AD9773BSVZ [ADI]
12-Bit, 160 MSPS, 2x/4x/8x Interpolating Dual TxDAC D/A Converter; 12位, 160 MSPS , 2倍/ 4倍/ 8倍内插双通道TxDAC D / A转换器![AD9773BSVZ](http://pdffile.icpdf.com/pdf1/p00199/img/icpdf/AD9773_1125378_icpdf.jpg)
型号: | AD9773BSVZ |
厂家: | ![]() |
描述: | 12-Bit, 160 MSPS, 2x/4x/8x Interpolating Dual TxDAC D/A Converter |
文件: | 总60页 (文件大小:1548K) |
中文: | 中文翻译 | 下载: | 下载PDF数据表文档文件 |
![](http://public.icpdf.com/style/img/ads.jpg)
12-Bit, 160 MSPS, 2×/4×/8× Interpolating
Dual TxDAC D/A Converter
AD9773
Versatile input data interface
FEATURES
Twos complement/straight binary data coding
Dual-port or single-port interleaved input data
Single 3.3 V supply operation
Power dissipation: typical 1.2 W @ 3.3 V
On-chip 1.2 V reference
12-bit resolution, 160 MSPS/400 MSPS
input/output data rate
Selectable 2×/4×/8× interpolating filter
Programmable channel gain and offset adjustment
fS/2, fS/4, fS/8 digital quadrature modulation capability
Direct IF transmission mode for 70 MHz + IFs
Enables image rejection architecture
Fully compatible SPI port
80-lead thin quad flat package, exposed pad (TQFP_EP)
APPLICATIONS
Communications
Excellent ac performance
Analog quadrature modulation architecture
3G, multicarrier GSM, TDMA, CDMA systems
Broadband wireless, point-to-point microwave radios
Instrumentation/ATE
SFDR −69 dBc @ 2 MHz to 35 MHz
WCDMA ACPR −69 dB @ IF = 19.2 MHz
Internal PLL clock multiplier
Selectable internal clock divider
Versatile clock input
Differential/single-ended sine wave or
TTL/CMOS/LVPECL compatible
FUNCTIONAL BLOCK DIAGRAM
IDAC
COS
AD9773
HALF-
BAND
FILTER1*
HALF-
BAND
HALF-
BAND
GAIN
DAC
OFFSET
DAC
FILTER2* FILTER3*
DATA
SIN
fDAC/2, 4, 8
SIN
ASSEMBLER
IMAGE
REJECTION/
DUAL DAC
MODE
BYPASS
MUX
12
16
16
16
16
16
I
I/Q DAC
GAIN/OFFSET
REGISTERS
LATCH
I AND Q
NONINTERLEAVED
OR INTERLEAVED
DATA
16
16
16
Q
LATCH
12
FILTER
BYPASS
MUX
COS
WRITE
MUX
CONTROL
I
IDAC
SELECT
OUT
/2
(fDAC)
CLOCK OUT
/2
/2
/2
SPI INTERFACE AND
CONTROL REGISTERS
PRESCALER
DIFFERENTIAL
CLK
PHASE DETECTOR
AND VCO
* HALF-BAND FILTERS ALSO CAN BE
CONFIGURED FOR ZERO STUFFING ONLY
PLL CLOCK MULTIPLIER AND CLOCK DIVIDER
Figure 1.
Rev. D
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 that may result from its use. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registeredtrademarks arethe property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
Fax: 781.461.3113
www.analog.com
©2007 Analog Devices, Inc. All rights reserved.
AD9773
TABLE OF CONTENTS
Features .............................................................................................. 1
Two-Port Data Input Mode ...................................................... 30
One-/Two-Port Input Modes.................................................... 30
PLL Enabled, Two-Port Mode.................................................. 30
DATACLK Inversion.................................................................. 31
DATACLK Driver Strength....................................................... 31
PLL Enabled, One-Port Mode .................................................. 31
ONEPORTCLK Inversion......................................................... 31
ONEPORTCLK Driver Strength.............................................. 32
IQ Pairing.................................................................................... 32
PLL Disabled, Two-Port Mode................................................. 32
PLL Disabled, One-Port Mode................................................. 32
Digital Filter Modes ................................................................... 33
Amplitude Modulation.............................................................. 33
Modulation, No Interpolation.................................................. 34
Modulation, Interpolation = 2× ............................................... 35
Modulation, Interpolation = 4× ............................................... 36
Modulation, Interpolation = 8× ............................................... 37
Zero Stuffing ............................................................................... 38
Interpolating (Complex Mix Mode)........................................ 38
Operations on Complex Signals............................................... 38
Applications....................................................................................... 1
Functional Block Diagram .......................................................... 1
Table of Contents.............................................................................. 2
Revision History........................................................................... 2
General Description......................................................................... 4
Product Highlights....................................................................... 4
Specifications..................................................................................... 5
DC Specifications ......................................................................... 5
Dynamic Specifications ............................................................... 6
Digital Specifications ................................................................... 7
Digital Filter Specifications......................................................... 8
Absolute Maximum Ratings............................................................ 9
Thermal Characteristics .............................................................. 9
ESD Caution.................................................................................. 9
Pin Configuration and Function Descriptions........................... 10
Typical Performance Characteristics ........................................... 12
Terminology .................................................................................... 17
Mode Control (Via SPI Port).................................................... 18
Register Description................................................................... 20
Functional Description.................................................................. 22
Serial Interface for Register Control........................................ 22
General Operation of the Serial Interface............................... 22
Instruction Byte .......................................................................... 23
Serial Interface Port Pin Descriptions ..................................... 23
MSB/LSB Transfers..................................................................... 23
Notes on Serial Port Operation ................................................ 25
DAC Operation........................................................................... 25
1R/2R Mode ................................................................................ 26
Clock Input Configurations...................................................... 27
Programmable PLL .................................................................... 27
Power Dissipation....................................................................... 29
Sleep/Power-Down Modes........................................................ 29
Complex Modulation and Image Rejection of Baseband
Signals .......................................................................................... 39
Image Rejection and Sideband Suppression of Modulated
Carriers........................................................................................ 41
Applying the Output Configurations........................................... 46
Unbuffered Differential Output, Equivalent Circuit ............. 46
Differential Coupling Using a Transformer............................ 46
Differential Coupling Using an Op Amp................................ 47
Interfacing the AD9773 with the AD8345 Quadrature
Modulator.................................................................................... 47
Evaluation Board ............................................................................ 48
Outline Dimensions....................................................................... 58
Ordering Guide .......................................................................... 58
REVISION HISTORY
Changes to Figure 108 .................................................................. 54
Updated Outline Dimensions ..................................................... 58
Changes to Ordering Guide......................................................... 58
10/07—Rev. C to Rev. D
Updated Formatting ........................................................ Universal
Changes to Figure 32 ....................................................................22
Rev. D | Page 2 of 60
AD9773
1/06—Rev. B to Rev. C
3/03—Data Sheet Changed from Rev. 0 to Rev. A.
Updated Formatting .........................................................Universal
Changes to Figure 32 .................................................................... 22
Changes to Figure 108.................................................................. 55
Updated Outline Dimensions ..................................................... 58
Changes to Ordering Guide......................................................... 58
Edits to Features ...............................................................................1
Edits to DC Specifications ..............................................................3
Edits to Dynamic Specifications ....................................................4
Edits to Pin Function Descriptions ...............................................7
Edits to Table I............................................................................... 14
Edits to Register Description—Address 02h Section............... 15
Edits to Register Description—Address 03h Section............... 16
Edits to Register Description—Address 07h, 0Bh Section...... 16
Edits to Equation 1........................................................................ 16
Edits to MSB/LSB Transfers Section........................................... 18
Changes to Figure 8 ...................................................................... 20
Edits to Programmable PLL Section........................................... 21
Added New Figure 14................................................................... 22
Renumbered Figures 15 through 69........................................... 22
Add Two-Port Data Input Mode Section................................... 23
Edits to PLL Enabled, Two-Port Mode Section ........................ 24
Edits to Figure 19 .......................................................................... 24
Edits to Figure 21 .......................................................................... 25
Edits to PLL Disabled, Two-Port Mode Section ....................... 25
Edits to Figure 22 .......................................................................... 25
Edits to Figure 23 .......................................................................... 26
Edits to Figure 26a ........................................................................ 27
Edits to Complex Modulation and Image Rejection of
4/04—Data Sheet Changed from Rev. A to Rev. B.
Update Layout....................................................................Universal
Changes to DC Specifications ....................................................... 5
Changes to Absolute Maximum Ratings...................................... 9
Changes to DAC Operation Section........................................... 25
Inserted Figure 38.......................................................................... 25
Changes to Figure 40 .................................................................... 26
Changes to Table 11 ...................................................................... 28
Changes to Programmable PLL Section..................................... 29
Changes to Power Dissipation Section....................................... 29
Changes to Figures 49, 50, and 51............................................... 29
Changes to PLL Enabled, One-Port Mode Section .................. 31
Changes to PLL Disabled, One-Port Mode Section ................. 32
Changes to Figure 102 .................................................................. 49
Changes to Figure 104 .................................................................. 50
Updated Ordering Guide ............................................................. 58
Updated Outline Dimensions...................................................... 58
Baseband Signals Section............................................................. 31
Changes to Figures 53 and 54...................................................... 38
Edits to Evaluation Board Section .............................................. 39
Changes to Figures 56 through 59 .............................................. 40
Replaced Figures 60 through 69.................................................. 42
Updated Outline Dimensions...................................................... 49
Rev. D | Page 3 of 60
AD9773
GENERAL DESCRIPTION
Dual high performance DAC outputs provide a differential
current output programmable over a 2 mA to 20 mA range. The
AD9773 is manufactured on an advanced 0.35 micron CMOS
process, operates from a single supply of 3.1 V to 3.5 V, and
consumes 1.2 W of power.
The AD97731 is the 12-bit member of the AD977x pin-
compatible, high performance, programmable 2×/4×/8×
interpolating TxDAC+® family. The AD977x family features
a serial port interface (SPI) that provides a high level of
programmability, thus allowing for enhanced system-level
options. These options include selectable 2×/4×/8× interpo-
lation filters; fS/2, fS/4, or fS/8 digital quadrature modulation
with image rejection; a direct IF mode; programmable channel
gain and offset control; programmable internal clock divider;
straight binary or twos complement data interface; and a single-
port or dual-port data interface.
Targeted at a wide dynamic range, multicarrier, and multi-
standard systems, the superb baseband performance of the
AD9773 is ideal for wide band CDMA, multicarrier CDMA,
multicarrier TDMA, multicarrier GSM, and high performance
systems employing high order QAM modulation schemes. The
image rejection feature simplifies and can help to reduce the
number of signal band filters needed in a transmit signal chain.
The direct IF mode helps to eliminate a costly mixer stage for a
variety of communications systems.
The selectable 2×/4×/8× interpolation filters simplify the
requirements of the reconstruction filters while simultaneously
enhancing the TxDAC+ family’s pass-band noise/distortion
performance. The independent channel gain and offset adjust
registers allow the user to calibrate LO feedthrough and side-
band suppression errors associated with analog quadrature
modulators. The 6 dB of gain adjustment range can also be
used to control the output power level of each DAC.
PRODUCT HIGHLIGHTS
1. The AD9773 is the 12-bit member of the AD977x pin
compatible, high performance, programmable 2×/4×/8×
interpolating TxDAC+ family.
2. Direct IF transmission is possible for 70 MHz + IFs
through a novel digital mixing process.
3. fS/2, fS/4, and fS/8 digital quadrature modulation and user
selectable image rejection simplify/remove cascaded SAW
filter stages.
4. A 2×/4×/8× user selectable interpolating filter eases data
rate and output signal reconstruction filter requirements.
5. User selectable twos complement/straight binary
data coding.
6. User programmable channel gain control over 1 dB range
in 0.01 dB increments.
7. User programmable channel offset control 10ꢀ over
the FSR.
8. Ultrahigh speed 400 MSPS DAC conversion rate.
9. Internal clock divider provides data rate clock for
easy interfacing.
10. Flexible clock input with single-ended or differential input,
CMOS, or 1 V p-p LO sine wave input capability.
11. Low power: Complete CMOS DAC operates on 1.2 W from a
3.1 V to 3.5 V single supply. The 20 mA full-scale current can
be reduced for lower power operation, and several sleep
functions reduce power during idle periods.
12. On-chip voltage reference: The AD9773 includes a 1.20 V
temperature compensated band gap voltage reference.
The AD9773 features the ability to perform fS/2, fS/4, and fS/8
digital modulation and image rejection when combined with an
analog quadrature modulator. In this mode, the AD9773
accepts I and Q complex data (representing a single or multi-
carrier waveform), generates a quadrature modulated IF signal
along with its orthogonal representation via its dual DACs, and
presents these two reconstructed orthogonal IF carriers to an
analog quadrature modulator to complete the image rejection
upconversion process. Another digital modulation mode (for
example, the direct IF mode) allows the original baseband
signal representation to be frequency translated such that pairs
of images fall at multiples of one-half the DAC update rate.
The AD977x family includes a flexible clock interface accepting
differential or single-ended sine wave or digital logic inputs.
An internal PLL clock multiplier is included and generates the
necessary on-chip high frequency clocks. It can also be disabled
to allow the use of a higher performance external clock source.
An internal programmable divider simplifies clock generation
in the converter when using an external clock source. A flexible
data input interface allows for straight binary or twos complement
formats and supports single-port interleaved or dual-port data.
1 Protected by U.S. Patent Numbers 5,568,145; 5,689,257; and 5,703,519. Other
patents pending.
13. 80-lead thin quad flat package, exposed pad (TQFP_EP).
Rev. D | Page 4 of 60
AD9773
SPECIFICATIONS
DC SPECIFICATIONS
TMIN to TMAX, AVDD = 3.3 V, CLKVDD = 3.3 V, DVDD = 3.3 V, PLLVDD = 3.3 V, IOUTFS = 20 mA, unless otherwise noted.
Table 1.
Parameter
Min
Typ
Max
Unit
RESOLUTION
DC Accuracy1
12
Bits
Integral Nonlinearity
Differential Nonlinearity
Monotonicity
−1.5
−1
0.4
0.2
+1.5
+1
LSB
LSB
Guaranteed over specified temperature range
ANALOG OUTPUT (for IR and 2R Gain Setting Modes)
Offset Error
Gain Error (with Internal Reference)
Gain Matching
Full-Scale Output Current2
Output Compliance Range
Output Resistance
−0.02
−1.0
−1.0
2
0.01
0.1
+0.02
+1.0
+1.0
20
% of FSR
% of FSR
% of FSR
mA
V
kΩ
−1.0
+1.25
200
Output Capacitance
3
pF
Gain, Offset Cal DACs, Monotonicity Guaranteed
REFERENCE OUTPUT
Reference Voltage
1.14
0.1
1.20
100
1.26
1.25
V
nA
Reference Output Current3
REFERENCE INPUT
Input Compliance Range
Reference Input Resistance
Small Signal Bandwidth
TEMPERATURE COEFFICIENTS
Offset Drift
Gain Drift (With Internal Reference)
Reference Voltage Drift
POWER SUPPLY
V
kΩ
MHz
7
0.5
0
50
50
ppm of FSR/°C
ppm of FSR/°C
ppm/°C
AVDD
Voltage Range
Analog Supply Current (IAVDD
IAVDD in Sleep Mode
CLKVDD
Voltage Range
Clock Supply Current (ICLKVDD
CLKVDD (PLL ON)
Clock Supply Current (ICLKVDD
DVDD
3.1
3.1
3.3
72.5
23.3
3.5
76
26
V
mA
mA
4
)
3.3
8.5
3.5
10.0
V
mA
4
)
)
23.5
mA
Voltage Range
Digital Supply Current (IDVDD
3.1
3.3
34
3.5
41
V
mA
4
)
Nominal Power Dissipation
PDIS
PDIS in PWDN
Power Supply Rejection Ratio—AVDD
OPERATING RANGE
380
1.75
6.0
0.4
410
mW
W
mW
% of FSR/V
°C
5
−40
+85
1 Measured at IOUTA driving a virtual ground.
2 Nominal full-scale current, IOUTFS, is 32× the IREF current.
3 Use an external amplifier to drive any external load.
4 100 MSPS fDAC with fOUT = 1 MHz, all supplies = 3.3 V, no interpolation, no modulation.
5 400 MSPS fDAC, fDATA = 50 MSPS, fS/2 modulation, PLL enabled.
Rev. D | Page 5 of 60
AD9773
DYNAMIC SPECIFICATIONS
TMIN to TMAX, AVDD = 3.3 V, CLKVDD = 3.3 V, DVDD = 3.3 V, PLLVDD = 0 V, IOUTFS = 20 mA, interpolation = 2×, differential
transformer-coupled output, 50 Ω doubly terminated, unless otherwise noted.
Table 2.
