16-Bit, 200 MSPS/500 MSPS TxDAC+® with
2×/4×/8× Interpolation and Signal Processing
AD9786
FEATURES
PRODUCT HIGHLIGHTS
16-bit resolution, 200 MSPS input data rate
IMD 90 dBc @10 MHz
Noise spectral density (NSD): −164 dBm/Hz @ 10 MHz
WCDMA ACLR = 80 dBc @ 40 MHz IF
DNL = ±0.3 LSB
INL = ±0.6 LSB
Selectable 2×/4×/8× interpolation filters
Selectable fDAC/2, fDAC/4, fDAC/8 modulation modes
Single- or dual-channel signal processing
Selectable image rejection Hilbert transform
Flexible calibration engine
Direct IF transmission features
Serial control interface
Versatile clock and data interface
3.3 V-compatible digital interface
On-chip 1.2 V reference
80-lead, thermally enhanced, TQFP_EP package
1.
16-bit, high speed, interpolating TxDAC+.
2.
2×/4×/8× user-selectable interpolating filter. The filter
eases data rate and output signal reconstruction filter
requirements.
3.
200 MSPS input data rate.
4.
Ultra high speed, 500 MSPS DAC conversion rate.
5.
Flexible clock with single-ended or differential input.
CMOS, 1 V p-p sine wave, and LVPECL capability.
6.
Complete CMOS DAC function. It operates from a 3.1 V
to 3.5 V single analog (AVDD) supply, 2.5 V digital supply,
and a 3.3 V digital (DRVDD) supply. The DAC full-scale
current can be reduced for lower power operation, and
a sleep mode is provided for low power idle periods.
7.
On-chip voltage reference. The AD9786 includes a
1.20 V temperature-compensated band gap voltage
reference.
8.
Multichip synchronization. Multiple AD9786 DACs can
be synchronized to a single master AD9786 to ease timing
design requirements and optimize image reject transmit
performance.
APPLICATIONS
Base stations: multicarrier WCDMA, GSM/EDGE, TD-SCDMA,
IS136, TETRA
Instrumentation
RF signal generators, arbitrary waveform generators
HDTV transmitters
Broadband wireless systems
Digital radio links
Satellite systems
2×
2×
2×
0
fDAC/2
fDAC/4
fDAC/8
0
90
CLK–
DATA PORT
SYNCHRONIZER
16-BIT DAC
REFIO
IOUTA
IOUTB
90
Q
2×
2×
CSB
SCLK
×1
LATCH
SDO
HILBERT
RESET
2×
CLOCK DISTRIBUTION AND CONTROL
03152-001
CLK+
ZERO
STUFF
FSADJ
SDIO
0
DATACLK
Δt
SPI
P2B[15:0]
DATA
ASSEMBLER
P1B[15:0]
90
Re()/Im()
I
REFERENCE
CIRCUITS
LATCH
CALIBRATION
FUNCTIONAL BLOCK DIAGRAM
Figure 1.
Rev. B
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
rights of third parties 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 registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
www.analog.com
Fax: 781.461.3113
© 2005 Analog Devices, Inc. All rights reserved.
AD9786
TABLE OF CONTENTS
Features .............................................................................................. 1
General Operation of the Serial Interface............................... 20
Applications....................................................................................... 1
Serial Interface Port Pin Descriptions ..................................... 20
Product Highlights ........................................................................... 1
MSB/LSB Transfers .................................................................... 21
Functional Block Diagram .............................................................. 1
Notes on Serial Port Operation ................................................ 21
Revision History ............................................................................... 3
Mode Control (via Serial Port) ..................................................... 22
General Description ......................................................................... 4
Digital Filter Specifications ........................................................... 26
Specifications..................................................................................... 5
Digital Interpolation Filter Coefficients.................................. 26
DC Specifications ......................................................................... 5
Clock/Data Timing .................................................................... 27
Dynamic Specifications ............................................................... 6
Real and Complex Signals......................................................... 32
Digital Specifications ................................................................... 7
Modulation Modes..................................................................... 33
Absolute Maximum Ratings............................................................ 8
Power Dissipation....................................................................... 38
Thermal Resistance ...................................................................... 8
Hilbert Transform Implementation......................................... 40
ESD Caution.................................................................................. 8
Operating the AD9786 Rev. F Evaluation Board ....................... 44
Pin Configuration and Function Descriptions............................. 9
Power Supplies ............................................................................ 44
Clock .............................................................................................. 9
PECL Clock Driver .................................................................... 44
Analog.......................................................................................... 10
Data Inputs.................................................................................. 45
Data .............................................................................................. 10
Serial Port .................................................................................... 45
Serial Interface ............................................................................ 11
Analog Output ............................................................................ 45
Terminology .................................................................................... 12
Outline Dimensions ....................................................................... 55
Typical Performance Characteristics ........................................... 14
Ordering Guide .......................................................................... 55
Serial Control Interface.................................................................. 20
Rev. B | Page 2 of 56
AD9786
REVISION HISTORY
10/05—Rev. A to Rev. B
Updated Format.................................................................. Universal
Changes to Figure 1...........................................................................1
Changes to Table 2 ............................................................................6
Changes to Table 3 ............................................................................7
Changes to External Sync Mode Section .....................................31
Updated Outline Dimensions........................................................58
Changes to Ordering Guide...........................................................58
2/05—Rev. 0 to Rev. A
Changed DRVDD Supply Range...................................... Universal
Changes to DC Specifications .........................................................4
Changes to Dynamic Specifications ...............................................5
Changes to Digital Specifications....................................................6
Changes to Absolute Maximum Ratings........................................7
Change to Figure 2 ............................................................................8
Replaced Figure 13 ..........................................................................14
Replaced Figure 14 ..........................................................................14
Replaced Figure 16 ..........................................................................15
Replaced Figure 21 ..........................................................................16
Replaced Figure 22 ..........................................................................16
Replaced Figure 26..........................................................................16
Replaced Figure 27..........................................................................17
Changes to Table 15 ........................................................................22
Change to Figure 44........................................................................26
Replaced Figure 45..........................................................................26
Change to Figure 47........................................................................27
Change to Figure 48........................................................................27
Change to Figure 51........................................................................29
Change to Figure 52........................................................................29
Change to Figure 53........................................................................30
Change to DATAADJUST Synchronization Section..................31
Changes to Power Dissipation Section.........................................40
Changes to Table 37 ........................................................................42
Changes to Data Inputs Section ....................................................46
Change to Figure 88........................................................................49
Replaced Figure 95..........................................................................55
Updated Outline Dimensions........................................................60
Changes to Ordering Guide...........................................................60
7/04—Revision 0: Initial Version
Rev. B | Page 3 of 56
AD9786
GENERAL DESCRIPTION
The AD9786 is a 16-bit, high speed, CMOS DAC with
2×/4×/8× interpolation and signal processing features tuned
for communications applications. It offers state-of-the-art
distortion and noise performance. The AD9786 was developed
to meet the demanding performance requirements of multicarrier
and third-generation base stations. The selectable interpolation
filters simplify interfacing to a variety of input data rates while
also taking advantage of oversampling performance gains. The
modulation modes allow convenient bandwidth placement and
selectable sideband suppression.
The flexible clock interface accepts a variety of input types such
as 1 V p-p sine wave, CMOS, and LVPECL in single-ended or
differential mode. Internal dividers generate the required data
rate interface clocks.
The AD9786 provides a differential current output, supporting
single-ended or differential applications; it provides a nominal
full-scale current from 10 mA to 20 mA. The AD9786 is
manufactured on an advanced, low cost, 0.25 μm CMOS process.
Rev. B | Page 4 of 56
AD9786
SPECIFICATIONS
DC SPECIFICATIONS
TMIN to TMAX; AVDD1, AVDD2, DRVDD = 3.3 V; ACVDD, ADVDD, CLKVDD, DVDD = 2.5 V; IOUTFS = 20 mA, unless otherwise noted.
Table 1.
Parameter
RESOLUTION
DC Accuracy1
Integral Nonlinearity
Differential Nonlinearity
ANALOG OUTPUT
Offset Error
Gain Error (with Internal Reference)
Full-Scale Output Current2
Output Compliance Range
Output Resistance
REFERENCE OUTPUT
Reference Voltage
Reference Output Current3
REFERENCE INPUT
Input Compliance Range
Reference Input Resistance (External Reference Mode)
Small Signal Bandwith
TEMPERATURE COEFFICIENTS
Unipolar Offset Drift
Gain Drift (with Internal Reference)
Reference Voltage Drift
POWER SUPPLY
AVDD1, AVDD2
Voltage Range
Analog Supply Current (IAVDD1 + IAVDD2)
IAVDD1 + IAVDD2 in Sleep Mode
ACVDD, ADVDD
Voltage Range
Analog Supply Current (IACVDD + IADVDD)
CLKVDD
Voltage Range
Clock Supply Current (ICLKVDD)
DVDD
Voltage Range
Digital Supply Current (IDVDD)
DRVDD
Voltage Range
Digital Supply Current (IDRVDD)
Nominal Power Dissipation4
OPERATING RANGE
Min
Typ
16
Max
±0.6
±0.3
±0.015
±1.5
10
–1.0
LSB
LSB
±0.0175
20
+1.0
10
1.15
1.23
1
V
μA
1.25
10
200
V
MΩ
kHz
0
±4
±30
ppm of FSR/°C
ppm of FSR/°C
ppm/°C
3.1
3.3
50
18
3.5
V
mA
mA
2.35
2.5
2.5
2.65
V
mA
2.35
2.5
12
2.65
V
mA
2.35
2.5
52.5
2.65
V
mA
3.1
3.3
5.3
1.25
3.5
V
μA
W
°C
–40
Measured at IOUTA driving a virtual ground.
Nominal full-scale current, IOUTFS, is 32× the IREF current.
3
Use an external amplifier to drive any external load.
4
Measured under the following conditions: fDATA = 125 MSPS, fDAC = 500 MSPS, 4× interpolation, fDAC/4 modulation, Hilbert off.
2
Rev. B | Page 5 of 56
% of FSR
% of FSR
mA
V
MΩ
1.30
0.1
1
Unit
Bits
+85
AD9786
DYNAMIC SPECIFICATIONS
TMIN to TMAX; AVDD1, AVDD2, DRVDD = 3.3 V; ACVDD, ADVDD, CLKVDD, DVDD = 2.5 V; IOUTFS = 20 mA; differential transformer
coupled output; 50 Ω doubly terminated, unless otherwise noted.
Table 2.
Parameter
DYNAMIC PERFORMANCE
Minimum DAC Output Update Rate
Maximum DAC Output Update Rate (fDAC)
AC LINEARITY/BASEBAND MODE
Spurious-Free Dynamic Range (SFDR) to Nyquist (fOUT = 0 dBFS)
fDATA = 100 MSPS; fOUT = 5 MHz, 4×, 2× Interpolation
fDATA = 200 MSPS; fOUT = 10 MHz
fDATA = 200 MSPS; fOUT = 25 MHz
fDATA = 200 MSPS; fOUT = 50 MHz
Two-Tone Intermodulation (IMD) to Nyquist (fOUT1 = fOUT2 = –6 dBFS)
fDATA = 200 MSPS; fOUT1 = 5 MHz; fOUT2 = 6 MHz
fDATA = 200 MSPS; fOUT1 = 15 MHz; fOUT2 = 16 MHz
fDATA = 200 MSPS; fOUT1 = 25 MHz; fOUT2 = 26 MHz
fDATA = 200 MSPS; fOUT1 = 45 MHz; fOUT2 = 46 MHz
fDATA = 200 MSPS; fOUT1 = 65 MHz; fOUT2 = 66 MHz
fDATA = 200 MSPS; fOUT1 = 85 MHz; fOUT2 = 86 MHz
Noise Power Spectral Density (NPSD)
fDATA = 156 MSPS; fOUT = 10 MHz; 0 dBFS, 8 Tones, Separation = 500 kHz
fDATA = 156 MSPS; fOUT = 50 MHz; 0 dBFS, 8 Tones, Separation = 500 kHz
Adjacent Channel Power Ratio (ACLR)
WCDMA ACLR with 3.84 MHz BW, Single Carrier
IF = 21 MHz, fDATA = 122.88 MSPS, 4× Interpolation
IF = 224.76 MHz, fDATA = 122.88 MSPS, 4× Interpolation, High-Pass Interpolation Filter Mode
Rev. B | Page 6 of 56
Min
Typ
500
Max
Unit
20
MHz
MSPS
93
85
78
78
dBc
dBc
dBc
dBc
85
85
84
80
78
75
dBc
dBc
dBc
dBc
dBc
dBc
−164
−161
dBm/Hz
dBm/Hz
80
72
dB
dB
AD9786
DIGITAL SPECIFICATIONS
TMIN to TMAX; AVDD1, AVDD2, DRVDD = 3.3 V; ACVDD, ADVDD, CLKVDD, DVDD = 2.5 V; IOUTFS = 20 mA, unless otherwise noted.
Table 3.
Parameter
DIGITAL INPUTS
Logic 1 Voltage
Logic 0 Voltage
Logic 1 Current
Logic 0 Current
Input Capacitance
CLOCK INPUTS1
Input Voltage Range
Common-Mode Voltage
Differential Voltage
Latch Pulse Width (tLPW)
Data Setup Time to DACCLK Out in Master Mode (tS)
Data Hold Time to DACCLK Out in Master Mode (tH)
1
Min
Typ
Max
Unit
0
0.9
+10
+10
V
V
μA
μA
pF
1.6
–10
–10
5
0
0.75
0.5
5
−0.5
2.9
See the Clock/Data Timing section for setup and hold times in various timing modes.
