Volume 46, Number 1

Volume 46, Number 1
Volume 46, Number 1, 2012
A forum for the exchange of circuits, systems, and software for real-world signal processing
In This Issue
2 Editors’ Notes; New Product Introductions
3 Robust Amplifiers Provide Integrated Overvoltage Protection
7 Modern DACs and DAC Buffers Improve System Performance, Simplify Design
12 DDS Devices Generate High-Quality Waveforms Simply, Efficiently, and Flexibly
17 Efficient FSK/PSK Modulator Uses Multichannel DDS to Switch at Zero Crossings
20 Low-Noise, Gain-Selectable Amplifier
21 Designing a Low-Power Toxic Gas Detector
www.analog.com/analogdialogue
Editors’ Notes
PRODUCT INTRODUCTIONS: VOLUME 46, NUMBER 1
Data sheets for all ADI products can be found by entering the part
number in the search box at www.analog.com.
IN THIS ISSUE
Robust Amplifiers Provide Integrated Overvoltage Protection
Faulty performance or damage can occur when an op amp’s
input voltage exceeds the specified input-voltage range. This
article discusses some common causes and effects of overvoltage
conditions, how cumbersome overvoltage protection can be added
to an unprotected amplifier, and how the integrated overvoltage
protection of newer amplifiers provides a compact, robust, costeffective solution. Page 3.
Modern DACs and DAC Buffers Improve System Performance,
Simplify Design
As the heart of many instrumentation and control systems, digitalto-analog converters (DACs) play a key role in determining system
performance and accuracy. This article looks at a new precision,
fast-settling, voltage-output 16-bit DAC and shows some ideas for
buffering the outputs of high-speed complementary current-output
DACs that can rival transformer performance. Page 7.
DDS Devices Generate High-Quality Waveforms Simply, Efficiently,
and Flexibly
Direct digital synthesis (DDS) technology is used to generate and
modify high-quality waveforms in a broad range of applications
in such diverse fields as communications, defense, medicine,
industry, and instrumentation. This article provides an overview
of the technology, describes its strengths and limitations, looks at
some application examples, and showcases some significant new
products. Page 12.
Efficient FSK/PSK Modulator Uses Multichannel DDS to Switch
at Zero Crossings
This article describes how two synchronized DDS channels can
implement a zero-crossing FSK or PSK modulator. In phasecoherent radar systems, zero-crossing switching reduces the
amount of postprocessing needed for target signature recognition;
and zero-crossing switching reduces PSK spectral splatter. Here,
the AD9958 two-channel complete DDS is used to switch at the
zero crossing. Page 17.
Low-Noise, Gain-Selectable Amplifier
Traditional gain-selectable amplifiers use switches in the feedback
loop to connect resistors to an amplifier, but the switch resistance
degrades the noise performance of the amplifier, adds significant
capacitance on the inverting input, and contributes to nonlinear
gain error. This article presents a gain-selectable amp that uses
an innovative switching technique that preserves the noise
performance while reducing the nonlinear gain error. Page 20.
Designing a Low-Power Toxic Gas Detector
Safety fi rst! Many industrial processes involve toxic compounds,
including chlorine, phosphine, arsine, and hydrogen cyanide, so it
is important to know when dangerous concentrations exist. This
article describes a portable carbon monoxide detector using an
electrochemical sensor. CO is relatively safe to handle, but it is
still lethal, so use extreme care and appropriate ventilation when
testing the circuit described here. Page 21.
Dan Sheingold [[email protected]]
Scott Wayne [[email protected]]
2
January
Amplifier, power, 2-W stereo, Class-D,
digital input ........................................................... SSM 2518
DAC, 16-bit, ultra-stable, voltage-output ................... AD5760
DAC, 18-bit, system-ready, voltage-output ................ AD5780
DAC, 20-bit, system-ready, voltage-output ................ AD5790
Transmitter, HDMI, 165-MHz,
high-performance .................................................. ADV7513
February
Processor, video, 10-bit, multiformat,
SDTV/HDTV .................................................... ADV7181D
Switch, analog crosspoint, 60-MHz,
16 × 16, G = 2 ...................................................... ADV3205
Transceivers, M-LVDS, 200-Mbps
half-/full-duplex ............................. ADN4696E/ADN4697E
March
ADCs, 16-bit, low-power, successiveapproximation ................................... AD7988-1/AD7988-5
Amplifier, difference, dual, short-to-battery
protection .......................................................... ADA4830-2
Amplifier, operational, dual, high-precision,
JFET-input ....................................................... ADA4610-2
Amplifier, operational, dual, low-noise,
JFET-input ....................................................... ADA4001-2
Comparator, fast, low-power, automotive ................. AD8469
Controller, power supply, 3-channel, digital ........... ADP1053
DAC, RF, 11-bit, 2.5-GSPS, current-output, ......... AD9737A
Demodulator, quadrature, 50-MHz to 525-MHz,
PLL, VCO ......................................................... ADRF6806
Micro-PMU, 1.2-A buck, two
300-mA LDOs ..................................................... ADP5040
Micro-PMU, 1.2-A buck, two 300-mA
LDOs, watchdog .................................................. ADP5041
Regulator, dc-to-dc, high-efficiency,
step-down, 4 A ...................................................... ADP2164
Sensor, temperature, 16-bit,
automotive, +175°C .............................................. ADT7312
Analog Dialogue, www.analog.com/analogdialogue, the technical
magazine of Analog Devices, discusses products, applications,
technology, and techniques for analog, digital, and mixed-signal
processing. Published continuously for 46 years—starting in 1967—it is
available in two versions. Monthly editions offer technical articles; timely
information including recent application notes, new-product briefs,
webinars and tutorials, and published articles; and Potpourri, a
universe of links to important and relevant information on the
Analog Devices website, www.analog.com. Printable quarterly
issues and ebook versions feature collections of monthly articles.
For history buffs, the Analog Dialogue archive, www.analog.com/
library/analogdialogue/archives.html, includes all regular editions,
star ting wit h Volume 1, Number 1 (1967), and t hree special
anniversary issues. To subscribe, please go to www.analog.com/library/
analogdialogue/subscribe.html. Your comments are always welcome:
Facebook: www.facebook.com/analogdialogue; Analog Diablog:
analogdiablog.blogspot.com; Email: [email protected] or
Dan Sheingold, Editor [[email protected]] or Scott Wayne,
Publisher and Managing Editor [[email protected]].
ISSN 0161-3626 ©Analog Devices, Inc. 2012
Robust Amplifiers Provide
Integrated Overvoltage
Protection
By Eric Modica and Michael Arkin
Faulty performance, or even damage, can occur when an op
amp’s input voltage exceeds the specified input-voltage range,
or—in extreme cases—the amplifier’s supply voltage. This
article discusses some common causes and effects of overvoltage
conditions, how cumbersome overvoltage protection can be added
to an unprotected amplifier, and how the integrated overvoltage
protection of newer amplifiers provides designers with a compact,
robust, transparent, cost-effective solution.
All electronic components have upper limits to the applied
voltages they can tolerate. When any of these upper limits are
exceeded, the effects can range from momentary interruption of
operation to system latch-up to permanent damage. The amount
of overvoltage a given component can tolerate depends on several
factors, including whether the part is installed or incidentally
contacted, the amplitude and duration of the overvoltage event,
and the robustness of the device.
Precision amplifiers, often the first component in sensor
measurement signal chains, are the most exposed to overvoltage
faults. When selecting a precision amplifier, system designers
must be aware of the common-mode input range of the amplifier. On
the data sheet, the common-mode input range may be specified
by the input voltage range (IVR), or in the test conditions for the
common-mode rejection ratio (CMRR), or both.
Real World Causes of Overvoltage Conditions
Amplifiers require: overvoltage protection to protect against faults
caused by power-supply sequencing, sleep-mode switching, and
voltage spikes; and ESD protection to protect against faults caused
by electrostatic discharge (ESD), even during handling. When
installed, the device can be subjected to system power sequence
conditions, which cause repetitive overvoltage stress. System
designers seek methods of routing the fault currents away from
sensitive components, or limiting those fault currents enough to
avoid damage.
In complex distributed power architecture (DPA) systems with multiple
supply voltages, power-supply sequencing allows the supplies
powering various portions of the system circuitry to turn on and off
at different times. Improper sequencing can cause overvoltage and
latch-up conditions to occur on any pin on any device.
With an increasing focus on energy efficiency, many systems
implement complex sleep and standby modes. This means that
some sections of a system may be powered down while others
may remain powered and active. As with supply sequencing,
these situations can cause unpredictable overvoltage events, but
primarily on input pins.
Many types of sensors can generate unexpected output spikes
that are unrelated to the physical phenomena they are meant to
measure. This type of overvoltage condition generally affects only
input pins.
Electrostatic discharge is a well-known overvoltage event that
often occurs before the component is installed. The damage it
can cause is so prevalent that industry-driven specifications,
such as JESD22-A114D, determine how to test and specify the
Analog Dialogue Volume 46 Number 1
semiconductor’s ability to withstand various types of ESD events.
Almost all semiconductor products incorporate some form of
integrated protection devices. The AN-397 Application Note,
“Electrically Induced Damage to Standard Linear Integrated
Circuits: The Most Common Causes and the Associated Fixes
to Prevent Reccurrence,” is a good reference that covers this topic
in detail. ESD cells are designed to go into a low-impedance state
after a high-energy pulse. This does not limit the input current,
but it does provide a low-impedance path to the supply rails.
A Simple Case Study: Power-Supply Sequencing
As mixed-signal circuits become ubiquitous, so too does the
need for multiple supplies on a single PCB. See the AN-932
Application Note, “Power Supply Sequencing,” for a look at some
subtle issues to consider in new designs, especially when several
unrelated power supplies are required.
Precision amplifiers can fall victim to this condition. Figure 1 shows
an op amp configured as a differential amplifier. The amplifier
senses the current through R SENSE and provides an output
proportional to the resultant voltage drop. Care must be taken that
the divider formed by R 3 and R4 biases the inputs somewhere within
the specified IVR. If the amplifier’s supply voltage is not derived
from VSY, and VCC comes up after VSY, the voltage at the inverting
input of A1 will be:
V– = VSY – (I– × R1)
(1)
where I– depends on the input impedance of A1 with no supply.
If the amplifier is not designed to handle overvoltage conditions,
the most likely current path will be through an ESD diode, clamp
diode, or parasitic diode to the power supply or ground. Damage
can occur if this voltage falls outside of the IVR or if the current
exceeds the data sheet maximum rating.
The ESD structures used on overvoltage-protected amplifiers,
such as the ADA4091 and ADA4096, are not diodes, but DIAC
(bidirectional “diode for alternating current”) devices, making these
amplifiers tolerant of overvoltage conditions, even without power.
VCC
R2
VSY
R1
V–
RSENSE R3
I–
A1
VOUT
R2
LOAD
R4
R1
=
VOUT =
R4
R3
R2
(I
× RSENSE)
R1 LOAD
Figure 1. Differential amplifier high-side current sensor. If
VSY powers up before VCC, the amplifier’s input voltage or
current can exceed the data sheet maximum.
Fault Conditions in Operational Amplifiers
Figure 2 shows an N-channel JFET input stage (J1, J2 , R1, and
R 2), followed by a secondary gain stage and output buffer (A1).
When the open-loop amplifier is within its specified IVR, the
differential input signal (V IN+ – V IN–) is 180 degrees out of phase
with V DIFF. When connected as a unity-gain buffer, as shown, if
the common-mode voltage at V IN+ exceeds the amplifier’s IVR,
J1’s gate-drain will un-pinch and conduct the entire 200-μA stage
current. As long as J1’s gate-drain voltage remains reverse-biased,
a further increase at V IN+ causes no change in V DIFF (VOUT stays
at the positive rail). Once J1’s gate-drain becomes forward-biased,
however, a further increase in V IN+ raises the voltage at A1’s
inverting input, causing an undesirable phase-reversal between
the input signal and V DIFF.
3
VCC
R1
10k𝛀
VCC
–
VOUT
A1
VIN–
D3ESD
+
J2
Figure 4. A bipolar input stage showing ESD, and
differential protection diodes.