Parameter
Min
Typ
Max
Unit
DYNAMIC PERFORMANCE
Maximum DAC Output Update Rate (fDAC
Output Settling Time (tST) (to 0.025%)
Output Rise Time (10% to 90%)1
Output Fall Time (10% to 90%)1
Output Noise (IOUTFS = 20 mA)
AC LINEARITY—BASEBAND MODE
)
400
MSPS
ns
ns
ns
pA√Hz
11
0.8
0.8
50
Spurious-Free Dynamic Range (SFDR) to Nyquist (fOUT = 0 dBFS)
fDATA = 100 MSPS, fOUT = 1 MHz
fDATA = 65 MSPS, fOUT = 1 MHz
fDATA = 65 MSPS, fOUT = 15 MHz
fDATA = 78 MSPS, fOUT = 1 MHz
fDATA = 78 MSPS, fOUT = 15 MHz
fDATA = 160 MSPS, fOUT = 1 MHz
fDATA = 160 MSPS, fOUT = 15 MHz
70
84.5
83
79
83
77
dBc
dBc
dBc
dBc
dBc
dBc
dBc
75
77
Spurious-Free Dynamic Range Within a 1 MHz Window
fOUT = 0 dBFS, fDATA = 100 MSPS, fOUT = 1 MHz
Two-Tone Intermodulation (IMD) to Nyquist (fOUT1 = fOUT2 = −6 dBFS)
fDATA = 65 MSPS, fOUT1 = 10 MHz; fOUT2 = 11 MHz
fDATA = 65 MSPS, fOUT1 = 20 MHz; fOUT2 = 21 MHz
fDATA = 78 MSPS, fOUT1 = 10 MHz; fOUT2 = 11 MHz
fDATA = 78 MSPS, fOUT1 = 20 MHz; fOUT2 = 21 MHz
fDATA = 160 MSPS, fOUT1 = 10 MHz; fOUT2 = 11 MHz
fDATA = 160 MSPS, fOUT1 = 20 MHz; fOUT2 = 21 MHz
Total Harmonic Distortion (THD)
72
92.6
dBc
80
75
80
75
80
75
dBc
dBc
dBc
dBc
dBc
dBc
fDATA = 100 MSPS, fOUT = 1 MHz; 0 dBFS
−70
−82.4
dB
Signal-to-Noise Ratio (SNR)
fDATA = 78 MSPS, fOUT = 5 MHz; 0 dBFS
fDATA = 160 MSPS, fOUT = 5 MHz; 0 dBFS
70
69
dB
dB
Adjacent Channel Power Ratio (ACPR)
WCDMA with 3.84 MHz BW, 5 MHz Channel Spacing
IF = Baseband, fDATA = 76.8 MSPS
IF = 19.2 MHz, fDATA = 76.8 MSPS
69
69
dBc
dBc
Four-Tone Intermodulation
21 MHz, 22 MHz, 23 MHz, and 24 MHz at −12 dBFS (fDATA = MSPS, Missing Center)
AC LINEARITY—IF MODE
Four-Tone Intermodulation at IF = 200 MHz
201 MHz, 202 MHz, 203 MHz, and 204 MHz at −12 dBFS (fDATA = 160 MSPS, fDAC = 320 MHz)
73
69
dBFS
dBFS
1 Measured single-ended into 50 Ω load.
Rev. D | Page 6 of 60
AD9773
DIGITAL SPECIFICATIONS
TMIN to TMAX, AVDD = 3.3 V, CLKVDD = 3.3 V, PLLVDD = 0 V, DVDD = 3.3 V, IOUTFS = 20 mA, unless otherwise noted.
Table 3.
Parameter
Min
Typ
Max
Unit
DIGITAL INPUTS
Logic 1 Voltage
Logic 0 Voltage
2.1
3
0
V
V
0.9
Logic 1 Current
Logic 0 Current
Input Capacitance
CLOCK INPUTS
−10
−10
+10
+10
μA
μA
pF
5
Input Voltage Range
Common-Mode Voltage
Differential Voltage
SERIAL CONTROL BUS
Maximum SCLK Frequency (fSLCK
0
0.75
0.5
3
2.25
V
V
V
1.5
1.5
)
15
30
30
MHz
ns
ns
Minimum Clock Pulse Width High (tPWH
Minimum Clock Pulse Width Low (tPWL
)
)
Maximum Clock Rise/Fall Time
Minimum Data/Chip Select Setup Time (tDS
1
ms
ns
ns
)
25
0
Minimum Data Hold Time (tDH
)
Maximum Data Valid Time (tDV
RESET Pulse Width
Inputs (SDI, SDIO, SCLK, CSB)
Logic 1 Voltage
)
30
ns
ns
1.5
2.1
3
0
V
V
Logic 0 Voltage
0.9
Logic 1 Current
Logic 0 Current
Input Capacitance
SDIO Output
−10
−10
+10
+10
μA
μA
pF
5
Logic 1 Voltage
DRVDD − 0.6
V
Logic 0 Voltage
0.4
V
Logic 1 Current
Logic 0 Current
30
30
50
50
mA
mA
Rev. D | Page 7 of 60
AD9773
20
0
DIGITAL FILTER SPECIFICATIONS
Table 4. Half-Band Filter No. 1 (43 Coefficients)
Tap
Coefficient
–20
–40
–60
–80
–100
–120
1, 43
2, 42
3, 41
4, 40
5, 39
6, 38
7, 37
8, 36
8
0
−29
0
67
0
−134
0
244
0
−414
0
673
0
−1079
0
1772
0
−3280
0
10,364
16,384
9, 35
0
0
0
0.5
1.0
1.5
2.0
2.0
8
10, 34
11, 33
12, 32
13, 31
14, 30
15, 29
16, 28
17, 27
18, 26
19, 25
20, 24
21, 23
22
fOUT (NORMALIZED TO INPUT DATA RATE)
Figure 2. 2× Interpolating Filter Response
20
0
–20
–40
–60
–80
–100
–120
Table 5. Half-Band Filter No. 2 (19 Coefficients)
0.5
1.0
1.5
Tap
Coefficient
fOUT (NORMALIZED TO INPUT DATA RATE)
1, 19
2, 18
3, 17
4, 16
5, 15
6, 14
7, 13
8, 12
9, 11
10
19
0
−120
0
438
0
−1288
0
5047
8192
Figure 3. 4× Interpolating Filter Response
20
0
–20
–40
–60
–80
–100
–120
Table 6. Half-Band Filter No. 3 (11 Coefficients)
Tap
1, 11
2, 10
3, 9
4, 8
5, 7
6
Coefficient
7
0
−53
0
302
512
2
4
6
fOUT (NORMALIZED TO INPUT DATA RATE)
Figure 4. 8× Interpolating Filter Response
Rev. D | Page 8 of 60
AD9773
ABSOLUTE MAXIMUM RATINGS
Table 7.
Parameter
With Respect To
AGND, DGND, CLKGND
AVDD, DVDD, CLKVDD
AGND, DGND, CLKGND
AGND
AGND
DGND
DGND
CLKGND
Min
−0.3
−4.0
−0.3
−0.3
−1.0
−0.3
−0.3
−0.3
−0.3
−0.3
Max
Unit
V
V
V
V
V
V
V
V
V
V
°C
°C
°C
AVDD, DVDD, CLKVDD
AVDD, DVDD, CLKVDD
AGND, DGND, CLKGND
REFIO, FSADJ1/FSADJ2
IOUTA, IOUTB
P1B11 to P1B0, P2B11 to P2B0, RESET
DATACLK, PLL_LOCK
CLK+, CLK−
+4.0
+4.0
+0.3
AVDD + 0.3
AVDD + 0.3
DVDD + 0.3
DVDD + 0.3
CLKVDD + 0.3
CLKVDD + 0.3
DVDD + 0.3
125
LPF
CLKGND
DGND
SPI_CSB, SPI_CLK, SPI_SDIO, SPI_SDO
Junction Temperature
Storage Temperature
Lead Temperature (10 sec)
−65
+150
300
Stresses above those listed under Absolute Maximum Ratings
may cause permanent damage to the device. This is a stress
rating only; 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 ratings for extended periods may affect
device reliability.
THERMAL CHARACTERISTICS
Thermal Resistance
80-lead thin quad flat package, exposed pad (TQFP_EP)
θJA = 23.5°C/W (with thermal pad soldered to PCB)
ESD CAUTION
Rev. D | Page 9 of 60
AD9773
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61
1
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
44
43
42
41
CLKVDD
LPF
FSADJ1
FSADJ2
REFIO
RESET
SPI_CSB
SPI_CLK
SPI_SDIO
SPI_SDO
DGND
DVDD
NC
PIN 1
2
3
CLKVDD
CLKGND
CLK+
4
5
6
CLK–
7
CLKGND
DATACLK/PLL_LOCK
DGND
AD9773
TxDAC+
TOP VIEW
(Not to Scale)
8
9
10
11
12
13
14
15
16
17
18
19
20
DVDD
P1B11 (MSB)
P1B10
NC
P1B9
NC
P1B8
NC
P1B7
P2B0 (LSB)
P2B1
P1B6
DGND
DGND
DVDD
P2B2
DVDD
P1B5
P1B4
P2B3
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
NC = NO CONNECT
Figure 5. Pin Configuration
Rev. D | Page 10 of 60
AD9773
Table 8. Pin Function Descriptions
Pin No.
Mnemonic
CLKVDD
LPF
Description
1, 3
2
Clock Supply Voltage.
PLL Loop Filter.
4, 7
5
6
CLKGND
CLK+
CLK−
Clock Supply Common.
Differential Clock Input.
Differential Clock Input.
8
DATACLK/PLL_LOCK
With the PLL enabled, this pin indicates the state of the PLL. A read of a Logic 1 indicates the
PLL is in the locked state. Logic 0 indicates the PLL has not achieved lock. This pin can also be
programmed to act as either an input or output (Address 02h, Bit 3) DATACLK signal running
at the input data rate.
9, 17, 25,
35, 44, 52
10, 18, 26,
36, 43, 51
11 to 16,
19 to 24,
27 to 30,
47 to 50
DGND
Digital Common.
Digital Supply Voltage.
Port 1 Data Inputs.
No Connect.
DVDD
P1B11 (MSB) to P1B0 (LSB)
NC
31
IQSEL/P2B11 (MSB)
In one-port mode, IQSEL = 1 followed by a rising edge of the differential input clock latches
the data into the I channel input register. IQSEL = 0 latches the data into the Q channel input
register. In two-port mode, this pin becomes the Port 2 MSB.
32
ONEPORTCLK/P2B10
With the PLL disabled and the AD9773 in one-port mode, this pin becomes a clock output
that runs at twice the input data rate of the I and Q channels. This allows the AD9773 to
accept and demux interleaved I and Q data to the I and Q input registers.
33, 34, 37 to P2B9 to P2B0 (LSB)
42, 45, 46
Port 2 Data Inputs.
53
SPI_SDO
In the case where SDIO is an input, SDO acts as an output. When SDIO becomes an output,
SDO enters a high-Z state. This pin can also be used as an output for the data rate clock. For
more information, see the Two-Port Data Input Mode section.
54
55
56
57
SPI_SDIO
SPI_CLK
SPI_CSB
RESET
Bidirectional Data Pin. Data direction is controlled by Bit 7 of Register Address 00h. The
default setting for this bit is 0, which sets SDIO as an input.
Data input to the SPI port is registered on the rising edge of SPI_CLK. Data output on the SPI
port is registered on the falling edge.
Chip Select/SPI Data Synchronization. On momentary logic high, resets SPI port logic and
initializes instruction cycle.
Logic 1 resets all of the SPI port registers, including Address 00h, to their default values. A
software reset can also be done by writing a Logic 1 to SPI Register 00h, Bit 5. However, the
software reset has no effect on the bits in Address 00h.
58
59
60
REFIO
Reference Output, 1.2 V Nominal.
Full-Scale Current Adjust, Q Channel.
Full-Scale Current Adjust, I Channel.
Analog Supply Voltage.
FSADJ2
FSADJ1
AVDD
61, 63, 65,
76, 78, 80
62, 64, 66,
67, 70, 71,
74, 75, 77,
79
AGND
Analog Common.
68, 69
72, 73
IOUTB2, IOUTA2
IOUTB1, IOUTA1
Differential DAC Current Outputs, Q Channel.
Differential DAC Current Outputs, I Channel.
Rev. D | Page 11 of 60
AD9773
TYPICAL PERFORMANCE CHARACTERISTICS
T = 25°C, AVDD = 3.3 V, CLKVDD = 3.3 V, DVDD = 3.3 V, IOUTFS = 20 mA, interpolation = 2×, differential transformer-coupled output,
50 Ω doubly terminated, unless otherwise noted.
10
10
0
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–10
–20
–30
–40
–50
–60
–70
–80
–90
0
65
130
0
50
100
150
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 6. Single-Tone Spectrum @ fDATA = 65 MSPS with fOUT = fDATA/3
Figure 9. Single-Tone Spectrum @ fDATA = 78 MSPS with fOUT = fDATA/3
90
90
85
85
0dBFS
0dBFS
80
80
75
75
–6dBFS
–6dBFS
70
70
65
–12dBFS
–12dBFS
65
60
55
50
60
55
50
0
10
20
30
0
10
20
30
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 7. In-Band SFDR vs. fOUT @ fDATA = 65 MSPS
Figure 10. In-Band SFDR vs. fOUT @ fDATA = 78 MSPS
90
85
80
75
70
65
60
55
50
90
85
80
75
70
65
60
55
50
0dBFS
–6dBFS
0dBFS
–6dBFS
–12dBFS
–12dBFS
0
10
20
30
0
10
20
30
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 8. Out-of-Band SFDR vs. fOUT @ fDATA = 65 MSPS
Figure 11. Out-of-Band SFDR vs. fOUT @ fDATA = 78 MSPS
Rev. D | Page 12 of 60
AD9773
90
85
80
75
70
65
60
55
50
10
0
–3dBFS
–6dBFS
–10
–20
–30
–40
–50
–60
–70
–80
–90
0dBFS
0
10
20
30
0
100
200
300
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 15. Third-Order IMD Products vs. fOUT @ fDATA = 65 MSPS
Figure 12. Single-Tone Spectrum @ fDATA = 160 MSPS with fOUT = fDATA/3
90
90
85
–6dBFS
85
80
75
70
65
60
55
50
0dBFS
0dBFS
–12dBFS
80
75
–3dBFS
70
–6dBFS
65
60
55
50
0
10
20
30
0
10
20
30
40
50
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 13. In-Band SFDR vs. fOUT @ fDATA = 160 MSPS
Figure 16. Third-Order IMD Products vs. fOUT @ fDATA = 78 MSPS
90
90
85
80
75
70
65
60
55
50
85
80
75
70
65
60
55
50
–3dBFS
–6dBFS
–6dBFS
0dBFS
0dBFS
–12dBFS
0
10
20
30
40
50
0
20
40
60
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 14. Out-of-Band SFDR vs. fOUT @ fDATA = 160 MSPS
Figure 17. Third-Order IMD Products vs. fOUT @ fDATA = 160 MSPS
Rev. D | Page 13 of 60
AD9773
90
85
80
75
70
65
60
55
50
90
85
80
75
70
65
60
55
–3dBFS
–6dBFS
8×
1×
4×
2×
0dBFS
50
0
3.1
3.2
3.3
AVDD (V)
3.4
3.5
20
40
60
FREQUENCY (MHz)
Figure 21. Third-Order IMD Products vs. AVDD @ fOUT = 10 MHz,
fDAC = 320 MSPS, fDATA = 160 MSPS
Figure 18. Third-Order IMD Products vs. fOUT and Interpolation Rate,
1× fDATA = 160 MSPS, 2× fDATA = 160 MSPS, 4× fDATA = 80 MSPS,
8× fDATA = 50 MSPS
90
4×
8×
90
85
80
75
70
65
60
55
50
85
80
75
70
65
60
55
50
PLL OFF
2×
1×
PLL ON
–15
–10
–5
0
0
50
100
150
A
(dBFS)
OUT
INPUT DATA RATE (MSPS)
Figure 19. Third-Order IMD Products vs. AOUT and Interpolation Rate,
fDATA = 50 MSPS for All Cases, 1× fDAC = 50 MSPS, 2× fDAC = 100 MSPS,
4× fDAC = 200 MSPS, 8× fDAC = 400 MSPS
Figure 22. SNR vs. Data Rate for fOUT = 5 MHz
90
90
85
80
75
70
65
60
55
50
85
0dBFS
–12dBFS
80
f
= 65MSPS
DATA
78MSPS
75
160MSPS
70
–6dBFS
65
60
55
50
3.1
3.2
3.3
3.4
3.5
–50
0
50
C)
100
AVDD (V)
TEMPERATURE (
°
Figure 20. SFDR vs. AVDD @ fOUT = 10 MHz,
fDAC = 320 MSPS, fDATA = 160 MSPS
Figure 23. SFDR vs. Temperature @ fOUT = fDATA/11
Rev. D | Page 14 of 60
AD9773
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–20
–40
–60
–80
–100
0
5
10
15
20
25
30
35
40
0
50
100
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 27. Two-Tone IMD Performance, fDATA = 150 MSPS, Interpolation = 4×
Figure 24. Single-Tone Spurious Performance, fOUT = 10 MHz,
fDATA = 150 MSPS, No Interpolation
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
50
100
150
200
250
0
10
20
30
40
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 25. Two-Tone IMD Performance, fDATA = 150 MSPS, No Interpolation
Figure 28. Single-Tone Spurious Performance, fOUT = 10 MHz,
fDATA = 80 MSPS, Interpolation = 4×
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–20
–40
–60
–80
0
50
100
150
200
250
–100
FREQUENCY (MHz)
0
5
10
15
20
25
FREQUENCY (MHz)
Figure 29. Two-Tone IMD Performance, fOUT = 10 MHz, fDATA = 50 MSPS,
Interpolation = 8×
Figure 26. Single-Tone Spurious Performance, fOUT = 10 MHz,
DATA = 150 MSPS, Interpolation = 2×
f
Rev. D | Page 15 of 60
AD9773
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
0
20
40
60
100
200
300
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 31. Eight-Tone IMD Performance, fDATA = 160 MSPS,
Interpolation = 8×
Figure 30. Single-Tone Spurious Performance, fOUT = 10 MHz,
fDATA = 50 MSPS, Interpolation = 8×
Rev. D | Page 16 of 60
AD9773
TERMINOLOGY
Adjacent Channel Power Ratio (ACPR)
A ratio, in dBc, between the measured power within a channel
relative to its adjacent channel.