Rev. B | Page 7 of 56
1.5
1.5
2.65
2.25
V
V
V
ns
ns
ns
AD9786
ABSOLUTE MAXIMUM RATINGS
Table 4.
Parameter
AVDD1, AVDD2,
DRVDD
ACVDD, ADVDD,
CLKVDD, DVDD
AGND1, AGND2,
ACGND, ADGND,
CLKGND, DGND
REFIO, FSADJ
IOUTA, IOUTB
P1B15 to P1B0,
P2B15 to P2B0, RESET
DATACLK
CLK+, CLK−
CSB, SCLK,
SDIO, SDO
Junction
Temperature Range
Storage
Temperature
Lead Temperature
(10 sec)
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
section of this specification is not implied. Exposure to absolute
maximum rating conditions for extended periods may affect
device reliability.
With Respect to
AGND1, AGND2,
ACGND, ADGND,
CLKGND, DGND
AGND1, AGND2,
ACGND, ADGND,
CLKGND, DGND
AGND1, AGND2,
ACGND, ADGND,
CLKGND, DGND
AGND1
AGND1
DGND
Rating
−0.3 V to +3.6 V
−0.3 to AVDD1 + 0.3
−1.0 to AVDD1 +0.3
−0.3 to DRVDD + 0.3
DGND
CLKGND
DGND
−0.3 to DRVDD + 0.3
−0.3 to CLKVDD + 0.3
−0.3 to DRVDD + 0.3
−0.3 V to +2.8 V
−0.3 V to +0.3 V
THERMAL RESISTANCE
θJA is specified for the worst-case conditions, that is, a device
soldered in a circuit board for surface-mount packages.
Table 5. Thermal Resistance
Package Type1
80-lead TQFP_EP (Thermally Enhanced)
`
1
With thermal pad soldered to PCB.
−65°C to +125°C
150°C
300°C
ESD CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate
on the human body and test equipment and can discharge without detection. Although this product
features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to
high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid
performance degradation or loss of functionality.
Rev. B | Page 8 of 56
θJA
23.5
Unit
°C/W
AD9786
DNC
ADVDD
ADGND
ACVDD
ACGND
AVDD2
AVDD1
AGND2
AGND1
IOUTA
IOUTB
AGND1
AVDD1
AGND2
AVDD2
ACGND
ADGND
ACVDD
ADVDD
DNC
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
CLKVDD 1
DNC 2
60 FSADJ
PIN 1
IDENTIFIER
59 REFIO
CLKVDD 3
58 RESET
CLKGND 4
57 CSB
CLK+ 5
56 SCLK
CLK– 6
55 SDIO
54 SDO
CLKGND 7
53 DGND
DGND 8
DVDD 9
52 DVDD
AD9786
P1B15 10
51 P2B0
TOP VIEW
(Not to Scale)
P1B14 11
50 P2B1
P1B13 12
49 P2B2
P1B12 13
48 P2B3
P1B11 14
47 P2B4
P1B10 15
46 P2B5
DGND 16
45 DGND
DVDD 17
44 DVDD
P1B9 18
P1B8 19
43 P2B6
42 P2B7
P1B7 20
41 P2B8
03152-002
P2B10
P2B9
P2B11
P2B12
DGND
DVDD
IQSEL/P2B15
ONEPORTCLOCK/P2B14
P2B13
DRVDD
DATACLK
P1B0
P1B2
P1B1
DVDD
DGND
P1B3
P1B4
P1B5
DNC = DO NOT CONNECT
P1B6
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Figure 2. Pin Configuration
CLOCK
Table 6. Clock Pin Function Descriptions
Pin
No.
5, 6
2
31
Mnemonic
CLK+, CLK–
DNC
DATACLK
1, 3
4, 7
CLKVDD
CLKGND
Direction
I
I/O
Description
Differential Clock Input.
Do Not Connect.
DCLKEXT
0x02[3]
Mode
0
Pin configured for input of channel data rate or synchronizer clock. Internal clock
synchronizer can be turned on or off with DCLKCRC (0x02[2]).
1
Pin configured for output of channel data rate or synchronizer clock.
Clock Domain 2.5 V.
Clock Domain 0 V.
Rev. B | Page 9 of 56
AD9786
ANALOG
Table 7. Analog Pin Function Descriptions
Pin No.
59
60
70, 71
61
62, 79
63, 78
64, 77
65, 76
66, 75
67, 74
68, 73
69, 72
80
Mnemonic
REFIO
FSADJ
IOUTB, IOUTA
DNC
ADVDD
ADGND
ACVDD
ACGND
AVDD2
AGND2
AVDD1
AGND1
DNC
Direction
A
A
A
Description
Reference.
Full-Scale Adjust.
Differential DAC Output Currents.
Do Not Connect.
Analog Domain Digital Content 2.5 V.
Analog Domain Digital Content 0 V.
Analog Domain Clock Content 2.5 V.
Analog Domain Clock Content 0 V.
Analog Domain Clock Switching 3.3 V.
Analog Domain Switching 0 V.
Analog Domain Quiet 3.3 V.
Analog Domain Quiet 0 V.
Do Not Connect.
DATA
Table 8. Data Pin Function Descriptions
Pin No.
10 to 15, 18 to
24, 27 to 29
Mnemonic
P1B15 to P1B0
Direction
I
32
IQSEL/P2B15
I
33
ONEPORTCLOCK/P2B14
I/O
34, 37 to 43,
46 to 51
30
9, 17, 26,
36, 44, 52
8, 16, 25,
35, 45, 53
P2B13 to P2B0
I
Description
Input Data Port 1.
ONEPORT
0x02[6]
Mode
0
Latched data routed for I channel processing.
1
Latched data demultiplexed by IQSEL and routed for
interleaved I/Q processing.
ONEPORT
IQPOL
IQSEL/
0x02[6]
0x02[1] P2B15 Mode (IQPOL = 0)
0
X
X
Latched data routed to Q channel Bit 15
(MSB) processing.
1
0
0
Latched data on Data Port 1 routed to Q
channel processing.
1
0
1
Latched data on Data Port 1 routed to I
channel processing.
1
1
0
Latched data on Data Port 1 routed to I
channel processing.
1
1
1
Latched data on Data Port 1 routed to Q
channel processing.
ONEPORT
0x02[6]
0
Latched data routed for Q channel Bit 14 processing.
1
Pin configured for output of clock at twice the channel
data route.
Input Data Port 2, Bit 13 to Bit 0.
DRVDD
DVDD
Digital Output Pin Supply, 3.3 V.
Digital Domain, 2.5 V.
DGND
Digital Domain, 0 V.
Rev. B | Page 10 of 56
AD9786
SERIAL INTERFACE
Table 9. Serial Interface Pin Function Descriptions
Pin No.
54
Mnemonic
SDO
Direction
O
55
SDIO
I/O
56
57
58
SCLK
CSB
RESET
I
I
I
Description
SDIODIR
CSB 0x00[7]
Mode
1
X
High impedance.
0
0
Serial data output.
0
1
High impedance.
SDIODIR
CSB 0x00[7]
Mode
1
X
High impedance.
0
0
Serial data output.
0
1
Serial data input/output depending on Bit 7 of the serial instruction byte.
Serial Interface Clock.
Serial Interface Chip Select.
Resets entire chip to default state.
Rev. B | Page 11 of 56
AD9786
TERMINOLOGY
Linearity Error (Integral Nonlinearity or 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 zero to full scale.
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.
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.
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-sec.
Monotonicity
A D/A converter is monotonic if the output either increases or
remains constant as the digital input increases.
Spurious-Free Dynamic Range (SFDR)
The difference between the rms amplitude of the output signal
and the amplitude of the peak spurious signal over the specified
bandwidth. The units are often in dBc (dB with respect to the
carrier).
Offset Error
The deviation of the output current from the ideal of zero is
called offset error. For IOUTA, 0 mA output is expected when the
inputs are all 0s. For IOUTB, 0 mA output is expected when all
inputs are set to 1.
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.
Output Compliance Range
The range of allowable voltage at the output of a current-output
DAC. Operation beyond the maximum compliance limits can
cause either output stage saturation or breakdown, resulting in
nonlinear performance.
Temperature Drift
Temperature drift 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 degree Celsius. For reference drift, the drift is
reported in ppm per degree Celsius.
Power Supply Rejection
The maximum change in the full-scale output as the supplies
are varied from minimum to maximum specified voltages.
Total Harmonic Distortion (THD)
THD is the ratio of the rms sum of the first six harmonic
components to the rms value of the measured fundamental. It is
expressed as a percentage or in decibels.
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.
Interpolation Filter
If the digital inputs to the DAC are sampled at a multiple rate of
fDATA (interpolation rate), a digital filter can be constructed that has
a sharp transition band near fDATA/2. Images that would typically
appear around fDAC (output data rate) can be greatly suppressed.
Pass Band
Frequency band in which any input applied therein passes
unattenuated to the DAC output.
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.
Rev. B | Page 12 of 56
AD9786
Group Delay
Number of input clocks between an impulse applied at the
device input and peak DAC output current. A half-band FIR
filter has constant group delay over its entire frequency range
Impulse Response
Response of the device to an impulse applied to the input.
Adjacent Channel Leakage Ratio (ACLR)
A ratio in dBc between the measured power within a channel
relative to its adjacent channel.
Complex Modulation
The process of passing the real and imaginary components of a
signal through a complex modulator (transfer function = ejwt =
coswt + jsinwt) and realizing real and imaginary components
on the modulator output.
Hilbert Transform
A function with unity gain over all frequencies, but with a phase
shift of 90° for negative frequencies and a phase shift of –90° for
positive frequencies. Although this function cannot be implemented ideally, it can be approximated with a short FIR filter
with enough accuracy to be very useful in single sideband radio
architectures.
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.
Rev. B | Page 13 of 56
AD9786
TYPICAL PERFORMANCE CHARACTERISTICS
TMIN to TMAX; AVDD1, AVDD2, DRVDD = 3.3 V; ACVDD, ADVDD, CLKVDD, DVDD = 2.5 V; IOUTFS = 20 mA; differential transformer
coupled output; 50 Ω doubly terminated, unless otherwise noted.
