifter for the serial data stream, and
±10kV ESD protection on all card
contacts.
For maximum flexibility, the level shifters
translate between any two logic voltages
in the 1.425V to 5.5V range. The translators form an ideal interface between
system controllers and plug-in cards with
1.8V, 2.5V, 3.3V, or 5V logic levels. To
prolong battery life, the quiescent current
is a low 1µ A and drops to only 0.1µ A
during shutdown.
The MAX1740 includes a shutdowncontrol pin to aid card insertion and
removal, and the MAX1741 includes a
system-side data driver to support microcontrollers (µ Cs) without open-drain
drivers. Both devices simplify power
management by automatically shutting
down when power for the system or card
is removed. These devices may be
combined with the MAX1686H regulated
charge pump, which provides SIM/smartcard power by generating 0V/3V/5V from
a +3V supply.
The MAX1740EUB/MAX1741EUB
come in a small 10-pin µMAX package,
which occupies half the area of an 8-pin
SO and is only 1.09mm high, specified for
the extended-industrial temperature range
(-40°C to +85°C). Prices start at $1.45
(1000-up, FOB USA). An evaluation kit
(MAX1741EVKIT) is available to help
speed designs.
19
Signal conditioner
enables highly
accurate sensors
and 4–20mA output
The MAX1459 signal conditioner
compensates for the nonlinearities of
piezoresistive transducers over a wide
temperature range (-40°C to +125°C),
using user-programmed compensation
coefficients stored in the IC’s internal
EEPROM. The MAX1459 also features
an uncommitted op amp that can be easily
configured to provide a 4–20mA loop
interface or the automotive diagnostic clip
features required in today’s automotive
systems. In addition to piezotransducers,
the MAX1459 can compensate nonbulk
micromachined sensors, strain gauges, and
similar resistive-based sensors.
modules simultaneously. This system is
free to users who commit to the purchase
of more than 10,000 units per year and is
available for lease to lower-volume users.
The MAX1459 includes a complete analog
signal path and four 12-bit DACs for integrated digital correction. The DACs are
user-programmed to correct for span, span
TC, offset, and offset TC. A fifth low-resolution DAC can adjust for course offset. All
can be programmed through a digital
interface, and the correction coefficients can
be stored in the chip’s internal EEPROM.
including low-power op amps, integrated
temperature sensors, and DSP engines.
For high-volume applications, Maxim can
quickly customize the MAX1459 to
optimize price and performance. Please
contact Maxim for more details.
The MAX1459 joins a family of sensorconditioning circuits ranging from the lowcost, laser-trimmed MAX1450 to the fully
digital MAX1460. The MAX1459 comes
in a space-saving 20-pin SSOP package.
Prices start at $2.95 (1000-up, FOB USA).
LabVIEW is a trademark of National Instruments.
The MAX1459 was designed with a
dedicated cell library of more than 200
sensor-specific functional blocks,
The MAX1459’s electronic trimming
simplifies sensor-module manufacturing
by eliminating the need for laser trimming
and other manual content in production. It
also allows efficient production flow by
consolidating the manufacturing steps of
pretest, calibration, compensation, and
final test. To speed development and
improve time to market, the MAX1459 is
supported by Maxim’s Pilot Production
System, which allows users to quickly set
up a manufacturing line. Running under
LabVIEW, a single Pilot Production
System can program up to 110 sensor
Resistor-programmable temperature
switches combine
flexibility with
simplicity
operate from a +2.7V to +5.5V supply,
making them ideal for use in portable and
space-constrained environments.
The MAX6509 is available in a 5-pin SOT23
package and has an open-drain output.
The MAX6510, available in a 6-pin SOT23
package, has an output programmable as
The MAX6509/MAX6510 temperature
switches are the newest additions to the
most complete line of such devices in the
industry. These two devices are the
smallest resistor-programmable temperature comparators available today.