Offset Error
The deviation of the output current from the ideal of 0 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 1.
Complex Image Rejection
In a traditional two-part upconversion, two images are created
around the second IF frequency. These images are redundant
and have the effect of wasting transmitter power and system
bandwidth. By placing the real part of a second complex
modulator in series with the first complex modulator, either
the upper or lower frequency image near the second IF can be
rejected.
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.
Pass Band
Complex Modulation
Frequency band in which any input applied therein passes
unattenuated to the DAC output.
The process of passing the real and imaginary components of a
signal through a complex modulator (transfer function = ejωt
=
Power Supply Rejection
The maximum change in the full-scale output as the supplies
are varied from minimum to maximum specified voltages.
cosωt + jsinωt) and realizing real and imaginary components
on the modulator output.
Differential Nonlinearity (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.
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.
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 1 minus the output when all inputs are set to 0.
Signal-to-Noise Ratio (SNR)
SNR is the ratio of the rms value of the measured output signal
to the rms sum of all other spectral components below the
Nyquist frequency, excluding the first six harmonics and dc.
The value for SNR is expressed in decibels.
Glitch Impulse
Asymmetrical switching times in a DAC give rise to undesired
output transients that are quantified by a glitch impulse. It is
specified as the net area of the glitch in pV-s.
Spurious-Free Dynamic Range
The difference, in dB, between the rms amplitude of the output
Group Delay
signal and the peak spurious signal over the specified bandwidth.
Number of input clocks between an impulse applied at the
device input and the peak DAC output current. A half-band FIR
filter has constant group delay over its entire frequency range.
Stop-Band Rejection
The amount of attenuation of a frequency outside the pass band
applied to the DAC, relative to a full-scale signal applied at the
DAC input within the pass band.
Impulse Response
Response of the device to an impulse applied to the input.
Temperature Drift
Interpolation Filter
If the digital inputs to the DAC are sampled at a multiple rate of
It is specified as the maximum change from the ambient (25°C)
value to the value at either TMIN or TMAX. For offset and gain
drift, the drift is reported in ppm of full-scale range (FSR) per
°C. For reference drift, the drift is reported in ppm per °C.
f
DATA (interpolation rate), a digital filter can be constructed with
a sharp transition band near fDATA/2. Images that would typically
appear around fDAC (output data rate) can be greatly suppressed.
Total Harmonic Distortion (THD)
Linearity Error
THD is the ratio of the rms sum of the first six harmonic com-
ponents to the rms value of the measured fundamental. It is
expressed as a percentage or in decibels (dB).
Also called integral nonlinearity (INL), linearity error is defined
as the maximum deviation of the actual analog output from the
ideal output, determined by a straight line drawn from 0 to full
scale.
Monotonicity
A DAC is monotonic if the output either increases or remains
constant as the digital input increases.
Rev. D | Page 17 of 60
AD9773
MODE CONTROL (VIA SPI PORT)
Table 9. Mode Control via SPI Port1
Address Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
00h
SDIO
LSB, MSB First
0 = MSB
1 = LSB
Software
Reset
on Logic 1
Sleep
Mode
Logic 1
Shuts
Down the
DAC
Power-Down
Mode Logic 1
Shuts Down
All Digital and
Analog
1R/2R Mode
DAC Output
Current Set by
One or Two
External
PLL_LOCK
Indicator
Bidirectional
0 = Input
1 = I/O
Functions
Resistors
Output
Currents
0 = 2R, 1 = 1R
0 = No Zero
Stuffing on
Interpolation
Filters, Logic 1
Enables Zero
Stuffing
0 = e–jωt
DATACLK/
PLL_LOCK2
Select
0 =
PLLLOCK
01h
02h
Filter
Interpolation
Rate
Filter
Interpolation
Rate
Modulation Modulation
Mode Mode
(None, fS/2, (None, fS/2,
1 = Real Mix
Mode
0 = Complex
Mix Mode
1 = e+jωt
(1×, 2×, 4×, 8×)
(1×, 2×, 4×, 8×) fS/4, fS/8)
fS/4, fS/8)
1 =
DATACLK
0 = Signed
Input Data
1 = Unsigned
0 = Two-Port
Mode
1 = One-Port
Mode
DATACLK
Driver
Strength
DATACLK
Invert
0 = No
ONEPORTCLK IQSEL
Invert Invert
0 = No Invert 0 = No
Q First
0 = I First
1 = Q First
1 = Invert
Invert
Invert
1 = Invert
1 = Invert
03h
04h
Data Rate2
Output Clock
PLL Divide
(Prescaler)
Ratio
PLL Divide
(Prescaler)
Ratio
0 = PLL OFF2
1 = PLL ON
0 = Automatic
Charge Pump
Control
PLL Charge
Pump
Control
PLL Charge PLL Charge
Pump
Control
Pump
Control
1 =
Programmable
05h
06h
IDAC
Fine Gain
Adjustment
IDAC
Fine Gain
Adjustment
IDAC
Fine Gain
IDAC
Fine Gain
IDAC
Fine Gain
IDAC
Fine Gain
Adjustment
IDAC
Fine Gain
Adjustment Adjustment
IDAC
Fine Gain
Adjustment Adjustment Adjustment
IDAC
Coarse Gain
Adjustment
IDAC
Coarse Gain
Adjustment
IDAC
Coarse
Gain
IDAC
Coarse
Gain
Adjustment Adjustment
07h
08h
IDAC Offset
Adjustment
Bit 9
IDAC Offset
Adjustment
Bit 8
IDAC Offset IDAC Offset IDAC Offset
Adjustment Adjustment Adjustment
IDAC Offset
Adjustment
Bit 4
IDAC Offset IDAC Offset
Adjustment Adjustment
Bit 7
Bit 6
Bit 5
Bit 3
Bit 2
IDAC IOFFSET
Direction
0 = IOFFSET
on IOUTA
IDAC Offset IDAC Offset
Adjustment Adjustment
Bit 1
Bit 0
1 = IOFFSET
on IOUTB
09h
QDAC
QDAC
QDAC
QDAC
QDAC
QDAC
QDAC
QDAC
Fine Gain
Adjustment
Fine Gain
Adjustment
Fine Gain
Fine Gain
Fine Gain
Fine Gain
Adjustment
Fine Gain
Adjustment Adjustment
Fine Gain
Adjustment Adjustment Adjustment
Rev. D | Page 18 of 60
AD9773
Address Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
0Ah
QDAC
QDAC
QDAC
QDAC
Coarse Gain
Adjustment
Coarse Gain
Adjustment
Coarse Gain Coarse Gain
Adjustment Adjustment
0Bh
0Ch
QDAC
Offset
Adjustment
Bit 9
QDAC
Offset
Adjustment
Bit 8
QDAC
Offset
QDAC
Offset
QDAC
Offset
QDAC
Offset
Adjustment
Bit 4
QDAC
Offset
QDAC
Offset
Adjustment Adjustment Adjustment
Bit 7
Adjustment Adjustment
Bit 3
Bit 6
Bit 5
Bit 2
QDAC IOFFSET
Direction
0 = IOFFSET
on IOUTA
QDAC
Offset
QDAC
Offset
Adjustment Adjustment
Bit 1
Bit 0
1 = IOFFSET
on IOUTB
0Dh
Version
Register
Version
Register
Version
Register
Version
Register
1 Default values are shown in bold.
2 See the Two-Port Data Input Mode section for more information.
Rev. D | Page 19 of 60
AD9773
Bit 3: Logic 1 enables zero stuffing mode for interpolation
filters.
REGISTER DESCRIPTION
Address 00h
Bit 2: Default (1) enables the real mix mode. The I and Q data
channels are individually modulated by fS/2, fS/4, or fS/8 after
the interpolation filters. However, no complex modulation is
done. In the complex mix mode (Logic 0), the digital modu-
lators on the I and Q data channels are coupled to create a
digital complex modulator. When the AD9773 is applied in
conjunction with an external quadrature modulator, rejection
can be achieved of either the higher or lower frequency image
around the second IF frequency (that is, the LO of the analog
quadrature modulator external to the AD9773) according to the
bit value of Register 01h, Bit 1.
Bit 7: Logic 0 (default) causes the SPI_SDIO pin to act as an
input during the data transfer (Phase 2) of the communications
cycle. When set to 1, SPI_SDIO can act as an input or output,
depending on Bit 7 of the instruction byte.
Bit 6: Logic 0 (default) determines the direction (LSB/MSB
first) of the communications and data transfer communications
cycles. Refer to the MSB/LSB Transfers section for more details.
Bit 5: Writing a 1 to this bit resets the registers to their default
values and restarts the chip. The RESET bit always reads back 0.
Register Address 00h bits are not cleared by this software reset.
However, a high level at the RESET pin forces all registers,
including those in Address 00h, to their default state.
Bit 1: Logic 0 (default) causes the complex modulation to be of
the form e−jωt, resulting in the rejection of the higher frequency
image when the AD9773 is used with an external quadrature
modulator. A Logic 1 causes the modulation to be of the form
e+jωt, which causes rejection of the lower frequency image.
Bit 4: Sleep Mode. A Logic 1 to this bit shuts down the DAC
output currents.
Bit 3: Power-Down Mode. Logic 1 shuts down all analog and
Bit 0: In two-port mode, a Logic 0 (default) causes Pin 8 to act
as a lock indicator for the internal PLL. A Logic 1 in this register
causes Pin 8 to act as a DATACLK. For more information, see
the Two-Port Data Input Mode section.
digital functions except for the SPI port.
Bit 2: 1R/2R Mode. The default (0) places the AD9773 in two-
resistor mode. In this mode, the IREF currents for the I and Q
DAC references are set separately by the RSET resistors on
FSADJ1 and FSADJ2 (Pin 59 and Pin 60). In 2R mode,
assuming the coarse gain setting is full scale and the fine gain
Address 02h
Bit 7: Logic 0 (default) causes data to be accepted on the inputs
as twos complement. Logic 1 causes data to be accepted as
straight binary.
setting is zero, IFULLSCALE1 = 32 × VREF/FSADJ1 and IFULLSCALE2
=
32 × VREF/FSADJ2. With this bit set to 1, the reference currents
for both I and Q DACs are controlled by a single resistor on
Pin 60. IFULLSCALE in one-resistor mode for both I and Q DACs is
half of what it would be in 2R mode, assuming all other
conditions (RSET, register settings) remain unchanged. The full-
scale current of each DAC can still be set to 20 mA by choosing
a resistor of half the value of the RSET value used in 2R mode.
Bit 6: Logic 0 (default) places the AD9773 in two-port mode. I
and Q data enters the AD9773 via Port 1 and Port 2,
respectively. A Logic 1 places the AD9773 in one-port mode in
which interleaved I and Q data is applied to Port 1. See Table 8
for detailed information on the DATACLK/PLL_LOCK, IQSEL,
and ONEPORTCLK modes.
Bit 1: PLL_LOCK Indicator. When the PLL is enabled, reading
this bit gives the status of the PLL. A Logic 1 indicates the PLL
is locked. A Logic 0 indicates an unlocked state.
Bit 5: DATACLK Driver Strength. With the internal PLL
disabled and this bit set to Logic 0, it is recommended that
DATACLK be buffered. When this bit is set to Logic 1,
DATACLK acts as a stronger driver capable of driving small
capacitive loads.
Address 01h
Bit 7 and Bit 6: This is the filter interpolation rate according to
Table 10.
Bit 4: Logic 0 (default). A value of 1 inverts DATACLK at Pin 8.
Bit 2: Logic 0 (default). A value of 1 inverts ONEPORTCLK
Table 10.
at Pin 32.
00
01
10
11
1×
2×
4×
8×
Bit 1: The Logic 0 (default) causes IQSEL = 0 to direct input
data to the I channel, while IQSEL = 1 directs input data to the
Q channel.
Bit 5 and Bit 4: This is the modulation mode according to
Table 11.
Bit 0: The Logic 0 (default) defines IQ pairing as IQ, IQ, ...
while programming a Logic 1 causes the pair ordering to
be QI, QI, ....
Table 11.
00
01
10
11
none
fS/2
fS/4
fS/8
Rev. D | Page 20 of 60
AD9773
Address 03h
Address 05h, 09h
Bit 7: Allows the data rate clock (divided down from the DAC
clock) to be output at either the DATACLK pin (Pin 8) or at the
SPI_SDO pin (Pin 53). The default of 0 in this bit enables the
data rate clock at DATACLK, while a 1 in this bit causes the data
rate clock to be output at SPI_SDO. For more information, see
the Two-Port Data Input Mode section.
Bit 7, Bit 6, Bit 5, Bit 4, Bit 3, Bit 2, Bit 1, and Bit 0: These bits
represent an 8-bit binary number (Bit 7 MSB) that defines the
fine gain adjustment of the I (05h) and Q (09h) DAC according
to Equation 1.
Address 06h, 0Ah
Bit 3, Bit 2, Bit 1, and Bit 0: These bits represent a 4-bit binary
number (Bit 3 MSB) that defines the coarse gain adjustment of
the I (06h) and Q (0Ah) DACs according to Equation 1.
Bit 1 and Bit 0: Setting this divide ratio to a higher number
allows the VCO in the PLL to run at a high rate (for best perform-
ance), while the DAC input and output clocks run substantially
slower. The divider ratio is set according to Table 12.
Address 07h, 0Bh
Bit 7, Bit 6, Bit 5, Bit 4, Bit 3, Bit 2, Bit 1, and Bit 0: These bits
are used in conjunction with Address 08h, Address 0Ch, Bits [1:0].
Table 12.
00
01
10
11
÷1
÷2
÷4
÷8
Address 08h, 0Ch
Bit 1 and Bit 0: The 10 bits from these two address pairs (07h,
08h and 0Bh, 0Ch) represent a 10-bit binary number that
defines the offset adjustment of the I and Q DACs according to
Equation 1: (07h, 0Bh: Bit 7 MSB; 08h, 0Ch: Bit 0 LSB).
Address 04h
Bit 7: Logic 0 (default) disables the internal PLL. Logic 1
enables the PLL.
Address 08h, 0Ch
Bit 7: This bit determines the direction of the offset of the I (08h)
and Q (0Ch) DACs. A Logic 0 applies a positive offset current to
Bit 6: Logic 0 (default) sets the charge pump control to
automatic. In this mode, the charge pump bias current is
controlled by the divider ratio defined in Address 03h, Bits 1
and 0. Logic 1 allows the user to manually define the charge
pump bias current using Address 04h, Bits 2, 1, and 0. Adjusting
the charge pump bias current allows the user to optimize the
noise/settling performance of the PLL.
IOUTA, while a Logic 1 applies a positive offset current to IOUTB.
The magnitude of the offset current is defined by the bits in
Addresses 07h, 0Bh, 08h, and 0Ch according to Equation 1.
Equation 1 shows IOUTA and IOUTB as a function of fine gain,
coarse gain, and offset adjustment when using 2R mode. In 1R
mode, the current IREF is created by a single FSADJ1 resistor
(Pin 60). This current is divided equally into each channel so
that a scaling factor of one-half must be added to these
equations for full-scale currents for both DACs and the offset.
Bit 2, Bit 1, and Bit 0: With the charge pump control set to
manual, these bits define the charge pump bias current
according to Table 13.
Table 13.