120
120
100
100
–6dBFS
–6dBFS
0dBFS
40
40
20
20
0
0
10
20
30
40
50
FREQUENCY (MHz)
60
70
80
0dBFS
60
0
0
10
20
30
40
50
FREQUENCY (MHz)
60
70
80
Figure 3. SFDR vs. Frequency, fDATA = 200 MSPS, 1× Interpolation
Figure 6. SFDR vs. Frequency, fDATA = 200 MSPS, 2× Interpolation
120
120
–3dBFS
100
03152-006
60
SFDR (dBc)
80
03152-003
100
–3dBFS
–6dBFS
80
SFDR (dBc)
80
–6dBFS
0dBFS
60
60
40
40
20
20
0
5
10
15
20
25
30
FREQUENCY (MHz)
35
40
45
0
03152-004
0
0dBFS
0
10
20
30
40
FREQUENCY (MHz)
50
60
Figure 4. SFDR vs. Frequency, fDATA = 100 MSPS, 4× Interpolation
Figure 7. SFDR vs. Frequency, fDATA = 125 MSPS, 4× Interpolation
120
120
–6dBFS
03152-007
SFDR (dBc)
80
SFDR (dBc)
–3dBFS
–3dBFS
–6dBFS
100
100
–3dBFS
80
60
0dBFS
60
–3dBFS
40
40
20
20
0
0
5
10
15
FREQUENCY (MHz)
20
25
Figure 5. SFDR vs. Frequency, fDATA = 50 MSPS, 8× Interpolation
0
0
5
10
15
20
FREQUENCY (MHz)
25
30
Figure 8. SFDR vs. Frequency, fDATA = 62.5 MSPS, 8× Interpolation
Rev. B | Page 14 of 56
03152-008
SFDR (dBc)
0dBFS
03152-005
SFDR (dBc)
80
AD9786
90
85
85
80
OUT OF BAND SFDR (dBc)
90
–3dBFS
–6dBFS
70
0dBFS
65
60
0dBFS
–3dBFS
75
70
–6dBFS
65
60
55
55
0
10
20
30
40
50
FOUT (MHz)
60
70
80
50
03152-009
50
0
10
20
30
ANALOG OUTPUT FREQUENCY (MHz)
03152-012
SFDR (dBc)
75
80
40
Figure 12. Out-of-Band SFDR, fDATA = 100 MSPS, 4× Interpolation
Figure 9. Out-of-Band SFDR, fDATA = 200 MSPS, 2× Interpolation
100
90
95
0dBFS
85
90
OUT OF BAND SFDR (dBc)
85
–6dBFS
75
70
65
–3dBFS
60
55
–6dBFS
75
–3dBFS
70
65
60
0
20
40
60
80 100 120 140 160 180 200 220 240 260
FOUT (MHz)
03152-010
55
50
50
0
5
10
15
20
ANALOG OUTPUT FREQUENCY (MHz)
25
Figure 10. Out-of-Band SFDR, fDATA = 125 MSPS, 4× Interpolation
Figure 13. Out-of-Band SFDR, fDATA = 50 MSPS, 8× Interpolation
100
100
95
–3dBFS
95
–3dBFS
03152-013
IMD (dBc)
80
0dBFS
80
90
90
85
85
–6dBFS
80
IMD (dBc)
75
70
0dBFS
–6dBFS
75
70
65
65
60
60
55
55
50
0
20
40
60
80 100 120 140 160 180 200 220 240 260
FOUT (MHz)
Figure 11. Out-of-Band SFDR, fDATA = 62.5 MSPS, 8× Interpolation
50
0
20
40
FOUT (MHz)
60
80
03152-014
0dBFS
03152-011
IMD (dBc)
80
Figure 14. Third-Order IMD vs. Frequency, fDATA = 160 MSPS, 1× Interpolation
Rev. B | Page 15 of 56
AD9786
100
100
95
95
–3dBFS
–3dBFS
90
80
80
IMD (dBc)
85
75
0dBFS
70
65
60
55
55
20
40
60
80
100
FOUT (MHz)
120
140
160
Figure 15. Third-Order IMD vs. Frequency, fDATA = 160 MSPS, 2× Interpolation
0dBFS
70
60
0
50
0
20
40
60
FOUT (MHz)
80
100
Figure 18. Third-Order IMD vs. Frequency, fDATA = 200 MSPS,1x Interpolation
100
100
–3dBFS
95
–3dBFS
95
90
90
–6dBFS
85
–6dBFS
85
80
IMD (dBc)
80
75
70
70
65
60
60
55
55
50
0
20
40
60
80
100 120
FOUT (MHz)
140
160
180
200
Figure 16. Third-Order IMD vs. Frequency, fDATA = 200 MSPS, 2× Interpolation
50
0
20
40
60
80
100 120
FOUT (MHz)
140
160
180
200
03152-019
65
75
0dBFS
0dBFS
03152-016
IMD (dBc)
75
65
50
–6dBFS
03152-018
–6dBFS
85
03152-015
IMD (dBc)
90
Figure 19. Third-Order IMD vs. Frequency, fDATA = 100 MSPS, 4× Interpolation
100
100
95
95
–3dBFS
0dBFS
90
90
85
85
–6dBFS
–6dBFS
80
IMD (dBc)
IMD (dBc)
80
75
70
75
70
0dBFS
65
–3dBFS
60
60
55
55
0
20
40
60
80 100 120 140 160 180 200 220 240 260
FOUT (MHz)
50
03152-017
50
Figure 17. Third-Order IMD vs. Frequency, fDATA = 125 MSPS, 4× Interpolation
0
20
40
60
80
100 120
FOUT (MHz)
140
160
180
200
03152-020
65
Figure 20. Third-Order IMD vs. Frequency, fDATA = 50 MSPS, 8× Interpolation
Rev. B | Page 16 of 56
AD9786
100
0.3
95
–3dBFS
0.2
90
0.1
85
–6dBFS
DNL (LSBs)
IMD (dBc)
80
75
70
0dBFS
65
0
–0.1
–0.2
60
–0.3
0
20
40
60
80 100 120 140 160 180 200 220 240 260
FOUT (MHz)
–0.4
03152-021
50
0
8192
16384 24576 32768 40960 49152 57344 65536
CODE
Figure 24. Typical DNL
Figure 21. Third-Order IMD vs. Frequency, fDATA = 62.5 MSPS, 8× Interpolation
1.25
–140
1.00
0.25
0
–0.50
0
8192
16384 24576 32768 40960 49152 57344 65536
CODE
03152-022
–0.25
FDATA = 78MSPS, 1× INTERPOLATION
–155
–160
–165
FDATA = 78MSPS, 2× INTERPOLATION
–170
–175
–180
0
10
20
30
40
50
60
ANALOG OUTPUT FREQUENCY (MHz)
70
80
Figure 25. Noise Spectral Density vs. Analog
Input Frequency, fDATA = 78 MSPS
Figure 22. Typical INL
–150
–140
–152
NOISE SPECTRAL DENSITY (dBm/Hz)
–145
–150
FDATA = 156MSPS, 1× INTERPOLATION
–155
–160
–165
FDATA = 156MSPS, 2× INTERPOLATION
–170
–180
0
20
40
60
80
100
120
ANALOG OUTPUT FREQUENCY (MHz)
140
160
03152-023
–175
–154
–156
–158
AIN = –3DBFS
–160 AIN = 0DBFS
–162
–164
–166
AIN = –6DBFS
–168
–170
0
10
20
30
40
50
60
ANALOG OUTPUT FREQUENCY (MHz)
70
80
Figure 26. Noise Spectral Density vs. Analog Input Frequency,
fDATA = 78 MSPS, 2x Interpolation
Figure 23. Noise Spectral Density vs. Analog
Input Frequency, fDATA = 156 MSPS
Rev. B | Page 17 of 56
03152-026
INL (LSBs)
0.50
–150
03152-025
NOISE SPECTRAL DENSITY (dBm/Hz)
–145
0.75
NOISE SPECTRAL DENSITY (dBm/Hz)
03152-024
55
AD9786
–150
–154
10
AIN = –3dBFS
–156
Ref Lv1
10 dBm
Marker 1 [T1]
RBW 10 kHz RF Att 20 dB
–87.73 dBm VBW 10 kHz
9.71442886 MHz SWT
5s
Unit
dBm
A
0
–10
–158
–20
AIN = 0dBFS
–160
1MA
–30 1AVG
–40
–162
–50
AIN = –6dBFS
–164
–60
–70
–166
–80
–168
1
–90
–170
0
20
40
60
80
100
120
ANALOG OUTPUT FREQUENCY (MHz)
140
160
–100
03152-027
NOISE SPECTRAL DENSITY (dBm/Hz)
–152
–110
START 100 kHz
19.9 MHz/
STOP 200 MHz
Figure 30. Two Tones Around 23 MHz, fDATA = 200 MSPS,
2× Interpolation, Low-Pass Digital Filter Mode
Figure 27. Noise Spectral Density vs. Analog Input Frequency,
fDATA = 156 MSPS, 2x Interpolation
–60
–65
10
Marker 1 [T1]
–87.95 dBm
11.71743487 MHz
RBW 10 kHz RF Att 20 dB
VBW 10 kHz
SWT
5s
Unit
dBm
A
0
–70
ACLR (dBc)
Ref Lv1
10 dBm
0dBFS
–10
–3dBFS
–20
–75
1MA
–30 1AVG
–40
–80
–50
–6dBFS
–60
–70
–85
–80
1
0
25
50
75
FOUT (MHz)
100
125
150
–100
–110
START 100 kHz
19.9 MHz/
STOP 200 MHz
03152-031
–90
03152-028
–90
Figure 31. Two Tones Around 177 MHz, fDATA = 200 MSPS,
2× Interpolation, High-Pass Digital Filter Mode
Figure 28. ACLR for First Adjacent Band vs. Frequency,
fDATA = 61.44 MSPS, 4× Interpolation
REF –29.82dBm
*AVG
Log
10dB/
*ATTEN 6dB
–60
–65
–6dBFS
AVERAGE
103
0dBFS
–75
–3dBFS
PAVG
22
W1 S2
–85
CENTER 51.44MHz
*RES BW 30kHz
VBW 300kHz
RMS RESULTS FREQ OFFSET REF BW
–90
0
20
40
60
80
100 120
FOUT (MHz)
140
160
180
200
CARRIER POWER 5.000MHz
–17.41dBm/
10.000MHz
3.84MHz
15.000MHz
20.000MHz
3.840MHz
3.840MHz
3.840MHz
3.840MHz
SPAN 43.84MHz
SWEEP 142.2ms (601 pts)
LOWER
dBc
dBm
0.15
–17.26
–74.24 –91.65
–75.73 –93.14
–75.67 –93.08
UPPER
dBc
dBm
–74.63 –92.05
–75.67 –93.08
–76.38 –93.79
–75.75 –93.17
Figure 32. ACLR for Two WCDMA Carriers @ 51.44 MHz,
fDATA = 61.44 MSPS, 4× Interpolation
Figure 29. ACLR for First Adjacent Band vs. Frequency,
fDATA = 76.8 MSPS, 4× Interpolation
Rev. B | Page 18 of 56
03152-032
–80
03152-029
ACLR (dBc)
–70
AD9786
REF –22.76dBm
*AVG
Log
10dB/
REF –33.3dBm
*AVG
Log
10dB/
*ATTEN 8dB
*ATTEN 6dB
AC-COUPLED
AC-COUPLED
AVERAGE
104
AVERAGE
22
PAVG
104
W1 S2
CENTER 46.40MHz
*RES BW 30kHz
VBW 300kHz
RMS RESULTS FREQ OFFSET REF BW
CARRIER POWER 5.000MHz
–10.38dBm/
10.000MHz
3.84 MHz
15.000MHz
3.840MHz
3.840MHz
3.840MHz
SPAN 33.84MHz
SWEEP 109.8ms (601 pts)
LOWER
dBc
dBm
–79.00 –89.38
–80.78 –91.16
–79.71 –90.09
UPPER
dBc
dBm
–79.63 –90.01
–81.77 –92.15
–81.45 –91.83
CARRIER POWER 5.000MHz
–20.32dBm/
10.000MHz
3.84MHz
15.000MHz
20.000MHz
25.000MHz
*ATTEN 6dB
AVERAGE
22
PAVG
22
W1 S2
VBW 300kHz
RMS RESULTS FREQ OFFSET REF BW
CARRIER POWER 5.000MHz
–15.30dBm/
10.000MHz
3.84MHz
15.000MHz
3.840MHz
3.840MHz
3.840MHz
SPAN 33.84MHz
SWEEP 109.8ms (601 pts)
LOWER
dBc
dBm
–72.33 –87.64
–72.41 –87.71
–72.67 –87.97
UPPER
dBc
dBm
–72.13 –87.43
–73.02 –88.32
–73.50 –88.88
03152-034
CENTER 142.88MHz
*RES BW 30kHz
3.840MHz
3.840MHz
3.840MHz
3.840MHz
3.840MHz
SPAN 53.84MHz
SWEEP 174.6ms (601 pts)
LOWER
dBc
dBm
0.22
–20.11
–0.60 –20.92
–72.68 –93.00
–72.74 –93.06
–73.05 –93.37
UPPER
dBc
dBm
–0.16
–20.48
–72.05 –92.37
–72.85 –93.18
–72.55 –92.88
–72.02 –92.35
Figure 35. ACLR for Four WCDMA Carriers Near 50 MHz,
fDATA = 61.44 MSPS, 4× Interpolation
Figure 33. ACLR for Single WCDMA Carrier @ 20 MHz,
fDATA = 61.44 MSPS, 4× Interpolation
REF –28.2dBm
*AVG
Log
10dB/
VBW 300kHz
RMS RESULTS FREQ OFFSET REF BW
03152-033
CENTER 20.00MHz
*RES BW 30kHz
Figure 34. ACLR for Single WCDMA Carrier @ 142.88 MHz,
fDATA = 61.44 MSPS, 4× Interpolation
Rev. B | Page 19 of 56
03152-035
PAVG
22
W1 S2
AD9786
SERIAL CONTROL INTERFACE
Instruction Byte
SDIO (PIN 55)
SCLK (PIN 56)
AD9786 SPI
PORT INTERFACE
CSB (PIN 57)
03152-036
SDO (PIN 54)
Figure 36. AD9786 SPI Port Interface
The AD9786 serial port is a flexible, synchronous serial communications port, allowing 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 AD9786. Singleor multiple-byte transfers are supported, as well as MSB-first or
LSB-first transfer formats. The AD9786 serial interface port can
be configured as a single pin I/O (SDIO), or as two unidirectional
pins for input/output (SDIO/SDO).
GENERAL OPERATION OF THE SERIAL INTERFACE
There are two phases to a communication cycle with the AD9786.
Phase 1 is the instruction cycle, which is the writing of an
instruction byte into the AD9786, coincident with the first eight
SCLK rising edges. The instruction byte provides the AD9786
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 upcoming data transfer is a
read or a 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 AD9786.
A logic high on the 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 midst 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 AD9786
and the system controller. Phase 2 of the communication cycle
is a transfer of 1, 2, 3, or 4 data bytes, as determined by the instruction byte. Using one multibyte transfer is the preferred method.
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.
R/W, Bit 7 of the instruction byte, determines whether a read or
a write data transfer occurs after the instruction byte write.
Logic high indicates a read operation; Logic 0 indicates a write
operation. N1 and N0, Bit 6 and Bit 5 of the instruction byte,
determine the number of bytes to be transferred during the data
transfer cycle (see Table 10).
Table 10. Bytes Transferred During Data Transfer Cycle
N1
0
0
1
1
N2
0
1
0
1
Description
Transfer 1 byte
Transfer 2 bytes
Transfer 3 bytes
Transfer 4 bytes
The bit decodes are shown as follows:
MSB
I7
R/W
I6
N1
I5
N0
I4
A4
I3
A3
I2
A2
I1
A1
LSB
I0
A0
A4, A3, A2, A1, and A0 (Bit 4, Bit 3, Bit 2, Bit 1, and Bit 0) of
the instruction byte determine which register is accessed during
the data transfer portion of the communication cycle. For multibyte
transfers, this address is the starting byte address. The remaining
register addresses are generated by the AD9786.
SERIAL INTERFACE PORT PIN DESCRIPTIONS
SCLK—Serial Clock. The serial clock pin is used to
synchronize data to and from the AD9786 and to run the
internal state machines. The maximum frequency of SCLK
is 20 MHz. All data input to the AD9786 is registered on the
rising edge of SCLK. All data is driven out of the AD9786
on the falling edge of SCLK.
CSB—Chip Select. Active low input starts and gates a communication cycle. It allows more than one device to be used on the
same serial communication lines. The SDO and SDIO pins go
to a high impedance state when this input is high. Chip select
should stay low during the entire communication cycle.
SDIO—Serial Data I/O. Data is always written into the
AD9786 on this pin. However, this pin can be used as a
bidirectional data line. The configuration of this pin is
controlled by Bit 7 of Register Address 0x00. The default is
Logic 0, which configures the SDIO pin as unidirectional.