VIN–
I1 = 200𝛍A
VEE
Figure 2. A conceptual N-channel, JFET-input op amp.
Figure 3 shows an example of phase reversal at the output of
A1. Unlike bipolar input amplifiers, JFET amplifiers are prone
to phase reversal because their inputs are not clamped. CMOS
amplifiers are typically immune to phase reversal because the
gates are electrically isolated from the drains. If phase inversion
doesn’t occur, op amp manufacturers will often state this on the
data sheet. Phase inversion is possible if: the amplifier inputs are
not CMOS, the maximum differential input is VSY, and the data
sheet does not claim immunity to phase inversion. Although phase
inversion by itself is nondestructive, it can cause positive feedback,
which leads to instability in servo loops.
15
Figure 5 shows the input current-voltage relationship of an
unprotected bipolar op amp with differential input and overvoltage
applied simultaneously. Once the applied voltage exceeds a diode
drop, the current can become destructive, degrading or even
destroying the op amp.
5
4
3
2
1
0
–1
–2
–3
VIN
10
–4
–5
5
VOLTAGE (V)
D4ESD
VEE
INPUT CURRENT (mA)
J1
D2ESD
VIN+
VDIFF
VIN+
CLAMP
D1ESD
R2
10k𝛀
–5
0
–4
–3
–2
–1
0
1
2
INPUT VOLTAGE (V)
3
4
5
OVERVOLTAGE CURVE TRACER TEST SETUP
COLLECTOR
CURVE
TRACER BASE
–5
3
2
7
DUT
4
–10
VOUT
–15
–20
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
Figure 5. Op amp input current as the differential input
voltage exceeds a diode drop.
1.5
1.6
1.7
1.8
TIME (ms)
Figure 3. When VIN exceeds the specified IVR, input
phase inversion causes the output of the amplifier to
become negative.
System designers also have to be concerned about what happens
when the amplifier inputs are pulled outside the power supplies.
Most often this fault condition occurs when power-supply
sequencing causes a source signal to be active before the amplifier
supplies turn on, or when a power supply spikes during turn-on,
turn-off, or in operation. This condition is destructive for most
amplifiers, especially if the overvoltage is greater than a diode drop.
Figure 4 shows a typical bipolar input stage with ESD protection
diodes and clamp diodes. In a buffer configuration, when V IN+
exceeds either rail, ESD and clamp diodes will be forward biased.
With very low source impedance, these diodes will conduct as
much current as the source will allow. Precision amplifiers, such
as the AD8622, provide a modicum of differential protection by
including 500-Ω resistors in series with the inputs to limit the input
current when a differential voltage is applied, but they protect only
as long as the maximum input current specification isn’t exceeded.
If the maximum input current is 5 mA, then the maximum allowed
differential voltage is 5 V. Note that these resistors are not in series
with the ESD diodes, so they cannot limit current to the rails (for
example, during an overvoltage condition).
4
External Input Overvoltage Protection
From the earliest days of semiconductor op amps, IC designers
have had to deal with trade-offs between chip architecture and
the external circuitry needed to deal with its weaknesses. Fault
protection has been among the most difficult of problem areas
(for examples, see MT-036, “Op Amp Output Phase-Reversal
and Input Over-Voltage Protection” and MT-069, “In-Amp Input
Overvoltage Protection”).
Two properties system designers need precision op amps for are
their low offset voltage (VOS) and high common-mode rejection
ratio (CMRR), both of which simplify calibration and minimize
dynamic error. To maintain these specifications in the presence
of electrical overstress (EOS), bipolar op amps often include
internal clamp diodes and small limiting resistors in series with
their inputs, but these cannot address fault conditions caused
when the inputs exceed the rails. To add protection, the system
designer can implement circuitry such as that shown in Figure 6.
VCC = +15V
RFB
VIN
ROVP
D2
D1
VEE = –15V
Figure 6. Precision op amp with external protection using a
current-limiting resistor and two Schottky diodes. RFB is set
equal to ROVP to balance offsets due to input bias currents.
Analog Dialogue Volume 46 Number 1
ROVP will limit the current into the op amp if the signal source at V IN
powers up first. Schottky diodes have a forward voltage 200 mV less
than typical small-signal diodes, so all of the overvoltage current
will be shunted through external diodes D1 and D2. However,
these diodes can degrade the op amp specifications. For example,
the reverse leakage plots from the 1N5711 (see Figure 7) can be
used to determine the CMRR penalty for a given OVP resistor.
The reverse leakage of the 1N5711 is 0 nA at 0 V and 60 nA at
30 V. With a common mode of 0 V, the additional IOS caused by
D1 and D2 depends on how well their leakages match. When V IN
is taken to +15 V, D1 will be reverse biased by 30 V, and D2 will
have 0 V bias. Thus, an additional 60 nA flows into ROVP. When
the input is taken to –15 V, D1 and D2 swap positions electrically,
and 60 nA flows out of ROVP. The additional IOS caused by the
protection diodes at any common mode is simply:
IOSaddr = ID1 – ID2
(2)
100
REVERSE CURRENT (𝛍A)
150°C
125°C
10
the supplies. If, for example, the positive supply cannot sink a
significant amount of current, the overvoltage current can force
the positive supply voltage to increase.
One way to prevent this is to use back-to-back Zener diodes
from the positive input to ground, as shown in Figure 8. When
the Zener voltage of either D1 or D2 is exceeded, the diode
shunts the overvoltage current to ground, protecting the power
supplies. This configuration prevents charge pumping during
overvoltage conditions, but Zener diodes have higher leakage
current and capacitance than small-signal diodes. In addition,
Zener diodes have a soft-knee characteristic in their leakage
current profi le. This, as described previously, adds an additional
CMRR penalty over the amplifier’s common-mode range. For
example, the BZB84-C24 is a back-to-back Zener diode pair
with a working voltage between 22.8 V and 25.6 V. The reverse
current is specified as 50 nA max at 16.8 V, but the manufacturer
doesn’t specify what the leakage is closer to the Zener voltage.
Also, to achieve a sharper breakdown characteristic, Zener diodes
are generally made of more highly doped diffusions than their
small-signal cousins. This causes a relative increase in parasitic
capacitance, which translates to increased distortion (especially
at higher amplitudes) and increased instability.
100°C
1
VCC
75°C
RFB
50°C
0.1
ROVP
VIN
D2
25°C
D1
0.01
0
10
20
30
40
50
60
VEE
Figure 8. Precision op amp with external protection using
a current-limiting resistor and two Zener diodes.
70
CONTINUOUS REVERSE VOLTAGE (V)
Early Integrated Overvoltage Protection
Figure 7. 1N5711 reverse current vs. continuous
reverse voltage.
From Equation 2, the VOS penalty can be derived at the extremes
of the common-mode range as follows:
VOSpenalty = IOSaddr × ROVP
(3)
The previous section discussed drawbacks to some commonly
used methods of external amplifier protection. Some of these
drawbacks could be avoided if the amplifier itself is designed
to tolerate a large input overvoltage. Figure 9 shows a common
integrated protection scheme used on a differential input pair.
VCC
Using 60 nA as the leakage of the 1N5711 at 30 V, and a 5-kΩ
protection resistor, VOS at each extreme would be increased by
300 μV, causing an additional 600 μV ∆VOS over the entire inputvoltage range. In data sheet terms, an op amp with 110-dB CMRR
would suffer a 17-dB reduction. Inserting a feedback resistor to
equalize source impedance only helps when the common mode
is 0 V, and does nothing to prevent additional IOS over the full
common-mode range. Table 1 shows the same calculation for
diodes commonly used for protecting precision amplifiers. For
CMRR penalty calculations, a 5-kΩ protection resistor is assumed.
All costs are recent quotes (2011) in USD from www.mouser.com.
Another possible drawback to the method shown in Figure 6
is that the protection diodes shunt the overvoltage current into
D1
VIN+
D5
R1
D6
D2
D3
R2
VIN–
D4
VEE
Figure 9. A differential input pair with resistive overvoltage
protection (ESD protection not shown).
This circuit includes input protection resistors on both amplifier
inputs. Although overvoltage protection is generally needed on
only one input, equalizing the parasitic capacitance and leakage
Table 1. Commonly Used Protection Diodes and Their Impact on a 110-dB CMRR Precision Op Amp
1N5711
BAV99
PAD5
BAS70-04
1N914
BZB84-C24
IOSaddr (nA)
60
10
<<0.005
8
40
50
VOSpenalty (μV)
600
100
0
80
400
500
CMRR Penalty (dB)
17
6
0
5
14
16
Cost @ 1k Quantities
$0.07
$0.015
$3.52
$0.095
$0.01
$0.034
Analog Dialogue Volume 46 Number 1
5
at each input reduces distortion and offset current. Furthermore,
the diodes do not have to handle ESD events, so they can be
relatively small.
Adding resistance, either external or internal, adds to the
amplifier’s root-sum-square (RSS) thermal noise (Equation 4):
En,total = (en,op amp)2 + (en,Rovp)2 + (RS × in,op amp)2
(4)
If a 1-kΩ resistor is used to protect an op amp with 4-nV/√Hz
of noise, the total voltage noise will increase by √2. Integrating
the protection resistors doesn’t change the fact that overvoltage
protection increases the input-referred voltage noise, but
integrating R1 and R 2 with the op amp ensures that the data sheet
noise specification covers the protective circuitry.
To avoid the noise-overvoltage trade-off requires a protective circuit
that presents a low resistance when the amplifier inputs are within
the specified range and a very high resistance when the amplifier
inputs exceed the rails. This characteristic would provide improved
overvoltage protection on-demand, thus lowering the overall noise
contribution under normal operating conditions. Figure 10 shows
one circuit implementation that behaves in this way.
The Benefits of Integration
Amplifiers such as the ADA4091 and ADA4096 demonstrate that
robust, overvoltage-tolerant op amp inputs can be achieved with
a minimal impact on precision (as in Figure 10). The ADA4096
provides 32-V protection, regardless of supply levels—eliminating
the need for external components that can either be inexpensive
but vastly degrade the amplifier’s precision, or precise but more
costly than the amplifiers themselves.
Figure 12 shows the ADA4096-2 in a 2-mm × 2-mm LFCSP
package next to a couple of discrete components often used
for external input protection. The ADA4096-2’s integrated
protection provides a significantly reduced PCB footprint; its
effects are included in the op amp’s specifications; and it protects
the amplifier even when power is not applied (see Figure 13). In
addition, the ADA4091 and ADA4096 have rail-to-rail inputs
and outputs (RRIO) and are free from phase-inversion over the
entire overvoltage protection range (see Figure 14). These benefits
allow system designers to worry a little less about power-supply
sequencing and latch-up.
VCC
J2B
J1A
J2A
BAS70-4-V
5k𝛀
VIN–
ADA4096-2
VEE
2mm
Figure 12. The ADA4096-2 in a 2-mm × 2-mm package
occupies less space than two components commonly
used for external voltage protection.
6
5
4
3
VCC = +15V
VEE = –15V
7
INPUT BIAS CURRENT (mA)
Figure 10. Input differential pair with active overvoltage
protection.
Jxy are all P-channel JFETs; they are depletion mode devices, so
the channel is the same polarity as the source and drain. When
the amplifier input levels are between the rails, J1A and J2A act
as simple resistors with resistance equal to R DSON because the
input bias currents are small enough that any potential difference
between the channel and gate won’t pinch the channel closed. If
V IN+ were to exceed the negative supply by a diode drop, current
would begin to flow through J1A , causing the drain to pinch
closed. This transition is actually J1A moving out of triode and
into the linear region. If V IN+ were to exceed the positive supply
voltage by a diode drop, J1A would act as a lateral PNP. V IN+ to
the gate would act as a forward-biased emitter-base junction,
with the other junction acting as the base-collector, standing
off overvoltage.