000
001
010
011
111
50 μA
100 μA
200 μA
400 μA
800 μA
6× I
COARSE +1
3× I
32
FINE
256
1024 DATA
⎡
⎤
⎡
⎤
⎛
⎞
⎟
⎛
⎜
⎝
⎞⎛
⎞
⎟
⎠
⎛
⎜
⎝
⎞⎛
⎟⎜
⎠⎝
⎞
⎟
⎠
⎛
⎜
⎝
⎞
⎟
⎠
REF
REF
REF
REF
IOUTA
=
=
−
−
×
×
(A)
⎜
⎜
⎟
212
⎢
⎣
⎥
⎦
⎢
⎣
⎥
⎦
8
16
24
⎠⎝
⎞⎛
⎝
⎠
12
⎡
⎤
⎥
⎛
⎞
6× I
COARSE +1
3× I
32
FINE
256
1024
24
2
− DATA −1
⎡
⎛
⎜
⎝
⎤
⎞
⎟
⎠
⎛
⎜
⎝
⎞⎛
⎟⎜
⎠⎝
⎞
⎟
⎠
⎛
⎜
⎝
⎞
⎟
⎠
⎜
⎜
⎟
⎟
IOUTB
(A)
(1)
⎜
⎟
⎢
212
⎢
⎣
⎥
⎦
8
16
⎠⎝
⎢
⎣
⎥
⎦
⎝
⎠
OFFSET
1024
⎛
⎜
⎞
⎠
IOFFSET = 4× IREF
(A)
⎟
⎝
Rev. D | Page 21 of 60
AD9773
FUNCTIONAL DESCRIPTION
The AD9773 dual interpolating DAC consists of two data
channels that can be operated completely independently or
coupled to form a complex modulator in an image reject
transmit architecture. Each channel includes three FIR filters,
making the AD9773 capable of 2×, 4×, or 8× interpolation.
High speed input and output data rates can be achieved within
the limitations shown in Table 14.
SDO (PIN 53)
SDIO (PIN 54)
AD9773 SPI PORT
INTERFACE
SPI_CLK (PIN 55)
CSB (PIN 56)
Figure 32. SPI Port Interface
Table 14.
Interpolation Rate
(MSPS)
SERIAL INTERFACE FOR REGISTER CONTROL
Input Data Rate
(MSPS)
DAC Sample Rate
(MSPS)
The AD9773 serial port is a flexible, synchronous serial
communications port that allows an easy interface to many
industry-standard microcontrollers and microprocessors.
The serial I/O is compatible with most synchronous transfer
formats, including both the Motorola SPI and Intel® SSR
protocols. The interface allows read/write access to all registers
that configure the AD9773. Single- or multiple-byte transfers
are supported as well as MSB first or LSB first transfer formats.
The AD9773’s serial interface port can be configured as a single
pin I/O (SDIO) or two unidirectional pins for I/O (SDIO/SDO).
1×
2×
4×
8×
160
160
100
50
160
320
400
400
Both data channels contain a digital modulator capable of
mixing the data stream with an LO of fDAC/2, fDAC/4, or fDAC/8,
where fDAC is the output data rate of the DAC. A zero stuffing
feature is also included and can be used to improve pass-band
flatness for signals being attenuated by the SIN(x)/x charac-
teristic of the DAC output. The speed of the AD9773, combined
with its digital modulation capability, enables direct IF
conversion architectures at 70 MHz and higher.
GENERAL OPERATION OF THE SERIAL INTERFACE
There are two phases to a communication cycle with the
AD9773. Phase 1 is the instruction cycle, which is the writing of
an instruction byte into the AD9773 coincident with the first
eight SCLK rising edges. The instruction byte provides the
AD9773 serial port controller with information regarding the
data transfer cycle, which is Phase 2 of the communication
cycle. The Phase 1 instruction byte defines whether the upcom-
ing data transfer is read or write, the number of bytes in the
data transfer, and the starting register address for the first byte
of the data transfer. The first eight SCLK rising edges of each
communication cycle are used to write the instruction byte into
the AD9773.
The digital modulators on the AD9773 can be coupled to form
a complex modulator. By using this feature with an external
analog quadrature modulator, such as Analog Devices’ AD8345,
an image rejection architecture can be enabled. To optimize the
image rejection capability, as well as LO feedthrough in this
architecture, the AD9773 offers programmable (via the SPI
port) gain and offset adjust for each DAC.
Also included on the AD9773 are a phase-locked loop (PLL)
clock multiplier and a 1.20 V band gap voltage reference. With
the PLL enabled, a clock applied to the CLK+/CLK− inputs is
frequency multiplied internally and generates all necessary
internal synchronization clocks. Each 12-bit DAC provides two
complementary current outputs whose full-scale currents can
be determined either from a single external resistor or inde-
pendently from two separate resistors (see the 1R/2R Mode
section). The AD9773 features a low jitter, differential clock
input that provides excellent noise rejection while accepting a
sine or square wave input. Separate voltage supply inputs are
provided for each functional block to ensure optimum noise
and distortion performance.
A Logic 1 on the SPI_CSB pin, followed by a logic low, resets
the SPI port timing to the initial state of the instruction cycle.
This is true regardless of the present state of the internal
registers or the other signal levels present at the inputs to the
SPI port. If the SPI port is in the middle of an instruction cycle
or a data transfer cycle, none of the present data is written.
The remaining SCLK edges are for Phase 2 of the communication
cycle. Phase 2 is the actual data transfer between the AD9773 and
the system controller. Phase 2 of the communication cycle is a
transfer of one to four data bytes, as determined by the instruction
byte. Normally, using one multibyte transfer is the preferred
method. However, single byte data transfers are useful to reduce
CPU overhead when register access requires one byte only.
Registers change immediately upon writing to the last bit of each
transfer byte.
Sleep and power-down modes can be used to turn off the DAC
output current (sleep) or the entire digital and analog sections
(power-down) of the chip. An SPI-compliant serial port is used
to program the many features of the AD9773. Note that in
power-down mode, the SPI port is the only section of the chip
still active.
Rev. D | Page 22 of 60
AD9773
The SPI_SDO and SPI_SDIO pins go to a high impedance state
when this input is high. Chip select should stay low during the
entire communication cycle.
INSTRUCTION BYTE
The instruction byte contains the information shown in
Table 15.
SPI_SDIO (Pin 54)—Serial Data I/O
Table 15.
Data is always written into the AD9773 on this pin. However,
this pin can be used as a bidirectional data line. The configura-
tion of this pin is controlled by Bit 7 of Register Address 00h.
The default is Logic 0, which configures the SDIO pin as
unidirectional.
N1
N0
0
1
0
1
Description
0
0
1
1
Transfer 1 Byte
Transfer 2 Bytes
Transfer 3 Bytes
Transfer 4 Bytes
SPI_SDO (Pin 53)—Serial Data Out
R/W
Data is read from this pin for protocols that use separate lines
for transmitting and receiving data. In the case where the
AD9773 operates in a single bidirectional I/O mode, this pin
does not output data and is set to a high impedance state.
Bit 7 of the instruction byte determines whether a read or a
write data transfer occurs after the instruction byte write.
Logic 1 indicates read operation. Logic 0 indicates a write
operation.
MSB/LSB TRANSFERS
N1, N0
The AD9773 serial port can support both most significant bit
(MSB) first or least significant bit (LSB) first data formats. This
functionality is controlled by the first LSB bit in Register 0. The
default is MSB first.
Bits 6 and 5 of the instruction byte determine the number of
bytes to be transferred during the data transfer cycle. The bit
decodes are shown in the following table.
MSB
I7
LSB
I0
When this bit is set active high, the AD9773 serial port is in LSB
first format. In LSB first mode, the instruction byte and data
bytes must be written from LSB to MSB. In LSB first mode, the
serial port internal byte address generator increments for each
byte of the multibyte communication cycle.
I6
I5
I4
I3
I2
I1
R/W
N1
N0
A4
A3
A2
A1
A0
A4, A3, A2, A1, and A0
Bits 4, 3, 2, 1, and 0 of the instruction byte determine which
register is accessed during the data transfer portion of the
communications cycle. For multibyte transfers, this address is
the starting byte address. The remaining register addresses are
generated by the AD9773.
When this bit is set default low, the AD9773 serial port is
in MSB first format. In MSB first mode, the instruction byte
and data bytes must be written from MSB to LSB. In MSB
first mode, the serial port internal byte address generator
decrements for each byte of the multibyte communication cycle.
SERIAL INTERFACE PORT PIN DESCRIPTIONS
When incrementing from 1Fh, the address generator changes to
00h. When decrementing from 00h, the address generator
changes to 1Fh.
SPI_CLK (Pin 55)—Serial Clock
The serial clock pin is used to synchronize data to and from the
AD9773 and to run the internal state machines. The SPI_CLK
maximum frequency is 15 MHz. All data input to the AD9773
is registered on the rising edge of SPI_CLK. All data is driven
out of the AD9773 on the falling edge of SPI_CLK.
SPI_CSB (Pin 56)—Chip Select
Active low input starts and gates a communication cycle. It
allows more than one device to be used on the same serial
communications lines.
Rev. D | Page 23 of 60
AD9773
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
CS
SCLK
SDIO
R/W
I6
I5
I4
I3
I2
I1
I0
D7
D7
D6
D6
D2
D2
D1
D1
D0
D0
(N)
(N)
N
N
N
0
0
0
0
0
SDO
0
N
Figure 33. Serial Register Interface Timing MSB First
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
CS
SCLK
SDIO
I0
I1
I2
I3
I4
I5
I6
R/W
D0
D0
D1
D1
D2
D6
D7
(N)
(N)
0
0
0
0
0
0
N
N
N
N
SDO
D2
D6
D7
Figure 34. Serial Register Interface Timing LSB First
tSCLK
tDS
CS
tPWH
tPWL
SCLK
SDIO
tDS
tDH
INSTRUCTION BIT 7
INSTRUCTION BIT 6
Figure 35. Timing Diagram for Register Write to AD9773
CS
SCLK
tDV
SDIO
SDO
DATA BIT N
DATA BIT N–1
Figure 36. Timing Diagram for Register Read from AD9773
Rev. D | Page 24 of 60
AD9773
NOTES ON SERIAL PORT OPERATION
OFFSET
CONTROL
REGISTERS
OFFSET
DAC
FINE
GAIN
DAC
The AD9773 serial port configuration bits reside in Bits 6 and 7
of Register Address 00h. It is important to note that the config-
uration changes immediately upon writing to the last bit of the
register. For multibyte transfers, writing to this register can
occur during the middle of the communication cycle. Care
must be taken to compensate for this new configuration for the
remaining bytes of the current communication cycle.
GAIN
CONTROL
REGISTERS
FINE
GAIN
DAC
I
I
IDAC
OUTA1
1.2VREF
REFIO
0.1μF
OUTB1
COARSE COARSE
QDAC
I
I
OUTA2
GAIN
DAC
GAIN
DAC
OUTB2
The same considerations apply to setting the reset bit in
Register Address 00h. All other registers are set to their default
values, but the software reset does not affect the bits in
Register Address 00h.
FSADJ1
OFFSET
CONTROL
REGISTERS
FSADJ2
OFFSET
DAC
R
1
SET
GAIN
CONTROL
R
2
SET
REGISTERS
Figure 37. DAC Outputs, Reference Current Scaling,
and Gain/Offset Adjust
It is recommended to use only single-byte transfers when changing
serial port configurations or initiating a software reset.
A write to Bit 1, Bit 2, and Bit 3 of Address 00h with the same
logic levels as for Bit 7, Bit 6, and Bit 5 (bit pattern: XY1001YX
binary) allows the user to reprogram a lost serial port config-
uration and to reset the registers to their default values. A
second write to Address 00h with reset bit low and serial port
configuration as previously specified (XY) reprograms the OSC
IN multiplier setting. A changed fSYSCLK frequency is stable after
a maximum of 200 fMCLK cycles (equals wake-up time).
AVDD
84μA
REFIO
7kΩ
0.7V
DAC OPERATION
Figure 38. Internal Reference Equivalent Circuit
The dual 12-bit DAC output of the AD9773, along with the
reference circuitry, gain, and offset registers, is shown in Figure 37
and Figure 38. Note that an external reference can be used by
simply overdriving the internal reference with the external
reference. Referring to the transfer functions in Equation 1, a
reference current is set by the internal 1.2 V reference, the
external RSET resistor, and the values in the coarse gain register.
The fine gain DAC subtracts a small amount from this and the
result is input to IDAC and QDAC, where it is scaled by an
amount equal to 1024/24. Figure 39 and Figure 40 show the
scaling effect of the coarse and fine adjust DACs. IDAC and
QDAC are PMOS current source arrays, segmented in a 5-4-3
configuration. The five most significant bits control an array of
31 current sources. The next four bits consist of 15 current
sources whose values are all equal to 1/16 of an MSB current
source. The three LSBs are binary weighted fractions of the
middle bits’ current sources. All current sources are switched to
either IOUTA or IOUTB, depending on the input code.
25
20
15
10
5
2R MODE
1R MODE
0
0
5
10
15
20
COARSE GAIN REGISTER CODE
(ASSUMING R 1, R 2 = 1.9kΩ)
SET SET
Figure 39. Coarse Gain Effect on IFULLSCALE
The fine adjustment of the gain of each channel allows for
improved balance of QAM modulated signals, resulting in
improved modulation accuracy and image rejection. In the
Interfacing the AD9773 with the AD8345 Quadrature
Modulator section, the performance data shows to what degree
image rejection can be improved when the AD9773 is used with
an AD8345 quadrature modulator from Analog Devices Inc.
Rev. D | Page 25 of 60
AD9773
0
–0.5
–1.0
–1.5
–2.0
–2.5
5
4
3
2
1
0
1R MODE
2R MODE
2R MODE
1R MODE
–3.0
0
200
400
600
800
1000
0
200
400
600
800
1000
COARSE GAIN REGISTER CODE
(ASSUMING RSET1, RSET2 = 1.9kΩ)
FINE GAIN REGISTER CODE
(ASSUMING R 1, R 2 = 1.9k
Ω
)
SET
SET
Figure 40. Fine Gain Effect on IFULLSCALE
Figure 41. DAC Output Offset Current
0
–10
–20
–30
–40
–50
–60
–70
–80
The offset control defines a small current that can be added
to IOUTA or IOUTB (not both) on the IDAC and QDAC. The
selection in which IOUT for this offset current is directed toward
is programmable via Register 08h, Bit 7 (IDAC) and Register
0Ch, Bit 7 (QDAC). Figure 41 shows the scale of the offset
current that can be added to one of the complementary outputs
on the IDAC and QDAC. Offset control can be used for
suppression of LO leakage resulting from modulation of dc
signal components. If the AD9773 is dc-coupled to an external
modulator, this feature can be used to cancel the output offset
on the AD9773 as well as the input offset on the modulator.
Figure 42 shows a typical example of the effect that the offset
control has on LO suppression.
OFFSET REGISTER 1 ADJUSTED
OFFSET REGISTER 2
ADJUSTED, WITH OFFSET
REGISTER 1 SET
TO OPTIMIZED VALUE
–1024 –768
–512
–256
0
256
512
768
1024
DAC1, DAC2 (OFFSET REGISTER CODES)
In Figure 42, the negative scale represents an offset added to
IOUTB, while the positive scale represents an offset added to IOUTA
of the respective DAC. Offset Register 1 corresponds to IDAC,
while Offset Register 2 corresponds to QDAC. Figure 42
represents the AD9773 synthesizing a complex signal that is
then dc-coupled to an AD8345 quadrature modulator with an
LO of 800 MHz. The dc coupling allows the input offset of the
AD8345 to be calibrated out as well. The LO suppression at the
AD8345 output was optimized first by adjusting Offset Register 1
in the AD9773. When an optimal point was found (roughly
Code 54), this code was held in Offset Register 1, and Offset
Register 2 was adjusted. The resulting LO suppression is
70 dBFS. These are typical numbers, and the specific code for
optimization varies from part to part.
Figure 42. Offset Adjust Control, Effect on LO Suppression
1R/2R MODE
In 2R mode, the reference current for each channel is set
independently by the FSADJ resistor on that channel. The
AD9773 can be programmed to derive its reference current
from a single resistor on Pin 60 by putting the part into 1R
mode. The transfer functions in Equation 1 are valid for 2R
mode. In 1R mode, the current developed in the single FSADJ
resistor is split equally between the two channels. The result is
that in 1R mode, a scale factor of 1/2 must be applied to the
formulas in Equation 1. The full-scale DAC current in 1R mode
can still be set to as high as 20 mA by using the internal 1.2 V
reference and a 950 Ω resistor instead of the 1.9 kΩ resistor
typically used in 2R mode.
Rev. D | Page 26 of 60
AD9773
These networks depend on the assumed transmission line
CLOCK INPUT CONFIGURATIONS
impedance and power supply voltage of the clock driver.
Optimum performance of the AD9773 is achieved when the
driver is placed very close to the AD9773 clock inputs, thereby
negating any transmission line effects such as reflections due to
mismatch.