SDO—Serial Data Out. Data is read from this pin for protocols
that use separate lines for transmitting and receiving data. In
the case where the AD9786 operates in a single bidirectional
I/O mode, this pin does not output data and is set to a high
impedance state.
Rev. B | Page 20 of 56
AD9786
MSB/LSB TRANSFERS
INSTRUCTION CYCLE
A4 A3
The same considerations apply to setting the software reset
SWRST (0x00[5]) bit. All other registers are set to their default
values, but the software reset does not affect the bits in Register
Address 0x00 and Register Address 0x04.
A0
D7 D6N D5N
D30 D20 D10 D00
D7 D6N D5N
D30 D20 D10 D00
Figure 37. Serial Register Interface Timing MSB First
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
CSB
SDIO
A0
A1 A2
A3 A4
N0 N1 R/W D00 D10 D20
D4N D5N D6N D7N
D00 D10 D20
D4N D5N D6N D7N
SDO
03152-038
SCLK
Figure 38. Serial Register Interface Timing LSB First
tDS
The AD9786 serial port configuration bits reside in Bit 6 and
Bit 7 of Register Address 0x00. Note that the configuration
changes immediately upon writing to the last bit of the register.
For multibyte transfers, writing to this register might occur
during the middle of a communication cycle. Care must be
taken to compensate for this new configuration for the
remaining bytes of the current communication cycle.
A2 A1
SDO
NOTES ON SERIAL PORT OPERATION
It is recommended to use only single-byte transfers when
changing serial port configurations or initiating a software
reset.
R/W N1 N0
03152-037
SDIO
tSCLK
CSB
tPWH
tPWL
SCLK
tDS
SDIO
tDH
INSTRUCTION BIT 7
INSTRUCTION BIT 6
03152-039
The AD9786 serial port controller address increments from 0x1F
to 0x00 for multibyte I/O operations if the MSB-first mode is
active. The serial port controller address decrements from 0x00 to
0x1F for multibyte I/O operations if the LSB-first mode is active.
SCLK
Figure 39. Timing Diagram for Register Write
CSB
SCLK
tDV
SDIO
SDO
DATA BIT n
DATA BIT n–1
Figure 40. Timing Diagram for Register Read
Rev. B | Page 21 of 56
03152-040
The AD9786 serial port can support both MSB-first or LSB-first
data formats. This functionality is controlled by register address
DATADIR (0x00[6]). The default is MSB first. When this bit is
set active high, the AD9786 serial port is in LSB-first format.
That is, if the AD9786 is in LSB-first mode, the instruction byte
must be written from least significant bit to most significant bit.
Multibyte data transfers in MSB-first format can be completed
by writing an instruction byte that includes the register address
of the most significant byte. In MSB-first mode, the serial port
internal byte address generator decrements for each byte
required of the multibyte communication cycle. Multibyte data
transfers in LSB-first format can be completed by writing an
instruction byte that includes the register address of the least
significant byte. In LSB-first mode, the serial port internal byte
address generator increments for each byte required of the
multibyte communication cycle.
DATA TRANSFER CYCLE
CSB
AD9786
MODE CONTROL (VIA SERIAL PORT)
Table 11.
Address
COMMS
FILTER
DATA
MODULATE
RESERVED
DCLKCRC
CALMEMCK
MEMRDWR
MEMADDR
MEMDATA
DCRCSTAT
00
01
02
03
04
05
06
07
08
09
0A
0B
0C
0D
0E
0F
10
11
12
Bit 7
SDIODIR
INTERP[1]
DATAFMT
CHANNEL
Reserved
DATAADJ[3]
Bit 6
DATADIR
INTERP[0]
ONEPORT
HILBERT
Reserved
DATAADJ[2]
CALSTAT
MEMADDR[7]
CALEN
MEMADDR[6]
Bit 5
SWRST
Bit 4
SLEEP
DCLKSTR
MODDUAL
Reserved
DATAADJ[1]
CALMEM[1]
XFERSTAT
MEMADDR[5]
MEMDATA[5]
Bit 3
PDN
ZSTUFF
DCLKPOL
DCLKEXT
SIDEBAND
MOD[1]
Reserved
Reserved
DATAADJ[0]
MODSYNC
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
CALMEN[0]
XFEREN
SMEMWR
MEMADDR[4]
MEMADDR[3]
MEMDATA[4]
MEMDATA[3]
Bit 2
Bit 1
HPFX8
DCLKCRC
MOD[0]
Reserved
MODADJ[2]
HPFX4
IQPOL
Bit 0
EXREF
HPFX2
CRAYDIN
Reserved
MODADJ[1]
Reserved
MODADJ[0]
CALCKDIV[2]
SMEMRD
MEMADDR[2]
MEMDATA[2]
DCRCSTAT[2]
CALCKDIV[2]
FMEMRD
MEMADDR[1]
MEMDATA[1]
DCRCSTAT[1]
CALCKDIV[2]
UNCAL
MEMADDR[0]
MEMDATA[0]
DCRCSTAT[0]
Table 12.
COMMS(00)
SDIODIR
Bit
7
Direction
I
Default
0
DATADIR
6
I
0
SWRST
SLEEP
PDN
EXREF
5
4
3
0
I
I
I
I
0
0
0
0
FILTER(01)
INTERP[1:0]
Bit
[7:6]
Direction
I
Default
00
ZSTUFF
HPFX8
3
2
I
I
0
0
HPFX4
1
I
0
HPFX2
0
I
0
Description
0: SDIO pin configured for input only during data transfer
1: SDIO configured for input or output during data transfer
0: Serial data uses MSB-first format
1: Serial data uses LSB-first format
1: Default all serial register bits, except Address 0x00 and Address 0x04
1: DAC output current off
1: All analog and digital circuitry, except serial interface, off
0: Internal band gap reference
1: External reference
Table 13.
Description
00: No interpolation
01: Interpolation 2×
10: Interpolation 4×
11: Interpolation 8×
1: Zero stuffing on
0: ×8 interpolation filter configured for low-pass
1: ×8 interpolation filter configured for high-pass
0: ×4 interpolation filter configured for low-pass
1: ×4 interpolation filter configured for high-pass
0: ×2 interpolation filter configured for low-pass
1: ×2 interpolation filter configured for high-pass
Rev. B | Page 22 of 56
AD9786
Table 14.
DATA(02)
DATAFMT
Bit
7
Direction
I
Default
0
ONEPORT
6
I
0
DCLKSTR
5
I
0
DCLKPOL
4
I
0
DCLKEXT
3
I
0
DCLKCRC
2
I
0
IQPOL
1
I
0
GRAYDIN
0
I
0
Description
0: Twos complement data format
1: Unsigned binary input data format
0: I and Q input data onto Port 1 and Port 2, respectively
1: I and Q input data interleaved onto Port 1
0: DATACLK pin, 12 mA drive strength
1: DATACLK pin, 24 mA drive strength
0: Input data latched on DATACLK/DACCLK rising edge (dependent on mode)
1: Input data latched on DATACLK/DACCLK falling edge (dependent on mode)
0: DATACLK pin inputs channel data rate or modulator synchronizer clock
1: DATACLK pin outputs channel data rate or modulator synchronizer clock
0: With DATACLK pin as input, DATACLK clock recovery off
1: With DATACLK pin as input, DATACLK clock recovery on
0: In one-port mode, IQSEL = 1 latches data into I channel, IQSEL = 0 latches data
into Q channel
1: In one-port mode, IQSEL = 0 latches data into I channel, IQSEL = 1 latches data
into Q channel
0: Gray decoder off
1: Gray decoder on
Table 15.
MODULATE(03)
CHANNEL
Bit
7
Direction
I
Default
0
HILBERT
MODDUAL
6
5
I
I
0
0
SIDEBAND
4
I
0
MOD[1:0]
[3:2]
I
00
Description
MODDUAL
CHANNEL
0x03[5]
0x03[7]
0
0
I channel processing routed to DAC
0
1
Q channel processing routed to DAC
1
0
Modulator real output routed to DAC
1
1
Modulator imaginary output routed to DAC
1: With MODDUAL on, Hilbert transform on
0: Modulator uses a single channel
1: Modulator uses both I and Q channels
0: With MODDUAL on, upper sideband rejected
1: With MODDUAL on, lower sideband rejected
00: No modulation
01: fS/2 modulation
10: fS/4 modulation
11: fS/8 modulation
Rev. B | Page 23 of 56
AD9786
Table 16.
DCLKCRC(05)
DATAADJ[3:0]
Bit
[7:4]
Direction
I
Default
0000
Description
DATACLK offset (twos complement representation)
0111: +7
:
0000: 0
:
1000: −8
0: Channel data rate clock synchronizer mode
1: State machine clock synchronizer mode
Modulator coefficient offset
fS/8
fS/4
fS/2
000
1
1
1
001
+1/√2
0
–1
010
0
–1
1
011
–1/√2
0
–1
100
–1
+1
+1
101
–1/√2
0
–1
110
0
–1
+1
111
+1/√2
0
–1
MODSYNC
3
I
00
MODADJ[2:0]
[2:0]
I
000
Bit
[3:0]
Direction
O
Default
Description
Hardware version identifier
CALMEMCK(OE)
CALMEM
Bit
[5:4]
Direction
O
Default
00
CALCKDIV[2:0]
[2:0]
I
00
Description
Calibration memory
00: Uncalibrated
01: Self-calibration
10: Factory calibration
11: User input
Calibration clock divide ratio from channel data rate
000: /32
001: /64
:
110: /2048
111: /4096
MEMRDWR(OF)
CALSTAT
Bit
7
Direction
O
Default
0
CALEN
XFERSTAT
6
5
I
O
0
0
XFEREN
SMEMWR
SMEMRD
FMEMRD
UNCAL
4
3
2
1
0
I
I
I
I
I
0
0
0
0
0
Table 17.
VERSION(0D)
VERSION[3:0]
Table 18.
Table 19.
Description
0: Self-calibration cycle not complete
1: Self-calibration cycle complete
1: Self-calibration in progress
0: Factory memory transfer not complete
1: Factory memory transfer complete
1: Factory memory transfer in progress
1: Write static memory data from external port
1: Read static memory to external port
1: Read factory memory data to external port
1: Use uncalibrated
Rev. B | Page 24 of 56
AD9786
Table 20.
MEMADDR(10)
MEMADDR [7:0]
Bit
[7:0]
Direction
I/O
Default
00000000
Description
Address of factory or static memory to be accessed
Bit
[5:0]
Direction
I/O
Default
000000
Description
Data or factory or static memory access
DCRCSTAT(12)
DCRCSTAT (2)
Bit
2
Direction
O
Default
0
DCRCSTAT(1)
1
O
0
DCRCSTAT(0)
0
O
0
Description
0: With DATACLK CRC on, lock has never been achieved
1: With DATACLK CRC on, lock has been achieved at least once
0: With DATACLK CRC on, system is currently not locked
1: With DATACLK CRC on, system is currently locked
0: With DATACLK CRC on, system is currently locked
1: With DATACLK CRC on, system lost lock due to jitter
Table 21.
MEMDATA(11)
MEMDATA [5:0]
Table 22.
Rev. B | Page 25 of 56
AD9786
DIGITAL FILTER SPECIFICATIONS
DIGITAL INTERPOLATION FILTER COEFFICIENTS
0
Table 23. Stage 1 Interpolation Filter Coefficients
Upper Coefficient
H(43)
H(42)
H(41)
H(40)
H(39)
H(38)
H(37)
H(36)
H(35)
H(34)
H(33)
H(32)
H(31)
H(30)
H(29)
H(28)
H(27)
H(26)
H(25)
H(24)
H(23)
Integer Value
9
0
–27
0
65
0
–131
0
239
0
–407
0
665
0
–1070
0
1764
0
–3273
0
10358
16384
–20
–40
–60
–80
–100
–120
–140
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
03152-041
Lower Coefficient
H(1)
H(2)
H(3)
H(4)
H(5)
H(6)
H(7)
H(8)
H(9)
H(10)
H(11)
H(12)
H(13)
H(14)
H(15)
H(16)
H(17)
H(18)
H(19)
H(20)
H(21)
H(22)
Figure 41. 2× Interpolation Filter Response
0
–20
–40
–60
–80
–100
Upper Coefficient
H(19)
H(18)
H(17)
H(16)
H(15)
H(14)
H(13)
H(12)
H(11)
Integer Value
19
0
–120
0
436
0
–1284
0
5045
8192
–140
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
03152-042
Lower Coefficient
H(1)
H(2)
H(3)
H(4)
H(5)
H(6)
H(7)
H(8)
H(9)
H(10)
0.5
03152-043
–120
Table 24. Stage 2 Interpolation Filter Coefficients
Figure 42. 4× Interpolation Filter Response
0
–20
–40
–60
–80
Table 25. Stage 3 Interpolation Filter Coefficients
Lower Coefficient
H(1)
H(2)
H(3)
H(4)
H(5)
H(6)
Upper Coefficient
H(11)
H(10)
H(9)
H(8)
H(7)
Integer Value
7
0
–53
0
302
512
–100
–120
–140
–0.5
Rev. B | Page 26 of 56
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
Figure 43. 8× Interpolation Filter Response
AD9786
CLOCK/DATA TIMING
Table 26. Data Port Synchronization
DCLKEXT
0x02, Bit 3
1
1
MODSYNC
0x05, Bit 3
0
1
DCLKCRC
0x02, Bit 2
X
X
Mode
DATACLK Master
Modulator Master
0
0
0
External Sync Mode
0
0
1
DATACLK Slave
0
1
0
Low Setup/Hold
0
1
1
Modulator Slave
Two-Port Data Input Mode (DATACLK Master)
With the interpolation set to 1×, the DATACLK output is a
delayed and inverted version of DACCLK at the same frequency.
