FUNCTIONAL BLOCK DIAGRAM
(MSB)
DB15
DB0
(LSB)
CLOCK NC REFCOM REFOUT IREFIN NR
V
EE
LADCOM
IOUTB
IOUTA
DCOM V
DD
MSBs: SEGMENTED
CURRENT SOURCES
AND SWITCHES
MSB
DECODER
AND
EDGE-
TRIGGERED
BIT
LATCHES
LSBs:
CURRENT SOURCES,
SWITCHES, AND
1kΩ R-2R
LADDERS
AD768
CONTROL
AMP
2.5V
BANDGAP
REFERENCE
1k
1k
REV. B
Information furnished by Analog Devices is believed to be accurate and
reliable. However, no responsibility is assumed by Analog Devices for its
use, nor for any infringements of patents or other rights of third parties
which may result from its use. No license is granted by implication or
otherwise under any patent or patent rights of Analog Devices.
a
FEATURES
30 MSPS Update Rate
16-Bit Resolution
Linearity: 1/2 LSB DNL @ 14 Bits
1 LSB INL @ 14 Bits
Fast Settling: 25 ns Full-Scale Settling to 0.025%
SFDR @ 1 MHz Output: 86 dBc
THD @ 1 MHz Output: 71 dBc
Low Glitch Impulse: 35 pV-s
Power Dissipation: 465 mW
On-Chip 2.5 V Reference
Edge-Triggered Latches
Multiplying Reference Capability
APPLICATIONS
Arbitrary Waveform Generation
Communications Waveform Reconstruction
Vector Stroke Display
16-Bit, 30 MSPS
D/A Converter
AD768
PRODUCT DESCRIPTION
The AD768 is a 16-bit, high speed digital-to-analog converter
(DAC) that offers exceptional ac and dc performance. The
AD768 is manufactured on ADI’s Advanced Bipolar CMOS
(ABCMOS) process, combining the speed of bipolar transistors,
the accuracy of laser-trimmable thin film resistors, and the effi-
ciency of CMOS logic. A segmented current source architecture
is combined with a proprietary switching technique to reduce
glitch energy and maximize dynamic accuracy. Edge triggered
input latches and a temperature compensated bandgap reference
have been integrated to provide a complete monolithic DAC
solution.
The AD768 is a current-output DAC with a nominal full-scale
output current of 20 mA and a 1 kΩ output impedance. Differ-
ential current outputs are provided to support single-ended
or differential applications. The current outputs may be tied
directly to an output resistor to provide a voltage output, or fed
to the summing junction of a high speed amplifier to provide a
buffered voltage output. Also, the differential outputs may be
interfaced to a transformer or differential amplifier.
The on-chip reference and control amplifier are configured for
maximum accuracy and flexibility. The AD768 can be driven by
the on-chip reference or by a variety of external reference volt-
ages based on the selection of an external resistor. An external
capacitor allows the user to optimally trade off reference band-
width and noise performance.
The AD768 operates on ±5 V supplies, typically consuming
465 mW of power. The AD768 is available in a 28-pin SOIC
package and is specified for operation over the industrial tem-
perature range.
PRODUCT HIGHLIGHTS
1. The low glitch and fast settling time provide outstanding
dynamic performance for waveform reconstruction or digital
synthesis requirements, including communications.
2. The excellent dc accuracy of the AD768 makes it suitable for
high speed A/D conversion applications.
3. On-chip, edge-triggered input CMOS latches interface
readily to CMOS logic families. The AD768 can support up-
date rates up to 40 MSPS.
4. A temperature compensated, 2.5 V bandgap reference is
included on-chip allowing for generation of the reference
input current with the use of a single external resistor. An ex-
ternal reference may also be used.
5. The current output(s) of the AD768 may be used singly or
differentially, either into a load resistor, external op amp
summing junction or transformer.
6. Proper selection of an external resistor and compensation
capacitor allow the performance-conscious user to optimize
the AD768 reference level and bandwidth for the target
application.
© Analog Devices, Inc., 1996
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 617/329-4700 Fax: 617/326-8703
AD768* PRODUCT PAGE QUICK LINKS
Last Content Update: 02/23/2017
COMPARABLE PARTS
View a parametric search of comparable parts.
EVALUATION KITS
•AD768 Evaluation Board
DOCUMENTATION
Application Notes
•AN-414: Low Cost, Low Power Devices for HDSL
Applications
•AN-912: Driving a Center-Tapped Transformer with a
Balanced Current-Output DAC
•AN-932: Power Supply Sequencing
Data Sheet
•AD768: 16-Bit, 30 MSPS D/A Converter Data Sheet
REFERENCE MATERIALS
Solutions Bulletins & Brochures
•Digital to Analog Converters ICs Solutions Bulletin
DESIGN RESOURCES
•AD768 Material Declaration
•PCN-PDN Information
•Quality And Reliability
•Symbols and Footprints
DISCUSSIONS
View all AD768 EngineerZone Discussions.
SAMPLE AND BUY
Visit the product page to see pricing options.
TECHNICAL SUPPORT
Submit a technical question or find your regional support
number.
DOCUMENT FEEDBACK
Submit feedback for this data sheet.
This page is dynamically generated by Analog Devices, Inc., and inserted into this data sheet. A dynamic change to the content on this page will not
trigger a change to either the revision number or the content of the product data sheet. This dynamic page may be frequently modified.
AD768–SPECIFICATIONS
Parameter Min Typ Max Units
RESOLUTION 16 Bits
DC ACCURACY
1
Linearity Error
T
A
= +25°C–8±4 +8 LSB
T
MIN
to T
MAX
–8 +8 LSB
Differential Nonlinearity
T
A
= +25°C–6±2 +6 LSB
T
MIN
to T
MAX
–8 +8 LSB
Monotonicity (13-Bit) GUARANTEED OVER RATED SPECIFICATION TEMPERATURE RANGE
ANALOG OUTPUT
Offset Error –0.2 +0.2 % of FSR
Gain Error –1.0 +1.0 % of FSR
Full-Scale Output Current
2
20 mA
Output Compliance Range –1.2 +5.0 V
Output Resistance 0.8 1.0 1.2 kΩ
Output Capacitance 3 pF
REFERENCE OUTPUT
Reference Voltage 2.475 2.5 2.525 V
Reference Output Current
3
+5.0 +15 mA
REFERENCE INPUT
Reference Input Current 1 5 7 mA
Reference Bandwidth
4
Small Signal, IREF = 5 mA ± 0.1 mA 28 MHz
Large Signal, IREF = 4 mA ± 2 mA 9 MHz
TEMPERATURE COEFFICIENTS
Unipolar Offset Drift –5 +5 ppm of FSR/
o
C
Gain Drift
5
–20 +20 ppm of FSR/
o
C
Gain Drift
6
–40 +40 ppm of FSR/
o
C
Reference Voltage Drift –30 +30 ppm/
o
C
DYNAMIC PERFORMANCE
7
Maximum Output Update Rate 30 40 MSPS
Output Settling Time (t
ST
) (to 0.025%) 25 35 ns
Output Propagation Delay (t
PD
)10ns
Glitch Impulse 35 pV-s
Output Rise Time (10% to 90%) 5 ns
Output Fall Time (10% to 90%) 5 ns
Output Noise (DB0–DB15 High, into 50 Ω) 3 nV/√Hz
Differential Gain Error 0.01 %
Differential Phase Error 0.01 Degree
DIGITAL INPUTS
Logic “1” Voltage 3.5 V
Logic “0” Voltage 1.5 V
Logic “1” Current –10 +10 µA
Logic “0” Current –10 +10 µA
Input Capacitance 10 pF
Input Setup Time (t
S
)10ns
Input Hold Time (t
H
)5ns
Latch Pulse Width (t
LPW
)10 ns
AC LINEARITY
7
Spurious-Free Dynamic Range (SFDR Within a Window)
F
OUT
= 1.002 MHz; CLOCK = 10 MHz; 2 MHz Span 86 79 dB
F
OUT
= 1.002 MHz; CLOCK = 20 MHz; 2 MHz Span 85 dB
F
OUT
= 5.002 MHz; CLOCK = 30 MHz; 10MHz Span 78 dB
Spurious-Free Dynamic Range (SFDR to Nyquist)
F
OUT
= 1.002 MHz; CLOCK = 10 MHz 74 70 dB
F
OUT
= 1.002 MHz; CLOCK = 20 MHz 73 dB
F
OUT
= 5.002 MHz; CLOCK = 30 MHz 67 dB
Total Harmonic Distortion (THD)
F
OUT
= 1.002 MHz; CLOCK = 10 MHz –71 –68 dB
F
OUT
= 1.002 MHz; CLOCK = 20 MHz –66 dB
F
OUT
= 5.002 MHz; CLOCK = 30 MHz –61 dB
REV. B
–2–
(TMIN to TMAX , VDD = +5.0 V, VEE = –5.0 V, LADCOM, REFCOM, DCOM = 0 V, IREFIN = 5 mA,
CLOCK = 10 MHz, unless otherwise noted)
Parameter Min Typ Max Units
POWER SUPPLY
Positive Voltage Range 4.75 5 5.25 V
Negative Voltage Range –5.25 –5 –4.75 V
Positive Supply Current 30 40 mA
Negative Supply Current 63 73 mA
Nominal Power Dissipation 465 600 mW
Power Supply Rejection Ratio (PSRR) –0.2 +0.2 % of FSR/V
OPERATING RANGE –40 +85 °C
NOTES
1
Measured at IOUTA, driving a virtual ground.