The current-voltage plot of Figure 11 shows the change in input
impedance of a FET-protected op amp when subjected to an
overvoltage sweep. The R DSON of the protection FET is 4.5 kΩ; as
the amplifier’s positive input is pulled above the rail, the protection
FET’s resistance increases rapidly to 22 kΩ at 30 V, limiting the
input current to 1.5 mA.
VCC = VEE = 0V
VIN+
J1B
2
1
0
–1
–2
–3
–4
–5
LOW RDSON SERIES FET
5k𝛀 SERIES RESISTOR
–6
–7
–48 –40 –32 –24 –16 –8
0
8
16
24
32
40
48
VIN (V)
Figure 13. ADA4096-2 input OVP current limiting with and
without power.
T
1.4
1.2
CURRENT (mA)
REFF = 12.5k𝛀
1.0
REFF = 22.0k𝛀
REFF = 4.5k𝛀
0.8
0.6
0.4
0.2
CH1 10.0V
0
0
5
10
15
20
25
30
OVERVOLTAGE (V)
Figure 11. Effective input impedance of a FET protected
op amp when subjected to a dc overvoltage sweep.
6
CH2 10.0V
M20.0ms
T 34.20%
A CH1
–3.6V
Figure 14. ADA4096-2 on ±10-V supplies with the inputs
pulled 30 V above and below the rails.
(continued on Page 23)
Analog Dialogue Volume 46 Number 1
Modern DACs and
DAC Buffers Improve
System Performance,
Simplify Design
By Padraic O’Reilly and Charly El-Khoury
At the heart of many control systems, digital-to-analog
converters (DACs) play a key role in determining system
performance and accuracy. This article looks at a new precision
16-bit DAC and shows some ideas for buffering the outputs of
high-speed complementary current-output DACs that can rival
transformer performance.
Voltage-Switching, 16-Bit DACs Provide Low
Noise, Fast Settling, Improved Linearity
The obvious advantage of using multiplying DACs in voltageswitching mode is that no signal inversion occurs, so a positive
reference voltage results in a positive output voltage. But the R-2R
ladder architecture also has a weakness when used in this mode.
The nonlinear resistance of the N-channel switches used in series
with the R-2R ladder will degrade the integral linearity (INL), as
compared to when the same DAC is used in current-steering mode.
Newer high-resolution DACs, such as the AD5541A, shown
in Figure 3, have been developed to overcome the limitations
of multiplying DACs while maintaining the benefits of voltage
switching. Using a partially segmented R-2R ladder network
and complementary switches, the AD5541A achieves ±1-LSB
accuracy at 16 bits without adjustment over the full specified
temperature range of −40°C to +125°C, 11.8-nV/√Hz noise, and
1-μs settling time.
R
2R
VREF
2R
2R . . . . .
2R
2R
2R . . . . .
2R
S0
S1 . . . . .
S11
E1
E2 . . . . .
E15
12-BIT R-2R LADDER
Resistance-ladder multiplying DACs, based on the game-changing
10-bit CMOS AD7520—introduced nearly 40 years ago—were, at
fi rst, used with inverting op amps, with the amplifier’s summing
point (IOUTA) providing a convenient virtual ground (Figure 1).
R
R
2R
2R
2R
2R
S1
S2
S3
S8
2R
R
VOUT
VREF
By Padraic O’Reilly
R
R
FOUR MSBs DECODED
INTO 15 EQUAL SEGMENTS
Figure 3. AD5541A architecture.
Performance Features
Settling Time: Figure 4 and Figure 5 compare the respective settling
times of a multiplying DAC in voltage mode and the AD5541A.
The AD5541A has a settling time of approximately 1 μs when
capacitive loading on the output is minimized.
RFBA
IOUTA
AGND
SCLK (0.5V/DIV)
DAC DATA LATCHES
AND DRIVERS
Figure 1. CMOS multiplying DAC architecture.
However, they can also be used, with some limitations, in a
voltage-switching confi guration that provides a noninverting
voltage output, with the op amp used as a voltage buffer (Figure 2).
Here the reference voltage, V IN, is applied to IOUT, and the output
voltage, VOUT, is available at V REF. A 12-bit version, optimized for
this purpose, soon became available.
VOUT (0.1V/DIV)
0
VDD
R1
2
4
6
8
10
12
14
16
18
20
TIME (μs)
R2
Figure 4. Multiplying DAC settling time.
RFB
VIN
IOUT
AGND
VDD
MULTIPLYING
DAC VREF
CS (2.5V/DIV)
VOUT
GND
10pF
NOTES
1. ADDITIONAL PINS OMITTED FOR CLARITY.
2. C1 PHASE COMPENSATION (1pF TO 2pF) MAY BE REQUIRED
IF A1 IS A HIGH SPEED AMPLIFIER.
50pF
100pF
Fast-forward to the present: As single-supply systems become
increasingly common, designers are faced with the challenge of
trying to maintain the level of performance achieved at higher
voltages while keeping power consumption in check. The need
has grown for devices with higher resolution (to 16 bits), capable
of being used in this mode.
Analog Dialogue Volume 46 Number 1
VOUT (0.25V/DIV)
200pF
Figure 2. Multiplying DAC in voltage-switching mode.
0
2
4
6
8
10
12
14
16
18
20
TIME (μs)
Figure 5. AD5541A settling time.
7
Noise Spectral Density: Table 1 compares noise spectral density of
the AD5541A and the multiplying DAC. The AD5541A exhibits
slightly better performance at 10 kHz and far better performance
at 1 kHz.
1.0
Table 1. Noise Spectral Density of AD5541A vs.
Multiplying DAC
0.4
VDD = 5V
0.8 VREF = 1.25V
TA = 25°C
0.6
INL (LSB)
0.2
0
NSD @ 10 kHz
(nV/√Hz)
NSD @ 1 kHz
(nV/√Hz)
AD5541A
12
12
–0.4
MDAC
30
140
–0.6
DAC
–0.2
–0.8
Integral Nonlinearity: Integral nonlinearity measures the maximum
deviation between the ideal output of a DAC and the actual output
after gain and offset errors have been removed. The switches used
in series with the R-2R network can affect the INL. Multiplying
DACs generally employ NMOS switches. When used in voltageswitching mode, the source of the NMOS switch is connected to
the reference voltage, the drain is connected to the ladder, and
the gate is driven by the internal logic (Figure 6).
–1.0
0
0.4k 0.8k 1.2k 1.6k 2.0k 2.4k 2.8k 3.2k 3.6k
4k
CODE
Figure 7. INL of IOUT multiplying DAC in reverse mode,
VDD = 5 V, VREF = 1.25 V.
1.0
VDD = 5V
0.8 VREF = 2.5V
TA = 25°C
0.6
0.4
D
G
S
VIN
INL (LSB)
VLOGIC
0.2
0
–0.2
–0.4
Figure 6. Multiplying DAC switch.
–0.6
For current to flow in an NMOS device, VGS must be greater than the
threshold voltage, VT. In voltage-switching mode, VGS = VLOGIC – VIN
must be greater than VT = 0.7 V.
RON
1
E (VGS VT )
–1.0
0
0.4k 0.8k 1.2k 1.6k 2.0k 2.4k 2.8k 3.2k 3.6k
4k
CODE
Figure 8. INL of IOUT multiplying DAC in reverse mode,
VDD = 5 V, VREF = 2.5 V.
200
VDD = 3V
180 VREF = 2.5V
TA = 25°C
160
140
INL (LSB)
The R-2R ladder of a multiplying DAC is designed to divide the
current evenly through each of the legs. This requires the overall
resistance to ground, seen from the top of each leg, to be exactly
the same. This can be accomplished by scaling the switches, where
the size of each switch is proportional to its on resistance. If the
resistance in one leg changes, the current flowing through that
leg will change, causing a linearity error. V IN cannot be so large
as to shut off the switch, but it must be large enough to keep the
switch resistance low, as changes in V IN affect VGS and, therefore,
cause a nonlinear change in on resistance as shown by:
–0.8
120
0
–120
–140
This change in on resistance will unbalance the currents and
degrade the linearity. Thus, the supply voltage on the multiplying
DAC cannot be reduced too much. Conversely, the reference
voltage should be no more than 1 V above AGND to maintain
linearity. With a 5-V supply, the linearity starts to degrade when
moving from a 1.25-V reference to a 2.5-V reference, as shown in
Figure 7 and Figure 8. The linearity falls apart altogether when
the supply voltage is decreased to 3 V, as shown in Figure 9.
8
–160
–180
–200
0
0.4k 0.8k 1.2k 1.6k 2.0k 2.4k 2.8k 3.2k 3.6k
4k
CODE
Figure 9. INL of multiplying DAC in reverse mode,
VDD = 3 V, VREF = 2.5 V.
Analog Dialogue Volume 46 Number 1
R
R
2R
2R
Figure 13 and Figure 14 show that the linearity changes very little
over a wide range of reference and supply voltages. The DNL
behavior is similar to that of the INL. The AD5541A linearity is
specified over temperature and supply voltage; and the reference
voltage can go from 2.5 V to the supply voltage.
0.50
VREF = 2.5V
TA = 25°C
0.25
LINEARITY ERROR (LSB)
To minimize this effect, the AD5541A uses complementary
NMOS/PMOS switches, as shown in Figure 10. Now, the total
on resistance of the switch comes from the parallel contribution
of the NMOS and PMOS switches. As previously shown, the gate
voltage of the NMOS switch is controlled by internal logic. An
internally generated voltage, VGN, sets the ideal gate voltage to
balance the on resistance of the NMOS with that of the PMOS.
The switches are sized to scale with code so the on resistance will
scale with code. Thus, the currents will scale, and accuracy will
be maintained. As the reference input sees an impedance that
varies with code, it should be driven from a low impedance source.
DNL
0
–0.25
–0.50
VGN
VLOGIC
INL
–0.75
VREF
Figure 10. Complementary NMOS/PMOS switches.
3
4
5
SUPPLY VOLTAGE (V)
6
7
Figure 13. AD5541A INL vs. supply voltage.
Figure 11 and Figure 12 show the INL performance of the
AD5541A with 5-V and 2.5-V references.
0.75
0.2
VDD = 5V
TA = 25°C
LINEARITY ERROR (LSB)
0.50
0.1
0
INL (LSB)
2
–0.1
–0.2
–0.3
DNL
0.25
0
INL
–0.25
–0.4
–0.5
–0.50
VDD = 5.5V
–0.6 VREF = 5V
0
5k 10k 15k 20k 25k 30k 35k 40k 45k 50k 55k 60k 65k
CODE
Figure 11. INL of AD5541A, VDD = 5.5 V, VREF = 5 V.
0.50
VDD = 5V
VREF = 2.5V
INL (LSB)
0.25
0
–0.25
–0.50
–0.75
0
8192
16384 24576 32768 40960 49152 57344 65536
CODE
Figure 12. INL of AD5541A, VDD = 5.5 V, VREF = 2.5 V.
Analog Dialogue Volume 46 Number 1
0
1
2
3
4
REFERENCE VOLTAGE (V)
5
6
Figure 14. AD5541A INL vs. reference voltage.
More About the AD5541A
The AD5541A serial-input, single-supply, voltage-output
nanoDAC+ digital-to-analog converter provides 16-bit resolution
with ±0.5-LSB typical integral- and differential nonlinearity.
It is well suited to applications that use multiplying DACs in
voltage-switching mode. It performs well over both temperature
and supply voltage, achieves excellent linearity, and can be used
in 3-V to 5-V systems where precision dc performance and quick
settling are required. Using an external reference voltage that
can range from 2 V to the supply voltage, the unbuffered voltage
output can drive a 60-kΩ load from 0 V to V REF. Featuring 1-μs
settling to ½ LSB, 11.8-nV/√Hz noise, and low glitch, the device is
well suited for deployment in a wide variety of medical, aerospace,
communications, and industrial applications. Its 3-wire, lowpower SPI-compatible serial interface can be clocked at up
to 50 MHz. Operating on a single 2.7-V to 5.5-V supply, the
AD5541A draws only 125 μA. Available in 8-lead and 10-lead
LFCSP and 10-lead MSOP packages, it is specified from –40°C
to +125°C and priced from $6.25 in 1000s.