Note that DACCLK refers to the differential clock inputs applied
at Pin 5 and Pin 6. As Figure 44 and Figure 45 show, there is a
constant delay between the edges of DACCLK and DATACLK.
The DCLKPOL bit (Register 0x02, Bit 4) allows the data to be
latched into the AD9786 upon either the rising or falling edge
of DACCLK. With DCLKPOL = 0, the data is latched in upon
the falling edge of DACCLK, as shown in Figure 44. With
DCLKPOL = 1, as shown in Figure 45, data is latched in upon
the rising edge of DACCLK. The setup and hold times are
always with respect to the latching edge of DACCLK.
With the interpolation set to 4× or 8×, the DACCLK input
runs at 4× or 8× the speed of the DATACLK output. The data
is latched in upon a rising edge of DACCLK, similar to the
2× interpolation mode.
DATACLKOUT
t12
tH = 2.9ns MIN
03152-044
tS = –0.5ns MIN
DATA
With the interpolation set to 2×, the DACCLK input runs at
twice the speed of the DATACLK. Data is latched into the digital
inputs of the AD9786 upon every other rising edge of DACCLK,
as shown in Figure 47 and Figure 48. With DCLKPOL = 0, as
shown in Figure 47, the latching edge of DACCLK is the rising
edge that occurs just before the falling edge of DATACLK. With
DCLKPOL = 1, as in Figure 48, the latching edge of DACCLK is
the rising edge of DACCLK that occurs just before the rising edge
of DATACLK. The setup and hold time values are identical to
those in Figure 44 and Figure 45.
Note that there is a slight difference in the delay from the rising
edge of DACCLK to the falling edge of DATACLK, and the
delay from the rising edge of DACCLK to the rising edge of
DATACLK. As Figure 46 shows, the DATACLK duty cycle is
slightly less than 50%. This is true in all modes.
DACCLKIN
tD = 6ns TYP
Figure 44. Data Timing, 1× Interpolation, DCLKPOL = 0
DACCLKIN
However, the latching edge is every fourth edge in 4× interpolation mode and every eighth edge in the 8× interpolation
mode. Similar to operation in the 2× interpolation mode, with
DCLKPOL = 0, the latching edge of DACCLK is the rising edge
that occurs just before the falling edge of DATACLK. With
DCLKPOL = 1, the latching edge of DACCLK is the rising
edge that occurs just before the rising edge of DATACLK.
The setup and hold time values are identical to those in 1×
and 2× interpolation.
DATACLKOUT
tH = 2.9ns MIN
DATA
03152-045
tD = 5.5ns TYP
tS = –0.5ns MIN
Function
Channel data rate clock output
Modulator synchronization
DATACLK output
DATACLK inactive, DACCLK
synchronous with external data
DATACLK input, data rate clock,
data recovery on
DATACLK input, input data
synchronous with DATACLK
Input modulator synchronizer
DATACLK input
Figure 45. Data Timing, 1× Interpolation, DCLKPOL = 1
Rev. B | Page 27 of 56
AD9786
Note that DCLKPOL (Register 0x02, Bit 4) can be used to select
the edge of DACCLK upon which the input data is latched.
DATACLKOUT
There are three status bits available for a read that allow the user
to verify DLL lock. These are Bit 0, Bit 1, and Bit 2 (DCRCSTAT) in
Register 0x12.
03152-046
DACCLKIN
There is a defined setup-and-hold window with respect to input
data and the latching edge of DACCLK. There is also a required
timing relationship between DATACLK and DACCLK. This is
referred to in Figure 49 and Figure 50 as tST and tHT (setup and
hold for transition). For example, with DCLKPOL set to Logic 0,
the input data latches upon the first rising edge of DACCLK
that occurs more than 1.5 ns before the falling edge of DATACLK.
DACCLK should not be given a rising edge in the window of
500 ps to 1.5 ns before the latching edge (falling edge when
DCLKPOL = 0, rising edge when DCLKPOL = 1) of DATACLK.
Failure to account for this timing relationship could result in
corrupt data.
Figure 46. DATACLK Duty Cycle
tS = –0.5ns MIN
tH = 2.9ns MIN
DACCLKIN
03152-047
tD = 6ns TYP
DATA
Figure 47. Data Timing, 2× Interpolation, DCLKPOL = 0
DATACLKIN
tHT = 1.5ns MIN
tST = –500ps MIN
tS = 0.0ns MIN
tH = 3.2ns MIN
03152-049
DACCLKIN
DATA
Figure 49. Slave Mode Timing, 2× Interpolation, DCLKPOL = 0
DATACLKOUT
DACCLKIN
tH = 2.9ns MIN
Figure 48. Data Timing, 2× Interpolation, DCLKPOL = 1
DATACLKIN
tST = –1.0ns MIN
DATACLK Slave Mode (Data Recovery On)
tHT = 2.0ns MIN
tS = 0.0ns MIN
DATACLK (Pin 31) can be used as an input to synchronize
multiple AD9786s. A clock generated by an AD9786 operating
in master mode, or a clock from an external source, can be used
to drive DATACLK.
In this mode, two clocks are required to be applied to the
AD9786. A clock running at the DAC sample rate, referred to as
DACCLK, must be applied to the differential inputs (Pin 5 and
Pin 6) of the AD9786. As described previously, a clock at the
input sample rate must also be applied to Pin 31 (DATACLK).
An internal DLL synchronizes the two applied clocks. The
timing relationships between the input data, DATACLK, and
DACCLK are given in Figure 49 and Figure 50.
Rev. B | Page 28 of 56
tH = 3.2ns MIN
DATA
Figure 50. Slave Mode Timing, 2× Interpolation, DCLKPOL = 1
03152-050
DATA
03152-048
tD = 5ns TYP
tS = –0.5ns MIN
AD9786
Low Setup/Hold Mode
(DATACLK Input, Data Recovery Off)
DACCLKIN
Some applications might require that digital input data be
synchronized with the DATACLK input, rather than DACCLK.
For these applications, the AD9786 can be programmed for low
setup/hold mode by entering the values in Table 26 into the SPI
registers. With data recovery off and the MODSYNC bit set to
Logic 1, the AD9786 latches data in upon the rising or falling
edge of DATACLK input, depending on the state of DCLKPOL.
tS = –300ps MIN
03152-053
DATA
DACCLKIN
Figure 53. External Sync Mode with 2× Interpolation
DATACLKIN
tHT = 0.0ns MIN
tS = –1.1ns MIN
tH = 2.8ns MIN
DATA
03152-051
tST = 3.0ns MIN
Figure 51. Low Setup and Hold Mode Timing, 1× Interpolation, DCLKPOL = 0
DACCLKIN
DATACLKIN
tHT = 1.0ns MIN
tS = –1.8ns MIN
tH = 3.1ns MIN
DATA
03152-052
tST = 2.0ns MIN
tH = 2.9ns MIN
Figure 52. Low Setup and Hold Mode Timing, 1× Interpolation, DCLKPOL = 1
External Sync Mode
In the external sync mode, the DATACLK is programmed as an
input but is not used. Applying a DATACLK input while in this
mode has no effect. The digital input data is synchronized solely
to the DACCLK input. With 1× interpolation, the data input is
latched upon every rising edge of DACCLK. The challenge is
that the user has no way of knowing exactly which edge is the
latching edge when the interpolating filters are in use. In 2×, 4×,
and 8× interpolation modes, the latching edge of DACCLK is
every 2nd, 4th, or 8th edge, respectively.
With the 2 ns keep-out window, shown in Figure 53, there is a
strong possibility of violating setup and hold times, especially at
high speeds. It is recommended that users sense the DAC output
noise floor for setup and hold violations. If setup and hold is violated,
DCLKPOL can be switched. The effect of switching the state of
DCLKPOL is that the latching edge is moved by one, two, or four
DACCLK cycles if the AD9786 is in 2×, 4×, or 8× interpolation
modes, respectively. Note that in this mode, the DATAADJ bits
have no effect.
Note that when using the AD9786 in external sync mode with
1× interpolation, that functionality is identical to master mode,
except that DATACLK out is not available. That is, with
DATACLKPOL = 0, data is latched on the falling edge of DACCLK,
and with DATACLKPOL = 1, data is latched on the rising edge
of DACCLK.
DATAADJUST Synchronization
When designing the digital interface for high speed DACs, care
must be taken to ensure that the DAC input data meets setup
and hold requirements. Often, compensation must be used in
the clock delay path to the digital engine driving the DAC. The
AD9786 has the on-chip capability to vary the latching edge of
DACCLK. With the interpolation function enabled, this allows
the user the choice of multiple edges upon which to latch the
data. For instance, if the AD9786 is using 8× interpolation, the
user can latch from one of eight edges before the rising edge of
DATACLK, or seven edges after this rising edge. The specific
edge upon which data is latched is controlled by SPI Register
0x05, Bits 7:4. Table 27 shows the relationship of the latching
edge of DACCLK and DATACLK with the various settings of
the DATAADJ bits.
Table 27. DATAADJ Values for Latching Edge Sync
SPI Register 0x05
Bit 7 Bit 6 Bit 5 Bit 4
0
0
0
0
0
0
0
1
0
0
1
0
0
0
1
1
0
1
0
0
0
1
0
1
0
1
1
0
0
1
1
1
1
0
0
0
1
0
0
1
1
0
1
0
1
0
1
1
1
1
0
0
1
1
0
1
1
1
1
0
1
1
1
1
Rev. B | Page 29 of 56
Latching Edge Write DATACLK
0
+1
+2
+3
+4
+5
+6
+7
–8
–7
–6
–5
–4
–3
–2
–1
AD9786
Figure 54, Figure 55, and Figure 56 show the alignment for the
latching edge of DACCLK with 4× interpolation and different
settings for DATAADJ. In Figure 54, the AD9786 is in
DATACLK master mode. DATAADJ is set to 0000, with
DCLKPOL set to 0 so that the latching edge of DACCLK is
immediately before the rising edge of DATACLK. The data
transitions shown in Figure 54 are synchronous with the
DACCLK, so that DACCLK and input data are constant with
respect to each other.
The only visible change when DATAADJ is altered is that
DATACLK moves, indicating the latching edge has moved as
well. Note that in DATACLK master mode, when DATAADJ is
altered, the latching edge with respect to DATACLK remains
the same.
Figure 55 shows the same conditions, but with DATAADJ set to
1111. This moves DATACLK to the left in the plot, indicating that
it occurs one DACCLK cycle before it did in Figure 54; therefore,
the latching edge of DACCLK also occurs one cycle earlier.
RISING EDGE OF DATACLK
CONCURRENT WITH
LATCHING EDGE OF DACCLK
DACCLK
LATCHING EDGE
DATA TRANSITION
03152-055
Note that the data in Figure 44 to Figure 53 was taken with the
DATAADJ default of 0000. Changing the DATAADJ values allows
the user to select the specific edge of DACCLK upon which the
input data is latched. This can be done in master mode, but it
is most useful in slave mode. For more information on using
DATAADJ and MODADJ to synchronize multiple AD9786s,
see Analog Devices Application Note 747. Table 27 lists the values
available for 8× interpolation, which, in turn, provides a choice of
16 edges to sync data. With 4× interpolation, there is
a choice of eight edges, and the relevant values from Table 27
are 0000, 0010, 0100, 0110, 1000, 1010, 1100, and 1110. These
options allow latching edge placement from +3 cycles to −4 cycles.
In 2× interpolation, four edges are available, and the relevant
values from Table 27 are 0000, 0100, 1000, and 1100. The
choices for DATAADJ are diminished to +1 cycle to –2 cycles.
Figure 55. DATAADJ = 1111
Figure 56 shows the same conditions, with DATAADJ set
to 0001; therefore, DATACLK moves to the right in the plot.
This indicates that it occurs one DACCLK cycle after it did in
Figure 54; therefore, the latching edge of DACCLK also occurs
one cycle later.
RISING EDGE OF DATACLK
CONCURRENT WITH
LATCHING EDGE OF DACCLK
RISING EDGE OF DATACLK
CONCURRENT WITH
LATCHING EDGE OF DACCLK
DACCLK
LATCHING EDGE
DATA TRANSITION
03152-054
Figure 56. DATAADJ = 0001
DATA TRANSITION
Figure 54. DATAADJ = 0000
Rev. B | Page 30 of 56
03152-056
DACCLK
LATCHING EDGE
AD9786
data images falling in the interpolation filter pass band are passed.
In band-limited applications, the images at the output |of the
DAC must be limited by an analog reconstruction filter. The
complexity of the analog reconstruction filter is determined by
the proximity of the closest image to the required signal band.
Higher interpolation rates yield larger stop-band regions,
suppressing more input images and resulting in a much relaxed
analog reconstruction filter.
Interpolation Modes
Table 28. Interpolation Modes
INTERP[0]
0
1
0
1
Mode
No interpolation
×2 interpolation
×4 interpolation
×8 interpolation
Interpolation is the process of increasing the number of points
in a time domain waveform by approximating points between
the input data points on a uniform time grid. This produces
a higher output data rate. Applied to an interpolation DAC,
a digital interpolation filter is used to approximate the interpolated
points, having an output data rate increased by the interpolation
factor. Interpolation filter responses are achieved by cascading
individual digital filter banks, whose filter coefficients are given in
Table 23, Table 24, and Table 25. Filter responses are shown in
Figure 57, which shows the interpolation filters of the AD9786
under different interpolation rates, normalized to the input data
rate, fSIN.