2
Nominal FS output current is 4 × the current at IREFIN. Therefore, nominal FS current is 20 mA when IREFIN = 5 mA.
3
Output current is defined as total current available for IREFIN and any external load.
4
Reference bandwidth is a function of external cap at NR pin. Refer to compensation section of data sheet for details.
5
Excludes internal reference drift.
6
Includes internal reference drift.
7
Measured as unbuffered voltage output (1 V range) with FS current into 50 Ω load at IOUTB.
Specifications subject to change without notice.
ABSOLUTE MAXIMUM RATINGS
*
Parameter with Respect to Min Max Units
Positive Supply Voltage (V
DD
) DCOM, REFCOM, LADCOM –0.5 +6.0 V
Negative Supply Voltage (V
EE
) DCOM, REFCOM, LADCOM –6.0 +0.5 V
Analog-to-Other Grounds (REFCOM) DCOM, LADCOM –0.5 +0.5 V
Digital-to-Other Grounds (DCOM) LADCOM, REFCOM –0.5 +0.5 V
Reference Output (REFOUT) REFCOM V
DD
+ 0.5 V
Reference Input Current (IREFIN) +7.5 mA
Digital Inputs (DB0–DB15, CLOCK) DCOM –0.5 V
DD
+ 0.5 V
Analog Outputs (IOUTA, IOUTB) LADCOM –2.0 +5.0 V
Maximum Junction Temperature +150 °C
Storage Temperature –65 +150 °C
Lead Temperature +300 °C
*Stresses above those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress
rating only and functional operation of the device at these or any other conditions above those indicated in the operational
sections of this specification is not implied. Exposure to absolute maximum rating for extended periods may affect device
reliability.
AD768
REV. B
ORDERING GUIDE
Package
Model Package Description Option
AD768AR 28-Pin 300 mil SOIC R-28
AD768ACHIPS Die
AD768-EB AD768 Evaluation Board
–3–
tH
tS
tLPW
tPD tST 0.025%
0.025%
DB0–DB15
CLOCK
IOUTA
OR
IOUTB
Timing Diagram
WARNING!
ESD SENSITIVE DEVICE
CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily
accumulate on the human body and test equipment and can discharge without detection.
Although the AD768 features proprietary ESD protection circuitry, permanent damage may
occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD
precautions are recommended to avoid performance degradation or loss of functionality.
REV. B
–4–
AD768
WAFER TEST LIMITS
1
Parameter AD768ACHIPS Limit Units
Integral Nonlinearity
2
±8 LSB max
Differential Nonlinearity
2
±6 LSB max
Offset Error ±0.2 % FSR max
Gain Error ±1.0 % FSR max
Reference Voltage ±1.0 % of nom. 2.5 V max
Positive Supply Current 40 mA max
Negative Supply Current 73 mA max
Power Dissipation 600 mW max
NOTES
1
Electrical test are performed at wafer probe to the limits shown. Due to variations in assembly methods and normal
yield loss, yield after packaging is not guaranteed for standard product dice.
2
Limits extrapolated from testing of individual bit errors.
3
Die offers latch control pad. Edge triggered latches become level triggered when latch control and clock pads are high.
4
Die substrate is connected to V
EE.
PIN DESCRIPTIONS
Pin No. Symbol Type Name and Function
1 IOUTA AO DAC Current Output. Full-scale current when all data bits are 1s.
2 NR AI Noise Reduction Node. Add capacitor for noise reduction.
3 REFOUT AO Reference Output Voltage. Nominal value is 2.5 V.
4 NC NC No Connect. Reserved for internal use.
5 REFCOM P Reference Ground.
6 IREFIN AI Reference Input Current. Nominal is 5 mA. DAC full-scale is 4× this current.
7 DB0 DI Data Bit 0 (LSB).
8–14 DB1–DB7 DI Data Bits 1–7.
15 DCOM P Digital Ground.
16 CLOCK DI Clock Input. Data latched on positive edge of clock.
17–23 DB8–DB14 DI Data Bits 8–14.
24 DB15 DI Data Bit 15 (MSB).
25 V
DD
P Positive Supply Voltage. Nominal is +5 V.
26 V
EE
P Negative Supply Voltage. Nominal is –5 V.
27 IOUTB AO Complementary DAC Current Output. Full-scale current when all data bits are 0s.
28 LADCOM P DAC Ladder Common.
Type: AI = Analog Input; DI = Digital Input; AO = Analog Output; P = Power.
(TA = +258C, VDD = +5.0 V, VEE = –5.0 V, IREFIN = 5 mA, unless otherwise noted)
V
DD
V
DD
DB15 DB14 DB13 DB12 DB11 DB10
REFCOM IREFIN DB0 DB1 DB2 DB3 DB4
V
EE
V
EE
IOUTB
LADCOM
IOUTA
NR
REFOUT
NC
DB9
DB8
CLOCK
LATCH CONTROL
DCOM
DB7
DB6
DB5
Die Size:
0.1106
×
0.1417 inch, 15,672 sq. mils
(2.81
×
3.60 mm, 10.116 sq. mm)
DICE CHARACTERISTICS
3, 4
PIN CONFIGURATION
14
13
12
11
17
16
15
20
19
18
10
9
8
1
2
3
4
7
6
5
TOP VIEW
(Not to Scale)
28
27
26
25
24
23
22
21
AD768
NC = NO CONNECT
IOUTA
V
DD
(+5V)
V
EE
(–5V)
IOUTB
LADCOM
NR
REFOUT
NC
DB13
DB14
DB15 (MSB)
REFCOM
IREFIN
(LSB) DB0
DB1
DB2
DB3 DB10
DB11
DB12
DB4
DB5
DB6
DB7
DB9
DCOM
CLOCK
DB8
AD768
REV. B –5–
Temperature Drift
Temperature drift is specified as the maximum change from the
ambient (+25°C) value to the value at either T
MIN
or T
MAX
. For
offset and gain drift, the drift is reported in ppm of full-scale
range (FSR) per degree C. For reference drift, the drift is re-
ported in ppm per degree C.
Power Supply Rejection
The maximum change in the full-scale output as the supplies
are varied from nominal to minimum and maximum specified
voltages.
Settling Time
The time required for the output to reach and remain within a
specified error band about its final value, measured from the
start of the output transition.