The clock inputs to the AD9773 can be driven differentially or
single-ended. The internal clock circuitry has supply and
ground (CLKVDD, CLKGND) separate from the other supplies
on the chip to minimize jitter from internal noise sources.
Figure 43 shows the AD9773 driven from a single-ended clock
source. The CLK+/CLK− pins form a differential input
(CLKIN) so that the statically terminated input must be dc-
biased to the midswing voltage level of the clock driven input.
The quality of the clock and data input signals is important in
achieving optimum performance. The external clock driver
circuitry should provide the AD9773 with a low jitter clock
input that meets the minimum/maximum logic levels while
providing fast edges. Although fast clock edges help minimize
any jitter that manifests itself as phase noise on a reconstructed
waveform, the high gain bandwidth product of the AD9773’s
clock input comparator can tolerate differential sine wave inputs
as low as 0.5 V p-p, with minimal degradation of the output
noise floor.
AD9773
R
SERIES
CLK+
CLKVDD
CLK–
V
THRESHOLD
0.1μF
PROGRAMMABLE PLL
CLKGND
CLKIN can function either as an input data rate clock (PLL
enabled) or as a DAC data rate clock (PLL disabled) according
to the state of Address 02h, Bit 7 in the SPI port register. The
internal operation of the AD9773 clock circuitry in these two
modes is illustrated in Figure 45 and Figure 46.
Figure 43. Single-Ended Clock Driving Clock Inputs
A configuration for differentially driving the clock inputs is
given in Figure 44. DC-blocking capacitors can be used to
couple a clock driver output whose voltage swings exceed
CLKVDD or CLKGND. If the driver voltage swings are within
the supply range of the AD9773, the dc-blocking capacitors and
bias resistors are not necessary.
The PLL clock multiplier and distribution circuitry produce the
necessary internal synchronized 1×, 2×, 4×, and 8× clocks for
the rising edge triggered latches, interpolation filters, modu-
lators, and DACs. This circuitry consists of a phase detector,
charge pump, voltage controlled oscillator (VCO), prescaler,
clock distribution, and SPI port control. The charge pump,
VCO, differential clock input buffer, phase detector, prescaler,
and clock distribution are all powered from CLKVDD. PLL lock
status is indicated by the logic signal at the DATACLK_PLL_LOCK
pin, as well as by the status of Bit 1, Register 00h. To ensure
optimum phase noise performance from the PLL clock
multiplier and distribution, CLKVDD should originate from a
clean analog supply. The VCO speed is a function of the input
data rate, the interpolation rate, and the VCO prescaler,
according to the following function:
AD9773
1kΩ
0.1μF
CLK+
1kΩ
0.1μF
0.1μF
ECL/PECL
CLKVDD
CLK–
1kΩ
1kΩ
CLKGND
Figure 44. Differential Clock Driving Clock Inputs
VCOSpeed
(
MHz
) =
A transformer, such as the T1-1T from Mini-Circuits®, can also be
used to convert a single-ended clock to differential. This method is
used on the AD9773 evaluation board so that an external sine wave
with no dc offset can be used as a differential clock.
Input DataRate
(
MHz)× Interpolation Rate ×Prescaler
Table 16 defines the minimum input data rates vs. the
interpolation and PLL divider setting. If the input data rate
drops below the defined minimum rates under these
conditions, VCO phase noise may increase significantly.
PECL/ECL drivers require varying termination networks, the
details of which are left out of Figure 43 and Figure 44 but can
be found in application notes such as the AND8020/D from On
Semiconductor®.
Rev. D | Page 27 of 60
AD9773
However, maximum rates of less than 160 MSPS and all
minimum fDATA rates are due to the maximum and minimum
speeds of the internal PLL VCO. Figure 48 shows typical
performance of the PLL lock signal (Pin 8 or Pin 53) when the
PLL is in the process of locking.
CLK+ CLK–
PLLVDD
PLL_LOCK
1 = LOCK
0 = NO LOCK
AD9773
INTERPOLATION
FILTERS,
MODULATORS,
AND DACS
PHASE
DETECTOR
CHARGE
PUMP
Table 16. PLL Optimization
LPF
Interpolation
Rate
Divider
Setting
Minimum
fDATA
Maximum
fDATA
2
4
8
1
1
1
1
2
2
2
2
4
4
4
4
8
8
8
8
1
2
4
8
1
2
4
8
1
2
4
8
1
2
4
8
32
16
8
160
160
112
56
160
112
56
28
100
56
28
14
50
28
14
7
1
CLOCK
DISTRIBUTION
CIRCUITRY
PRESCALER
VCO
INPUT
DATA
LATCHES
PLL DIVIDER
(PRESCALER)
CONTROL
4
INTERNAL SPI
CONTROL
REGISTERS
24
12
6
INTERPOLATION
RATE
CONTROL
PLL
CONTROL
(PLL ON)
MODULATION
RATE
CONTROL
SPI PORT
3
Figure 45. PLL and Clock Circuitry with PLL Enabled
24
12
6
CLK+ CLK–
PLL_LOCK
1 = LOCK
0 = NO LOCK
3
AD9773
24
12
6
INTERPOLATION
FILTERS,
MODULATORS,
AND DACS
PHASE
DETECTOR
CHARGE
PUMP
3
2
4
8
1
Table 17. Required PLL Prescaler Ratio vs. fDATA
CLOCK
DISTRIBUTION
CIRCUITRY
PRESCALER
VCO
INPUT
DATA
fDATA
PLL
Prescaler Ratio
LATCHES
125 MSPS
125 MSPS
100 MSPS
75 MSPS
50 MSPS
Disabled
Enabled
Enabled
Enabled
Enabled
PLL DIVIDER
(PRESCALER)
CONTROL
INTERNAL SPI
CONTROL
REGISTERS
Div 1
Div 2
Div 2
Div 4
INTERPOLATION
RATE
CONTROL
PLL
CONTROL
(PLL ON)
MODULATION
RATE
CONTROL
SPI PORT
Figure 46. PLL and Clock Circuitry with PLL Disabled
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
In addition, if the zero stuffing option is enabled, the VCO
doubles its speed again. Phase noise may be slightly higher with
the PLL enabled. Figure 47 illustrates typical phase noise
performance of the AD9773 with 2× interpolation and various
input data rates. The signal synthesized for the phase noise
measurement was a single carrier at a frequency of fDATA/4. The
repetitive nature of this signal eliminates quantization noise and
distortion spurs as a factor in the measurement. Although the
curves blend together in Figure 47, the different conditions are
given for clarity in Table 17. Table 16 details PLL divider
settings vs. interpolation rate and maximum and minimum
fDATA rates. Note that the maximum fDATA rates of 160 MSPS are
due to the maximum input data rate of the AD9773.
–110
0
1
2
3
4
5
FREQUENCY OFFSET (MHz)
Figure 47. Phase Noise Performance
Rev. D | Page 28 of 60
AD9773
76.0
75.5
75.0
74.5
74.0
73.5
73.0
72.5
72.0
4×, (MOD. ON)
8×, (MOD. ON)
2×, (MOD. ON)
4×
8×
2×
1×
0
50
100
fDATA (MHz)
150
200
Figure 48. PLL_LOCK Output Signal (Pin 8) in the
Process of Locking (Typical Lock Time)
Figure 50. IAVDD vs. fDATA vs. Interpolation Rate, PLL Disabled
It is important to note that the resistor/capacitor needed for the
PLL loop filter is internal on the AD9773. This suffices unless the
input data rate is below 10 MHz, in which case an external series
RC is required between the LPF and CLKVDD pins.
35
30
25
20
15
10
5
8
×
4
×
2×
POWER DISSIPATION
The AD9773 has three voltage supplies: DVDD, AVDD, and
CLKVDD. Figure 49, Figure 50, and Figure 51 show the current
required from each of these supplies when each is set to the
3.3 V nominal specified for the AD9773. Power dissipation (PD)
can easily be extracted by multiplying the given curves by 3.3.
As Figure 49 shows, IDVDD is very dependent on the input data
rate, the interpolation rate, and the activation of the internal
digital modulator. IDVDD, however, is relatively insensitive to the
modulation rate by itself. In Figure 50, IAVDD shows the same
type of sensitivity to the data, the interpolation rate, and the
modulator function but to a much lesser degree (<10ꢀ). In
Figure 51, ICLKVDD varies over a wide range yet is responsible for
only a small percentage of the overall AD9773 supply current
requirements.
1
×
0
0
50
100
150
200
fDATA (MHz)
Figure 51. ICLKVDD vs. fDATA vs. Interpolation Rate, PLL Disabled
SLEEP/POWER-DOWN MODES
(Control Register 00h, Bit 3 and Bit 4)
The AD9773 provides two methods for programmable
reduction in power savings. The sleep mode, when activated,
turns off the DAC output currents but the rest of the chip
remains functioning. When coming out of sleep mode, the
AD9773 immediately returns to full operation. Power-down
mode, on the other hand, turns off all analog and digital
circuitry in the AD9773 except for the SPI port. When
returning from power-down mode, enough clock cycles must
be allowed to flush the digital filters of random data acquired
during the power-down cycle.
400
8×, (MOD. ON)
2×, (MOD. ON)
350
300
250
200
150
100
50
4×, (MOD. ON)
8×
4×
2×
1×
0
0
50
100
150
200
fDATA (MHz)
Figure 49. IDVDD vs. fDATA vs. Interpolation Rate, PLL Disabled
Rev. D | Page 29 of 60
AD9773
TWO-PORT DATA INPUT MODE
ONE-PORT/TWO-PORT INPUT MODES
The digital data input ports can be configured as two
independent ports or as a single (one-port mode) port. In the
two-port mode, data at the two input ports is latched into the
AD9773 on every rising edge of the data rate clock (DATACLK).
Also, in the two-port mode, the AD9773 can be programmed to
generate an externally available DATACLK for the purpose of
data synchronization. This data rate clock can be programmed to
be available at either Pin 8 (DATACLK/PLL_LOCK) or Pin 53
(SPI_SDO). Because Pin 8 can also function as a PLL lock
indicator when the PLL is enabled, there are several options for
configuring Pin 8 and Pin 53. The following information
describes the options.
The digital data input ports can be configured as two
independent ports or as a single (one-port mode) port. In two-
port mode, the AD9773 can be programmed to generate an
externally available data rate clock (DATACLK) for the purpose
of data synchronization. Data at the two input ports can be
latched into the AD9773 on every rising clock edge of
DATACLK. In one-port mode, P2B10 and P2B11 from Input
Data Port 2 are redefined as IQSEL and ONEPORTCLK,
respectively. The input data in one-port mode is steered to one
of the two internal data channels based on the logic level of
IQSEL. A clock signal, ONEPORTCLK, is generated by the
AD9773 in this mode for the purpose of external data
synchronization. ONEPORTCLK runs at the input interleaved
data rate, which is 2× the data rate at the internal input to either
channel.
PLL Off (Register 4, Bit 7 = 0)
Register 4, Bit 7 = 0; DATACLK out of Pin 8.
Register 4, Bit 7 = 1; DATACLK out of Pin 53.
Test configurations showing the various clocks required and
produced by the AD9773 in the PLL and one-port/two-port
modes are given in Figure 101 to Figure 104. Jumper positions
needed to operate the AD9773 evaluation board in these modes
are given as well.
PLL On (Register 4, Bit 7 = 1)
Register 4, Bit 7 = 0, Register 1, Bit 0 = 0; PLL lock indicator out
of Pin 8.
Register 4, Bit 7 = 1, Register 1, Bit 0 = 0; PLL lock indicator out
of Pin 53.
Register4, Bit 7 = 0, Register 1, Bit 0 = 1; DATACLK out of Pin 8.
Register 4, Bit 7 = 1, Register 1, Bit 0 = 1; DATACLK out of Pin 53.
PLL ENABLED, TWO-PORT MODE
(Control Register 02h, Bits [6:0] and 04h, Bits [7:1]
With the phase-locked loop (PLL) enabled and the AD9773 in
two-port mode, the speed of CLKIN is inherently that of the
input data rate. In two-port mode, Pin 8 (DATACLK/PLL_
LOCK) can be programmed (Control Register 01h, Bit 0) to
function as either a lock indicator for the internal PLL or as a
clock running at the input data rate. When Pin 8 is used as a
clock output (DATACLK), its frequency is equal to that of
CLKIN. Data at the input ports is latched into the AD9773 on
the rising edge of the CLKIN. Figure 52 shows the delay, tOD,
inherent between the rising edge of CLKIN and the rising edge
of DATACLK, as well as the setup and hold requirements for
the data at Ports 1 and 2. The setup and hold times given in
Figure 52 are the input data transitions with respect to CLKIN.
Note that in two-port mode (PLL enabled or disabled), the data
rate at the interpolation filter inputs is the same as the input
data rate at Ports 1 and 2.
In one-port mode, P2B14 and P2B15 from input data port two
are redefined as IQSEL and ONEPORTCLK, respectively. The
input data in one-port mode is steered to one of the two internal
data channels based on the logic level of IQSEL. A clock signal,
ONEPORTCLK, is generated by the AD9773 in this mode for
the purpose of data synchronization. ONEPORTCLK runs at
the input interleaved data rate, which is 2× the data rate at the
internal input to either channel.
Test configurations showing the various clocks that are required
and generated by the AD9773 with the PLL enabled/disabled
and in the one-port/two-port modes are given in Figure 101 to
Figure 104. Jumper positions needed to operate the AD9773
evaluation board in these modes are given as well.
The DAC output sample rate in two-port mode is equal to the
clock input rate multiplied by the interpolation rate. If zero
stuffing is used, another factor of 2 must be included to
calculate the DAC sample rate.
Rev. D | Page 30 of 60
AD9773
DATACLK INVERSION
PLL ENABLED, ONE-PORT MODE
(Control Register 02h, Bit 4)
(Control Register 02h, Bits [6:1] and 04h, Bits [7:1]
By programming this bit, the DATACLK signal shown in
Figure 52 can be inverted. With inversion enabled, tOD refers to
the time between the rising edge of CLKIN and the falling edge
of DATACLK. No other effect on timing occurs.
In one-port mode, the I and Q channels receive their data from
an interleaved stream at Digital Input Port 1. The function of
Pin 32 is defined as an output (ONEPORTCLK) that generates a
clock at the interleaved data rate, which is 2× the internal input
data rate of the I and Q channels. The frequency of CLKIN is
equal to the internal input data rate of the I and Q channels.
The selection of the data for the I or Q channel is determined by
the state of the logic level at Pin 31 (IQSEL when the AD9773 is
in one-port mode) on the rising edge of ONEPORTCLK. Under
these conditions, IQSEL = 0 latches the data into the I channel
on the clock rising edge, while IQSEL = 1 latches the data into
the Q channel. It is possible to invert the I and Q selection by
setting Control Register 02h, Bit 1 to the invert state (Logic 1).
Figure 54 illustrates the timing requirements for the data inputs
as well as the IQSEL input. Note that the 1× interpolation rate is
not available in the one-port mode.
tOD
CLKIN
DATACLK
DATA AT PORTS
1 AND 2
The DAC output sample rate in one port mode is equal to
CLKIN multiplied by the interpolation rate. If zero stuffing is
used, another factor of 2 must be included to calculate the DAC
sample rate.
tS = 0.0ns (MAX)
tH = 2.5ns (MAX)
tS tH
Figure 52. Timing Requirements in Two-Port
Input Mode with PLL Enabled
ONEPORTCLK INVERSION
(Control Register 02h, Bit 2)
DATACLK DRIVER STRENGTH
By programming this bit, the ONEPORTCLK signal shown in
Figure 54 can be inverted. With inversion enabled, tOD refers to
the delay between the rising edge of the external clock and the
falling edge of ONEPORTCLK. The setup and hold times, tS
and tH, are with respect to the falling edge of ONEPORTCLK.
There is no other effect on timing.
(Control Register 02h, Bit 5)
The DATACLK output driver strength is capable of driving
>10 mA into a 330 Ω load while providing a rise time of 3 ns.
Figure 53 shows DATACLK driving a 330 Ω resistive load at a
frequency of 50 MHz. By enabling the drive strength option
(Control Register 02h, Bit 5), the amplitude of DATACLK under
these conditions increases by approximately 200 mV.
3.0
2.5
2.0
1.5
1.0
0.5
0
DELTA APPROX. 2.8ns
–0.5
0
10
20
30
40
50
TIME (ns)
Figure 53. DATACLK Driver Capability into 330 Ω at 50 MHz
Rev. D | Page 31 of 60
AD9773
ONEPORTCLK DRIVER STRENGTH
PLL DISABLED, TWO-PORT MODE
The drive capability of ONEPORTCLK is identical to that of
DATACLK in the two-port mode. Refer to Figure 53 for
performance under load conditions.
With the PLL disabled, a clock at the DAC output rate must be
applied to CLKIN. Internal clock dividers in the AD9773
synthesize the DATACLK signal at Pin 8, which runs at the
input data rate and can be used to synchronize the input data.