The digital filter’s frequency domain response exhibits symmetry
about half the output data rate and dc. It causes images of the
input data to be shaped by the interpolation filter’s frequency
response. This has the advantage of causing input data images
that fall in the stop band of the digital filter to be rejected by
the stop-band attenuation of the interpolation filter, while input
A DAC shapes its output with a sinc function, having a null at
the sampling frequency of the DAC. The higher the DAC sampling rate compared to the input signal bandwidth, the less the
DAC sinc function shapes the output. The higher the
interpolation rate, the more input data images fall in the
interpolation filter stop band and are rejected; the bandwidth
between passed images is larger with higher interpolation
factors. The sinc function shaping is also reduced with a higher
interpolation factor.
Table 29. Sinc Shaping at Band Edge of Interpolation Filters
Mode
No interpolation
×2 interpolation
×4 interpolation
×8 interpolation
Sinc Shaping
@ 0.43 fSIN (dB)
–2.8241
–0.6708
–0.1657
–0.0413
Bandwidth to First Image
fSIN
2 fSIN
4 fSIN
8 fSIN
SINC RESPONSE
NO INTERPOLATION
0
–50
INTERP[1] = 0
INTERP[0] = 0
–100
–150
–8
–6
–4
–2
–0
2
4
6
8 fSIN
×2 INTERPOLATION
0
–50
INTERP[1] = 0
INTERP[0] = 1
–100
–150
–8
–6
–4
–2
0
2
4
6
8 fSIN
×4 INTERPOLATION
0
–50
INTERP[1] = 1
INTERP[0] = 0
–100
–150
–8
–6
–4
–2
0
2
4
6
8 fSIN
×8 INTERPOLATION
0
–50
INTERP[1] = 1
INTERP[0] = 1
–100
–150
–8
–6
–4
–2
0
2
Figure 57. Interpolation Modes
Rev. B | Page 31 of 56
4
6
8 fSIN
03152-057
INTERP[1]
0
0
1
1
AD9786
REAL AND COMPLEX SIGNALS
A complex signal contains both magnitude and phase
information. Given two signals at the same frequency, if the
leading signal in phase is cosinusoidal and the lagging signal is
sinusoidal, information pertaining to the magnitude and phase of
a combination of the two signals can be derived; the combination
of the two signals can be considered a complex signal. The cosine
and sine can be represented as a series of exponentials, recalling
that a multiplication by j is a counterclockwise rotation about
the Re/Im plane. The phasor representation of a complex signal
with Frequency f is shown in Figure 58.
Im
Im
Re
C
A/2
A/2
2πft
Re
A/2
A
–f
0
+f
FREQUENCY
A/2
C = Ae2πft = Acos(2πft) + jAsin(2πft)
Asin(2πft) = A
e+j2πft + e–j2πft
2
e+j2πft + e–j2πft
2j
=
=
A
2
A
2
[e+j2πft + e–j2πft]
[ je+j2πft + e–j2πft]
The AD9786 has two channels of interpolation filters, allowing
both I and Q components to be shaped by the same filter transfer
function. The interpolation filter’s frequency response is a real
transfer function. Two DACs are required to represent a complex
signal. A single DAC can only synthesize a real signal. When a
DAC synthesizes a real signal, negative frequency components
fold onto the positive frequency axis. If the input to the DAC is
mirrored symmetrically about dc, the negative frequency
components fold directly onto the positive frequency components in phase-producing, constructive signal summation. If
the input to the DAC is not mirrored symmetrically about dc,
negative frequency components might not be in phase with
positive frequency components, causing destructive signal
summation. Different applications might benefit from either
type of signal summation.
03152-058
Acos(2πft) = A
The cosine term—referred to as the real in-phase, or I component,
of a complex signal—represents a signal on the real plane with
mirror symmetry about dc. The sine term—referred to as the
imaginary quadrature, or Q complex signal component—
represents a signal on the imaginary plane with mirror
asymmetry about dc.
Figure 58. Complex Phasor Representation
Rev. B | Page 32 of 56
AD9786
MODULATION MODES
Table 30. Single-Channel Modulation
MODDUAL
0
0
0
0
0
0
0
0
CHANNEL
0
0
0
0
1
1
1
1
MOD[1]
0
0
1
1
0
0
1
1
MOD[0]
0
1
0
1
0
1
0
1
Mode
I channel, no modulation
I channel, modulation by fDAC/2
I channel, modulation by fDAC/4
I channel, modulation by fDAC/8
Q channel, no modulation
Q channel, modulation by fDAC/2
Q channel, modulation by fDAC/4
Q channel, modulation by fDAC/8
Either channel of the AD9786 interpolation filter channels can
be routed to the DAC and modulated. In single-channel
operation, the input data can be modulated by a real sinusoid;
the input data and the modulating sinusoid contain both
positive and negative frequency components. A double sideband output results when modulating two real signals. At the
DAC output, the positive and negative frequency components
add in phase, resulting in constructive signal summation.
Table 31. Synthesis Bandwidth vs. Interpolation Modes
As the modulating sinusoidal frequency becomes a larger
fraction of the DAC update rate, fDAC, the sinc function of the
DAC shapes the modulated signal bandwidth more, and the
first image moves closer.
Table 32. Modulated Pass-Band Edges Sinc Shaping
(Lower/Upper)
Because the AD9786 interpolation filter pass band represents a
large portion of the input data Nyquist band, it is possible for
modulated signal bands to touch or overlap images if sufficient
interpolation is not used under certain modulation and
interpolation modes.
Figure 59 shows the effects of fDAC/8 modulation when using 8×
interpolation. Figure 60 to Figure 62 show the effects of real
modulation under all interpolation modes. The sinc shaping at
the corners of the modulated signal band and the bandwidth to
the first image for those cases whose pass bands do not touch or
overlap are tabulated.
Modulation
None
fDAC/2
fDAC/4
fDAC/8
Modulation
None
None
fSIN
fSIN
Overlap
Overlap
fDAC/4
None
0
–2.8241
–0.0701
–22.5378
Overlap
fDAC/8
Overlap
fDAC/2
Rev. B | Page 33 of 56
Interpolation
×2
×4
2 fSIN
4 fSIN
2 fSIN
4 fSIN
Touching
2 fSIN
Overlap
Touching
Interpolation
×2
×4
0
0
–0.6708
–0.1657
–1.1932
–2.3248
–9.1824
–6.1190
Touching
–0.2921
–1.9096
Overlap
Touching
×8
8 fSIN
8 fSIN
4 fSIN
6 fSIN
×8
0
–0.0413
–3.0590
–4.9337
–0.5974
–1.3607
–0.0727
–0.4614
AD9786
fDAC/4
3fDAC/8
fDAC/2
5fDAC/8
3fDAC/4
7fDAC/8
fDAC/4
3fDAC/8
fDAC/2
5fDAC/8
3fDAC/4
7fDAC/8
fDAC
fDAC/8
fDAC/8
0
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–7fDAC/8
–fDAC
FILTERED INTERPOLATION IMAGES
fDAC
03152-059
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–7fDAC/8
Figure 59. Double Sideband Modulation
NO INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 0
MOD[1] = 0
MOD[0] = 1
–50
–100
–150
–8
–6
–4
–2
0
2
4
6
8 fSIN
×2 INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 1
MOD[1] = 0
MOD[0] = 1
–50
–100
–150
–8
–6
–4
–2
0
2
4
6
8 fSIN
×4 INTERPOLATION
0
INTERP[1] = 1
INTERP[0] = 0
MOD[1] = 0
MOD[0] = 1
–50
–100
–150
–8
–6
–4
–2
0
2
4
6
INTERP[1] = 1
INTERP[0] = 1
MOD[1] = 0
MOD[0] = 1
–50
–100
–150
–8
8 fSIN
×8 INTERPOLATION
0
–6
–4
–2
0
2
4
Figure 60. Real Modulation by fDAC/2 Under All Interpolation Modes
Rev. B | Page 34 of 56
6
8 fSIN
03152-060
–fDAC
fS/8 MODULATION
AD9786
NO INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 0
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
6
8 fSIN
×2 INTERPOLATION
0
INTERP[1] = 0
–50
INTERP[0] = 1
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
6
8 fSIN
×4 INTERPOLATION
0
INTERP[1] = 1
–50
INTERP[0] = 0
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
6
8 fSIN
×8 INTERPOLATION
0
INTERP[1] = 1
INTERP[0] = 1
–50
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
6
8 fSIN
03215-061
MOD[1] = 1
–100
Figure 61. Real Modulation by fDAC/4 Under All Interpolation Modes
NO INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 0
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
6
8 fSIN
×2 INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 1
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
0
8 fSIN
6
×4 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 0
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
0
8 fSIN
6
×8 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 1
–50
–150
8–
MOD[0] = 0
–6
–4
–2
0
2
4
Figure 62. Real Modulation by fDAC/8 Under All Interpolation Modes
Rev. B | Page 35 of 56
6
8 fSIN
03152-062
MOD[1] = 1
–100
AD9786
Table 33. Dual-Channel Complex Modulation
MODDUAL
0
0
0
0
0
0
0
0
CHANNEL
0
0
0
0
1
1
1
1
MOD[1]
0
0
1
1
0
0
1
1
MOD[0]
0
1
0
1
0
1
0
1
Mode
Real output, no modulation
Real output, modulation by fDAC/2
Real output, modulation fDAC/4
Real output, modulation fDAC/8
Image output, no modulation
Image output, modulation by fDAC/2
Image output, modulation by fDAC/4
Image output, modulation by fDAC/8
Table 34. Complex Modulated Pass-Band Edges Sinc Shaping
(Lower/Upper)
In dual-channel mode, the two channels can be modulated by
a complex signal, with either the real or imaginary modulation
result directed to the DAC. Assume initially, as in Figure 63,
that the complex modulating signal is defined for a positive
frequency only. This causes the output spectrum to be translated in frequency by the modulation factor only. No additional
sidebands are created as a result of the modulation process;
therefore, the bandwidth to the first image from the baseband
bandwidth is the same as the output of the interpolation filters.
Furthermore, pass bands do not overlap or touch. The sinc
shaping at the corners of the modulated signal band is tabulated
in Table 34. Figure 64, Figure 65, and Figure 66 show the effects
of complex modulation with varying interpolation rates.
Modulation
None
None
0
–2.8241
–0.0701
–22.5378
–0.4680
–6.0630
–1.3723
–4.9592
fDAC/2
fDAC/4
fDAC/8
Interpolation
×2
×4
0
0
–0.6708
–0.1657
–1.1932
–2.3248
–9.1824
–6.1190
–0.0175
–0.2921
–3.3447
–1.9096
–0.1160
–0.0044
–1.7195
–0.7866
×8
0
–0.0413
–3.0590
–4.9337
–0.5974
–1.3607
–0.0727
–0.4614
Figure 63. Complex Modulation
Rev. B | Page 36 of 56
fDAC/2
5fDAC/8
3fDAC/4
7fDAC/8
fDAC
fDAC/2
5fDAC/8
3fDAC/4
7fDAC/8
fDAC
3fDAC/8
fDAC/4
03152-063
NO NEGATIVE
SIDEBAND
3fDAC/8
fDAC/8
0
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–7fDAC/8
–fDAC
fS/8 MODULATION
fDAC/4
fDAC/8
0
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–7fDAC/8
–fDAC
FILTERED INTERPOLATION IMAGES
AD9786
×2 INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 1
–50
MOD[1] = 0
–100
–150
–8
MOD[0] = 1
–6
–4
–2
0
2
4
0
6
8f
SIN
×4 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 0
–50
MOD[1] = 0
–100
–150
–8
MOD[0] = 1
–6
–4
–2
0
2
4
0
6
8f
SIN
×8 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 1
–50
MOD[1] = 0
–100
–6
–4
–2
0
2
4
6
8f
SIN
03152-064
–150
–8
MOD[0] = 1
Figure 64. Complex Modulation by fDAC/2 Under All Interpolation Modes
×2 INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 1
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
0
6
8 fSIN
×4 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 0
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 0
–6
–4
–2
0
2
4
0
6
8 fSIN
×8 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 1
–50
MOD[1] = 1
–100
–6
–4
–2
0
2
4
6
8 fSIN
03152-065
–150
–8
MOD[0] = 0
Figure 65. Complex Modulation by fDAC/4 Under All Interpolation Modes
×2 INTERPOLATION
0
INTERP[1] = 0
INTERP[0] = 1
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 1
–6
–4
–2
0
2
4
0
6
8 fSIN
×4 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 0
–50
MOD[1] = 1
–100
–150
–8
MOD[0] = 1
–6
–4
–2
0
2
4
0
6
8 fSIN
×8 INTERPOLATION
INTERP[1] = 1
INTERP[0] = 1
–50
–150
–8
MOD[0] = 1
–6
–4
–2
0
2
4
Figure 66. Complex Modulation by fDAC/8 Under All Interpolation Modes
Rev. B | Page 37 of 56
6
8 fSIN
03152-066
MOD[1] = 1
–100
AD9786
POWER DISSIPATION
60
The AD9786 has seven power-supply domains: two 3.3 V
analog domains (AVDD1 and AVDD2), two 2.5 V analog
domains (ADVDD and ACVDD), one 2.5 V clock domain
(CLKVDD), and two digital domains (DVDD, which runs
from 2.5 V; and DRVDD, which runs from 3.3 V).
4×
2×
30
1×
20
10
0
0
25
50
75
100 125 150
FDATA (MSPS)
175
200
225
250
03152-068
The current for the 2.5 V analog supplies and the digital
supplies varies depending on speed and mode of operation.
Figure 67, Figure 68, and Figure 69 show this variation. Note
that CLKVDD, ADVDD, and ACVDD vary with clock speed
and interpolation rate, but not with modulation rate.