9
High Speed Current Output DAC Buffers
In Figure 16, this circuit is applied with a quad high-speed, lowpower, 14-bit DAC, where the complementary current-output stage
increases the speed and reduces the distortion of low-power DACs.
Transformers are often considered to be the best option for
converting the complementary output of a high-speed currentoutput DAC to a single-ended voltage output, as transformers
do not add noise or consume power. Although transformers
operate well with high-frequency signals, they cannot handle
the low-frequency signals required for many instrumentation
and medical applications. These applications require a lowpower, low-distortion, low-noise, high-speed amplifier to
convert the complementary current to a single-ended voltage.
The three circuits presented here accept the complementary
output currents from the DAC and provide a single-ended
output voltage. Distortion for the last two is compared with
a transformer solution.
Difference Amplifi er: The AD8129 and AD8130 differentialto-single-ended amplifi ers (Figure 15) are used in the fi rst
circuit (Figure 16). They feature extremely high commonmode rejection at high frequency. The AD8129 is stable for
gains of 10 or more, whereas the AD8130 is stable with unity
gain. Their user-adjustable gain can be set by the ratio of two
resistors, R F and RG . The AD8129 and AD8130 have very
high input impedance on Pin 1 and Pin 8, regardless of the
gain setting. A reference voltage (V REF, Pin 4) can be used
to set a bias voltage that is multiplied by the same gain as the
differential input voltage.
PD
7
VO
6
VREF
4
2
5
RG
RF
AD8129 ±12V, CG = 3.3pF, CF = 33pF
AD8129 ±5V, CG = 3.3pF, CF = 33pF
–40
–45
–50
–55
–60
0.1
1
10
100
FREQUENCY (MHz)
Op Amp at Unity Gain: The second circuit (Figure 18) uses a
high-speed amplifier with two RT resistors. The amplifier simply
transforms the complementary currents, I1 and I2, into a singleended output voltage, VO, through RT. This simple circuit does
not allow signal amplification using the amplifier as a gain block.
1
8
–35
Figure 17. Distortion of the DAC and AD8129 with VO = 8 V p-p.
+VS
3
VIN
Figure 17 shows the spurious-free dynamic range (SFDR) of the
circuit, as a function of frequency, using the DAC and the AD8129,
with R F = 2 kΩ, RG = 221 Ω, RT = 100 Ω, and VO = 8 V p-p, at two
values of the supply voltage. Here, the AD8129 was chosen because
it provides a large output signal, is stable for G = 10, and has a high
gain-bandwidth product compared to the AD8130. The SFDR is
generally better than 55 dB for both cases, to beyond 10 MHz, with
approximately >3-dB improvement at the lower supply voltage.
SFDR (dBc)
By Charly El-Khoury
–VS
CF
RT
Figure 15. AD8129/AD8130 difference amplifiers.
+VS
+VS
D11 TO D0
DAC
I1
D11 TO D0
RT
I2
RT
VREF
AD8129/
AD8130
I2
OP AMP
VO
–VS
RT
CF
RF
Figure 18. Simple differential-to-single-ended converter
using an op amp.
Figure 16. DAC buffer using AD8129/AD8130.
Equation 1 and Equation 2 show the relation between the output
voltage of the amplifier and the complementary output current of
the DAC. The termination resistors, RT, perform a current-tovoltage conversion; the ratio of R F and RG determines the gain.
V REF is set to 0 in Equation 2.
VIN I1RT I2RT RTI1 I2
(1)
§
RF ·¸
¨
VO ¨
¸ V VREF
©
RG ¹ IN
§
RF ·¸
¨¨ ¸ R I I ©
RG ¹ T 1 2
(2)
10
I1
VO
–VS
RG
DAC
Equation 3 shows the relationship between VO and the DAC output
current. Distortion data was measured with 5-pF capacitors in
parallel with RT.
VO I1RT I2RT RTI1 I2
(3)
To demonstrate the performance of this circuit, the DAC
was paired with the A DA4857 and A DA4817 op amps,
with R T = 125 Ω (and CT = CF = 5 pF in parallel with RT for
stability and low-pass fi ltering). The single ADA4857-1 and dual
ADA4857-2 are unity-gain stable, high-speed, voltage-feedback
amplifiers with low distortion, low noise, and high slew rate. An
ideal solution for a variety of applications, including ultrasound,
ATE, active filters, and ADC drivers, it features 850-MHz
Analog Dialogue Volume 46 Number 1
bandwidth, 2800-V/μs slew rate, and 10-ns settling time to
0.1%—all while operating on 5 mA of quiescent current. With
a wide supply voltage range (5 V to 10 V), the ADA4857-1 and
ADA4857-2 are ideal candidates for systems that require wide
dynamic range, precision, high speed, and low power.
The single ADA4817-1 and dual ADA4817-2 FastFET ™ amplifiers
are unity-gain stable, ultrahigh-speed, voltage-feedback op amps
with FET inputs. Developed on ADI’s proprietary eXtra Fast
Complementary Bipolar (XFCB) process, they achieve ultralow
noise (4 nV/√Hz and 2.5 fA/√Hz) and very high input impedance.
With 1.3-pF input capacitance, 2-mV maximum offset voltage,
low power dissipation (19 mA), and wide −3-dB bandwidth
(1050 MHz), they are ideal for data acquisition front ends,
photodiode preamps, and other wideband transimpedance
applications. With a 5-V to 10-V supply voltage range and the
ability to operate on either single or dual supplies, they are
designed to work in a variety of applications, including active
fi ltering, ADC driving, and DAC buffering.
–20
–30
§
RT1uRF
¨
¨
VO I1 © R R R
T1
I2
TRANSFORMER
ADA4857
ADA4817
–40
–50
§
¨
¨1
©
RF ·
R R ¸¸¹
G
T2
§ RT2uRF ·
¨
¸
¨ R R ¸
© G
T2 ¹
(4)
TRANSFORMER
ADA4857
ADA4817
–60
–70
–40
SFDR (dBc)
F
·
¸
¸
G¹
Figure 21 compares the distortion of the amplifiers in this
configuration with that of the transformer circuit. RT1 = 143 Ω,
RT2 = 200 Ω, R F = RG = 499 Ω, CF = 5 pF—for stability and high
frequency fi ltering—and R L = 1 kΩ. Here, the performance of
the ADA4817 is comparable to that of the transformer at high
frequency, maintaining better than −70 dBc SFDR up to 70 MHz.
Both op amps maintain excellent low-frequency fidelity compared
to the transformer.
SFDR (dBc)
Figure 19 compares the distortion vs. frequency of this circuit at
VO = 500 mV p-p with a circuit using a transformer.The transformer
has less distortion than the amplifier, which has decreasing gain, at
high frequencies, but its distortion becomes increasingly worse at
low frequencies. Here, SFDRs of nearly 90 dB are achievable over
a limited range, with better than 70 dB to 10 MHz.
Equation 4 defines the relationship between the DAC output current
and the amplifier output voltage for VREF1 = VREF2 = 0. To match the
input impedance of the amplifier network looking out of the DAC,
the two termination resistors, RT1 and RT2, must be set individually,
taking into consideration the characteristics of the amplifier.
–80
–50
–90
–60
–100
0.1
–70
1
10
100
FREQUENCY (MHz)
–80
Figure 21. Distortion of the DAC, ADA4817, and ADA4857
with VO = 500 mV p-p.
–90
–100
0.1
1
10
100
FREQUENCY (MHz)
Figure 19. Distortion of the DAC, ADA4857, and ADA4817
with VO = 500 mV p-p, RL = 1 kΩ.
Op Amp with Gain: The third circuit (Figure 20) also uses the same
high-speed op amps but includes a resistive network that distances
the amplifier from the DAC, allows for gain setting, and has the
flexibility to adjust the output bias voltage using either of the two
reference voltages, V REF1 and V REF2.
This article showed some of the advantages of using lowdistortion, low-noise, high-speed amplifiers as DAC buffers
comparing their performance with that of transformers. It also
compares three types of application circuits using two different
amplifi er architectures, while giving examples of measured
data with a DAC and the AD8129, ADA4857-1/ADA4857-2,
and ADA4817-1/ADA4817-2 amplifiers. The data show that the
amplifier outperforms the transformer at frequencies less than
1 MHz and can closely match its performance up to 80 MHz.
Amplifier selection is important when considering trade-offs in
terms of power dissipation and distortion.
CF
VREF1
RT1
D11 TO D0
DAC
+VS
RG
I1
RG
I2
Authors
RF
RT2
VREF2
OP AMP
VO
–VS
RF
CF
Figure 20. Differential-to-single-ended with gain and
bias capability.
Analog Dialogue Volume 46 Number 1
Padraic O’Reilly [[email protected]] is
an applications engineer in the Precision Digitalto-Analog Converters Group. He has worked at
ADI since graduating with a BEng in electronic
engineering from the University of Limerick in 2007.
Charly El-Khoury [[email protected]]
is an applications engineer in the High Speed
Amplifier Group. He has worked at ADI since
graduating with a master’s in ECE from Worcester
Polytechnic Institute (WPI) in 2006.
11
DDS Devices Generate HighQuality Waveforms Simply,
Efficiently, and Flexibly
PHASE ACCUMULATOR
N-BIT CARRY
TUNING
WORD
M
24 TO 48
BITS
By Brendan Cronin
N
PHASE
REGISTER
DAC
SYSTEM
CLOCK
Abstract
Direct digital synthesis (DDS) technology is used to generate and modify
high-quality waveforms in a broad range of applications in such diverse
fields as medicine, industry, instrumentation, communications, and
defense. This article provides an overview of the technology, describes its
benefits and limitations, and takes a look at some application examples—
and new products that make the technology more readily available.
Figure 1. Functional block diagram of a DDS system.
For sine-wave outputs, the phase-to-digital converter is usually
a sine lookup table (Figure 2). The phase accumulator counts by
N, to generate a frequency related to fC according to the equation,
fOUT =
Introduction
A key requirement across a multitude of industries is to accurately
produce, easily manipulate, and quickly change waveforms of
various frequencies and types. Whether a wideband transceiver
requires an agile low-phase-noise frequency source with excellent
spurious-free dynamic performance or an industrial measurement
and control system needs a stable frequency stimulus, the ability to
quickly, easily, and cost effectively generate an adjustable waveform
while maintaining phase continuity is a critical design criterion
that direct digital frequency synthesis can fulfi ll.
The Task of Frequency Synthesis
Increasing spectrum congestion, coupled with the insatiable need
for lower power, higher quality measurement equipment, calls for
the use of new frequency ranges and better exploitation of existing
ones. As a result, better control of frequency generation is being
sought—in most cases with the assistance of frequency synthesizers.
These devices use a given frequency, f C , to generate a waveform
at a related desired frequency (and phase), fOUT. The general
relationship can be written simply as
fOUT = εx × fC
where the scaling factor, εx, is sometimes called the normalized frequency.
The equation is always implemented using algorithms for stepby-step approximations of real numbers. When the scaling factor
is a rational number, a ratio of two relatively prime integers, the
output frequency and the reference frequency will be harmonically
related. In most cases, however, εx can belong to a much broader
set of real numbers, and the approximation process is truncated
as soon as it falls within an acceptable limit.
Direct Digital Frequency Synthesis
One such practical realization of a frequency synthesizer is direct
digital frequency synthesis (DDFS), often shortened to direct digital
synthesis (DDS). The technique uses digital data processing to
generate a frequency- and phase-tunable output related to a fi xed
frequency reference, or clock source, fC. In a DDS architecture,
the reference or system clock frequency is divided down by the
scaling factor, set by a programmable binary tuning word.
Simply stated, a direct digital frequency synthesizer translates a
train of clock pulses into an analog waveform, typically a sine,
triangular, or square wave. As Figure 1 shows, its essential parts
are: a phase accumulator, which produces a number corresponding
to a phase angle of the output waveform, a phase-to-digital
converter, which generates the instantaneous digital fraction of
the output amplitude occurring at a particular phase angle, and
a digital-to-analog converter (DAC), which converts that digital
value to a sampled analog data point.