Spurious-Free Dynamic Range
The difference, in dB, between the rms amplitude of the
input signal and the peak spurious signal over the specified
bandwidth.
Total Harmonic Distortion
THD is the ratio of the rms sum of the first six harmonic com-
ponents to the rms value of the measured input signal. It is ex-
pressed as a percentage or in decibels (dB).
Glitch Impulse
Asymmetrical switching times in a DAC give rise to undesired
output transients which are quantified by a glitch impulse. It is
specified as the net area of the glitch in pV-sec.
DEFINITIONS OF SPECIFICATIONS
Linearity Error (Also Called Integral Nonlinearity or INL)
Linearity error is defined as the maximum deviation of the ac-
tual analog output from the ideal output, determined by a
straight line drawn from zero to full scale.
Differential Nonlinearity (or DNL)
DNL is the measure of the variation in analog value, normalized
to full scale, associated with a 1 LSB change in digital input code.
Monotonicity
A D/A converter is monotonic if the output either increases or
remains constant as the digital input increases.
Offset Error
The deviation of the output current from the ideal of zero is
called offset error. For IOUTA, 0 mA output is expected when
the inputs are all 0s. For IOUTB, 0 mA output is expected
when all inputs are set to 1s.
Gain Error
The difference between the actual and ideal output span. The
actual span is determined by the output when all inputs are set
to 1s minus the output when all inputs are set to 0s. The ideal
output current span is 4×the current applied to the IREFIN pin.
Output Compliance Range
The range of allowable voltage at the output of a current-output
DAC. Operation beyond the maximum compliance limits may
cause either output stage saturation or breakdown, resulting in
nonlinear performance.
5mA
6
C
REFCOMP
1µF R
REF
500Ω
REFINREFOUT
NC
V
DD
34
+2.5V REF
25
5
REFCOM
1µF
C
NR
2
NR
V
EE
26
15
1µF
1µF
+5V
–5V
DCOM
R
LAD
1kΩR
LAD
1kΩ
IOUTA
IOUTB
LADCOM
1
28
27
50ΩR
LOAD
50Ω
IOUTA
IOUTB
SEGMENTED
CURRENT
SOURCES CURRENT SOURCES
AND R-2R LADDER
CLOCK
21222324 19
20 13141718 7
89101112
LATCHES – LOWER 12 BITS
16
MSB DECODE
& LATCHES
CLOCK
DB15
DB14
DB13
DB12
DB11
DB10
DB9
DB8
DB7
DB6
DB5
DB4
DB3
DB2
DB1
DB0
AD768
Figure 1. Functional Block Diagram and Basic Hookup
FUNCTIONAL DESCRIPTION
The AD768 is a current-output DAC with a nominal full-scale
current of 20 mA and a 1 kΩ output impedance. Differential
outputs are provided to support single-ended or differential
applications. The DAC architecture combines segmented cur-
rent sources for the top four bits (MSBs) and a 1kΩ R-2R lad-
der for the lower 12 bits (LSBs). The DAC current sources are
implemented with laser-trimmable thin film resistors for excel-
lent dc linearity. A proprietary switching technique is utilized to
reduce glitch energy and maximize dynamic accuracy.
The digital interface offers CMOS compatible edge-triggered
input latches that interface readily to CMOS logic and supports
clock rates up to 40 MSPS. A temperature compensated 2.5 V
bandgap reference is integrated on-chip to drive the AD768 ref-
erence input current with the use of a single external resistor.
The functional block diagram in Figure 1 is a simple representa-
tion of the internal circuitry to aid the understanding of the
AD768’s operation. The DAC transfer function is described,
and followed by a detailed description of each key portion of the
circuit. Typical circuit configurations are shown in the section
APPLYING THE AD768.
REV. B
–6–
AD768
I
REFIN
– mA
POWER – mW
550
500
300
1.0 7.02.0 3.0 4.0 5.0 6.0
450
400
350
Figure 3. Power Dissipation vs. I
REFIN
Current
Note the AD768 is optimized for operation at an input current
of 5 mA. Both linearity and dynamic performance at other input
currents may be somewhat degraded. Figure 4 shows typical dc
linearity over a range of input currents. Figure 5 shows typical
SFDR (to Nyquist) performance over a range of input currents
and CLOCK input rates for a 1 MHz output frequency.
ERROR – LSB
I
REFIN
– mA
10
0
1.0 7.02.0 3.0 4.0 5.0 6.0
9
6
3
2
1
8
7
5
4INL
DNL
Figure 4. INL/DNL vs. I
REFIN
Current
I
REFIN
– mA
SFDR – dB
–85
–551.0 7.02.0 3.0 4.0 5.0 6.0
–80
–75
–70
–65
–60
CLOCK = 10 MSPS
CLOCK = 20 MSPS
CLOCK = 30 MSPS
CLOCK = 40 MSPS
Figure 5. SFDR (to Nyquist) vs. I
REFIN
@ F
OUT
= 1 MHz
DAC TRANSFER FUNCTION
The AD768 may be used in either current-output mode with the
output connected to a virtual ground, or voltage-output mode
with the output connected to a resistive load.
In current output mode,
I
OUT
= (DAC CODE/65536) × (I
REFIN
× 4)
In voltage output mode,
V
OUT
= I
OUT
× R
LOAD
iR
LAD
where:
DAC CODE is the decimal representation of the DAC inputs;
an integer between 0 and 65535.
I
REFIN
is the current applied at the IREFIN
pin, determined by
V
REF
/R
REF
.
Substituting for I
OUT
and
I
REFIN
,
V
OUT
= –V
REF
× (DAC CODE/65536) × 4 × [(R
LOAD
iR
LAD
)/R
REF
]
These equations clarify an important aspect of the AD768
transfer function; the full-scale current output of the DAC is
proportional to a current input. The voltage output is then a
function of the ratio of (R
LOAD
iR
LAD
)/R
REF
, allowing for cancel-
lation of resistor drift by selection of resistors with matched
characteristics.
REFERENCE INPUT
The IREFIN
pin is a current input node with low impedance to
REFCOM. This input current sets the magnitude of the DAC
current sources such that the full-scale output current is exactly
four times the current applied at IREFIN. For the nominal in-
put current of 5 mA, the nominal full-scale output current is
20 mA.
The 5 mA reference input current can be generated from the
on-chip 2.5 V reference with an external resistor of 500 Ω from
REFOUT to IREFIN. If desired, a variety of external reference
voltages may be used based on the selection of an appropriate
resistor. However, to maintain stability of the reference ampli-
fier, the external impedance at IREFIN must be kept below
1kΩ.
5
6
REFCOM
IREFIN
VEE VEE
IFB
5mA
Figure 2. Equivalent Reference Input Circuit
The I
REFIN
current can be varied from 1 mA to 7 mA which
subsequently will result in a proportional change in the DAC
full-scale. Since the operating currents within the DAC vary
with I
REFIN
, so does the power dissipation. Figure 3 illustrates
that relationship.
AD768
REV. B –7–
REFERENCE OUTPUT
The internal 2.5 V bandgap reference is provided for generation
of the I
REFIN
current, and must be compensated externally with
a capacitor of 0.1 µF or greater from REFOUT to REFCOM. If
an external reference is used, REFOUT should be tied directly
to the positive supply voltage, V
DD
. This effectively turns off the
internal reference, eliminating the need for the external capaci-
tor at REFOUT. The reference is specified to drive a nominal
load of 5 mA with a maximum of 15 mA. Operation with a
heavier load will result in degradation of supply rejection and
reference voltage accuracy. Therefore, the reference output
should be buffered with an amplifier when additional load cur-
rent is required. A properly sized pull-up resistor can also be
used to source additional current to the load. The resistors value
should be selected such that REFOUT will always source a
minimum of 5 mA to IREFIN and the additional load.
AD768
IREFIN
REFOUT
REFCOM
C
REFCOMP
1µF
500Ω
5
3
6
Figure 6. Typical Reference Hookup
TEMPERATURE CONSIDERATIONS
Note that the reference plays a key role in the overall tempera-
ture performance of the AD768. Any drift of I
REFIN
shows up
directly in I
OUT.