One external resistor sets the temperature
threshold between -40°C and +125°C. Trippoint accuracy is ±0.5°C (typ) and ±4°C
(max) over the -40°C to +125°C temperature range. Hysteresis is pin-selectable at
2°C or 10°C. These low-power devices
20
active-high, active-low, or open-drain with
internal pullup. Both have versions whose
outputs become active as the temperature
rises above its threshold (SET HOT) or falls
below its threshold (SET COLD). Prices start
at $0.70 for the MAX6509 and $0.72 for the
MAX6510 (1000-up, FOB USA).
0.5%, SOT23
series references
at shunt-reference
prices
The MAX6101–MAX6105 family of lowcost, low-dropout, low-power, seriesmode voltage references are available in a
3-pin SOT23 package. Intended for costsensitive portable power systems requiring
precision and small size, their proprietary
Highly integrated
transmitter ICs
serve dual-band
cellular phones
curvature-correction circuits and lasertrimmed thin-film resistors provide 0.4%
initial accuracy along with a 75ppm/°C
temperature coefficient over the extendedindustrial temperature range (-40°C to
+85°C).
MAX6101–MAX6105 have supply
currents virtually independent of supply
voltage, and they require no external
resistor. Furthermore, they require no
external capacitor and are stable for load
capacitances up to 1µF.
Although they draw only 150µA of supply
current, these internally compensated references can source 5mA and sink 2mA of
load current. Unlike conventional shuntmode (two-terminal) references, the
MAX6101–MAX6105 references accept
input voltages from (VOUT + 200mV) to
12.6V and offer voltage options of 1.25V,
2.5V, 3V, 4.096V, and 5V. Prices start at
$0.55 (1000-up, FOB USA).
The MAX2360 dual-band transmitter fits
applications such as CDMA, TDMA, and
EDGE dual-band phones. The MAX2362
suits single-band PCS and W-CDMA
applications, and the MAX2364 suits
single-band, dual-mode cellular applications. All are available in a 48-pin
TQFP package. Prices start at $5.85
(1000-up, FOB USA).
The MAX2360/MAX2362/MAX2364 ICs
contain complete baseband-to-PA transmitters for dual-band cellular phones.
Designed for use in dual-band, tri-mode,
and single-mode cellular phones, they are
compatible with N-CDMA, TDMA,
EDGE, and W-CDMA systems. High-level
integration in these ICs dramatically
reduces the size and number of components.
The MAX2360 includes an I/Q modulator,
IF VGA, dual IF VCOs, upconverters,
dual synthesizers, RF VGA, and three RF
power-amplifier drivers. The MAX2360’s
high level of linear output power (+7dBm)
allows it to drive the power amplifiers
directly. Because of the image-reject
upconverters, two SAW filters and two
external PA drivers can be eliminated.
These devices’ wide frequency range and
dual IF ports are ideal for a variety of
dual-band and dual-mode radio architectures.
21
3.5GHz, SiGe
active mixers
deliver +11dBm
IP3 performance
The MAX2683/MAX2684 SiGe mixers are
the first active mixers to provide high IP3
performance. Until now, high IP3 performance has been achieved only with
expensive and bulky passive mixers that
exhibit 6dB insertion loss (typ) and require
high-level drive for the LO inputs. The
MAX2683/MAX2684 match the IP3
performance of passive mixers but reduce
cost by 40% and board space by 80%.
They deliver up to +11dBm IP3 at 3.5GHz,
10dB higher than other active mixers. To
allow the setting of minimum IP3 performance without wasting supply current, the
IP3 value is user-programmable with an
external bias resistor.
An on-chip LO buffer and filter provide
37dB isolation between the LO and RF
ports, 20dB better than that obtained in
passive-mixer designs. The improved
3V, 622Mbps,
low-power
transimpedance
amplifier provides
high sensitivity
The MAX3665 transimpedance amplifier
for SDH/SONET applications features
3.3V, 622Mbps operation and ultra-low
power consumption. It was designed for
PIN-photodiode preamplifiers and
receivers in SDH/SONET OC-12 transmission systems. When combined with the
MAX3681 1:4 deserializer and the
MAX3676 CDR and limiting amplifier, the
MAX3665 forms a complete, high-performance 622Mbps receiver.
isolation helps to reduce front-end filter
complexity and cost. These mixers are
ideal for use in wireless local loop (WLL),
wireless broadband access, and digital
microwave radios.