12
14 TO 16
BITS
fOUT
PHASE-TODIGITAL
CONVERTER
N
f
2M C
where:
M is the resolution of the tuning word (24 bits to 48 bits).
N is the number of pulses of fC corresponding to the smallest
incremental phase change of the phase accumulator’s output word.
REFERENCE
CLOCK
DDS CIRCUITRY
M
PHASE
ACCUMULATOR
AMPLITUDE/SINE
CONV. ALGORITHM
DAC
TUNING WORD SPECIFIES
OUTPUT FREQUENCY AS A
FRACTION OF REFERENCE
CLOCK FREQUENCY
IN DIGITAL DOMAIN
sin(x)/x
Figure 2. Typical DDS architecture and signal path with DAC.
Since changes to N result in immediate changes in the output phase
and frequency, the system is inherently phase-continuous, a critical
attribute in many applications. No loop settling time is required, in
contrast to analog-type systems, such as phase-locked loops (PLLs).
The DAC is usually a high-performance circuit specifically designed
to work with the DDS core (phase accumulator and phase-toamplitude converter). In most cases, the resulting device, often a
single chip, is commonly referred to as a complete DDS or C-DDS.
Practical DDS devices often integrate multiple registers to allow
various frequency- and phase-modulation schemes to be realized.
When included, the phase register’s contents are added after
the phase accumulator. This enables the output sine wave to
be phase-delayed in correspondence with a phase tuning word.
This is extremely useful for phase-modulation applications in
communication systems. The resolution of the adder circuit
determines the number of bits in the phase tuning word and,
therefore, the resolution of the delay.
Integrating a DDS engine and a DAC in a single device has
advantages and disadvantages, but whether integrated or not,
a DAC is required to create a high quality analog signal of
exceptional purity. The DAC converts the digital sine output
into an analog sine wave and may be either single-ended or
differential. A few of the key requirements are low phase noise,
excellent wideband (WB-) and narrow-band (NB-) spurious-free
dynamic range (SFDR), and low power consumption. If it is an
external component, the DAC needs to be fast enough to process
the signal—so devices with a parallel port are common.
Analog Dialogue Volume 46 Number 1
DDS vs. Other Solutions
Other possibilities for frequency generation include analog phaselocked loops (PLLs), clock generators, and using an FPGA to
dynamically program the output of a DAC. A simple comparison of
the technologies can be made by examining spectral performance
and power consumption, qualitatively demonstrated in Table 1.
Table 1. DDS vs. Competing Technologies—
High Level Comparison
Power
Consumption
Spectral
Purity
Comment
Low
Medium
Easy to tune
Discrete Medium
DAC +
FPGA
Medium-high
Ability to
tune
Analog
PLL
High
Difficult to
tune
DDS
Medium
A phase-locked loop is a feedback loop comprising: a phase
comparator, a divider, and a voltage-controlled oscillator (VCO).
The phase comparator compares a reference frequency with the
output frequency (usually divided down by a factor, N), The error
voltage generated by the phase comparator is applied to the VCO,
which generates the output frequency. When the loop has settled,
the output will bear an accurate relationship to the reference
in frequency and/or phase. PLLs have long been recognized as
superior devices for low phase noise and high spurious-free dynamic
range (SFDR) applications requiring high fidelity and stable signals
in a specific band of interest.
Their inability to accurately and quickly tune the frequency output
and waveform and their slow response limits their suitability for
applications such as agile frequency hopping and some frequencyand phase-shift keying applications.
Other approaches, including field-programmable gate arrays
(FPGAs) with embedded DDS engines—in combination with
off-the-shelf DACs to synthesize output sine waves—solve the
frequency-hopping difficulties of PLLs, but have their own
weaknesses. The main system disadvantages include higher
operating and interface power requirements, higher cost, large
size, and additional software-, hardware-, and memory overhead
for the system developer. For example, up to 72 kB of memory
are required to generate a 10-MHz output signal with 60-dB
dynamic range using the DDS engine option on modern FPGAs.
In addition, the designer needs to be comfortable and familiar with
subtle trade-offs and the architecture of the DDS core.
As a practical matter (see Table 2), rapid advances in
CMOS processing, together with modern digital design
techniques and improved DAC topologies, have resulted in
the DDS technology achieving power consumption, spectral
performance, and cost levels that were previously unattainable
for a wide range of applications. While complete DDS products
will never match the highest performance and design flexibility
achievable with custom combinations of high-end DAC
technology and FPGAs, the size-, power- and cost benefits,
coupled with the simplicity of DDS devices, may make them
easily the first choice for many applications.
Also note that since a DDS device fundamentally embodies
a digital method of generating an output waveform, it can
simplify the architecture of some solutions or make it possible
to digitally program the waveform. While a sine wave is
normally used to explain the function and operation of a DDS,
it is easily possible to generate triangular or square (clock)
wave outputs from modern DDS ICs, avoiding the need for a
lookup table in the former case, and for a DAC in the latter
case, where the integration of a simple yet precise comparator
will suffice.
DDS Performance and Limitations
Images and Envelopes: Sin(x)/x Roll-Off
The actual output of the DAC is not a continuous sine wave
but a train of pulses with a sinusoidal time envelope. The
corresponding frequency spectrum is a set of images and aliases.
The images lie along a sin(x)/x envelope (see |amplitude| plot in
Figure 3). Filtering is necessary to suppress frequencies outside
the band of interest, but it cannot suppress higher-order aliases
(due to DAC nonlinearities, for example) appearing within the
pass band.
The Nyquist Criterion dictates that a minimum of two samples
per cycle are required to reconstruct a desired output waveform.
Image responses are created in the sampled output spectrum at
K fCLOCK ± fOUT. In this example, where fCLOCK = 25 MHz and
fOUT = 5 MHz, the fi rst and second images occur (see Figure 3)
at fCLOCK ± fOUT, or 20 MHz and 30 MHz. The third and fourth
images appear at 45 MHz and 55 MHz. Note that the sin(x)/x
nulls appear at multiples of the sampling frequency. In the case
where fOUT is greater than the Nyquist bandwidth (1/2 fCLOCK), the
fi rst image response will appear within the Nyquist bandwidth as
an aliased image (a 15-MHz signal will alias down to 10 MHz, for
example). The aliased image cannot be fi ltered from the output
with a traditional Nyquist antialiasing fi lter.
Table 2. Benchmark Analysis Summary—Frequency-Generation Technologies (<50 MHz)
Phase-Locked Loops
DAC + FPGA
DDS
Spectral Performance
High
Medium-high
Medium
System Power Requirements
High
High
Low
No
Yes
Yes
Tuning Response Time
High
Low
Low
Solution Size/Footprint
Medium
High
Low
Low
Medium
High
Medium
High
Low
Digital Frequency Tuning
Waveform Flexibility
Cost
Design Reuse
Medium
Low
High
Implementation Complexity
Medium
High
Low
Analog Dialogue Volume 46 Number 1
13
So choosing a stable reference clock oscillator with low jitter and
sharp edges is critical. Higher frequency reference clocks allow
greater oversampling, and jitter can be somewhat ameliorated by
frequency division, since dividing the frequency of the signal yields
the same amount of jitter across a longer period, and so reduces
the percentage of jitter on the signal.
NYQUIST BANDWITH
sin(x)/x ENVELOPE
–10dB
–20dB
Noise—Including Phase Noise
0
5
fOUT
12.5 20
FUNDAMENTAL
25
fCLOCK – fOUT
FIRST IMAGE
NYQUIST
LIMIT
30
fCLOCK + fOUT
SECOND IMAGE
fCLOCK
45
50
55
2fCLOCK
2fCLOCK – fOUT
THIRD IMAGE
70
75
MHz
3fCLOCK – fOUT
FIFTH IMAGE
2fCLOCK + fOUT
FOURTH IMAGE
3fCLOCK
Figure 3. Sin(x)/x roll-off in a DDS.
In typical DDS applications, a low-pass filter is utilized to suppress
the effects of the image responses in the output spectrum. To
keep the cutoff requirements of the low-pass fi lter reasonable and
the fi lter design simple, an accepted guideline is to limit the fOUT
bandwidth to approximately 40% of the fCLOCK frequency using
an economical low-pass output fi lter.
T he a mplit ude of a ny g iven i mage i n response to t he
fundamental can be calculated using the sin(x)/x formula.
Because the function rolls off with frequency, the amplitude of
the fundamental output will decrease inversely with its tuned
frequency; in a DDS system, the decrease will be –3.92 dB over
the dc to Nyquist bandwidth.
The amplitude of the fi rst image is substantial—within 3 dB of
the fundamental. To simplify fi ltering requirements for DDS
applications, it is important to generate a frequency plan and
analyze the spectral considerations of the image and the sin(x)/x
amplitude responses at the desired fOUT and fCLOCK frequencies.
Online interactive design tools supporting the Analog Devices
DDS product family allow for quick and easy simulation of where
images lie and allow the user to choose frequencies where images
are outside the band of interest. See the Further Information and
Useful Links section for additional useful information.
Other anomalies in the output spectrum, such as integral and
differential linearity errors of the DAC, glitch energy associated
with the DAC, and clock feedthrough noise, will not follow
the sin(x)/x roll-off response. These anomalies will appear as
harmonics and spurious energy in many places in the output
spectrum—but will generally be much lower in amplitude than
the image responses. The general noise floor of a DDS device is
determined by the cumulative combination of substrate noise,
thermal noise effects, ground coupling, and other sources of signal
coupling. The noise floor, performance spurs, and jitter of a DDS
device are greatly influenced by circuit board layout, the quality
of the power supplies, and—most importantly—the quality of the
input reference clock.
Jitter
A perfect clock source would have edges occurring at precise
intervals in time that would never vary. This, of course, is
impossible; even the best oscillators are constructed from nonideal components and have noise and other imperfections. A
high-quality, low-phase-noise crystal oscillator will have jitter
on the order of picoseconds, accumulated over many millions of
clock edges. Jitter is caused by thermal noise, instabilities in the
oscillator’s electronic circuitry, and external interference through
the power, ground, and output connections—all contributing to
disturbances in the oscillator’s timing. In addition, oscillators
are influenced by external magnetic or electric fields, and RF
interference from nearby transmitters. A simple amplifier, inverter,
or buffer in the oscillator circuitry will also add jitter to a signal.
14
Noise in a sampled system depends on many factors, starting
with reference clock jitter, which shows up as phase noise on the
fundamental signal. In a DDS system, truncation of the phase
register output may introduce code-dependent system errors.
Binary-coded words don’t cause truncation errors. For nonbinary
coded words, however, the phase-noise truncation error produces
spurs in the spectrum. The frequency/magnitude of the spurs
is determined by the code word. The DAC’s quantization and
linearity errors will also add harmonic noise in the system. Timedomain errors, such as undershoot/overshoot and code glitches,
all contribute distortion to the output signal.
Applications
DDS applications can be segmented into two primary categories:
• Communication and radar systems that require agile frequency
sources for data encoding and modulation applications
• Measurement, industrial, and optical applications that require
a generic frequency synthesis function with programmable
tuning, sweeping, and excitation
In both cases, an increasing trend towards higher spectral purity
(lower phase noise and higher spurious-free dynamic range) is
coupled with low operating power and size requirements for remote
or battery-operated equipment.
DDS in Modulation/Data Encoding and Synchronization
From its exclusive origins in radar and military applications, some
of the advances in DDS product characteristics (performance
improvements, cost, and size) have combined to make DDS
technology very popular in modulation and data encoding
applications. This section will discuss two data encoding schemes
and their proposed implementation with a DDS system.
Binary frequency shift keying (BFSK, or simply FSK) is one
of the simplest forms of data encoding. The data is transmitted
by shifting the frequency of a continuous carrier between one
(binary 1, or mark) and the other (binary 0, or space) of two discrete
frequencies. Figure 4 shows the relationship between the data and
the transmitted signal.
DATA
–30dB
1
0
TIME
SIGNAL
AMPLITUDE
SIGNAL AMPLITUDE
0dB
TIME
MARK
SPACE
f0
f1
Figure 4. Binary FSK modulation.