When the output is taken as a current, the drift
of I
REFIN
(which depends on both V
REF
and R
REF
) must be mini-
mized. This can be done by using the internal temperature com-
pensated reference for V
REF
and a low temperature coefficient
resistor for R
REF.
If the output is taken as a voltage, it is a func-
tion of a resistor ratio, not an absolute resistor value.
By select-
ing resistors with matched temperature coefficients for R
REF
and R
LOAD,
the drift in the resistor values will cancel, providing
optimal drift performance.
REFERENCE NOISE REDUCTION AND MULTIPLYING
BANDWIDTH
For application flexibility and multiplying capabilities, the refer-
ence amplifier is designed to offer adjustable bandwidth that can
be reduced by connecting an external capacitor from the NR
node to the negative supply pin, V
EE
. This capacitor limits the
bandwidth and acts as a filter to reduce the noise contribution
from the reference amplifier.
The noise reduction capacitor, C
NR
, is not required for stability
and does not affect the settling time of the DAC output. With-
out this capacitor, the I
REFIN
bandwidth is 15 MHz allowing
high frequency modulation of the DAC full-scale range through
the reference input node. Figure 7 shows the relationship be-
tween the external noise reduction capacitor and the –3 dB
bandwidth of the reference amplifier.
NOISE REDUCTION CAPACITOR – F
BANDWIDTH – Hz
100M
1k
10p 1µ100p 1n 10n 100n
10M
1M
100k
10k
Figure 7. External Noise Reduction Capacitor vs. –3 dB
Bandwidth
The sensitivity of the NR node requires that care be taken in
capacitor placement. The capacitor should be located as physi-
cally close to the package pins as possible and lead lengths
should be minimized. For this purpose, the use of a chip
capacitor is recommended. For applications that do not require
high frequency modulation at IREFIN, it is recommended that
a capacitor on the order of 1 µF be connected from NR to V
EE
.
If the reference input is purely dc, noise may be minimized with
multiple capacitors, such as 1 µF and 0.1 µF, to more effectively
filter both high and low frequency disturbances.
ANALOG OUTPUTS
The AD768 offers two analog outputs; IOUTA is trimmed for
optimal INL and DNL performance and has a full-scale output
when all bits are high. For applications that require the specified
dc accuracy, IOUTA should be used. IOUTB is the comple-
mentary output with full-scale output when all bits are low.
Both IOUTA and IOUTB provide similar dynamic perfor-
mance. Refer to Figures 8 and 9 for typical INL and DNL per-
formance curves. The outputs can also be used differentially.
Refer to the section “Applying the AD768” for examples of vari-
ous output configurations.
DIGITAL INPUT CODE – k
8
–806510 20 30 40
4
–2
–4
–6
6
0
2
INL ERROR – LSB
50 6051525354555
Figure 8. Typical INL Performance
REV. B
–8–
AD768
DIGITAL INPUT CODE – k
8
–806510 20 30 40
4
–2
–4
–6
6
0
2
DNL ERROR – LSB
50 6051525354555
Figure 9. Typical DNL Performance
The outputs have a compliance range of –1.2 V to +5.0 V with
respect to LADCOM. The current steering output stages will
remain functional over this range. Operation beyond the maxi-
mum compliance limits may cause either output stage saturation
or breakdown, resulting in nonlinear performance. The rated dc
and ac performance specifications are for an output voltage of
0 V to –1 V.
The current in LADCOM is proportional to I
REFIN
and has been
carefully configured to be independent of digital code when the
output is connected to a virtual ground. This minimizes any det-
rimental effects of ladder ground resistance on linearity. For
optimal dc linearity, IOUTA should be connected directly to a
virtual ground, and IOUTB should be grounded. An example of
this configuration is provided in the section “Buffered Voltage
Output.” If IOUTA is driving a resistive load directly, then
IOUTB should be terminated with an equal impedance. This
will ensure the current in LADCOM remains constant with digi-
tal code, and is recommended for improved dc linearity in the
unbuffered voltage output configuration.
As shown in Figure 10, there is an equivalent output impedance
of 1 kΩ in parallel with 3 pF at each output terminal. If the out-
put voltage deviates from the ladder common voltage, an error
current flows through this 1 k Ω impedance. This is a linear effect
which does not change with input code, so it appears as a gain
error. With 50 Ω output termination, the resulting gain error is
approximately –5%. An example of this configuration is pro-
vided in the section Unbuffered Voltage Output.
1
26
27
28
1kΩ1kΩ
3pF 3pF
I
OUT
I
OUT
IREFIN
x2.75
V
EE
LADCOM IOUTB
IOUTA
Figure 10. Equivalent Analog Output Circuit
DIGITAL INPUTS
The AD768 digital inputs consist of 16 data input pins and a
clock pin. The 16-bit parallel data inputs follow standard posi-
tive binary coding, where DB15 is the most significant bit
(MSB) and DB0 is the least significant bit (LSB). IOUTA pro-
duces full-scale output current when all data bits are at logic 1.
IOUTB is the complementary output, with full-scale when all
data bits are at logic 0. The full-scale current is split between
the two outputs as a function of the input code.
The digital interface is implemented using an edge-triggered
master slave latch. The DAC output is updated following the
rising edge of the clock, and is designed to support a clock rate
as high as 40 MSPS. The clock can be operated at any duty
cycle that meets the specified minimum latch pulse width. The
setup and hold times can also be varied within the clock cycle as
long as the specified minimums are met, although the location
of these transition edges may affect digital feedthrough. The
digital inputs are CMOS compatible with logic thresholds set to
approximately half the positive supply voltage. The small input
current requirements allow for easy interfacing to unbuffered
CMOS logic. Figure 11 shows the equivalent digital input
circuit.
V
CC
V
EE
DIGITAL
INPUT
V
CC
DCOM
Figure 11. Equivalent Digital Input Circuit
Digital input signals to the DAC should be isolated from the
analog output as much as possible. Interconnect distances to the
DAC inputs should be kept as short as possible. Termination
resistors may improve performance if the digital lines become
too long. To minimize digital feedthrough, the inputs should be
free from glitches and ringing, and may be further improved
with a reduction of edge speed.
AD768
REV. B –9–
Typical Performance Curves––AD768
0.2
0.0
–0.2
–0.4
–0.6
–0.8
–1.0
VOLTS
TIME – 5ns/Div
R
L
= 50Ω
V
OUT
= 0 TO –1V
Figure 14. Rise and Fall
Characteristics
10k 2.0M1.0M
OUTPUT – dB
FREQUENCY – Hz
CLOCK = 10MHz
F
OUT
= 1.002MHz
AMPLITUDE = 0dBm
0
–20
–40
–60
–80
–100
Figure 17. Typical SFDR
(With a Window)
FREQUENCY – Hz
OUTPUT – dB
0
–100
150 350
190 230 270 310
–60
–80
–40
–20
CLOCK = 10 MHz
F
OUT1
= 240 kHz
F
OUT2
= 260 kHz
AMPLITUDE = –6.0 dB FS
PER TONE
OUTPUT IOUTB
R
L
= 50Ω
Figure 20. Intermodulation
Distortion
R
L
= 50Ω
V
OUT
= 0 TO –1V
–0.470
–0.472
–0.474
–0.476
–0.478
–0.480
–0.482
OUTPUT – V
TIME – 2ns/Div
Figure 13. Glitch Impulse at Major
Carry
50.1 3241
OUTPUT – dB
FREQUENCY – MHz
0
–20
–40
–60
–80
–100
CLOCK = 10MHz
F
OUT
= 1.002MHz
AMPLITUDE = 0dBm
OUTPUT IOUTB
R
L
= 50Ω
Figure 16. Typical Spectral
Performance
FREQUENCY – MHz
THD – dB
–80
–75
–50
0.01 0.1 10
1
–65
–60
–55
–70
CLOCK = 1MHz
CLOCK = 10MHz
CLOCK = 20MHz
CLOCK = 30MHz
CLOCK = 40MHz
Figure 19. THD vs. F
OUT
Figure 12. Settling Time
FREQUENCY – MHz
–80
–75
–5042040
THD – dB
–65
–60
–55
–70
5678910 30
R
L
= 50Ω
VOUT = 0 TO –1V
Figure 15. THD vs. Clock Frequency
at F
OUT
= 1 MHz
110
FREQUENCY – MHz
CLOCK = 30MHz
CLOCK = 20MHz
–86
–82
–50
–78
–74
–70
–66
–62
–58
–54
SFDR
CLOCK = 10MHz
Figure 18. SFDR (Within a Window)
vs. F
OUT
–0.9405
–0.9443
–0.9415
–0.9424
–0.9433
–0.9453
–0.9500
–0.9462
–0.9472
–0.9498
–0.9481
12 24 3614 16 18 20 22 3426 28 30 32
1.00
0.60
0.90
0.80
0.70
0.50
0.10
0.40
0.30
0.20
0.025
0.05
OUTPUT – dB
ERROR BAND – %
TIME – ns
R
L
= 50Ω
V
OUT
= 0 TO –1V
REV. B
–10–
AD768
APPLYING THE AD768
OUTPUT CONFIGURATIONS
The following sections illustrate some typical output configura-
tions for the AD768. While most figures take the output at
IOUTA, IOUTB can be interchanged in all cases. Unless other-
wise noted, it is assumed that I
REFIN
and full-scale currents are
set to nominal values.