Data is latched into input Ports 1 and 2 of the AD9773 on the
rising edge of DATACLK. DATACLK speed is defined as the
speed of CLKIN divided by the interpolation rate. With zero
stuffing enabled, this division increases by a factor of 2. Figure 55
illustrates the delay between the rising edge of CLKIN and the
rising edge of DATACLK, as well as tS and tH in this mode.
tOD
tOD = 4.0ns (MIN)
TO 5.5ns (MAX)
CLKIN
tS = 3.0ns (MAX)
tH = –0.5ns (MAX)
tIQS = 3.5ns (MAX)
tIQH = –1.5ns (MAX)
ONEPORTCLK
The programmable modes DATACLK inversion and DATACLK
driver strength described in the previous section (PLL Enabled,
Two-Port Mode) have identical functionality with the PLL
disabled.
I AND Q INTERLEAVED
INPUT DATA AT PORT 1
The data rate clock created by dividing down the DAC clock in
this mode can be programmed (via Register 03h, Bit 7) to be
output from the SPI_SDO pin, rather than the DATACLK pin.
In some applications, this may improve complex image
rejection. When SPI_SDO is used as data rate clock out, tOD
increases by 1.6 ns.
tS tH
IQSEL
tOD
tIQS
tIQH
Figure 54. Timing Requirements in One-Port
Input Mode with the PLL Enabled
CLKIN
IQ PAIRING
(Control Register 02h, Bit 0)
DATACLK
In one-port mode, the interleaved data is latched into the
AD9773 internal I and Q channels in pairs. The order of how
the pairs are latched internally is defined by this control register.
The following is an example of the effect this has on incoming
interleaved data.
DATA AT PORTS
1 AND 2
tOD = 6.5ns (MIN) TO 8.0ns (MAX)
tS
tH
tS = 5.0ns (MAX)
tH = –3.2ns (MAX)
Given the following interleaved data stream, where the data
indicates the value with respect to full scale:
Figure 55. Timing Requirements in Two-Port
Input Mode with PLL Disabled
I
Q
0.5
I
1
Q
1
I
Q
0.5
I
0
Q
0
I
Q
0.5
0.5
0.5
0.5
PLL DISABLED, ONE-PORT MODE
With the control register set to 0 (I first), the data appears at the
internal channel inputs in the following order in time:
In one-port mode, data is received into the AD9773 as an
interleaved stream on Port 1. A clock signal (ONEPORTCLK),
running at the interleaved data rate, which is 2× the input data
rate of the internal I and Q channels, is available for data
synchronization at Pin 32.
I Channel
Q Channel
0.5
0.5
1
1
0.5
0.5
0
0
0.5
0.5
With the control register set to 1 (Q first), the data appears at
the internal channel inputs in the following order in time:
With PLL disabled, a clock at the DAC output rate must be
applied to CLKIN. Internal dividers synthesize the ONEPORTCLK
signal at Pin 32. The selection of the data for the I or Q channel
is determined by the state of the logic level applied to Pin 31
(IQSEL when the AD9773 is in one-port mode) on the rising
edge of ONEPORTCLK.
I Channel
Q Channel
0.5
y
1
0.5
0.5
1
0
0.5
0.5
0
x
0.5
The values x and y represent the next I value and the previous
Q value in the series.
Rev. D | Page 32 of 60
AD9773
Under these conditions, IQSEL = 0 latches the data into the I
channel on the clock rising edge, while IQSEL = 1 latches the
data into the Q channel. It is possible to invert the I and Q
selection by setting Control Register 02h, Bit 1 to the invert
state (Logic 1). Figure 56 illustrates the timing requirements for
the data inputs as well as the IQSEL input. Note that the
1× interpolation rate is not available in the one-port mode.
AMPLITUDE MODULATION
Given two sine waves at the same frequency, but with a 90°
phase difference, a point of view in time can be taken such that
the waveform that leads in phase is cosinusoidal and the
waveform that lags is sinusoidal. Analysis of complex variables
states that the cosine waveform can be defined as having real
positive and negative frequency components, while the sine
waveform consists of imaginary positive and negative frequency
images. This is shown graphically in the frequency domain in
Figure 57.
One-port mode is very useful when interfacing with devices
such as the Analog Devices AD6622 or AD6623 transmit signal
processors, in which two digital data channels have been
interleaved (multiplexed).
–jωt
e
/2j
The programmable modes’ ONEPORTCLK inversion,
ONEPORTCLK driver strength and IQ pairing described in the
PLL Enabled, Two-Port Mode section have identical
functionality with the PLL disabled.
SINE
DC
–jωt
e
/2j
tOD
–jωt
/2
–jωt
e
e
/2
COSINE
CLKIN
DC
Figure 57. Real and Imaginary Components of
Sinusoidal and Cosinusoidal Waveforms
Amplitude modulating a baseband signal with a sine or a cosine
convolves the baseband signal with the modulating carrier in
the frequency domain. Amplitude scaling of the modulated
signal reduces the positive and negative frequency images by a
factor of 2. This scaling is very important in the discussion of
the various modulation modes. The phase relationship of the
modulated signals is dependent on whether the modulating
carrier is sinusoidal or cosinusoidal, again with respect to the
reference point of the viewer. Examples of sine and cosine
modulation are given in Figure 58.
ONEPORTCLK
I AND Q INTERLEAVED
INPUT DATA AT PORT 1
tS tH
IQSEL
tOD = 4.0ns (MIN)
TO 5.5ns (MAX)
–jωt
Ae
/2j
SINUSOIDAL
MODULATION
tS = 3.0ns (MAX)
tH = –1.0ns (MAX)
tIQS = 3.5ns (MAX)
tIQH = –1.5ns (MAX)
tIQS
tIQH
DC
Figure 56. Timing Requirements in One-Port
Input Mode with DLL Disabled
–jωt
–jωt
Ae
Ae
/2j
/2
–jωt
Ae
/2
DIGITAL FILTER MODES
COSINUSOIDAL
MODULATION
The I and Q data paths of the AD9773 have their own
independent half-band FIR filters. Each data path consists of
three FIR filters, providing up to 8× interpolation for each
channel. The rate of interpolation is determined by the state of
Control Register 01h, Bits 7 and 6. Figure 2 to Figure 4 show the
response of the digital filters when the AD9773 is set to 2×, 4×,
and 8× modes. The frequency axes of these graphs have been
normalized to the input data rate of the DAC. As the graphs
show, the digital filters can provide greater than 75 dB of
out-of-band rejection.
DC
Figure 58. Baseband Signal, Amplitude Modulated
with Sine and Cosine Carriers
An online tool is available for quick and easy analysis of the
AD9773 interpolation filters in the various modes.
Rev. D | Page 33 of 60
AD9773
By comparing the digital domain spectrum to the DAC SIN(x)/x
roll-off, an estimate can be made for the characteristics required
for the DAC reconstruction filter. Note also, per the previous
discussion on amplitude modulation, that the spectral com-
ponents (where modulation is set to fS/4 or fS/8) are scaled by a
factor of 2. In the situation where the modulation is fS/2, the
modulated spectral components add constructively, and there is
no scaling effect.
MODULATION, NO INTERPOLATION
With Control Register 01h, Bit 7 and Bit 6 set to 00, the
interpolation function on the AD9773 is disabled. Figure 59 to
Figure 62 show the DAC output spectral characteristics of the
AD9773 in the various modulation modes, all with the
interpolation filters disabled. The modulation frequency is
determined by the state of Control Register 01h, Bits 5 and 4.
The tall rectangles represent the digital domain spectrum of a
baseband signal of narrow bandwidth.
The Effects of the Digital Modulation on the DAC Output Spectrum, Interpolation Disabled
0
0
–20
–20
–40
–60
–80
–40
–60
–80
–100
–100
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
fOUT
(×
fDATA
)
fOUT
(×fDATA)
Figure 59. No Interpolation, Modulation Disabled
Figure 61. No Interpolation, Modulation = fDAC/4
0
–20
–40
0
–20
–40
–60
–60
–80
–80
–100
–100
0
0.2
0.4
0.6
0.8
1.0
0
0.2
0.4
0.6
0.8
1.0
fOUT
(×fDATA)
fOUT
(×fDATA)
Figure 60. No Interpolation, Modulation = fDAC/2
Figure 62. No Interpolation, Modulation = fDAC/8
Rev. D | Page 34 of 60
AD9773
Another significant point is that the interpolation filtering is
MODULATION, INTERPOLATION = 2×
done previous to the digital modulator. For this reason, as
Figure 63 to Figure 66 show, the pass band of the interpolation
filters can be frequency shifted, giving the equivalent of a high-
pass digital filter.
With Control Register 01h, Bit 7 and Bit 6 set to 01, the inter-
polation rate of the AD9773 is 2×. Modulation is achieved by
multiplying successive samples at the interpolation filter output
by the sequence (+1, −1). Figure 63 to Figure 66 represent
the spectral response of the AD9773 DAC output with 2×
interpolation in the various modulation modes to a narrow
band baseband signal (again, the tall rectangles in the graphic).
The advantage of interpolation becomes clear in Figure 63 to
Figure 66, where it can be seen that the images that would
normally appear in the spectrum around the input data rate
frequency are suppressed by >70 dB.
Note that when using the fS/4 modulation mode, there is no
true stop band as the band edges coincide with each other. In
the fS/8 modulation mode, amplitude scaling occurs over only
a portion of the digital filter pass band due to constructive
addition over just that section of the band.
The Effects of the Digital Modulation on the DAC Output Spectrum, Interpolation = 2×
0
0
–20
–20
–40
–60
–80
–40
–60
–80
–100
–100
0
0.5
1.0
1.5
2.0
0
0.5
1.0
1.5
2.0
fOUT
(×
fDATA
)
fOUT
(×fDATA)
Figure 63. 2x Interpolation, Modulation = Disabled
Figure 65. 2x Interpolation, Modulation = fDAC/4
0
0
–20
–20
–40
–60
–80
–40
–60
–80
–100
–100
0
0.5
1.0
1.5
2.0
0
0.5
1.0
1.5
2.0
fOUT
(×fDATA)
fOUT
(×fDATA)
Figure 64. 2x Interpolation, Modulation = fDAC/2
Figure 66. 2x Interpolation, Modulation = fDAC/8
Rev. D | Page 35 of 60
AD9773
Figure 67 to Figure 70 represent the spectral response of the
AD9773 DAC output with 4× interpolation in the various
modulation modes to a narrow band baseband signal.
MODULATION, INTERPOLATION = 4×
With Control Register 01h, Bit 7 and Bit 6 set to 10, the inter-
polation rate of the AD9773 is 4×. Modulation is achieved by
multiplying successive samples at the interpolation filter output
by the sequence (0, +1, 0, −1).
The Effects of the Digital Modulation on the DAC Output Spectrum Interpolation = 4×
0
0
–20
–20
–40
–60
–80
–40
–60
–80
–100
–100
0
1
2
3
4
0
1
2
3
4
fOUT
(×
fDATA
)
fOUT
(×fDATA)
Figure 67. 4x Interpolation, Modulation Disabled
Figure 69. 4x Interpolation, Modulation = fDAC/4
0
0
–20
–20
–40
–60
–80
–40
–60
–80
–100
–100
0
1
2
3
4
0
1
2
3
4
fOUT
(×fDATA)
fOUT
(×fDATA)
Figure 68. 4x Interpolation, Modulation = fDAC/2
Figure 70. 4x Interpolation, Modulation = fDAC/8
Rev. D | Page 36 of 60
AD9773
Looking at Figure 63 to Figure 74, the user can see how higher
interpolation rates reduce the complexity of the reconstruction
filter needed at the DAC output. It also becomes apparent that
the ability to modulate by fS/2, fS/4, or fS/8 adds a degree of
flexibility in frequency planning.
MODULATION, INTERPOLATION = 8×
With Control Register 01h, Bit 7 and Bit 6 set to 11, the
interpolation rate of the AD9773 is 8×. Modulation is achieved
by multiplying successive samples at the interpolation filter
output by the sequence (0, +0.707, +1, +0.707, 0, −0.707, −1,
+0.707). Figure 71 to Figure 74 represent the spectral response
of the AD9773 DAC output with 8× interpolation in the various
modulation modes to a narrow band baseband signal.
The Effects of the Digital Modulation on the DAC Output Spectrum, Interpolation = 8×
0
0
–20
–20
–40
–60
–40
–60
–80
–80
–100
–100
0
1
2
3
4
5
6
7
8
0
1
2
3
4
fOUT (×fDATA
)
fOUT
(×fDATA)
Figure 71. 8x Interpolation, Modulation Disabled
Figure 73. 8x Interpolation, Modulation = fDAC/4
0
0
–20
–20
–40
–60
–80
–40
–60
–80
–100
–100
0
1
2
3
4
0
1
2
3
4
5
6
7
8
fOUT
(×fDATA)
fOUT (×fDATA
)
Figure 72. 8x Interpolation, Modulation = fDAC/2
Figure 74. 8x Interpolation, Modulation = fDAC/8
Rev. D | Page 37 of 60
AD9773
The net effect is to increase the DAC output sample rate by a
factor of 2× with the 0 in the SIN(x)/x DAC transfer function
occurring at twice the original frequency. A 6 dB loss in
amplitude at low frequencies is also evident, as can be seen in
Figure 76.
ZERO STUFFING
(Control Register 01h, Bit 3)
As shown in Figure 75, a 0 or null in the output frequency
response of the DAC (after interpolation, modulation, and DAC
reconstruction) occurs at the final DAC sample rate (fDAC). This
is due to the inherent SIN(x)/x roll-off response in the digital-
to-analog conversion. In applications where the desired
frequency content is below fDAC/2, this may not be a problem.
Note that at fDAC/2, the loss due to SIN(x)/x is 4 dB. In direct RF
applications, this roll-off may be problematic due to the
increased pass-band amplitude variation as well as the reduced
amplitude of the desired signal.
It is important to realize that the zero stuffing option by itself
does not change the location of the images but rather their
amplitude, pass-band flatness, and relative weighting. For
instance, in the previous example, the pass-band amplitude
flatness of the image at 3 × fDATA/4 is now improved to 0.59 dB
while the signal level has increased slightly from −10.5 dBFS to
−8.1 dBFS.
INTERPOLATING (COMPLEX MIX MODE)
Consider an application where the digital data into the AD9773
represents a baseband signal around fDAC/4 with a pass band of
(Control Register 01h, Bit 2)
f
DAC/10. The reconstructed signal out of the AD9773 would
In the complex mix mode, the two digital modulators on the
AD9773 are coupled to provide a complex modulation function.
In conjunction with an external quadrature modulator, this
complex modulation can be used to realize a transmit image
rejection architecture. The complex modulation function can
be programmed for e+jωt or e−jωt to give upper or lower image
rejection. As in the real modulation mode, the modulation
frequency ω can be programmed via the SPI port for fDAC/2,
fDAC/4, and fDAC/8, where fDAC represents the DAC output rate.
experience only a 0.75 dB amplitude variation over its pass
band. However, the image of the same signal occurring at
3 × fDAC/4 suffers from a pass-band flatness variation of 3.93 dB.
This image may be the desired signal in an IF application using
one of the various modulation modes in the AD9773. This roll-
off of image frequencies can be seen in Figure 59 to Figure 74,
where the effect of the interpolation and modulation rate is
apparent as well.
10
OPERATIONS ON COMPLEX SIGNALS
Truly complex signals cannot be realized outside of a computer
simulation. However, two data channels, both consisting of real
data, can be defined as the real and imaginary components of a
complex signal. I (real) and Q (imaginary) data paths are often
defined this way. By using the architecture defined in Figure 76,
a system can be realized that operates on complex signals,
giving a complex (real and imaginary) output.
ZERO STUFFING
ENABLED
0
–10
–20
ZERO STUFFING
–30
DISABLED
If a complex modulation function (e+jωt) is desired, the real and
imaginary components of the system correspond to the real and
imaginary components of e+jωt or cosωt and sinωt. As Figure 77
shows, the complex modulation function can be realized by
applying these components to the structure of the complex
system defined in Figure 76.
–40
–50
0
0.5
1.0
1.5
2.0
fOUT, NORMALIZED TO fDATA WITH ZERO STUFFING DISABLED (Hz)
Figure 75. Effect of Zero Stuffing on DAC’s SIN(x)/x Response
a(t)
c(t) × b(t) + d × b(t)
INPUT
OUTPUT
To improve upon the pass-band flatness of the desired image,
the zero stuffing mode can be enabled by setting the control
register bit to Logic 1. This option increases the ratio of
fDAC/fDATA by a factor of 2 by doubling the DAC sample rate and
inserting a midscale sample (that is, 1000 0000 0000 0000) after
every data sample originating from the interpolation filter. This
is important as it affects the PLL divider ratio needed to keep
the VCO within its optimum speed range. Note that the zero
stuffing takes place in the digital signal chain at the output of
the digital modulator, before the DAC.