Binary 1s and 0s are represented as two different frequencies, f 0
and f 1, respectively. This encoding scheme is easily implemented
with a DDS device. The DDS frequency tuning word representing
the output frequencies is changed so that f0 and f 1 are generated
from 1s and 0s to be transmitted. In at least two members of
Analog Devices complete DDS product families (the AD9834
and the AD9838—see also the Appendix), the user can simply
Analog Dialogue Volume 46 Number 1
program the two current FSK frequency tuning words into the
IC’s embedded frequency registers. To shift output frequency,
a dedicated pin, FSELECT, selects the register containing the
appropriate tuning word (see Figure 5).
DATA
1
0
between multiple DDS devices can be predictably shifted by
means of the phase offset register. The AD983x series of DDS
products have 12 bits of phase resolution, providing an effective
resolution of 0.1°.
MCLK
AD9834
MUX
TUNING
WORD 1
RESET
DDS
DAC
FSK
TUNING
WORD 2
AD9834
CLOCK
PHASE SHIFT
Figure 5. FSK encoding using the tuning-word selector of
an AD9834 or AD9838 DDS.
Phase-shift keying (PSK) is another simple form of data
encoding. In PSK, the frequency of the carrier remains constant,
and the phase of the transmitted signal is varied to convey the
information. Several schemes can be used to accomplish PSK. The
simplest method, commonly known as binary PSK (or BPSK), uses
only two signal phases: 0° (Logic 1) and 180° (Logic 0). The state
of each bit is determined according to the state of the preceding
bit. If the phase of the wave does not change, the signal state stays
the same (low or high). If the phase of the wave changes by 180°,
that is, if the phase reverses—the signal state changes (low to high,
or high to low). PSK encoding is easily implemented with a DDS
product as most of the devices have a separate input register (a
phase register) that can be loaded with a phase value. This value
is directly added to the phase of the carrier without changing its
frequency. Changing the contents of this register modulates the
phase of the carrier, generating a PSK output. For applications
that require high-speed modulation, the AD9834 and AD9838,
which have pairs of phase registers, allow signals on a PSELECT
pin to alternate between the preloaded phase registers to modulate
the carrier as required.
More complex forms of PSK employ four or eight wave phases.
This allows binary data to be transmitted at a faster rate per phase
change than is possible with BPSK modulation. In four-phase
modulation (quadrature PSK), the possible phase angles are 0°,
+90°, −90°, and +180°; each phase shift can represent two signal
elements. The AD9830, AD9831, AD9832, and AD9835 provide
four phase registers to allow complex phase modulation schemes to
be implemented by continuously updating different phase offsets
to the registers.
I/Q Capability Using Multiple DDS Components in Synchronous Mode
Many applications require the generation of two or more sinusoidal
or square wave signals having a known phase relationship. A
popular example is in-phase and quadrature modulation (I/Q), a
technique wherein signal information is derived from a carrier
frequency at its 0° and 90° phase angles. Two single DDS
components can be run from the same source clock to output
signals whose phase relationship can be directly controlled and
manipulated. In Figure 6, the AD9838 devices are programmed
using one reference clock; the same RESET pin is used to update
both devices. In this way, simple I/Q modulation can be achieved.
A reset must be initiated after power-up and before transferring
any data to the DDS. This establishes the DDS output in a
known phase, which becomes the common reference angle that
allows synchronization of multiple DDS devices. When new data
is sent simultaneously to multiple DDS devices, a coherent phase
relationship can be maintained—or the relative phase offset
Analog Dialogue Volume 46 Number 1
Figure 6. Synchronizing two DDS components.
For more information about synchronizing multiple DDS devices,
see AN-605 Application Note, Synchronizing Multiple AD9852
DDS-Based Synthesizers.
Network Analysis
Many applications in the electronic world involve the gathering and
decoding of data from networks such as analog measurement and
optical communications systems. Normally, the system analysis
requirement is to stimulate a circuit or system with a frequency of
known amplitude and phase, and analyze the signal characteristics
of the response signal through the system.
The information gathered on the response signal is used to
determine key system information. The range of networks being
tested (see Figure 7) can be quite wide, including cable integrity
testing, biomedical sensing, and flow-rate measurement systems.
Wherever the basic requirement is to generate frequency-based
signals and compare phase and amplitude of the response signal(s)
to the original signal, or if a range of frequencies needs to be
excited through the system, or if test signals with different phase
relationships (as in systems with I/Q capability) are required, direct
digital synthesis ICs can be highly useful for digitally controlling
stimulus frequency and phase through software with simplicity
and elegance.
DDS
MCU
GAIN
ADC
RESPONSE
NETWORK
GAIN
Figure 7. Typical network analysis architecture using
frequency stimulus.
Cable Integrity/Loss Measurement
Cable integrity measurement is a noninvasive method of analyzing
cables in applications such as airplane wiring, local area networks
(LANs), and telephone lines. One way to determine performance
is to see how much signal is lost through the cable. By injecting a
signal of known frequency and amplitude, the user can calculate
cable attenuation by measuring the amplitude and phase at
remote portions of the cable. Parameters such as dc resistance
and characteristic impedance will affect a particular cable’s
attenuation. The result is usually expressed in decibels below the
signal source (0 dB) over the frequency range of the test. The
frequencies of interest depend on the cable type. DDS devices,
with their ability to generate a wide range of frequencies, can be
used as a stimulus with the necessary frequency resolution.
15
Flow Meter
Evaluation Kit
A related application area is in water, other liquids, and gas flow
analysis in pipelines. An example is ultrasonic flow measurement,
which operates on a phase-shift principle, as shown in Figure 8.
Basically, a signal is transmitted from one side of the channel where
the liquid is flowing and a transducer sensor is positioned on the
opposite side to measure the phase response—which depends on
the flow rate. There are many variations on this technique. Test
frequencies depend on the substance being measured; in general, the
output signal is often transmitted over a range of frequencies. DDS
provides the flexibility to set and change the frequency seamlessly.
The AD983x series of products come with a fully functional
evaluation kit with schematics and layout. The software provided
in the evaluation kit allows the user to easily program, configure,
and test the device (see Figure 9).
RECEIVED SIGNAL
LIQUID FLOW
ULTRASOUND SIGNAL (DDS)
ULTRASOUND SIGNAL
RECEIVED SIGNAL
Figure 9. AD9838 evaluation software interface.
PHASE SHIFT PROPORTIONAL TO LIQUID VELOCITY
Figure 8. Ultrasonic flow meter.
Other useful DDS information can be found on the DDS website.
Author
See also:
Brendan Cronin [[email protected]] is
a product marketing engineer working with the Core
Products and Technologies (CPT) Group within
Analog Devices. Brendan joined ADI in 1998 and
spent six years working as a mixed-signal design
engineer in the Industrial and Automotive Products
Group. Brendan’s focus today is on linear and related technologies.
Murphy, Eva and Colm Slattery. “All About Direct Digital
Synthesis.” Ask The Applications Engineer—33. Analog Dialogue.
Volume 38, No. 3, (2004): 8–12.
A Technical Tutorial on Digital Signal Synthesis. 1999. Analog
Devices, Inc.
APPENDIX
FURTHER INFORMATION AND USEFUL LINKS
Interactive Design Tool
The AD9838 in Brief: A block diagram of the AD9838 DDS
appears in Figure 10. Built on a fi ne-line CMOS process, the
device is an ultralow power (11-mW), complete DDS. The
28-bit frequency registers permit 0.06-Hz frequency resolution
with a 16-MHz clock and 0.02-Hz frequency resolution with a
5-MHz clock. Phase- and frequency modulation are configured
via on-chip registers using software or pin selection. The device
features −68-dBc wideband and −97-dBc narrow-band SFDR
and operates over the extended temperature range of –40°C to
+125°C. The device is housed in a small 4-mm × 4-mm, 20-lead
LFCSP (lead-frame chip-scale package).
What is it? An online interactive design tool for DDS is an assistant
for selecting tuning words, given a reference clock and desired
output frequencies and/or phases. The tool shows the tuning word
and other configuring bits encoded as a sequence of codes for use
in programming the part via its serial interface. Idealized output
harmonics can be shown for the selected reference clock and output
frequency after an external reconstruction fi lter has been applied.
Links to ADI’s design tools can be found on the Interactive Design
Tools home page. One example is the AD9834 design tool.
AVDD
AGND
DGND
DVDD
CAP/2.5V
REFOUT
REGULATOR
MCLK
VCC
2.5V
ON-BOARD
REFERENCE
FULL-SCALE
CONTROL
FSELECT
28-BIT FREQ0
REG
12
PHASE
ACCUMULATOR
(28-BIT)
MUX
28-BIT FREQ1
REG
FSADJUST
SIN
ROM
10-BIT
DAC
MUX
COMP
IOUT
IOUTB
MSB
12-BIT PHASE0 REG
12-BIT PHASE1 REG
MUX
MUX
DIVIDE
BY 2
16-BIT CONTROL
REGISTER
MUX
SIGN BIT OUT
SERIAL INTERFACE
AND
CONTROL LOGIC
COMPARATOR
VIN
AD9838
FSYNC
SCLK
SDATA
PSELECT
SLEEP RESET
Figure 10. Block diagram of the AD9838 DDS.
16
Analog Dialogue Volume 46 Number 1
Efficient FSK/PSK Modulator
Uses Multichannel DDS to
Switch at Zero Crossings
This article describes how two synchronized DDS channels can
implement a zero-crossing FSK or PSK modulator. Here, the
AD9958 two-channel, 500-MSPS, complete DDS (see Appendix)
is used to switch frequencies or phases at the zero-crossing point,
but any two-channel synchronized solution should be capable of
accomplishing this function. In phase-coherent radar systems,
zero-crossing switching reduces the amount of post processing
needed for signature recognition of the target; and implementing
PSK at the zero crossing reduces spectral splatter.
By David Brandon and Jeff Keip
Frequency-shift keying (FSK) and phase-shift keying (PSK)
modulation schemes are used in digital communications,
radar, RFID, and numerous other applications. The simplest
form of FSK uses two discrete frequencies to transmit binary
information, with Logic 1 representing the mark frequency and
Logic 0 representing the space frequency. The simplest form of
PSK is binary (BPSK), which uses two phases separated by 180°.
Figure 1 illustrates the two types of modulation.
1
0
1
1
Although both of the AD9958 DDS-channel outputs are
independent, they share an internal system clock and reside on
a single piece of silicon, so they should provide more reliable
channel-to-channel tracking over temperature and power-supply
deviations than the outputs of multiple, single-channel devices
synchronized together. The process variability that may exist
between distinct devices is also larger than any process variability
you might see between two channels fabricated in a single piece
of silicon, making a multichannel DDS preferable for use as a
zero-crossing FSK or PSK modulator.
0
ZERO-CROSSING FSK/PSK MODULATOR
AD9520
AD9958
DATA
LVPECL
CARRIER SIGNAL
XTAL
LVPECL
500MSPS
REF CLK
REF CLK
CH0
LPF
CH1
PS2
CARRIER
MODULATED SIGNAL
DATA
SOURCE
CLK
MODULATED SIGNAL
(a)
Figure 2. Setup for zero-crossing FSK or PSK modulator.
(b)
A critical element of any DDS is the phase accumulator, which,
in this implementation, is 32 bits wide. When the accumulator
overflows, it retains any excess value. When the accumulator
overflows with no remainder (see Figure 3), the output is precisely
at Phase 0, and the DDS engine starts over from where it was at
Time 0. The rate at which the zero-overflow is experienced is
referred to as the grand-repetition rate (GRR) of the DDS.
Figure 1. Binary FSK (a) and PSK (b) modulation.
The modulated output of a direct digital synthesizer (DDS) can
switch frequency and/or phase in a phase-continuous or phase-coherent
manner, as shown in Figure 1, and as described in “Multichannel
DDS Enables Phase-Coherent FSK Modulation,” making DDS
technology well suited for both FSK and PSK modulation.
ACCUMULATOR
ANGLE-TO-AMPLITUDE
NCO
N
FREQUENCY
TUNING WORD
SYSTEM CLOCK
N-BIT
ACCUM.