For application that require the specified dc accuracies, proper
resistor selection is required. In addition to absolute resistor tol-
erances, resistor self-heating can result in unexpected errors. For
optimal INL, the buffered voltage output is recommended as
shown in Figure 23. In this configuration, self-heating of R
FB
may cause a change in gain, producing a bow in the INL curve.
This effect can be minimized by selection of a low temperature
coefficient resistor.
UNBUFFERED VOLTAGE OUTPUT CONFIGURATIONS
Figure 21 shows the AD768 configured to provide a unipolar
output range of approximately 0 V to –1 V. The nominal full-
scale current of 20 mA flows through the parallel combination
of the 50 Ω R
L
resistor and the 1 kΩ DAC output resistance
(from the R-2R ladder), for a combined 47.6 Ω. This produces
an ideal full-scale voltage of –0.952 V with respect to LADCOM.
In addition, the 1 kΩ DAC output resistance has a tolerance of
±20% which may vary the full-scale gain by ±1%. This linear
variation results in a gain error which can be easily compensated
for by adjusting I
REFIN
.
1
27
28
VA
VB
AD768
R
L
49.9Ω
R
L
49.9Ω
IOUTA
LADCOM
IOUTB
Figure 21. 0 V to –1 V Unbuffered Voltage Output
In this configuration, it is important to note the restrictions from
the output compliance limits. The maximum negative voltage
compliance is –1.2 V, prohibiting use of a 100 Ω load to produce
a 0 V to –2 V output swing. One additional consideration for
operation in this mode is integral nonlinearity. As the voltage at
the output node changes, the finite output impedance of the
DAC current steering switches gives rise to small changes in the
output current that vary with output voltage, producing a bow
(up to 8 LSBs) in the INL. For optimal INL performance, the
buffered voltage output mode is recommended.
The INL is also slightly dependent on the termination of the
unused output (IOUTB) as described in the ANALOG OUT-
PUT section. To eliminate this effect, IOUTB should be termi-
nated with the same impedance as IOUTA, so both outputs see
the same resistive divider to ground. This will keep the current
in LADCOM constant, minimizing any code-dependent IR
drops within the DAC ladder that may give rise to additional
nonlinearities.
AC-Coupled Output
Configuring the output as shown in Figure 22 provides a bipolar
output signal from the AD768 without requiring the use of a
summing amplifier. The ac load impedance presented to the
DAC output is the parallel combination of the AD768’s output
impedance, R
L
, and bias resistor R
B
. The nominal output swing
with the values given in Figure 22 is ±0.5 V assuming R
B
>> R
L
.
The gain of the circuit will be a function of the tolerances of the
impedances R
LAD
, R
B
, and R
L
.
Choosing the value of R
B
and C will depend primarily on the
desired –3 dB high pass cutoff frequency and the bias current,
I
B
, of the subsequent stage connected to R
B
. The –3 dB fre-
quency can be approximated by the equation,
f
–3 dB
= 1/[2 × π × (R
B
+ R
L
iR
LAD
) × C].
The dc offset of the output is a function of the bias current of
the subsequent stage and the value of R
B
. For example, if
C = 390 pF, R
B
= 20 kΩ, and I
B
= 1.0 µA, the –3 dB frequency
is approximately 20.4 kHz and the dc offset would be 20mV.
1
27
28
R
B
AD768
R
L
49.9Ω
IOUTA
LADCOM
IOUTB R
L
49.9Ω
I
B
C
Figure 22. 0.5 V to –0.5 V Unbuffered AC-Coupled Output
BUFFERED VOLTAGE OUTPUT CONFIGURATIONS
Unipolar Configuration
For positive output voltages, or voltage ranges greater than
allowed by output compliance limits, some type of external
buffer is needed. A wide variety of amplifiers may be selected
based on considerations such as speed, accuracy and cost. The
AD9631 is an excellent choice when dynamic performance is
important, offering low distortion up to 10MHz. Figure 23
shows the implementation of 0 V to +2 V full-scale unipolar
buffered voltage output. The amplifier establishes a summing
node at ground for the DAC output. The buffered output volt-
age results from the DAC output current flowing through the
amplifier’s feedback resistor, R
FB
. In this case, the 20 mA full-
scale current across R
FB
(100 Ω) produces an output voltage
range of 0 V through +2 V. The same configuration using a pre-
cision amplifier such as the AD845 is recommended for optimal
dc linearity.
1
27
28
AD768
R
FB
100Ω
IOUTA
LADCOM
IOUTB
A1
Figure 23. Unipolar 0 V to +2 V Buffered Voltage Output
Buffered Output Using a Current Divider
The configuration shown in Figure 23 may not be possible in
cases where the amplifier cannot supply the requisite 20mA
feedback current. As an alternative, Figure 24 shows amplifier
A1 in conjunction with a resistive current divider. The values of
R
FF
and R
L
are chosen to limit the current, I
3
, which must be
supplied by A1. Current, I
2
, is shunted to ground through resis-
tor, R
L
. The parallel combination of R
FF
and R
L
should not ex-
ceed 60 Ω to avoid exceeding the specified compliance voltage.
AD768
REV. B –11–
For the values given in Figure 24, I
3
equals 4 mA, which results
in a nominal unipolar output swing of 0 V to 2 V. Note, since
A1 has an inverting gain of approximately –4 and a noise gain of
+5, A1’s distortion and noise performance should be considered.
1
27
28
AD768
R
FB
500Ω
IOUTA
LADCOM
IOUTB
R
FF
100Ω
R
L
24.9Ω
R
P
20Ω
I
1
I
2
I
3
A1
Figure 24. 0 V to 2 V Buffered Unipolar Output Using a
Current Divider
Bipolar Configuration
Bipolar mode is accomplished by providing an offset current,
I
BIPOLAR
, to the I/V amplifier’s (A1) summing junction. By set-
ting I
BIPOLAR
to exactly half the full-scale current flowing
through R
FB
, the resulting output voltage will be symmetrical
about the summing junction voltage, typically ground. Figure 25
shows the implementation for a bipolar ±2.5 V buffered voltage
output. The resistor divider sets the full-scale current for I
DAC
to
5 mA. The internal 2.5 V reference generates a 2.5 mA I
BIPOLAR
current across R
BIP
. An output voltage of 0 V is produced when
the DAC is set to half scale (100. . .0) such that the 2.5 mA cur-
rent, I
DAC
, is exactly offset by I
BIPOLAR
. As the DAC is varied
from zero to full-scale, the output voltage swings from –2.5V to
+2.5 V. Note, in configurations that require more than 15 mA
of total current from REFOUT, an external buffer is required.