ICLKVDD (mA)
40
The current needed for the 3.3 V analog supplies, AVDD1 and
AVDD2, is consistent across speed and varying modes of the
AD9786. Nominally, the current for AVDD1 is 29 mA across all
speeds and modes, whereas the current for AVDD2 is 20 mA.
Figure 68. CLKVDD Supply Current vs. Clock Speed and Interpolation Rates
30
2× fs/8
4× fs/8
8× fs/8
4× fs/4
25
8× fs/4
2× fs/4
IADVDD AND IACVDD (mA)
425
400
375
350
325
300
275
250
225
200
175
150
125
100
75
50
25
0
4×
8×
2×
1×
8×
2×
4×
20
15
1×
10
0
25
50
75
100 125 150
FDATA (MSPS)
175
200
225
250
0
0
25
50
75
100 125 150
FDATA (MSPS)
175
200
225
250
Figure 69. ADVDD and ACVDD Supply Current vs. Clock Speed
and Interpolation Rates
Figure 67. DVDD Supply Current vs. Clock Speed,
Interpolation, and Modulation Rates
Rev. B | Page 38 of 56
03152-069
5
03152-067
IDVDD (mA)
8×
50
AD9786
fDAC
7fDAC/8
3fDAC/4
5fDAC/8
fDAC/2
3fDAC/8
fDAC/4
fDAC/8
0
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–fDAC
–7fDAC/8
FILTERED INTERPOLATION IMAGES
fDAC
7fDAC/8
3fDAC/4
5fDAC/8
fDAC/2
3fDAC/8
fDAC/4
fDAC/8
0
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–fDAC
–7fDAC/8
fS/8 MODULATION
fDAC
03152-070
7fDAC/8
3fDAC/4
5fDAC/8
fDAC/2
3fDAC/8
fDAC/4
fDAC/8
0
–fDAC/8
–fDAC/4
–3fDAC/8
–fDAC/2
–5fDAC/8
–3fDAC/4
–fDAC
–7fDAC/8
fS/4 MODULATION
Figure 70. Complex Modulation with Negative Frequency Aliasing
Table 35. Dual Channel Complex Modulation with Hilbert
Hilbert
0
1
Mode
Hilbert transform off
Hilbert transform on
When complex modulation is performed, the entire spectrum is
translated by the modulation factor. If the resulting modulated
spectrum is not mirrored symmetrically about dc when the
DAC synthesizes the modulated signal, negative frequency
components fall on the positive frequency axis and can cause
destructive summation of the signals, as shown in Figure 70. For
some applications, this can distort the modulated output signal.
X = Ae j2π(f + fm)t
Y = Ae j2π(f + fm)t – π/2
Im
Im
Re
Z = HILBERT(Y)
C=X–Z
Im
Re
0
Im
Re
A/2
The operation of the Hilbert transform (Figure Z) rotates the
negative frequency components of Figure Y by +π/2, and the
positive frequency components of Figure Y by −π/2. The result
of the Hilbert transform output is then summed with the complex
signal in the main signal path. The result is that negative frequencies are cancelled and, therefore, do not fold back into the
positive side of the frequency spectrum. The Δt block in the main
signal path offsets the delay inherent in the Hilbert transform
(nine DAC clock cycle delay). When the DAC synthesizes the
modulated output, there are no negative frequency components to
fold onto the positive frequency axis out of phase; consequently,
no distortion is produced as a result of the modulation process.
ALIASED NEGATIVE FREQUENCY INTERPOLATION IMAGES
Re
A/2
A/2
f
A/2
A/2
A
A/2
00
A/2
f
A/2
–50
00
A/2
f
A/2
f
A
dBFS
A/2
03152-071
A/2
–100
In Figure 71, Figure X represents a complex signal typically
found in the AD9786 signal path. Figure Y is identical to Figure X,
but it is shifted by π/2. The phase shifting in the AD9786 occurs
because the digital LO driving the digital quadrature modulator
in the Hilbert transform path is phase shifted by π/2.
Rev. B | Page 39 of 56
–150
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
Figure 72. Negative Frequency Aliasing Distortion
0.5
03152-072
Figure 71. Negative Frequency Image Rejection
AD9786
Figure 72 shows this effect at the DAC output for a signal
mirrored asymmetrically about dc that is produced by complex
modulation without a Hilbert transform. The signal bandwidth
was narrowed to show the aliased negative frequency
interpolation images.
The transfer function of an ideal Hilbert transform has a +90°
phase shift for negative frequencies, and a –90° phase shift for
positive frequencies. Because of the discontinuities that occur at
0 Hz and at 0.5 × the sample rate, any real implementation of
the Hilbert transform trades off bandwidth vs. ripple.
In contrast, Figure 73 shows the same waveform with the
Hilbert transform applied. Clearly, the aliased interpolation
images are not present.
Figure 74 and Figure 75 show the gain of the Hilbert transform
vs. frequency. Gain is essentially flat, with a pass-band ripple of
0.1 dB over the frequency range of 0.07 × the sample rate to
0.43 × the sample rate.
0
dBFS
–50
–150
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
03152-073
–100
Figure 76 shows the phase response of the Hilbert transform
implemented in the AD9786. The phase response for positive
frequencies begins at –90° at 0 Hz, followed by a linear phase
response (pure time delay) equal to nine filter taps (nine
DACCLK cycles). For negative frequencies, the phase response
at 0 Hz is +90°, followed by a linear phase delay of nine filter
taps. To compensate for the unwanted 9-cycle delay, an equal
delay of nine taps is used in the AD9786 digital signal path
opposite the Hilbert transform. This delay block is shown as Δt
in the Functional Block Diagram (Figure 1).
10
0
Figure 73. Effects of Hilbert Transform
–10
If the output of the AD9786 is used with a quadrature modulator,
negative frequency images are cancelled without the need for
a Hilbert transform.
–20
–30
–40
–50
HILBERT TRANSFORM IMPLEMENTATION
–60
The Hilbert transform on the AD9786 is implemented as a
19-coefficient FIR. The coefficients are given in Table 36.
–70
–80
Table 36.
03152-074
–90
Integer Value
–6
0
–17
0
–40
0
–91
0
–318
0
+318
0
+91
0
+40
0
+17
0
+6
–100
100
200
300
400
500
600
700
800
900 1000
Figure 74. Hilbert Transform Gain
1.0
0.8
0.6
0.4
0.2
0
–0.2
–0.4
–0.6
–0.8
–1.0
100
200
300
400
500
600
700
800
Figure 75. Hilbert Transform Ripple
Rev. B | Page 40 of 56
900 1000
03152-075
Coefficient
H(1)
H(2)
H(3)
H(4)
H(5)
H(6)
H(7)
H(8)
H(9)
H(10)
H(11)
H(12)
H(13)
H(14)
H(15)
H(16)
H(17)
H(18)
H(19)
AD9786
A baseband double sideband signal modulated to IF increases
IF filter complexity and reduces power efficiency. If the baseband signal is complex, a single sideband IF modulation can be
used, relaxing IF filter complexity and increasing power
efficiency.
4
3
2
1
The AD9786 has the ability to place the baseband single sideband complex signal either above or below the IF frequency.
Figure 78, Figure 79, and Figure 80 illustrate this.
0
–1
0
–2
200
400
600
800
1000
1200
–50
dBFS
–4
100
03152-076
–3
Figure 76. Phase Response of Hilbert Transform
–100
Re()
AD9786
–0.1
0
0.1
0.3
0.4
0.5
0.4
0.5
0
0
90
Figure 77. AD9786 Driving Quadrature Modulator
dBFS
–50
The AD9786 can be configured to drive a quadrature modulator,
as in Figure 77. When two AD9786s are used with one AD9786
producing the real output, the second AD9786 produces the
imaginary output. By configuring the AD9786 as a complex
modulator coupled to a quadrature modulator, IF image
rejection is possible. The quadrature modulator acts as the real
part of a complex modulation, producing a double sideband
spectrum at the local oscillator (LO) frequency with mirror
symmetry about dc.
–fIF
0
–fIF
0
BASEBAND
–100
–150
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
IF
SIDEBAND = 1
0
Figure 80. IF Quadrature Modulation of Real and Complex Baseband Signals
Rev. B | Page 41 of 56
0.2
0.3
Figure 79. Lower IF Sideband Rejected
SIDEBAND = 0
–fIF
0.2
Figure 78. Upper IF Sideband Rejected
03152-080
Q
Im()
–0.2
fIF
LO
–0.3
fIF
AD9786
–0.4
fIF
I
–150
–0.5
03152-078
Mode
Upper IF sideband rejected
Lower IF sideband rejected
03152-077
Sideband
0
1
03152-079
Table 37. Dual Channel Complex Modulation Sideband Selection
AD9786
The second master mode, DATACLK master mode, generates a
reference clock that is at the channel data rate. In this mode, the
slave devices align their internal channel data rate clock to the
master. If modulator phase alignment is needed, a concurrent
serial write to all slave devices is necessary. To achieve this, the
CSB pin on all slaves must be connected together, and a group
serial write to the MODADJ register bits must be performed.
Following a successful serial write, the modulator coefficient
alignment is updated upon the next rising edge of the internal
state machine (see Figure 81). Modulator master mode does not
need a concurrent serial write, because slaves lock to the master
phase automatically.
Master/Slave, Modulator/DATACLK Master Modes
In applications where two or more AD9786s are used to synthesize several digital data paths, it might be necessary to ensure
that the digital inputs to each device are latched synchronously.
In complex data processing applications, digital modulator phase
alignment might be required between two AD9786s. To allow
data synchronization and phase alignment, only one AD9786
should be configured as a master device, providing a reference
clock for another slave-configured AD9786.
With synchronization enabled, a reference clock signal is
generated on the DATACLK pin of the master. The DATACLK
pins on the slave devices act as inputs for the reference clock
generated by the master. The DATACLK pin on the master and
all slaves must be directly connected. All master and slave devices
must have the same clock source connected to their respective
CLK+/CLK– pins.
In a slave device, the local channel data rate clock and the
digital modulator clock are created from the internal state
machine. The local channel data rate clock is used by the slave
to latch digital input data. At high data rates, the delay inherent
in the signal path from master to slave can cause the slave to lag
the master when acquiring synchronization. To accommodate
for this, an integer number of the DAC update clock cycles can
be programmed into the slave device as an offset. The value in
DATAADJ allows the local channel data rate clock in the slave
device to advance by up to eight cycles of the DAC clock, or to
be delayed by up to seven cycles (see Figure 82).
When configured as a master, the reference clock generated can
take one of two forms. In modulator master mode, the reference
clock is a square wave with a period equal to 16 cycles of the
DAC update clock. Internal to the AD9786 is a 16-state, finite
state machine, running at the DAC update rate. This state machine
generates all internal and external synchronization clocks and
modulator phasings. The rising edge of the master reference clock
is time aligned to state zero of the internal state machine. Slave
devices use the master reference clock to synchronize data latching
and align modulator phase by aligning state zero of the local
state machine to the master.
The digital modulator coefficients are updated at the DAC clock
rate and decoded in sequential order from the state machine
according to Figure 83. The MODADJ bits can be used to align
a different coefficient to the finite state machine’s zero state, as
shown in Figure 84.
DAC
CLOCK
STATE
MACHINE
0
1
2
3
4
5
6
7
8
9
10
11 12 13 14 15
0
1
2
3
4
5
6
7
8
9
10
11 12 13 14 15
MODULATOR
COEFFICIENT
1
0
–1
0
1
0
–1
0
1
0
–1
0
–1
0
1
0
–1
0
1
0
–1
0
1
0
MODADJ
1
0
–1
0
000
–1
0
1
0
000
03152-081
STATE MACHINE
CYCLE CLOCK
CHANNEL DATA
RATE CLOCK
Figure 81. Synchronous Serial Modulator Phase Alignment
Rev. B | Page 42 of 56
AD9786
DATADJ[3:0]
0000
1111
0001
DAC CLOCK
RECEIVED CHANNEL
DATA RATE CLOCK
–1
03152-082
LOCAL CHANNEL
DATA RATE CLOCK
+1
Figure 82. Local Channel Data Rate Clock Synchronized with Offset
STATE
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
DECODE
1
0
1/ 2
0
0
0
–1/ 2
0
–1
0
–1/ 2
0
0
0
–1/ 2
0
fs/8
0
0
0
2
3
4
1
5
6
2
7
03152-083
fs/4
fs/2
1
3
1
Figure 83. Digital Modulator State Machine Decode
MODADJ[2:0]
000
010
101
DAC CLOCK
14
15
0
1
2
3
15
0
1
15
0
1
2
MODULATOR
COEFFICIENT
–1
0
1
0
–1
0
0
–1
0
1
0
–1
0
03152-084
STATE
MACHINE
STATE MACHINE
CYCLE CLOCK
Figure 84. Local Modulator Coefficient Synchronized with Offset
Rev. B | Page 43 of 56
AD9786
OPERATING THE AD9786 REV. F EVALUATION BOARD
This section provides information to power up the board and
verify correct operation; a description of more advanced modes
of operation has been omitted.
POWER SUPPLIES
The AD9786 Rev. F evaluation board has five power supply
connectors, labeled AVDD1, AVDD2, ACVDD/ADVDD,
CLKVDD, and DVDD, whereas the AD9786 has seven power
supply domains. To reconcile the power supply domains on the
chip with the power supply connectors on the evaluation board,
use Table 38.
Additionally, the DRVDD power supply on the AD9786 is used
to supply power for the digital input bus. DRVDD should be
run from 3.3 V. On the evaluation board, DRVDD is jumperselectable by JP1, which is just to the left of the chip on the
evaluation board. With the jumper set to the 3.3 V position, the
DRVDD chip receives its power from VDD3IN.