N
FTW
DAC
N
N
P
x
y
D
ANALOG
SINE WAVE
DAC
FS
PHASE TRUNCATION
(NĺP)
C = 2N
N
n × FTW
REMAINDER
FTW
REMAINDER
FTW
N
N
3 × FTW
2 × FTW
FREQUENCY
TUNING WORD
FTW
FTW
ROLLOVER
FTW
N
ROLLOVER
(n – 1) × FTW
N-BIT
ACCUM.
FTW
0
FS
SYSTEM
CLOCK
SYSTEM
TS = 1/FS
Figure 3. Basic DDS with overflowing accumulator.
Analog Dialogue Volume 46 Number 1
17
The GRR is determined by the rightmost nonzero bit of the
DDS frequency tuning word (FTW), as established by the
following equation:
GRR = FS /2 n
2
where:
FS is the sampling frequency of the DDS.
n is the rightmost nonzero bit of the FTW.
For example, suppose a DDS with a 1-GHz sampling frequency
employs 32-bit mark and space FTWs with the binary values
shown. In this case, the rightmost nonzero bit of either FTW
is the 19th bit, so GRR = 1 GHz/219, or approximately 1907 Hz.
Mark (CH0) 00101010 00100110 10100000 00000000
Space (CH0) 00111010 11110011 11000000 00000000
GRR
(CH1) 00000000 00000000 00100000 00000000
A DDS inherently switches frequency in a phase-continuous
manner. This means that no instantaneous phase change occurs
when the frequency tuning word changes. That is, the accumulator
starts accumulating the new FTW from whatever phase position
it was at when the new FTW was applied. Phase coherence, on
the other hand, requires an instantaneous transition to the phase
of the new frequency as if the new frequency had been present
all along. Therefore, in order for a standard DDS to implement a
phase-coherent FSK switch, the change from a mark frequency to
a space frequency must occur when both frequencies have the same
absolute phase. To implement a zero-crossing switch in a phasecoherent manner, the DDS must make the frequency transition
at 0 degrees (that is, when the accumulator overflows with zero
excess). Therefore, we must determine the instants at which phase
coherent zero-crossings occur. If the GRR of the mark and space
FTWs are known, the smaller of the two GRRs (if different) will
indicate the desired phase coherent zero-crossing point.
3
CH3 1.00V 𝛀
CH2 100mV 𝛀 M100𝛍s
CH1
D 100ns RUNS AFTER
720mV
Figure 4. Phase-continuous FSK transition.
2
3
CH3 1.00V 𝛀
Three criteria are necessary for implementing a phase-coherent
zero-crossing switch:
CH2 100mV 𝛀 M100𝛍s
CH1
D 100ns RUNS AFTER
720mV
Figure 5. Zero-crossing FSK transition.
1. The ability to determine the smaller GRR of the mark and
space FTWs associated with CH0 of Figure 2.
2. A second DDS channel (CH1 of Figure 2) synchronized to
CH0 of Figure 2 and programmed with an FTW having all
zeros except for the one bit corresponding to the smaller GRR.
3. The capability to use the rollover of the second channel to
trigger a frequency change on CH0 of Figure 2.
2
Unfortunately, the latency between when a DDS accumulator
hits zero and when that zero phase is represented at the output
further complicates the solution. Fortunately, this latency is
constant. The ideal solution necessitates that the auxiliary channel
be phase adjusted to compensate for this latency. Both channels
on the AD9958 have a phase-offset word that can be used to fi x
this problem.
The AD9958 two-channel DDS produced the results shown
in Figure 4, Figure 5, and Figure 6. Figure 4 and Figure 5
exhibit phase-continuous FSK switching vs. zero-crossing FSK
switching. Figure 5 shows both phase continuous switching and
phase coherent switching. Figure 6 shows the results from a
pseudorandom sequence (PRS) data stream that toggles between
multiple frequencies.
18
CH2 100mV 𝛀 M400ns
CH3
D 200ns RUNS AFTER
540mV
Figure 6. Zero-crossing with multi-FSK transitions.
The AD9958 two-channel DDS produced the results shown in
Figure 7 and Figure 8. These figures exhibit phase-continuous
BPSK switching vs. zero-crossing BPSK switching.
Analog Dialogue Volume 46 Number 1
Authors
David Brandon [[email protected]nalog.com]
has supported DDS products since the fi rst DDS
released back in 1995. His career spans 28 years
at ADI, with the last 11 years as an applications
engineer in the Clock and Signal Synthesis
Group. He has authored a number of application
notes and a couple of magazine articles.
2
Jeff Keip [[email protected]] has nearly
20 years of experience in the semiconductor
industry; over 15 of those have been spent
working on and with frequency synthesis
products. For the past nine years, Jeff has had
primary responsibility for the high-speed DDS
product portfolio at ADI.
3
CH3 1.00V 𝛀
CH2 100mV 𝛀 M100𝛍s
CH1
D 100ns RUNS AFTER
720mV
Figure 7. Phase-continuous BPSK transition.
APPENDIX
Two-Channel, 10-Bit, 500-MSPS Direct Digital Synthesizer
The AD9958 two-channel direct digital synthesizer (DDS)
comes complete with two 10-bit, 500-MSPS current-output
DACs, as shown in Figure 9. Both channels share a common
system clock, providing inherent synchronization; additional
packages can be used if more than two channels are required.
The frequency, phase, and amplitude of each channel can be
independently controlled, allowing them to provide correction
for system-related mismatches. These parameters can be
swept linearly; or 16 levels can be chosen for FSK, PSK, or
ASK modulation. Output sine waves can be tuned with 32-bit
frequency resolution, 14-bit phase resolution, and 10-bit
amplitude resolution. Operating with a 1.8-V core supply, plus a
3.3-V I/O supply for logic compatibility, the AD9958 consumes
315 mW with all channels on, and 13 mW in power-down mode.
Specified from –40°C to +85°C, it is available in a 56-lead LFCSP
package and priced at $20.24 in 1000s.
2
3
CH3 1.00V 𝛀
CH2 100mV 𝛀 M100𝛍s
CH1
D 100ns RUNS AFTER
720mV
Figure 8. Zero-crossing BPSK transition.
DDS CORE
AD9958
𝚺
32
32
𝚺
𝚺
15
COS(X)
10
10
DAC
10
10
DAC
CH0_IOUT
CH0_IOUT
DDS CORE
𝚺
32
FTW
𝚫FTW
32
32
SYNC_IN
SYNC_OUT
I/O_UPDATE
SYNC_CLK
𝚺
15
PHASE/
𝚫PHASE
COS(X)
14
AMP/
𝚫AMP
SCALABLE
DAC REF
CURRENT
10
TIMING AND CONTROL LOGIC
BUFFER/
XTAL
OSCILLATOR
REF CLOCK
MULTIPLIER
4× TO 20×
CLK_MODE_SEL
CH1_IOUT
CH1_IOUT
DAC_RSET
PWR_DWN_CTL
MASTER_RESET
SYSTEM CLK
÷4
REF_CLK
REF_CLK
𝚺
CONTROL
REGISTERS
MUX
I/O
PORT
BUFFER
CHANNEL
REGISTERS
PROFILE
REGISTERS
1.8V
1.8V
AVDD
DVDD
P0 P1
P2
P3
SCLK
CS
SDIO_0
SDIO_1
SDIO_2
SDIO_3
DVDD_I/O
Figure 9. AD9958 block diagram.
Analog Dialogue Volume 46 Number 1
19
Low-Noise, Gain-Selectable
Amplifier
The following derivation shows that sampling at V1 yields the
desired signal gain without gain error. R S denotes the switch
resistance. V2 can be derived using the same method.
§
R + RS1 ·
¸
V O1 = V IN × ¨¨1 + F1
¸
R
G1
©
¹
By Nathan Carter and Chilann Chan
Data acquisition, sensor signal conditioning, and other
applications with input signals that vary over a wide range
require gain-selectable amplifiers. Traditional gain-selectable
amplifiers use switches in the feedback loop to connect resistors
to the inverting input, but the switch resistance degrades the
noise performance of the amplifier, adds significant capacitance
on the inverting input, and contributes to nonlinear gain error.
The additional noise and capacitance are especially bothersome
when working with low-noise amplifiers, and the nonlinear gain
error is problematic in precision applications.
RF1
75𝛀
6
8
S1B
ADA4896-2
VIN
3
4
–5V
1
VO1
ADG633
+5V
+5V
2
D3
D1
S1A
ADA4896-2
S2B
V1
8
D2 5
V2
RBALANCE
150𝛀
S2A
R
·
(3)
Note that if VO1 yields the desired signal gain without gain error,
the buffered output, VO2, will also be free from gain error. Figure 2
shows the normalized frequency response of the circuit at VO2.
6
RL
4
–5V
ADG633
Figure 1. Using the ADA4896-2 and the ADG633 to construct a low-noise, gain-selectable amplifier to drive a lowresistance load.
The gain-selectable amplifier shown in Figure 1 uses an
innovative switching technique that preserves the 1-nV/√Hz noise
performance of the ADA4896-2 while reducing the nonlinear
gain error. With this technique, the user can choose switches with
minimal capacitance to maximize the bandwidth of the circuit.
The switches, implemented with an ADG633 triple SPDT CMOS
switch, are configured such that either S1A and S2A are on, or
S1B and S2B are on. Switch S1 connects to the output end of the
feedback resistors, and Switch S2 samples at a point (V1 or V2)
where the nonlinear switch resistance does not affect the gain. This
reduces the gain error while preserving the noise performance.
With the values shown, the fi rst stage amplifier gain is 4 V/V when
the “A” switches are on or 2 V/V when the “B” switches are on.
The number of switched gains can be extended with additional
switch packages or by using a multiplexer such as the 4:1 ADG659
or 8:1 ADG658.
Note that an offset is created by the input bias current of the
output buffer flowing through the nonlinear on resistance of the S2
sampling switch. To compensate for this offset, place the unused
switch (S3B) in the feedback path of the output buffer.
In addition, the bias current of the input amplifier causes a gaindependent offset. Because the input amplifier and output buffer
are built on the same chip, the relative matching of their bias
currents can be used to cancel out the varying offset. Placing a
resistor equal to the difference between R F2 and R F1, in series with
Switch S2A, results in less offset-voltage variation.
20
§
7
VO2
(2)
V1 = V IN × ¨1 + F1 ¸
¨
RG1 ¸¹
©
NORMALIZED CLOSED-LOOP GAIN (dB)
S3B
·
¸
¸
¹
Substituting Equation 1 into Equation 2,
S3B IS OPTIONAL
RF2
225𝛀
RG1
75𝛀
§ R F1 + R G1
V1 = V O1 × ¨¨
© R F1 + R G1 + R S1
(1)
3
0
–3
G = +2
G = +4
–6
–9
–12
–15
–18
–21
–24
VS = ±5V
–27 VIN = 100mV p-p
RL = 1k𝛀
–30
0.1
1
10
100
500
FREQUENCY (MHz)
Figure 2. Frequency response of VO2/VIN.
Authors
Nathan Carter [[email protected]] is
a design engineer in the Linear and RF Group,
where he has been working for more than 10 years.
He has degrees from California Polytechnic State
University and Worcester Polytechnic Institute.
Chilann Chan [[email protected]]
joined Analog Devices in August 2008 and is
presently an applications engineer in the HighSpeed Amplifi er Group. Chilann fi nished her
bachelor of engineering degree at Dartmouth
College. She earned her master’s in electrical
e n g i ne e r i n g f r om Wor ce ste r Poly te c h n ic
Institute, where she worked on applying the “Split-ADC”
architecture to a 16-bit, 1-MSPS differential successiveapproximation analog-to-digital converter.
Analog Dialogue Volume 46 Number 1
Designing a Low-Power
Toxic Gas Detector
By Luis Orozco
Safety fi rst! Many industrial processes involve toxic compounds,
including chlorine for manufacturing plastics, agrochemical, and
pharmaceutical products; phosphine and arsine for producing
semiconductors; and hydrogen cyanide, released when burning
consumer-packaging materials. It is important to know when
dangerous concentrations exist.