Op amps such as the AD811, AD8001, and AD9631 are good
selections for superior dynamic performance. In dc applications,
op amps such as the AD845 or AD797 may be more appropriate.
27
AD768
R
FB
1kΩ
IOUTA
LADCOM
IOUTB
75Ω
24.9Ω
R
P
20Ω
A1
C
3I
BIPOLAR
R
BIP
1kΩ
REFOUT
I
DAC
28
1
Figure 25. Bipolar
±
2.5 V Buffered Voltage Output
DIFFERENTIAL OUTPUT CONFIGURATIONS
AC Coupling via a Transformer
Applications that do not require baseband operation typically
use transformer coupling. Transformer coupling the comple-
mentary outputs of the AD768 to a load has the inherent benefit
of providing electrical isolation while consuming no additional
power. Also, a properly applied transformer should not degrade
the AD768’s output signal with respect to noise and distortion,
since the transformer is a passive device. Figure 26 shows a
center-tapped output transformer that provides the necessary dc
load conditions at the outputs IOUTA and IOUTB to drive a
±0.5 V signal into a 50 Ω load. In this particular circuit, the cen-
ter-tapped transformer has an impedance ratio of 4 that corre-
sponds to a turns ratio of 2. Hence, any load, R
L
, referred to the
primary side is multiplied by a factor of 4 (i.e., in this case
200 Ω). To avoid dc current from flowing into the R-2R ladder
of the DAC, the center tap of the transformer should be con-
nected to LADCOM.
In order to comply with the minimum voltage compliance of
–1.2 V, the maximum differential resistance seen between
IOUTA and IOUTB should not exceed 240 Ω. Note that the
differential resistance consists of the load R
L
, referred to the
primary side of the transformer in parallel with any added differ-
ential resistance, R
DIFF
, across the two outputs. R
DIFF
is typically
added to the primary side of the transformer to match the effec-
tive primary source impedance to the load (i.e., in this case
200 Ω).
1
27
AD768
IOUTA
LADCOM
IOUTB
RDIFF
200ΩT1 RL
50Ω
4:1 IMPEDANCE
RATIO
T1 = MINI-CIRCUITS T4-6T
28
Figure 26. Differential Output Using a Transformer
DC COUPLING VIA AN AMPLIFIER
A dc differential to single-ended conversion can be easily ac-
complished using the circuit shown in Figure 27. This circuit
will attenuate both ac and dc common-mode error sources due
to the differential nature of the circuit. Thus, common-mode
noise (i.e., clock feedthrough) as well as dc unipolar offset errors
will be significantly reduced. Also, excellent temperature stabil-
ity can be obtained by using temperature tracking, thin film
resistors for R and R
REF
. The design equations for the circuit are
provided such that the voltage output swing and IREF can be
optimized for a given application.
1
27
28
AD768
R*
IOUTA
LADCOM
IOUTB
R
REF
*I
REF
A1
R*
6
3
REFOUT
REFIN
V
OUT
V
OUT
= ±4 I
REF
R
WHERE I
REF
= 2.5V
R
REF
V
OUT
± 2V
R = 200Ω
R
REF
= 5 x 200Ω
*OHMTEK TDP-1403
Figure 27. DC Differential to Single-Ended Conversion
POWER AND GROUNDING CONSIDERATIONS
In systems seeking to simultaneously achieve high speed and
high accuracy, the implementation and construction of the
printed circuit board design is often as important as the circuit
design. Proper RF techniques must be used in device selection,
placement and routing, and supply bypassing and grounding.
Maintaining low noise on power supplies and ground is critical
to obtaining optimum results from the AD768. Figure 28 pro-
vides an illustration of the recommended printed circuit board
ground plane layout which is implemented on the AD768 evalu-
ation board.
REV. B
–12–
AD768
Figure 28. Printed Circuit Board Ground Plane Layout
Figure 29. Printed Circuit Board Power Plane Layout
AD768
REV. B –13–
If properly implemented, ground planes can perform a host of
functions on high speed circuit boards: bypassing, shielding,
current transport, etc. In mixed signal design, the analog and
digital portions of the board should be distinct from each other,
with the analog ground plane confined to the areas covering
analog signal traces and the digital ground plane confined to
areas covering the digital interconnects.
All analog ground pins of the DAC, reference, and other analog
output components, should be tied directly to the analog ground
plane. The two ground planes should be connected by a path
1/4 to 1/2 inch wide underneath or within 1/2 inch of the DAC
as shown in Figure 28. Care should be taken to ensure that the
ground plane is uninterrupted over crucial signal paths. On the
digital side, this includes the digital input lines running to the
DAC as well as any clock signals. On the analog side, this in-
cludes the DAC output signal, reference signal, and the supply
feeders.
The use of wide runs or planes in the routing of power lines is
also recommended. This serves the dual role of providing a low
series impedance power supply to the part, as well as, providing
some “free” capacitive decoupling to the appropriate ground
plane. Figure 29 illustrates the power plane layout used in the
AD768 evaluation board. The AD768 evaluation board uses a
four layer P.C. board which illustrates good layout practices as
discussed above.
It is essential that care be taken in the layout of signal and
power ground interconnects to avoid inducing extraneous volt-
age drops in the signal ground paths. It is recommended that all
connections be short, direct and as physically close to the pack-
age as possible, in order to minimize the sharing of conduction
paths between different currents. When runs exceed an inch in
length, some type of termination resistor should be considered.
The necessity and value of this resistor will be dependent upon
the logic family used.
For maximum ac performance, the DAC should be mounted
directly to the circuit board; sockets should be avoided since
they introduce unwanted capacitive coupling between adjacent
pins of the device.
POWER SUPPLY AND DECOUPLING
One of the most important external components associated with
high speed designs are the capacitors used to bypass the power
supplies. Both selection and placement of these capacitors can
be critical and, to a large extent, dependent upon the specifics of
the system configuration. The dominant consideration in the
selection of bypass capacitors for the AD768 is the minimization
of the series resistance and inductance. Many capacitors will
begin to look inductive at 20 MHz and above. Ceramic and film
type capacitors generally feature lower series inductance than
tantalum or electrolytic types.
It is recommended that each power supply to the AD768 be de-
coupled by a 0.1 µF capacitor located as close to the device pins
as possible. Surface-mount chip capacitors, by virtue of their
low parasitic inductance, are preferable to through-hole types.
Some series inductance between the DAC supply pins and the
power supply plane may help to provide additional filtering of
high frequency power supply noise. This inductance can be gen-
erated by using small ferrite beads.
A clean digital supply may be generated using the circuit shown
in Figure 30. The circuit consists of a differential LC filter with
separate power supply and return lines. Lower noise can be at-
tained using low ESR (Equivalent Series Resistance) type elec-
trolytic and tantalum capacitors.
FERRITE
BEADS
100µF
ELECT. 10–20µF
TANT. 0.1µF
CER.
V
DD
DCOM
+5V DGND
+5V
POWER SUPPLY
TTL/CMOS
LOGIC
CIRCUITS
Figure 30. Differential LC Filter for Single +5 V
Applications
APPLICATIONS
USING THE AD768 AS A MULTIPLYING DAC
The AD768 can be easily configured as a multiplying DAC
since I
REFIN
can be modulated from 1 mA to 7 mA. The refer-
ence amplifier sets the maximum multiplying bandwidth to 15
MHz, while any external capacitor to the NR node serves to
limit the bandwidth according to Figure 7. I
REFIN
can be easily
modulated by properly scaling and summing into the IREFIN
node the modulating signal. Figure 31 demonstrates how the
modulating signal VMOD can be properly scaled and converted
to a current via R
REFMOD
such that its peak current does not ex-
ceed 3.0 mA. Figure 32 shows the AD768’s typical distortion
versus the reference channel frequency.