COMPLEX FILTER
= (c + jd)
IMAGINARY
b(t)
INPUT
OUTPUT
b(t) × a(t) + c × b(t)
Figure 76. Realization of a Complex System
Rev. D | Page 38 of 60
AD9773
INPUT
INPUT
(REAL)
(REAL)
OUTPUT
(REAL)
OUTPUT
INPUT
(IMAGINARY)
INPUT
(IMAGINARY)
SINωt
90°
90°
COSωt
Figure 78. Quadrature Modulator
OUTPUT
(IMAGINARY)
The entire upconversion from baseband to transmit frequency is
represented graphically in Figure 79. The resulting spectrum
shown in Figure 79 represents the complex data consisting of the
baseband real and imaginary channels, now modulated onto
orthogonal (cosine and negative sine) carriers at the transmit
frequency. It is important to remember that in this application (two
baseband data channels) the image rejection is not dependent on
the data at either of the AD9773 input channels. In fact, image
rejection still occurs with either one or both of the AD9773 input
channels active. Note that by changing the sign of the sinusoidal
multiplying term in the complex modulator, the upper sideband
image could be suppressed while passing the lower one. This is
easily done in the AD9773 by selecting the e+jωt bit (Register 01h, Bit
1). In purely complex terms, Figure 79 represents the two-stage
upconversion from complex baseband to carrier.
–jωt
e
= COSωt + jSINωt
Figure 77. Implementation of a Complex Modulator
COMPLEX MODULATION AND IMAGE REJECTION
OF BASEBAND SIGNALS
In traditional transmit applications, a two-step upconversion is
done in which a baseband signal is modulated by one carrier to
an intermediate frequency (IF) and then modulated a second
time to the transmit frequency. Although this approach has
several benefits, a major drawback is that two images are
created near the transmit frequency. Only one image is needed,
the other being an exact duplicate. Unless the unwanted image
is filtered, typically with analog components, transmit power is
wasted and the usable bandwidth available in the system is
reduced.
A more efficient method of suppressing the unwanted image
can be achieved by using a complex modulator followed by a
quadrature modulator. Figure 78 is a block diagram of a
quadrature modulator. Note that it is in fact the real output
half of a complex modulator. The complete upconversion can
actually be referred to as two complex upconversion stages,
the real output of which becomes the transmitted signal.
Rev. D | Page 39 of 60
AD9773
REAL CHANNEL (OUT)
A/2 A/2
1
–fC
fC
REAL CHANNEL (IN)
A
–B/2J
B/2J
fC
DC
–
fC
COMPLEX
MODULATOR
TO QUADRATURE
MODULATOR
IMAGINARY CHANNEL (OUT)
–A/2J A/2J
IMAGINARY CHANNEL (IN)
–
fC
–fC
B
DC
B/2
B/2
fC
–
fC
A/4 + B/4J A/4 – B/4J
A/4 + B/4J A/4 – B/4J
2
–
fQ
fQ
–fQ
–
fC
–
fQ
+
fC
fQ
–
fC
fQ + fC
OUT
REAL
–A/4 – B/4J A/4 – B/4J
A/4 + B/4J –A/4 + B/4J
QUADRATURE
MODULATOR
–fQ
fQ
IMAGINARY
REJECTED IMAGES
A/2 + B/2J
A/2 – B/2J
–fQ
fQ
1fC = COMPLEX MODULATION FREQUENCY
2fQ = QUADRATURE MODULATION FREQUENCY
Figure 79. Two-Stage Upconversion and Resulting Image Rejection
Rev. D | Page 40 of 60
AD9773
COMPLEX BASEBAND
SIGNAL
A system in which multiple baseband signals are complex
modulated and then applied to the AD9773 real and imaginary
inputs followed by a quadrature modulator is shown in
Figure 82, which also describes the transfer function of this
system and the spectral output. Note the similarity of the
transfer functions given in Figure 82 and Figure 80. Figure 82
adds an additional complex modulator stage for the purpose of
summing multiple carriers at the AD9773 inputs. Also, as in
Figure 79, the image rejection is not dependent on the real or
imaginary baseband data on any channel. Image rejection on a
channel occurs if either the real or imaginary data, or both, is
present on the baseband channel.
1
j(ω1 + ω2)t
×
OUTPUT = REAL
1/2
e
1/2
= REAL
–ω1 – ω2
DC
ω1 + ω2
FREQUENCY
Figure 80. Two-Stage Complex Upconversion
IMAGE REJECTION AND SIDEBAND SUPPRESSION
OF MODULATED CARRIERS
As shown in Figure 79, image rejection can be achieved by
applying baseband data to the AD9773 and following the
AD9773 with a quadrature modulator. To process multiple
carriers while still maintaining image reject capability, each
carrier must be complex modulated. As Figure 81 shows, single
or multiple complex modulators can be used to synthesize
complex carriers. These complex carriers are then summed and
applied to the real and imaginary inputs of the AD9773.
It is important to remember that the magnitude of a complex
signal can be 1.414× the magnitude of its real or imaginary
components. Due to this 3 dB increase in signal amplitude, the
real and imaginary inputs to the AD9773 must be kept at least
3 dB below full scale when operating with the complex
modulator. Overranging in the complex modulator results in
severe distortion at the DAC output.
BASEBAND CHANNEL 1
R(1)
REAL INPUT
COMPLEX
MULTICARRIER
REAL OUTPUT =
R(1) + R(2) + . . .R(N)
(TO REAL INPUT OF AD9773)
MODULATOR 1
R(1)
IMAGINARY INPUT
BASEBAND CHANNEL 2
REAL INPUT
R(2)
R(2)
COMPLEX
MODULATOR 2
MULTICARRIER
IMAGINARY OUTPUT =
I(1) + I(2) + . . .I(N)
IMAGINARY INPUT
(TO IMAGINARY INPUT OF AD9773)
R(N) = REAL OUTPUT OF N
I(N) = IMAGINARY OUTPUT OF N
BASEBAND CHANNEL N
REAL INPUT
R(N)
R(N)
COMPLEX
MODULATOR N
IMAGINARY INPUT
Figure 81. Synthesis of Multicarrier Complex Signal
MULTIPLE
BASEBAND
CHANNELS
REAL
REAL
REAL
REAL
MULTIPLE
COMPLEX
MODULATORS
AD9773
COMPLEX
MODULATOR
FREQUENCY = ω
QUADRATURE
MODULATOR
FREQUENCY = ω
IMAGINARY
IMAGINARY
IMAGINARY
Q
FREQUENCY = ω , ω ...ω
1
2
N
C
COMPLEX BASEBAND
SIGNAL
×
OUTPUT = REAL
j(ω + ω + ω )t
e
N
C
Q
–ω – ω – ω
ω
+ ω + ω
Q
DC
REJECTED IMAGES
1
C
Q
1
C
Figure 82. Image Rejection with Multicarrier Signals
Rev. D | Page 41 of 60
AD9773
The complex carrier synthesized in the AD9773 digital modu-
lator is accomplished by creating two real digital carriers in
quadrature. Carriers in quadrature cannot be created with the
modulator running at fDAC/2. As a result, complex modulation
only functions with modulation rates of fDAC/4 and fDAC/8.
Region C
Region C is most accurately described as a downconversion, as
the modulating carrier is −ejωt. If viewed as a complex signal, only
the images in Region C remain. This image appears on the real
and imaginary outputs of the AD9773, as well as on the output of
the quadrature modulator, where the center of the spectral plot
now represents the quadrature modulator LO and the horizontal
scale represents the frequency offset from this LO.
Regions A and B of Figure 83 to Figure 88 are the result of the
complex signal described previously, when complex modulated
in the AD9773 by +ejωt. Regions C and D are the result of the
complex signal described previously, again with positive fre-
quency components only, modulated in the AD9773 by −ejωt.
The analog quadrature modulator after the AD9773 inherently
modulates by +ejωt.
Region D
Region D is the image (complex conjugate) of Region C. If a
spectrum analyzer is used to view the real or imaginary DAC
outputs of the AD9773, Region D appears in the spectrum.
However, on the output of the quadrature modulator, Region D
is rejected.
Region A
Region A is a direct result of the upconversion of the complex
signal near baseband. If viewed as a complex signal, only the
images in Region A remain. The complex Signal A, consisting
of positive frequency components only in the digital domain,
has images in the positive odd Nyquist zones (1, 3, 5, …), as
well as images in the negative even Nyquist zones. The
appearance and rejection of images in every other Nyquist
zone becomes more apparent at the output of the quadrature
modulator. The A images appear on the real and the imaginary
outputs of the AD9773, as well as on the output of the quadrature
modulator, where the center of the spectral plot now represents
the quadrature modulator LO and the horizontal scale now
represents the frequency offset from this LO.
Figure 89 to Figure 96 show the measured response of the AD9773
and AD8345 given the complex input signal to the AD9773 in
Figure 89. The data in these graphs was taken with a data rate of
12.5 MSPS at the AD9773 inputs. The interpolation rate of 4× or 8×
gives a DAC output data rate of 50 MSPS or 100 MSPS. As a result,
the high end of the DAC output spectrum in these graphs is the
first null point for the SIN(x)/x roll-off, and the asymmetry of the
DAC output images is representative of the SIN(x)/x roll-off over
the spectrum. The internal PLL was enabled for these results. In
addition, a 35 MHz third-order low-pass filter was used at the
AD9773/AD8345 interface to suppress DAC images.
An important point can be made by looking at Figure 91 and
Figure 93. Figure 91 represents a group of positive frequencies
modulated by complex +fDAC/4, while Figure 93 represents a
group of negative frequencies modulated by complex −fDAC/4.
When looking at the real or imaginary outputs of the AD9773,
as shown in Figure 91 and Figure 93, the results look identical.
However, the spectrum analyzer cannot show the phase
relationship of these signals. The difference in phase between
the two signals becomes apparent when they are applied to
the AD8345 quadrature modulator, with the results shown in
Figure 92 and Figure 94.
Region B
Region B is the image (complex conjugate) of Region A. If a
spectrum analyzer is used to view the real or imaginary DAC
outputs of the AD9773, Region B appears in the spectrum.
However, on the output of the quadrature modulator, Region B
is rejected.
Rev. D | Page 42 of 60
AD9773
0
–20
–40
–60
–80
0
–20
–40
–60
–80
D
A
B
C
D
A
B
C
D
A
B
C D
A
B
C
–100
–100
–2.0
–1.5
–1.0
–0.5
0
0.5
1.0
1.5
2.0
–2.0
–1.5
–1.0
–0.5
0
0.5
1.0
1.5
2.0
(LO)
(LO)
fOUT (×fDATA
fOUT (×fDATA
)
)
Figure 83. 2x Interpolation, Complex fDAC/4 Modulation
Figure 86. 2x Interpolation, Complex fDAC/8 Modulation
0
–20
–40
–60
–80
0
–20
–40
–60
–80
D
A
B
C
D
A
B
C
D
A
B
C
D
A
B
C
–100
–100
–4.0
–3.0
–2.0
–1.0
0
1.0
2.0
3.0
4.0
–4.0
–3.0
–2.0
–1.0
0
1.0
2.0
3.0
4.0
(LO)
(LO)
fOUT (×fDATA
fOUT (×fDATA
)
)
Figure 84. 4x Interpolation, Complex fDAC/4 Modulation
Figure 87. 4x Interpolation, Complex fDAC/8 Modulation
0
–20
–40
–60
–80
0
–20
–40
–60
–80
D
A
B
C
D
A
B
C
D A
B C
D A
B C
–100
–100
–8.0
–6.0
–4.0
–2.0
0
2.0
4.0
6.0
8.0
–8.0
–6.0
–4.0
–2.0
0
2.0
4.0
6.0
8.0
(LO)
(LO)
fOUT (×fDATA
)
fOUT (×fDATA
)
Figure 85. 8x Interpolation, Complex fDAC/4 Modulation
Figure 88. 8x Interpolation, Complex fDAC/8 Modulation
Rev. D | Page 43 of 60
AD9773
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
0
10
20
30
40
10
20
30
40
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 91. AD9773 Real DAC Output of Complex Input Signal Near
Baseband (Positive Frequencies Only), Interpolation = 4x,
Complex Modulation in AD9773 = +fDAC/4
Figure 89. AD9773 Real DAC Output of Complex Input Signal Near Baseband
(Positive Frequencies Only), Interpolation = 4x, No Modulation in AD9773
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
750
760
770
780
790
800
810
820
830
750
760
770
780
790
800
810
820
830
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 90. AD9773 Complex Output from Figure 89, Now Quadrature Modulated
by AD8345 (LO = 800 MHz)
Figure 92. AD9773 Complex Output from Figure 91, Now Quadrature Modulated
by AD8345 (LO = 800 MHz)
Rev. D | Page 44 of 60
AD9773
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
0
20
40
60
80
0
10
20
30
40
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 95. AD9773 Real DAC Output of Complex Input Signal Near
Baseband (Positive Frequencies Only), Interpolation = 8x,
Complex Modulation in AD9773 = +fDAC/8
Figure 93. AD9773 Real DAC Output of Complex Input Signal Near
Baseband (Negative Frequencies Only), Interpolation = 4x,
Complex Modulation in AD9773 = −fDAC/4
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
700
–100
750
720
740
760
780
800
820
840
860
760
770
780
790
800
810
820
830
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 96. AD9773 Complex Output from Figure 95, Now Quadrature Modulated
by AD8345 (LO = 800 MHz)
Figure 94. AD9773 Complex Output from Figure 93, Now Quadrature Modulated
by AD8345 (LO = 800 MHz)
Rev. D | Page 45 of 60
AD9773
APPLYING THE OUTPUT CONFIGURATIONS
For the typical situation, where IOUTFS = 20 mA and RA and RB
both equal 50 Ω, the equivalent circuit values become
The following sections illustrate typical output configurations
for the AD9773. Unless otherwise noted, it is assumed that
IOUTFS is set to a nominal 20 mA. For applications requiring
optimum dynamic performance, a differential output
VSOURCE = 2 V p-p
ROUT = 100 Ω
configuration is suggested. A simple differential output can be
achieved by converting IOUTA and IOUTB to a voltage output by
terminating them to AGND via equal value resistors. This type of
configuration may be useful when driving a differential voltage
input device such as a modulator. If a conversion to a single-
ended signal is desired and the application allows for ac coupling,
an RF transformer may be useful, or if power gain is required, an
op amp may be used. The transformer configuration provides
optimum high frequency noise and distortion performance. The
differential op amp configuration is suitable for applications
requiring dc coupling, signal gain, and/or level shifting within the
bandwidth of the chosen op amp.
Note that the output impedance of the AD9773 DAC itself
is greater than 100 kΩ and typically has no effect on the
impedance of the equivalent output circuit.
DIFFERENTIAL COUPLING USING A
TRANSFORMER
An RF transformer can be used to perform a differential-to-
single-ended signal conversion, as shown in Figure 98. A
differentially coupled transformer output provides the optimum
distortion performance for output signals whose spectral content
lies within the transformer’s pass band. An RF transformer such as
the Mini-Circuits T1-1T provides excellent rejection of common-
mode distortion (that is, even-order harmonics) and noise over a
wide frequency range. It also provides electrical isolation and the
ability to deliver twice the power to the load. Transformers with
different impedance ratios may also be used for impedance
matching purposes.
A single-ended output is suitable for applications requiring a
unipolar voltage output. A positive unipolar output voltage
results if IOUTA and/or IOUTB is connected to a load resistor, RLOAD
referred to AGND. This configuration is most suitable for a
single-supply system requiring a dc-coupled, ground-referred
output voltage. Alternatively, an amplifier could be configured
as an I-V converter, thus converting IOUTA or IOUTB into a
negative unipolar voltage. This configuration provides the best
DAC dc linearity as IOUTA or IOUTB are maintained at ground or
virtual ground.
,
MINI-CIRCUITS
T1-1T
I
OUTA
R
DAC
LOAD
I
OUTB
Figure 98. Transformer-Coupled Output Circuit
UNBUFFERED DIFFERENTIAL OUTPUT,
EQUIVALENT CIRCUIT
The center tap on the primary side of the transformer must be
connected to AGND to provide the necessary dc current path
for both IOUTA and IOUTB. The complementary voltages appearing
at IOUTA and IOUTB (that is, VOUTA and VOUTB) swing symmetrically
around AGND and should be maintained within the specified
output compliance range of the AD9773. A differential resistor,
RDIFF, may be inserted in applications where the output of the
transformer is connected to the load, RLOAD, via a passive
reconstruction filter or cable. RDIFF is determined by the
transformer’s impedance ratio and provides the proper source
termination that results in a low VSWR. Note that approximately
In many applications, it may be necessary to understand the
equivalent DAC output circuit. This is especially useful when
designing output filters or when driving inputs with finite input
impedances. Figure 97 illustrates the output of the AD9773 and
the equivalent circuit. A typical application where this information
may be useful is when designing an interface filter between
the AD9773 and the Analog Devices AD8345 quadrature
modulator.