Both devices operate with RF input
frequencies between 3.4GHz and 3.8GHz.
IF frequencies range from 100MHz to
400MHz (MAX2683) and from 800MHz to
1000MHz (MAX2684). At 3.5GHz, the
conversion gain is 6dB for the MAX2683
and 1dB for the MAX2684. The LO input
can be injected at full or half frequency. For
the half-frequency input, a logic-enabled
doubler increases the LO signal to full
frequency. An internal LO filter reduces LO
harmonics and mixer spurious products.
The MAX2683/MAX2684 operate from a
single 2.7V to 5.5V supply. Both are
available in a 16-pin TSSOP package,
with an exposed paddle for optimum
performance at 3.5GHz. Prices start at
$1.90 (1000-up, FOB USA). Fully
assembled evaluation kits (MAX2683EVKIT/MAX2684EVKIT) are available
to help reduce design time.
by providing true differential output
swings over a wide range of input current
levels. Typical deterministic jitter is just
100ps. The differential outputs are backterminated with 50Ω per side.
Available as dice and in an 8-pin µMAX
package specified for the extended
temperature range (-40°C to +85°C), the
MAX3665E/D is priced from $5.80
(1000-up, FOB USA). Evaluation kits are
available to reduce design time.
Direct I/Q
transmitters
for broadband
wireless are highly
integrated
The MAX2720/MAX2721 are the
industry’s most highly integrated direct
I/Q transmitters for broadband wireless
applications. Operating from 1.7GHz to
2.5GHz, they include differential amplifiers
for modulation inputs, two matched
double-balanced mixers, a selectable LO
doubler, LO buffer, wideband 90° quadrature
generator, RF variable-gain amplifier
(VGA) with 35dB control range, and
+13dBm PA driver amplifier.
Compared with traditional IF-based
designs, the direct-conversion approach
eliminates an IF SAW filter, VCO, PLL,
and upconverter (i.e., one IF stage). This
simplification reduces component costs by
35% and board space by 50%. The
MAX2720/MAX2721 are ideal for applications in wideband CDMA systems,
wireless local loops, PCS/DCS basestations, LMDS/MMDS, and 2.4GHz broadband ISM-band radios.
The MAX2720 converts baseband signals
directly to RF outputs in the 1.7GHz to
2.1GHz frequency range. The MAX2721
output range is 2.1GHz to 2.5GHz. A
logic-enabled frequency doubler allows the
user to inject the LO signal at half
frequency, resulting in a 33dB carrier
suppression. The I/Q inputs are amplitude
matched to within ±0.2dB and phase
matched to within ±1°. Excellent amplitude
and phase matching results in sideband
suppression of 40dB at 1.9GHz (MAX2720)
and 35dB at 2.3GHz (MAX2721).
To help reduce transmitter noise at the RF
output, I/O pins are available for insertion
of a bandpass filter between the VGA
output and the PA driver input. The
MAX2720/MAX2721 transmitters offer a
low-power shutdown mode and 2.7V to 3.3V
single-supply operation. Both devices are
available in 20-pin TSSOP packages, with
an exposed paddle. Prices start at $4.55
(1000-up, FOB USA). Fully assembled
evaluation kits (MAX2720EVKIT/
MAX2721EVKIT) are available to help
reduce design time.