6
3
AD768
R
REF
625Ω
IREFIN
REFOUT
1µF
R
REFMOD
VMOD
VMOD
R
REFMOD
≤ ±3.0mA
Figure 31. Typical Multiplying DAC Application
FREQUENCY – kHz
THD – dB
–75
–40
250 2500500 750 1000 1250 1500 1750 2000 2250
–70
–60
–55
–50
–45
–65
I
REF
= 4.0+/–3 mA
I
REF
= 4.0+/–2 mA
I
REF
= 5.0+/–1 mA
Figure 32. Reference Channel Distortion vs. Frequency
REV. B
–14–
AD768
AD768 IN MULTITONE TRANSMITTERS (FOR ADSL)
Communications applications frequently require aspects of
component performance that differ significantly from the
simple, single tone signals used in typical SNR and THD tests.
This is particularly true for spread-spectrum and frequency divi-
sion multiplexed (FDM) type signals, where information con-
tent is held in a number of small signal components spread
across the frequency band. In these applications, a combination
of wide dynamic range, good fine-scale linearity, and low inter-
modulation distortion is required. Unfortunately, a part’s full
scale SNR and THD performance may not be a reliable indica-
tor of how it will perform in these multitone applications.
One example of an FDM communications system is the DMT
(discrete multitone) ADSL (Asymmetrical Digital Subscriber
Line) standard currently being considered by ANSI. Figure 33
shows a block diagram of a transmitter function.
The digital bits are used to QAM modulate each of approxi-
mately 200 discrete tones. An inverse FFT turns this modu-
lated frequency domain information into 512 time points at a
2.2 MSPS sample rate. These time points are then put through
an FIR interpolation filter to upsample (in this case to 4.4 MSPS).
The bit stream is run through the AD768, which is followed by
a 4th order analog smoothing filter, then run to the line-driving
circuitry
TO
TRANSMITTER
BIT
STREAM QAM
ENCODER
256
MODULATED
FREQUENCY
BINS
512 TIME
POINTS
@ 2.2MSPS
1024 TIME
POINTS
@ 4.4MSPS
4TH ORDER
SMOOTHING
FILTER
AD768
+BUFFER
2X
INTERPOLATOR
FIR
INVERSE
FFT
Figure 33. Typical DMT ADSL Transmit Chain
Figure 34a shows a frequency domain representation of a test
vector run through this system, while 34b shows the time do-
main representation. (Clearly the frequency domain picture is
more informative.) We wish to optimize the SINAD of each
4 kHz frequency band: this is a function of both noise
(wideband and quantization) and distortion (simple harmonic
and intermod).
OUTPUT – dB
FREQUENCY – Hz
0
–20
–40
–60
01.1M
–80
Figure 34a. Output Spectrum of ADSL Test Vector
TIME – 25µs/DIV
1.5
1.0
0.5
0
–0.5
–1.0
–1.5
OUTPUT – Volts
Figure 34b. Time Domain Output Signal of ADSL Test
Vector
Table I and II show the available SNR and THD at the output
of the filter vs. frequency bin for the ADSL application. The
AD768’s combination of 16-bit dynamic range and 14-bit lin-
earity provides excellent performance for the DMT signal. Its
fast input rate would support even faster rates of oversampling,
if one were interested in trading off digital filter complexity in
the interpolator for a simplified analog filter.
Table I. SNR vs. Frequency
Frequency SNR
151 kHz 70.1 dB
349 kHz 69.7 dB
500 kHz 69.4 dB
1 MHz 69.8 dB
Table II. THD vs. Frequency
Frequency THD
160 kHz –68.9 dBc
418 kHz –64.0 dBc
640 kHz –64.3 dBc
893 kHz –63.8 dBc
AD768
REV. B –15–
AD768 EVALUATION BOARD
GENERAL DESCRIPTION
The AD768-EB is an evaluation board for the AD768 16-bit
30 Msps D/A converter. Careful attention to layout and circuit
design combined with analog and digital prototyping areas al-
lows the user to easily and effectively evaluate the AD768 in any
application where high resolution, high speed conversion is
required.
The digital inputs to the AD768-EB may be driven directly us-
ing the standard 40-pin IDC connector. An external clock is
also required. These signals may be applied from a user’s
bench, or they can be generated from a circuit built on the
prototyping area. The analog outputs from the AD768-EB are
available on BNC connectors. These outputs may be configured
to use either resistors, op amps, or a transformer.
OPERATING PROCEDURE AND FUNCTIONAL
DESCRIPTION
Power
Power may be supplied to the AD768-EB by applying either
wires or banana plugs to the metal binding posts included on the
printed circuit board.
DGND. Digital Ground. The digital ground and the analog
ground are connected together underneath the AD768. Optimal
performance can be obtained with separate analog and digital
supplies. For evaluation purposes, a single-supply which makes
a second analog and digital ground connection at the supply is
acceptable.
+5D.The +5 V (±5%) digital supply should be capable of sup-
plying 50 mA.
–5A. The –5 V (±5%) analog supply should be capable of sup-
plying –75 mA.
AGND.Analog ground. The analog ground and the digital
ground are connected together underneath the AD768. Optimal
performance can be obtained with separate analog and digital
supplies. For evaluation purposes, a single-supply which makes
a second analog and digital ground connection at the supply is
acceptable.
–V
EE
. Negative analog supply; typically –5 V to –15 V. This
supply is used as the negative supply rail for the external op
amps. For the AD811 supplied with the AD768-EB, a supply
capable of supplying –20 mA (excluding external load require-
ments) is required.
+V
CC
. Positive analog supply; typically +5 V to +15 V. This
supply is used as the positive supply rail for the external op
amps. For the AD811 supplied with the AD768-EB, a supply
capable of supplying +20 mA (excluding external load require-
ments) is required.
Analog Outputs
The analog output(s) from the AD768-EB are available on BNC
jacks “A” and “B.” The complementary current outputs from
the AD768 can be configured using either resistors, op amps, or
a transformer. Only the “A” portion of the AD768-EB is popu-
lated and shipped from the factory. The “B” side, or comple-
mentary output, may be populated by the user if so desired.
JP1. Buffered op amp output “A”. Jumper JP1 should be
installed if the buffered op amp output is desired. When JP1
is installed, JP2 and JP3 must be removed for proper
operation. The output, available on the “A” connector, has a
nominal voltage swing of 0 V to 2 V and is in-phase with the
digital input. This is the factory default setting.
JP2. Bipolar 50 Ω transformer output. If jumper JP2 is in-
stalled, a transformer coupled output is available on the “A”
connector. When JP2 is installed, JP1 and JP3 must be re-
moved for proper operation. The transformer acts both as a
differential-to-single-ended converter and as an impedance
transformer. For proper operation, the transformer must be
terminated with a 50 Ω resistor. R2 must be replaced with
the 100 Ω resistor, R7. An additional 100 Ω resistor and the
transformer are included with the AD768-EB. The additional
100 Ω resistor must be soldered into the appropriate position la-
beled “R3” and the transformer must be inserted into the
socket labeled “T1.” The nominal output voltage into a 50Ω
load is 1 V
p-p
centered on a common-mode voltage of 0V.
JP3. Resistor output “A.” JP3 is used to connect the resistor
R2 to the “A” output. U2 should be removed from its socket.
Using a 24.9 Ω resistor for R2, the output is an unbuffered 0 V
to –0.5 V output that is out of phase with the digital input. Re-
sistor R2 may be replaced with other values, but careful atten-
tion to the recommended output compliance range should be
observed. When JP3 is installed, JP1 and JP2 must be re-
moved for proper operation.
JP4. Resistor output “B.” JP4 is used to connect the resistor
R3 to the “A” output. U3 should be removed from its socket.