AD9773
I
V
+
OUT
OUTA
half the signal power dissipates across RDIFF
.
I
V
–
OUTB
OUT
R
B
R
A
R
+ R
B
A
V
=
SOURCE
V
OUT
I
× (R + R
)
OUTFS
A
B
(DIFFERENTIAL)
p-p
Figure 97. DAC Output Equivalent Circuit
Rev. D | Page 46 of 60
AD9773
DAC Compliance Voltage/Input Common-Mode Range
DIFFERENTIAL COUPLING USING AN OP AMP
The dynamic range of the AD9773 is optimal when the DAC
outputs swing between 1.0 V. The input common-mode range
of the AD8345, at 0.7 V, allows optimum dynamic range to be
achieved in both components.
An op amp can also be used to perform a differential-to-single-
ended conversion, as shown in Figure 99. This has the added
benefit of providing signal gain as well. In Figure 99, the
AD9773 is configured with two equal load resistors, RLOAD, of
25 Ω. The differential voltage developed across IOUTA and IOUTB is
converted to a single-ended signal via the differential op amp
configuration. An optional capacitor can be installed across
IOUTA and IOUTB, forming a real pole in a low-pass filter. The
addition of this capacitor also enhances the op amp’s distortion
performance by preventing the DAC’s fast slewing output from
overloading the input of the op amp.
Gain/Offset Adjust
The matching of the DAC output to the common-mode input
of the AD8345 allows the two components to be dc-coupled,
with no level shifting necessary. The combined voltage offset of
the two parts can therefore be compensated via the AD9773
programmable offset adjust. This allows excellent LO cancel-
lation at the AD8345 output. The programmable gain adjust
allows for optimal image rejection as well.
500Ω
225Ω
I
The AD9773 evaluation board includes an AD8345 and
recommended interface (Figure 105 and Figure 106). On the
output of the AD9773, R9 and R10 convert the DAC output
current to a voltage. R16 may be used to execute a slight
common-mode shift if necessary. The (now voltage) signal is
applied to a low-pass reconstruction filter to reject DAC images.
The components installed on the AD9773 provide a 35 MHz
cutoff but may be changed to fit the application. A balun (Mini-
Circuits ADTL1-12) is used to cross the ground plane boundary
to the AD8345. Another balun (Mini-Circuits ETC1-1-13) is
used to couple the LO input of the AD8345. The interface
requires a low ac impedance return path from the AD8345, so a
single connection between the AD9773 and AD8345 ground
planes is recommended.
OUTA
AD8021
DAC
I
OUTB
C
225Ω
500Ω
OPT
AVDD
R
OPT
225Ω
25Ω
25Ω
Figure 99. Op Amp-Coupled Output Circuit
The common-mode (and second-order distortion) rejection of
this configuration is typically determined by the resistor
matching. The op amp used must operate from a dual supply
since its output is approximately 1.0 V. A high speed amplifier,
such as the AD8021, capable of preserving the differential
performance of the AD9773 while meeting other system level
objectives (for example, cost, power) is recommended. The op
amp’s differential gain, its gain setting resistor values, and full-
scale output swing capabilities should all be considered when
optimizing this circuit. ROPT is necessary only if level shifting is
required on the op amp output. In Figure 99, AVDD, which is
the positive analog supply for both the AD9773 and the op amp,
is also used to level shift the differential output of the AD9773
to midsupply (for example, AVDD/2).
The performance of the AD9773 and AD8345 in an image
reject transmitter, reconstructing three WCDMA carriers, can
be seen in Figure 100. The LO of the AD8345 in this application
is 800 MHz. Image rejection (50 dB) and LO feedthrough
(−78 dBFS) have been optimized with the programmable
features of the AD9773. The average output power of the digital
waveform for this test was set to −15 dBFS to account for the
peak-to-average ratio of the WCDMA signal.
INTERFACING THE AD9773 WITH THE AD8345
QUADRATURE MODULATOR
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
–100
The AD9773 architecture was defined to operate in a transmit
signal chain using an image reject architecture. A quadrature
modulator is also required in this application and should be
designed to meet the output characteristics of the DAC as much
as possible. The AD8345 from Analog Devices meets many of
the requirements for interfacing with the AD9773. As with any
DAC output interface, there are a number of issues that have to
be resolved. The following sections list some of these major issues.
762.5
782.5
802.5
822.5
842.5
FREQUENCY (MHz)
Figure 100. AD9773/AD8345 Synthesizing a Three-Carrier
WCDMA Signal at an LO of 800 MHz
Rev. D | Page 47 of 60
AD9773
EVALUATION BOARD
DAC DIFFERENTIAL OUTPUTS
The AD9773 evaluation board allows easy configuration of the
various modes, programmable via the SPI port. Software is
available for programming the SPI port from Windows 95®,
Windows 98®, or Windows NT®/2000. The evaluation board
also contains an AD8345 quadrature modulator and support
circuitry that allows the user to optimally configure the AD9773
in an image reject transmit signal chain.
Transformers T2 and T3 should be in place. Note that the lower
band of operation for these transformers is 300 kHz to 500 kHz.
Jumpers 4, 8, 13 to 17, and 28 to 30 should remain unsoldered.
The outputs are taken from S3 and S4.
USING THE AD8345
Remove Transformers T2 and T3. Jumpers JP4 and Jumpers 28
to 30 should remain unsoldered. Jumpers 13 to 16 should be
soldered. The desired components for the low-pass interface
filters L6, L7, C55, and C81 should be in place. The LO drive is
connected to the AD8345 via J10 and the balun T4; AD8345
output is taken from J9.
Figure 101 through Figure 104 describe how to configure the
evaluation board in the one-port and two-port input modes
with the PLL enabled and disabled. Refer to Figure 105 through
Figure 114, the schematics, and the layout for the AD9773
evaluation board for the jumper locations described below. The
AD9773 outputs can be configured for various applications by
referring to the following instructions.
DAC SINGLE-ENDED OUTPUTS
Remove transformers T2 and T3. Solder jumper link JP4 or
JP28 to look at the DAC1 outputs. Solder jumper link JP29 or
JP30 to look at the DAC2 outputs. Jumper 8 and Jumpers 13 to
17 should remain unsoldered. Jumpers JP35 to JP38 may be
used to ground one of the DAC outputs while the other is
measured single-ended. Optimum single-ended distortion
performance is typically achieved in this manner. The outputs
are taken from S3 and S4.
Rev. D | Page 48 of 60
AD9773
LECROY
PULSE
GENERATOR
SIGNAL GENERATOR
TRIG
INP
DATACLK
CLK+/CLK–
INPUT CLOCK
AWG2021
OR
DG2020
40-PIN RIBBON CABLE
DAC1, DB11–DB0
DAC2, DB11–DB0
AD9773
JUMPER CONFIGURATION FOR TWO-PORT MODE, PLL ON
SOLDERED/IN UNSOLDERED/OUT
JP1 –
JP2 –
×
×
JP3 –
JP5 –
×
×
JP6 –
×
×
×
JP12 –
JP24 –
JP25 –
JP26 –
JP27 –
JP31 –
JP32 –
JP33 –
×
×
×
×
×
×
NOTES
1. TO USE PECL DRIVER (U8), SOLDER JP41 AND JP42 AND REMOVE TRANSFORMER T1.
2. IN TWO-PORT MODE, IF DATACLK/PLL_LOCK IS PROGRAMMED TO OUTPUT PIN 8, JP25
AND JP39 SHOULD BE SOLDERED. IF DATACLK/PLL_LOCK IS PROGRAMMED TO OUTPUT
PIN 53, JP46 AND JP47 SHOULD BE SOLDERED. FOR MORE INFORMATION, SEE THE
TWO-PORT DATA INPUT MODE SECTION.
Figure 101. Test Configuration for AD9773 in Two-Port Mode with PLL Enabled, Signal Generator Frequency = Input Data Rate,
DAC Output Data Rate = Signal Generator Frequency × Interpolation Rate
LECROY
PULSE
GENERATOR
SIGNAL GENERATOR
TRIG
INP
ONEPORTCLK
CLK+/CLK–
INPUT CLOCK
AWG2021
OR
DG2020
DAC1, DB11–DB0
DAC2, DB11–DB0
AD9773
JUMPER CONFIGURATION FOR ONE-PORT MODE, PLL ON
SOLDERED/IN UNSOLDERED/OUT
JP1 –
JP2 –
×
×
JP3 –
×
JP5 –
JP6 –
×
×
×
×
JP12 –
JP24 –
JP25 –
JP26 –
JP27 –
JP31 –
JP32 –
JP33 –
×
×
×
×
×
×
NOTES
1. TO USE PECL DRIVER (U8), SOLDER JP41 AND JP42 AND REMOVE TRANSFORMER T1.
Figure 102. Test Configuration for AD9773 in One-Port Mode with PLL Enabled, Signal Generator Frequency = One-Half Interleaved Input Data Rate,
ONEPORTCLK = Interleaved Input Data Rate, DAC Output Data Rate = Signal Generator Frequency × Interpolation Rate
Rev. D | Page 49 of 60
AD9773
LECROY
PULSE
GENERATOR
SIGNAL GENERATOR
TRIG
INP
DATACLK
CLK+/CLK–
INPUT CLOCK
AWG2021
OR
DG2020
40-PIN RIBBON CABLE
DAC1, DB11–DB0
DAC2, DB11–DB0
AD9773
JUMPER CONFIGURATION FOR TWO-PORT MODE, PLL OFF
SOLDERED/IN UNSOLDERED/OUT
JP1 –
JP2 –
×
×
JP3 –
JP5 –
×
×
JP6 –
×
×
×
JP12 –
JP24 –
JP25 –
JP26 –
JP27 –
JP31 –
JP32 –
JP33 –
×
×
×
×
×
×
NOTES
1. TO USE PECL DRIVER (U8), SOLDER JP41 AND JP42 AND REMOVE TRANSFORMER T1.
2. IN TWO-PORT MODE, IF DATACLK/PLL_LOCK IS PROGRAMMED TO OUTPUT PIN 8, JP25
AND JP39 SHOULD BE SOLDERED. IF DATACLK/PLL_LOCK IS PROGRAMMED TO OUTPUT
PIN 53, JP46 AND JP47 SHOULD BE SOLDERED. FOR MORE INFORMATION, SEE THE
TWO-PORT DATA INPUT MODE SECTION.
Figure 103. Test Configuration for AD9773 in Two-Port Mode with PLL Disabled, DAC Output Data Rate = Signal Generator Frequency,
DATACLK = Signal Generator Frequency/Interpolation Rate
LECROY
PULSE
GENERATOR
SIGNAL GENERATOR
TRIG
INP
ONEPORTCLK
CLK+/CLK–
INPUT CLOCK
AWG2021
OR
DG2020
DAC1, DB11–DB0
DAC2, DB11–DB0
AD9773
JUMPER CONFIGURATION FOR ONE-PORT MODE, PLL OFF
SOLDERED/IN UNSOLDERED/OUT
JP1 –
JP2 –
×
×
JP3 –
×
JP5 –
JP6 –
×
×
×
×
JP12 –
JP24 –
JP25 –
JP26 –
JP27 –
JP31 –
JP32 –
JP33 –
×
×
×
×
×
×
NOTES
1. TO USE PECL DRIVER (U8), SOLDER JP41 AND JP42 AND REMOVE TRANSFORMER T1.
Figure 104. Test Configuration for AD9773 in One-Port Mode with PLL Disabled, DAC Output Data Rate = Signal Generator Frequency,
ONEPORTCLK = Interleaved Input Data Rate = 2x Signal Generator Frequency/Interpolation Rate
Rev. D | Page 50 of 60
AD9773
0 6 C 0 R 3
0 6 C 0 R 3
G 2
G 3
N E B L
V P S 1
T
V O U
L O I P
L O I N
G 1 B
G 1 A
V P S 2
G 4 A
G 4 B
Q B B N I B B N
Q B B P I B B P
A D T L 1 - 1 2
0 6 C 0 C 3
A D T L 1 - 1 2
0 8 C 0 C 5
Figure 105. AD8345 Circuitry on AD9773 Evaluation Board
Rev. D | Page 51 of 60
AD9773
C C 0 6 0 3
R C 1 2 0 6
C C 0 6 0 3
R C 0 6 0 3
R C 0 6 0 3
C C 0 6 0 5
C C 0 6 0 3
C C 0 8 0 5
Figure 106. AD9773 Clock, Power Supplies, and Output Circuitry
Rev. D | Page 52 of 60
AD9773
Figure 107. AD9773 Evaluation Board Input (A Channel) and Clock Buffer Circuitry
Rev. D | Page 53 of 60
AD9773
Figure 108. AD9773 Evaluation Board Input (B Channel) and SPI Port Circuitry
Rev. D | Page 54 of 60
AD9773
Figure 109. AD9773 Evaluation Board Components, Top Side
Figure 110. AD9773 Evaluation Board Components, Bottom Side
Rev. D | Page 55 of 60
AD9773
Figure 111. AD9773 Evaluation Board Layout, Layer One (Top)
Figure 112. AD9773 Evaluation Board Layout, Layer Two (Ground Plane)
Rev. D | Page 56 of 60
AD9773
Figure 113. AD9773 Evaluation Board Layout, Layer Three (Power Plane)
Figure 114. AD9773 Evaluation Board Layout, Layer Four (Bottom)
Rev. D | Page 57 of 60
AD9773
OUTLINE DIMENSIONS
14.20
14.00 SQ
13.80
12.20
1.20
MAX
12.00 SQ
11.80
0.75
0.60
0.45
61
80
61
80
1
60
1
60
PIN 1
EXPOSED
PAD
6.00
BSC SQ
TOP VIEW
(PINS DOWN)
BOTTOM VIEW
(PINS UP)
0° MIN
1.05
1.00
0.95
0.20
0.09
7°
3.5°
20
20
41
41
40
40
21
21
VIEW A
0.15
0.05
0.50 BSC
0°
SEATING
PLANE
0.27
0.22
0.17
LEAD PITCH
0.08 MAX
COPLANARITY
VIEW A
ROTATED 90° CCW
COMPLIANT TO JEDEC STANDARDS MS-026-ADD-HD
`
Figure 115. 80-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
(SV-80-1)
Dimensions shown in millimeters
ORDERING GUIDE
Model
Temperature Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
Package Option
SV-80-1
SV-80-1
SV-80-1
SV-80-1
AD9773BSV
AD9773BSVRL
AD9773BSVZ1
AD9773BSVZRL1
AD9773-EB
80-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
80-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
80-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
80-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
Evaluation Board
1 Z = RoHS Compliant Part.
Rev. D | Page 58 of 60
AD9773
NOTES
Rev. D | Page 59 of 60
AD9773
NOTES
©2007 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D02857-0-10/07(D)
Rev. D | Page 60 of 60
相关型号:
![](http://pdffile.icpdf.com/pdf2/p00307/img/page/AD9774ASRL_1851564_files/AD9774ASRL_1851564_1.jpg)
![](http://pdffile.icpdf.com/pdf2/p00307/img/page/AD9774ASRL_1851564_files/AD9774ASRL_1851564_2.jpg)
AD9774ASRL
PARALLEL, WORD INPUT LOADING, 0.035 us SETTLING TIME, 14-BIT DAC, PQFP44, MQFP-44
ROCHESTER
![](http://pdffile.icpdf.com/pdf2/p00308/img/page/AD9775BSVZ_1855625_files/AD9775BSVZ_1855625_1.jpg)
![](http://pdffile.icpdf.com/pdf2/p00308/img/page/AD9775BSVZ_1855625_files/AD9775BSVZ_1855625_2.jpg)
AD9775BSV
PARALLEL, WORD INPUT LOADING, 0.011 us SETTLING TIME, 14-BIT DAC, PQFP80, PLASTIC, MO-026-ADD-HD, TQFP-80
ROCHESTER
![](http://pdffile.icpdf.com/pdf1/p00075/img/page/AD9775_396022_files/AD9775_396022_1.jpg)
![](http://pdffile.icpdf.com/pdf1/p00075/img/page/AD9775_396022_files/AD9775_396022_2.jpg)
AD9775BSVRL
IC PARALLEL, WORD INPUT LOADING, 0.011 us SETTLING TIME, 14-BIT DAC, PQFP80, PLASTIC, MO-026-ADD-HD, TQFP-80, Digital to Analog Converter
ADI
![](http://pdffile.icpdf.com/pdf2/p00308/img/page/AD9775BSVZ_1855625_files/AD9775BSVZ_1855625_1.jpg)
![](http://pdffile.icpdf.com/pdf2/p00308/img/page/AD9775BSVZ_1855625_files/AD9775BSVZ_1855625_2.jpg)
AD9775BSVZ
PARALLEL, WORD INPUT LOADING, 0.011 us SETTLING TIME, 14-BIT DAC, PQFP80, LEAD FREE, PLASTIC, MO-026-ADD-HD, TQFP-80
ROCHESTER
©2020 ICPDF网 联系我们和版权申明