In the United States, the National Institute for Occupational
Safety and Health (NIOSH) and the American Conference of
Government Industrial Hygienists (ACGIH) have established
short- and long-term exposure limits for many toxic industrial
gases. The Threshold Limit Value-Time Weighted Average
(TLV-TWA) is the TWA concentration to which most workers
can be exposed repeatedly in an 8-hour day without adverse
effect; the Threshold Limit Value-Short Term Exposure Limit
(TLV-STEL) is the concentration to which most workers can be
exposed continuously for a short period of time without irritation,
damage, or impairment; and the Immediately Dangerous to Life
or Health Concentration (IDLHC) is a limiting concentration
that poses an immediate or delayed threat to life, that would cause
irreversible adverse health effects, or that would interfere with an
individual’s ability to escape unaided. Table 1 shows limits for a
few common gases.
Electrochemical sensors offer several advantages for instruments
that detect or measure the concentration of toxic gases. Most
sensors are gas-specific, have usable resolutions under one part per
million of gas concentration, and operate with very small amounts
of current, making them well suited for portable, battery powered
instruments. One important characteristic of electrochemical
sensors is their slow response: when fi rst powered up, the sensor
may take several minutes to settle to its fi nal output value; and
when exposed to a mid-scale step in gas concentration, the sensor
may take 25 to 40 seconds to reach 90% of its fi nal output value.
This article describes a portable carbon monoxide (CO) detector
using an electrochemical sensor. The IDLH concentration for CO
is much higher than for most other toxic gases, making it relatively
safe to handle. Nevertheless, CO is still lethal, so use extreme care
and appropriate ventilation when testing the circuit described here.
Figure 1. CO-AX carbon monoxide sensor.
Figure 1 shows a CO-AX sensor from Alphasense. Table 2 shows
a summary of the CO-AX sensor’s specifications.
Table 2. CO-AX Sensor Specifications
Sensitivity
55 nA/ppm to 90 nA/ppm (65 typ)
Response Time
(T90 from 0 ppm to
400 ppm CO)
< 30 s
Range (Guaranteed 0 ppm to 2,000 ppm
Performance)
Overgas Limit
4,000 ppm
Achieving the longest possible battery life is the most important
goal for portable instruments in this application, so keeping
power consumption to a minimum is crucial. In typical lowpower systems, the measurement circuitry powers up to make
a measurement, then shuts down for a long standby period. In
this application, however, the measurement circuit must remain
continuously powered due to the electrochemical sensor’s long
time constants. Fortunately, the slow response allows the use of
micropower amplifiers, high value resistors, and low-frequency
fi lters that minimize Johnson noise and 1/f noise. In addition,
single-supply operation avoids the wasted power of a bipolar supply.
Figure 2 shows the circuit of the portable gas detector. An
ADA4505-2 dual micropower amplifier is used in a potentiostat
configuration (U2-A), and a transimpedance configuration
(U2-B). The amplifier is a good choice for both potentiostat
and transimpedance sections because its power dissipation and
input bias current are extremely low. Consuming only 10 μA per
amplifier, it will permit a very long battery life.
Table 1. Exposure Limits for Some Common Industrial Toxic Gases
Long-Term Exposure
Limit (TLV-TWA)
(ppm)
Short-Term Exposure
Limit (TLV-STEL)
(ppm)
Immediate Danger to Life and
Health Concentration (IDLHC)
(ppm)
50
200
1,200
5,000
30,000
40,000
Chlorine
0.5
1
10
Phosphine
0.3
1
50
Hydrogen Sulfide
10
20
100
Toxic Gas
Carbon Monoxide
Carbon Dioxide
Analog Dialogue Volume 46 Number 1
21
In 3-electrode electrochemical sensors, the target gas diffuses
into the sensor through a membrane before interacting with
the working electrode (WE). The potentiostat circuit senses
the voltage at the reference electrode (RE) and supplies the
counter electrode (CE) with the current required to maintain a
constant voltage between the RE and WE terminals. No current
f lows in or out of the RE terminal, so the current f lowing out
of the CE terminal f lows into the WE terminal. This current
is directly proportional to the target gas concentration. The
current through the WE terminal can be positive or negative,
depending on whether reduction or oxidation takes place in
the sensor. For carbon monoxide, oxidation occurs, causing
the CE terminal current to be negative (current f lows into the
output of the potentiostat op amp). Resistor R4 is typically
very small, so the voltage at the WE terminal is approximately
equal to V REF.
The current that flows into the WE terminal results in a negative
voltage at the output of U2-A with respect to the WE terminal.
This voltage is typically a few hundred millivolts for a carbon
monoxide sensor, but can be as high as 1 V for other sensor
types. To run from a single supply, an ADR291 micropower
reference, U1, raises the entire circuit 2.5 V above ground.
The ADR291 consumes only 12 μA; it can also provide the
reference voltage for an analog-to-digital converter to digitize
the output of this circuit.
The output voltage of the transimpedance amplifier is simply:
VO = 2.5 V + IWE × Rf
(1)
where:
IWE is the current into the WE terminal.
Rf is the transimpedance resistor (shown as U4 in Figure 2).
The maximum response of the sensor is 90 nA/ppm, as shown in
Table 2, and its maximum input range is 2,000 ppm. This results in a
maximum output current of 180 μA, and a maximum output voltage
determined by the transimpedance resistor, as shown in Equation 2.
VO = 2.5 V + 2,000 ppm × 90 nA/ppm × Rf =
2.5 V + 180ȝA × Rf
(2)
Sensors for different gases or from different manufacturers
will have different current output ranges. Using an AD5271
programmable rheostat for U4 instead of a fi xed resistor makes
it possible to maintain a single assembly and bill of materials for
different gas sensors. In addition, it allows the product to have
interchangeable sensors, since a microcontroller can set the
AD5271 to the appropriate resistance value for each different gas
sensor. The AD5271’s 5-ppm/°C temperature coefficient is better
than most discrete resistors, and its 1-μA supply current is a very
small contributor to the system power consumption.
Figure 2. Portable gas detector using an electrochemical sensor.
22
Analog Dialogue Volume 46 Number 1
When operating from a single 5-V supply, a 2.5-V range is available
at the output of transimpedance amplifier U2-B, according to
Equation 1. Setting the AD5271 to 12.5 kΩ takes advantage of
the available range for worst-case sensor sensitivity, and allows for
approximately 10% overrange capability.
Using the typical 65-nA/ppm sensor response, the output voltage
can be translated to ppm of carbon monoxide as follows:
VO = 2.5 V + 813 μV/ppm
(3)
With a differential input ADC, simply connect the 2.5-V reference
output to the ADC’s A IN– terminal, eliminating the 2.5 V term
in Equation 3.
Resistor R4 keeps the transimpedance amplifier’s noise gain at
reasonable level. The value of R4 is a compromise between the
magnitude of the noise gain and the sensor’s settling time error when
exposed to high concentrations of gas. For this circuit, R4 = 33 Ω,
which results in a noise gain of 380, as shown in Equation 4.
NG = 1 +
12.5 kȍ
ȍ = 380
(4)
The input noise of the transimpedance amplifier is multiplied by
this gain. The ADA4505-2’s 0.1 Hz to 10 Hz input voltage noise
is 2.95 μV p-p, so the noise seen at the output will be
Voutput_noise = 2.95 ȝV × NG = 1.1 mV p-p
(5)
The output noise is equivalent to over 1.3 ppm p-p of gas
concentration. This low-frequency noise is difficult to fi lter out.
Fortunately, the sensor response is very slow, so the low-pass fi lter
formed by R5 and C6 can have a cutoff frequency of 0.16 Hz.
This fi lter has a time constant of one second, which is negligible
compared to the sensor’s 30-second response time.
Q1 is a P-channel JFET. When the circuit turns on, the gate is
at VCC, and the transistor is off. When the system powers off, the
gate drops to 0 V, and the JFET turns on to maintain the RE and
WE terminals at the same potential. This greatly improves the
turn-on settling time of the sensor when the circuit turns on again.
Two AAA batteries power the circuit. Using a diode for reversevoltage protection would waste precious energy, so this circuit uses
a P-channel MOSFET (Q2) instead. The MOSFET protects the
circuit by blocking reverse voltages, and turns on when a positive
voltage is applied. The MOSFET’s on-resistance is less than
100 mΩ, causing a much smaller voltage drop than a diode. The
ADP2503 buck-boost regulator allows the use of an external supply
of up to 5.5 V as well as the AAA batteries. When operating in
power-save mode, the ADP2503 consumes only 38 μA. The fi lter
formed by L2, C12, and C13 removes any switching noise from
the analog power rail. Rather than using a circuit to disconnect
the batteries when an external power supply is connected, a jack
that mechanically disconnects the batteries when plugging in the
external power connector avoids waste of power.
The total current drawn from the AAA batteries is approximately
100 μA under normal conditions (no gas detected), and 428 μA
under worst-case conditions (2,000 ppm CO detected). When
the instrument is connected to a microcontroller that can enter a
low-power standby mode while not making measurements, battery
life can extend to over one year.
References
NIOSH Pocket Guide to Chemical Hazards
http://www.cdc.gov/niosh/npg/
Alphasense CO-AX data sheet
http://www.alphasense.com/pdf/COAX.pdf
Author
Luis Orozco [[email protected]] is a system
applications engineer in ADI’s industrial and
instrumentation segment. He focuses on precision
instrumentation, chemical analysis, and environmental
monitoring applications. Luis joined ADI in February 2011.
(continued from Page 6)
Conclusions
Authors
In summar y, integrated over voltage protection provides
many benefits:
1. Improved robustness and precision in analog signal chains.
2. Reduced time-to-market (TTM), shorter design time, and
reduced testing requirements.
3. Reduced bill of materials (BOM) cost.
4. Fewer components required in approved component lists.
5. Reduced PCB footprint/higher density.
6. Lower failure rates.
Eric Modica [[email protected]] graduated
from San Jose State University in 2002 with
a BSEE. Responsible for process models and
precision amplifier design, he has worked for
Analog Devices for nine years.
References
1N914 data sheet available at www.fairchildsemi.com.
1N5711 data sheet available at www.st.com.
BAV99, BAS70-04, and BZB84-C24 data sheets available at
www.nxp.com.
PAD5 data sheet available at www.vishay.com.
JESD22-A114D standard available at www.jedec.org.
Analog Dialogue Volume 46 Number 1
Michael Arkin [[email protected]] is
a product marketing manager for the Precision
Operational Amplifier Group. He received his
BSEE from West Coast University and an MBA
from the University of Texas. Michael has more
than 15 years’ experience marketing electronics
products with companies such as ADI, TI, Pulse,
and Lineage Power.
The authors would like to thank Derek Bowers and Harry Holt for
their technical contributions to this article.
23
Analog Devices, Inc.
Worldwide Headquarters
Analog Devices, Inc.
One Technology Way
P.O. Box 9106
Norwood, MA 02062-9106
U.S.A.
Tel: 781.329.4700
(800.262.5643,
U.S.A. only)
Fax: 781.461.3113
Analog Devices, Inc.
Europe Headquarters
Analog Devices, Inc.
Wilhelm-Wagenfeld-Str. 6
80807 Munich
Germany
Tel: 49.89.76903.0
Fax: 49.89.76903.157
Analog Devices, Inc.
Japan Headquarters
Analog Devices, KK
New Pier Takeshiba
South Tower Building
1-16-1 Kaigan, Minato-ku,
Tokyo, 105-6891
Japan
Tel: 813.5402.8200
Fax: 813.5402.1064
Analog Devices, Inc.
Southeast Asia
Headquarters
Analog Devices
22/F One Corporate Avenue
222 Hu Bin Road
Shanghai, 200021
China
Tel: 86.21.2320.8000
Fax: 86.21.2320.8222
©2012 Analog Devices, Inc. All rights reserved.
Trademarks and registered trademarks are the
property of their respective owners.
M02000461-0-3/12
www.analog.com/analogdialogue
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