The MAX3665 converts small photodiode
currents to a measurable differential
voltage. Operating from a single +3.3V to
+5.0V supply, it consumes only 70mW of
power at 3.3V. The overall transimpedance gain is nominally 8kΩ, with a
bandwidth of 470MHz. For 1300nm,
622Mbps receivers, the MAX3665’s low
input noise (55nARMS) provides a typical
sensitivity of -33.2dBm. Its DC-cancellation circuit reduces pulse-width distortion
22
2.125Gbps, 3.3V
Fibre Channel
repeaters tolerate
0.7UI jitter
The MAX3770/MAX3771 repeaters serve
Fibre Channel arbitrated-loop applications
in storage systems, disk arrays, hubs, and
switches. The 2.125Gbps MAX3770 and
the 1.0625Gbps MAX3771 tolerate total
input jitter of 0.7UI, and operate without
need for a reference clock. The devices’
outputs have only 0.10UI jitter.
All inputs and outputs have internally
terminated current-mode logic that provides
low jitter and is tolerant of mismatched
circuit board traces and inductive connectors. The MAX3770/MAX3771 repeaters
provide outputs for the recovered clock and
a PLL lock indicator.
Both devices come in small 16-pin QSOP
packages and operate from a +3.3V
supply. The MAX3770 consumes
215mW, and the MAX3771 consumes
190mW. Prices start at $3.95 for the
MAX3771 and at $8.55 for the MAX3770
(1000-up, FOB USA).
3.3V, 2.5Gbps
clock-and-datarecovery IC
surpasses SDH/
SONET regenerator
specs
3.3V, 2.5Gbps laser
driver includes
current monitors
and APC
The MAX3869 is a 3.3V, 2.5Gbps
SDH/SONET laser driver with current
monitors and automatic power control
(APC). It is ideal for applications in
2.5Gbps SDH/SONET transmission
systems such as add/drop multiplexers,
digital cross-connects, section regenerators, and optical transmitters. When
combined with the MAX3890 16:1 serializer, the MAX3869 forms a complete,
low-power 2.5Gbps transmitter.
control and a failure-monitor output to
indicate when the APC loop is unable to
maintain the average optical power. Ease
of programming both the wide modulation-current range (5mA to 60mA) and the
bias-current range (1mA to 100mA) make
this device ideal for use in a variety of
SDH/SONET applications.
The MAX3869 is available as dice and in a
small (5mm x 5mm) 32-pin TQFP package
specified for the extended-industrial temperature range (-40°C to +85°C). Prices start
at $18.00 (1000-up, FOB USA). An evaluation kit is available to speed designs.
The MAX3869 requires only 211mW of
power. It accepts differential PECL data
and clock inputs while providing bias and
modulation currents for driving a laser.
An APC feedback loop is incorporated to
maintain constant average optical power
vs. temperature and lifetime. The chip
includes bias- and modulation-current
monitors that are directly proportional to
the laser bias and modulation currents.
The MAX3869 also provides enable
SONET specification by 6.3mUI RMS ,
and the typical jitter tolerance at 100kHz
exceeds the SDH/SONET specification by
1.8UIp-p.
Differential CML outputs are provided for
clock and data signals, and an extra
2.488Gbps serial input is available for
system loopback diagnostic testing. The
The MAX3876 compact, low-power, 3.3V
clock-recovery and data-retiming IC with
CML outputs is ideal for 2.488Gbps
SDH/SONET applications. It is designed
for section-regenerator, terminal-receiver,
and switch core applications in OC48/STM-16 transmission systems. In
conjunction with the MAX3831 2.5Gbps
interconnect mux/demux, the MAX3876
forms a complete backplane transceiver.
Operating from a single +3.3V to +5.0V
supply, the MAX3876 consumes only
445mW at +3.3V. The fully integrated
phase-locked loop recovers a synchronous
clock signal from the serial NRZ data
input, which is also retimed by the
recovered clock. Low jitter generation in
the MAX3876 surpasses the SDH/
23
MAX3876 also includes a TTL-compatible
loss-of-lock (LOL) monitor. Available as
dice and in a 32-pin TQFP package
specified for the extended temperature
range (-40°C to +85°C), the MAX3876 is
priced at $40.99 (1000-up, FOB USA). An
evaluation kit is available to speed designs.
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

advertisement