The AD768-EB is shipped from the factory with resistor R3
shorted to ground. A different value selected by the user can be
installed for R3 to generate an unbuffered output that is in-
phase with the digital input. Careful attention to the recom-
mended output compliance range should be observed when
selecting the value of R3. When JP4 is installed, JP5 must
be removed for proper operation.
JP5. Buffered op amp output “B.” Jumper JP5 should be in-
stalled if the buffered op amp output is desired. When JP5 is
installed, JP4 must be removed for proper operation.
The output is available on the “B” connector and has a nominal
voltage swing determined by the combination of resistors R3,
R9, and R10. This op amp is not provided with the AD768-EB.
Reference
Either the internal reference of the AD768 or an external refer-
ence may be selected on the AD768-EB. R12 is used to adjust
the full-scale output current of the AD768.
SW2. Internal/External reference select switch. When SW2 is
in position 1, the internal reference of the AD768 is selected.
When SW2 is in position 2, an external reference must be pro-
vided by the user.
Level-Shifting the Analog Output
Resistor sockets R8 and R6 can be populated with an appropri-
ately valued resistor to add dc offset current to an output which
uses the op amp configuration. As an example, to generate a
bipolar output signal, a 1.25 kΩ resistor installed into the “R8”
socket level-shifts the normally unipolar output by –1V. The
factory defaults for R8 and R6 are open circuits.
REV. B
–16–
AD768
Table III. Summary of Jumper Functionality
Installed Jumper Function
Jumper
JP1 Buffered Output A
JP2 50 Ω Transformer Output
JP3 (STBY) Unbuffered Output A
JP4 Unbuffered Output B
JP5 Buffered Output B
Table IV. AD768-EB Parts List
Reference Value / Part Type Package Qty/Bd
U1 AD768 28-Pin SOIC 1
U2 AD811 8-Pin DIP 1
T1 Mini-Circuits Not Installed 1
T4–6T
A, B, CLOCK BNC JACKs, Small, Vertical 3
Small
JP1–5 Header 2-Pin 5
SW1, 2 SPDT, Secme 0.1" × 0.3" 2
J1 40-Pin IDC R.A., Male, 1
Connector w/ Latches
R1 500 Ω1/4 W, 0.01%, 1
Vishay
R2 25 Ω1/4 W, 0.01%, 1
Vishay
R3, R13–21, &
R23–29 Wire Jumpers 17
R5 500 Ω1/4 W, 0.01%, 1
Vishay
R7 100 Ω1/4 W, 0.01%, 1
Vishay
R11 51 Ω1/8 W, 5%, Carbon 1
R12 10 kΩ Pot. 3266 W 1
C1–4 1 µF Ceram. Cap. Leaded 4
C5–8, C10, 12, 0.1 µF Chip Cap,
14, & C16–19 1206 11
C9, 11, 13, 15 22 µF Tant. Cap., Teardrop, 4
25 V 0.1" Spacing
Clock Input
An external sample clock must be provided to either the BNC
connector labeled “CLOCK” or on Pin 33 of the IDC connec-
tor. This clock must comply with the logic levels outlined in the
AD768 data sheet. The “CLOCK” input is terminated with a
removable 51 Ω resistor. The IDC connector clock connection
is unterminated.
SW1. Clock source select switch. When SW1 is in position 1, Pin
33 of the IDC connected is applied to the CLOCK input of the
AD768. When SW2 is in position 2, the “CLOCK” BNC connec-
tor is applied to the CLOCK input of the AD768.
Digital Inputs
The digital inputs of the AD768, DB0–DB15, are available via
J1, a 40-pin IDC connector. These inputs should comply with
the specifications given in the AD768 data sheet.
Layout Considerations
Figures 28 and 29 show the AD768-EB ground and power
plane layouts. Figures 35–38 show the schematic diagram, trace
routing, silk screening, and component layout for the AD768 4
layer evaluation board.
Separate ground and power planes have several advantages for
high speed layouts. (For further information outlining these
advantages, see the application note “Design and Layout of a
Video Graphics System for Reduced EMI” [E1309] available
from Analog Devices [(617) 461-3392].) A solid ground plane
can be used if the digital return current can be routed such that
it does not modulate the analog ground plane. If this is not pos-
sible, it may be necessary to split the ground plane in order to
force currents to flow in a controlled direction. This type of
grounding scheme is shown in the Figure 28. The ground plane
is separated into analog and digital planes that are joined
together under the AD768. In any case, the AD768 should be
treated as an analog component and a common ground connec-
tion should be made underneath the AD768 despite some pins
being labeled “digital” ground and some as “analog” ground.
A complete parts list for the AD768 evaluation board is given in
Table IV.
AD768
REV. B –17–
IOUTA
NR
REFOUT
NC
REFCOM
IREFIN
DB0 (LSB)
DB1
DB2
DB3
DB4
DB5
DB6
DB7
LADCOM
IOUTB
(–5V) V
EE
(+5V) V
DD
(MSB) DB15
DB14
DB13
DB12
DB11
DB10
DB9
DB8
CLOCK
DCOM
AD768
1
2
3
4
5
6
7
8
9
10
11
12
13
14
28
27
26
25
24
23
22
21
20
19
18
17
16
15
C18
0.1µF
C2
1µF
C1
1µF
R1
499
A
A
R12
10k
31
29
27
25
23
21
19
17
R13
R14
R15
R16
R17
R18
R19
R20
0
0
0
0
0
0
0
0
TP1 J1
J1
J1
J1
J1
J1
J1
J1
1
2
SW2
C4
1µF C19
0.1µF
C3
1µF C17
0.1µF
A
C15
47µF, 25V C16
0.1µF
C13
47µF, 25V C14
0.1µF
A
R21
R23
R24
R25
R26
R27
R28
R29
0
0
0
0
0
0
0
0
1
3
5
7
9
11
13
15
J1
J1
J1
J1
J1
J1
J1
J1
–5A
+5D
33
J1
CLOCK
1
2
1
2
SW1
2
4
6
8
10
12
14
J1
J1
J1
J1
J1
J1
J1
16
18
20
22
24
26
28
J1
J1
J1
J1
J1
J1
J1
30
32
34
36
38
40
J1
J1
J1
J1
J1
J1
AGND CONNECTED TO DGND
ON GND PLANE UNDER U1
IN BETWEEN PINS 5 AND 25
DGND
AGND
A
C11
47µF, 25V C12
0.1µF
C9
47µF, 25V C10
0.1µF
A
–V
EE
+V
CC
A
C7
0.1µF
+V
CC
–V
EE
A
AD811
76
2
34
U3
U1
R10
499
50
JP5
R6
∞
C5
0.1µF
A
+V
CC
–V
EE
A
AD811
76
2
34
U2
R5
499
C6
0.1µF
A
C8
0.1µF
AB
11
22
JP1
JP2
50
4:1
31
T1
45 6 JP3
JP4
R9
100
R3
24.9
R2
24.9
R8
∞
A
R7
100
R11
50
A A
Figure 35. AD768 Evaluation Board Schematic
REV. B
–18–
AD768
Figure 36. Silkscreen Layer (Not to Scale)
AD768
REV. B –19–
Figure 37. Component Side PCB Layout (Not to Scale)
Figure 38. Solder Side PCB Layout (Not to Scale)
REV. B
–20–
AD768
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
R-28
300 Mil 28-Pin SOIC
SEATING
PLANE
0.0118 (0.30)
0.0040 (0.10) 0.0192 (0.49)
0.0138 (0.35)
0.1043 (2.65)
0.0926 (2.35)
0.0500
(1.27)
BSC 0.0125 (0.32)
0.0091 (0.23)
0.0500 (1.27)
0.0157 (0.40)
8°
0°
0.0291 (0.74)
0.0098 (0.25) x 45°
0.7125 (18.10)
0.6969 (17.70)
0.4193 (10.65)
0.3937 (10.00)
0.2992 (7.60)
0.2914 (7.40)
PIN 1
28 15
141
PRINTED IN U.S.A. C1941a–5–6/96