CLKVDDS
PECL CLOCK DRIVER
The AD9786 system clock is driven from an external source via
Connector S1. The AD9786 evaluation board includes an
ON Semiconductor® MC100EPT22 PECL clock driver. In the
factory, the evaluation board is set to use this PECL driver as a
single-ended-to-differential clock receiver. The PECL driver can
be set to run from 2.5 V from the CLKVDD power connector
or 3.3 V from the VDD3IN power connector. This setting is
done via Jumper JP2, situated next to the CLKVDD power
connector, and by setting Input Bias Resistor R23 and Input
Bias Resistor R4 on the evaluation board. The factory default is
for the PECL driver to be powered from CLKVDD at 2.5 V
(R23 = 90.9 Ω, R4 = 115 Ω). To operate the PECL driver with
a 3.3 V supply, R23 must be replaced with a 115 Ω resistor; R4
must be replaced with a 90.9 Ω resistor; and the position of JP2
must be changed. The schematic of the PECL driver section of
the evaluation board is shown in Figure 85. A low jitter sine
wave should be used as the clock source. Care must be taken to
ensure that the clock amplitude does not exceed the power
supply rails for the PECL driver.
CLKVDDS
CLK+
ACLKX
R23
115Ω
7
R5
50Ω
MC100EPT22
1
COND;5
U2
CLKVDDS;8
2
R4
90.9Ω
R6
50Ω
R7
50Ω
CLK–
03152-085
C32
0.1μF
Figure 85. PECL Driver on AD9786 Rev. F Evaluation Board
Table 38. Power Supply Domains on AD9786 Rev. F Evaluation Board
Evaluation Board Label/PS Domain on Chip
DVDD
CLKVDD
ACVDD/ADVDD
AVDD2
AVDD1
Nominal Power
Supply Voltage (V)
2.5
2.5
2.5
3.3
3.3
Description
SPI port
Clock circuitry
Analog circuitry containing clock and digital interface circuitry
Switching analog circuitry
Analog output circuitry
Rev. B | Page 44 of 56
AD9786
DATA INPUTS
ANALOG OUTPUT
Digital data inputs to the AD9786 are accessed on the evaluation
board through Connector J1 and Connector J2. These are 40-pin,
right-angle connectors that are intended to be used with standard
ribbon cable connectors. The input level should be 3.3 V. The
data format is selectable through Register 0x02, Bit 7 (DATAFMT).
With this bit set to a default 0, the AD9786 assumes that the input
data is in twos complement format. With this bit set to 1, data
should be input in offset binary format.
The analog output of the AD9786 is accessed via Connector S3.
Once all settings are selected and the current levels and SPI port
functionality are verified, the analog signal at S3 can be viewed.
For most of the AD9786 applications, a spectrum analyzer is the
preferred instrument to verify proper performance. A typical
spectral plot is shown in Figure 86, with the AD9786 synthesizing
a two-tone signal in the default mode with a 200 MSPS sample
rate. A single-tone CW signal should provide output power of
approximately +0.5 dBm to the spectrum analyzer.
When the evaluation board is first powered up and the clock
and data are running, it is recommended that the proper operating
current be verified. Press Reset Switch SW1 to ensure that the
AD9786 is in default mode. The default mode for the AD9786 is
for the interpolation set to 1×. The modulator is turned off in
default mode. The nominal operating currents for the evaluation
board in the power-up default mode are shown in Table 39.
Table 39. Nominal Operating Currents in Power-Up Default
Mode
Evaluation Board
Power Supply
DVDD
CLKVDD
ACVDD/ADVDD
AVDD1
AVDD2
Nominal Current @ Speed (mA)
50
100
150
200
MSPS
MSPS
MSPS
MSPS
26
49
74
99
78
83
87
92
1
4
6
8
30
30
30
30
27
27
27
27
If the spectrum does not look correct at this point, the data
input might be violating setup and hold times with respect to
the input clock. To correct this, the user should vary the input
data timing. If this is not possible, SPI Register 0x02, Bit 4
(DCLKPOL), can be inverted. This bit controls the clock edge
upon which the data is latched. If neither of these methods corrects
the spectrum, it is unlikely that the issue is timing related. In this
case, verify that all instructions have been followed correctly and
that the SPI port readback indicates the correct values.
MARKER 1 [T1]
REF LVL
0
0dBm
RBW 30kHz
–84.96dBm
VBW 30kHz
193.00170300MHz
SWT 560ms
RF ATT
20dB
UNIT
dBm
A
–10
–20
–30
–40 1AVG
1MA
–50
–60
Table 40. SPI Registers
Bit 7
INTERP[1]
Bit 6
INTERP[0]
Bit 5
Bit 4
Bit 3
–70
Bit 0
–80
–90
–100
SERIAL PORT
–110
SW1 is a hard reset switch that sets the AD9786 to its default
state. It should be used every time the AD9786 power supply is
cycled, the clock is interrupted, or new data is to be written via
the SPI port. Set the SPI software to read back data from the
AD9786, and then verify that the expected values are read back
when the software is run.
–120
Rev. B | Page 45 of 56
START 100MHz
19.9MHz/
STOP 200MHz
Figure 86. Typical Spectral Plot
03152-086
Register
0x01
Rev. B | Page 46 of 56
Figure 87. Power Supply Distribution, Rev. F Evaluation Board
03152-087
S11
3.3VQ
2.5VQ
CGND;3,4,5
SMAEDGE
CLKVDD_IN
2
AGND; 3,4,5
SMAEDGE
1
S10
3.3V
AGND2; 3,4,5
AVDD_IN
S9
SMAEDGE
2.5VN
DGND; 3,4,5
ADVDD3_IN
S5
SMAEDGE
2.5V
AGND2; 3,4,5
DVDD_IN
S7
SMAEDGE
ADVDD2_IN
L2
FERRITE
C65
22μF
16V
TP6
RED
TP4
RED
TP2
RED
TP18
BLK
TP13
RED
TP1
RED
C69
0.1μF
C68
0.1μF
AVD1
C67
0.1μF
C48
0.1μF
C47
0.1μF
POWER INPUT FILTERS
FERRITE
C63
+
22μF
16V
L1
FERRITE
+ C64
22μF
16V
+
L3
FERRITE
+ C46
22μF
16V
TP17
BLK
L9
FERRITE
+ C45
22μF
16V
L8
L11
FERRITE
L12
TP7
BLK
JP5
CVD
TP5
BLK
JP10
TP3
BLK
AVD2
JP9
TP16
BLK
VDD
JP34
1
2
3
A
B
JP33
JP30
C76
0.1μF
C34
0.1μF
JP6
JP8
JP7
CLKVDDS
DRVDD
AVDD2
ACVDD
ADVDD
BLK
BLK
BLK
BLK
ACLKX
BLK
TP30 TP31 TP32 TP33 TP34
FERRITE
+
L6
JP1
2
1
3
A B
C29
22μF
16V
JP36
C75
0.1μF
JP2
CLKVDD
AVDD
AVDD2
DVDDS
DVDD
DVDD
TP12 FERRITE
BLK
AVD3
C32
0.1μF
L7
VAL
L10
VAL
L13
VAL
L14
VAL
BLK
TP36
BLK
TP35
C35
0.1μF
CLKVDDS
R23
90.9Ω MC100EPT22
1
7
CGND; 5
U2
CLKVDDS; 8
2
R4
115Ω
CLKVDDS
+ C28
4.7μF
6.3V
50Ω
R6
50Ω
R5
50Ω
R7
6
4
CLKVDDS; 8
CGND; 5
U2
MC100EPT22
3
CLK–
CLK+
AD9786
Figure 88. AD9786 Local Circuitry, Rev. F Evaluation Board
Rev. B | Page 47 of 56
IQ
B
A
S6
1
2
3
DGND; 3,4,5
OPCLK_3
JP28
BD15
03152-088
TP14
WHT
C33
0.1μF
OPCLK
JP27
BD14
+ C7
10μF
6.3V
DVDD
+ C8
10μF
6.3V
DVDD
+ C9
10μF
6.3V
DVDD
OPCLK
S4
DATACLK
S2
C54
0.001μF
DGND; 3,4,5
+ C31
10μF
6.3V
DRVDD
+ C10
10μF
6.3V
C26
0.001μF
C23
0.001μF
C24
0.001μF
C25
0.001μF
C36
0.1μF
C39
0.1μF
C41
0.1μF
C40
0.1μF
AD15
RESET 58
SPI_CSB 57
22 P1B5
23 P1B4
24 P1B3
AD05
AD04
AD03
DVDD6 52
29 P1B0LSB
AD00
P2B2 49
P2B3 48
32 P2B15MSB-IQSEL
33 P2B14-OPCLK
DVDD5 44
P2B6 43
P2B7 42
37 P2B12
38 P2B11
39 P2B10
BD12
BD11
BD10
AD9786BTSP
P2B8 41
DCOM5 45
36 DVDD4
40 P2B9
P2B5 46
35 DCOM4
BD09
P2B4 47
U1
BD01
P2B1 50
31 DCLK-PLLL
BD08
BD07
BD06
BD05
BD04
BD03
BD02
BD00
SPSDO
SPSDI
SPCLK
SPCSB
RESET
TP11
WHT
C37
0.1μF
+ C30
10V
10μF
C17
0.1μF
C19
0.1μF
C15
0.1μF
C66
10μF
6.3V
C2
10μF
6.3V
+
C3
10μF
6.3V
C22
0.001μF
DVDD
+ C6
10μF
6.3V
C38
0.1μF
4
6
5
4
3
DVDD
4
3
R42
49.9Ω
2
1
DRVDD
AGND; 3,4,5
S8
TP29
BLK
SW1
FLOAT; 5
RESET
AVDD2
AGND; 3,4,5
S3
OUT1
C61
0.001μF
S
P
1
TTWB-1-B
T2B
NC = 5
R9
49.9Ω
R10
49.9Ω
C18
0.001μF
+ C5
C21
10μF
0.001μF
6.3V
6
T3
S
1
C4
0.1μF
3
P
ADVDD
AVDD
C62
0.1μF
R8
C16
2kΩ
0.1μF 0.01%
TP8
WHT
TP10
WHT
+
C55
0.001μF
C14
0.1μF
C20
0.001μF
ACVDD
C49
0.1μF
P2B0LSB 51
34 P2B13
BD13
DCOM6 53
28 P1B1
AD01
30 DRVDD1
SP-SDO 54
27 P1B2
AD02
SP-SDI 55
REFIO 59
21 P1B6
AD06
SP-CLK 56
DNC1 61
FSADJ 60
20 P1B7
AD07
26 DVDD3
ADVDDP2 62
19 P1B8
AD08
25 DCOM3
ADCOMP2 63
18 P1B9
AD09
15 P1B10
AD10
ACVDDP2 64
AVDD2P2 66
14 P1B11
AD11
17 DVDD2
AVDD1P1 68
ACOM2P2 67
13 P1B12
AD12
ACCOM2P2 65
ACOM1P21 69
12 P1B13
AD13
16 DCOM2
IOUTB 70
11 P1B14
AD14
IOUTA 71
ACOM1P11 72
9 DVDD1
10 P1B15MSB
AVDD1P2 73
8 DCOM1
ACOM2P12 74
7 CLKCOM2
S1
CGND; 3,4,5
CLK+
AVDD2P1 75
6 CLK–
ACLKX
JP23
ACOM2P1 76
DNC2 80
5 CLK+
1 CLKVDD1
C27
1pF
ACVDDP1 77
CLK–
DVDD
C42
0.1μF
4 CLKCOM1
3
T1
T1-1T
4
C11
0.1μF
ACCOMP1 78
R1
50Ω
C12
0.1μF
3 CLKVDD2
2
+C1
10μF
6.3V
CLKVDD
ADVDDP1 79
5
JP22
CLKVDD
2 LPF
1
6
TP15
WHT
C13
0.1μF
ADTL1-12
R3
10kΩ
+
R2
10kΩ
AD9786
AD9786
R29
100Ω
1
4
3
6
5
8
7
10
9
12
11
14
13
16
15
18
17
20
19
22
21
24
23
26
25
28
27
30
29
32
31
34
33
36
35
38
37
40
39
3
4
5
6
7
8
9
10
2
AX13
3
AX12
4
AX11
5
AX10
6
AX09
7
AX08
8
AX07
1
AX06
2
AX05
3
AX04
4
AX03
5
AX02
6
AX01
7
AX00
8
1
2
3
4
5
6
7
8
9 10
RP1 22
RP1 22
RP1 22
RP1 22
RP1 22
RP1 22
RP1 22
RP1 22
RP2 22
RP2 22
RP2 22
RP2 22
RP2 22
RP2 22
RP2 22
RP2 22
RP6
DNP
R1 R2 R3 R4 R5 R6 R7 R8 R9
AX00
R38
100Ω
AX04
2
RP5
DNP
1
AX07
AX05
R1 R2 R3 R4 R5 R6 R7 R8 R9
AX14
RIBBON
J1
AX06
JP12
AX11
AX15
RCOM
2
JP3
AX10
R33
100Ω
1
DATA-A
AX09
R32
100Ω
RCOM
AX12
R28
100Ω
AX08
R31
100Ω
R39
100Ω
R40
100Ω
R34
100Ω
R41
100Ω
2
3
4
5
6
7
8
9
10
RP7
DNP
16
AD15
15
AD14
14
AD13
13
AD12
12
AD11
11
AD10
10
AD09
9
AD08
16
AD07
15
AD06
14
AD05
13
AD04
12
AD03
11
AD02
10
AD01
9
AD00
1
2
3
4
5
6
7
8
9 10
RP8
DNP
R1 R2 R3 R4 R5 R6 R7 R8 R9
JP21
R44
100Ω
R43
100Ω
1
R1 R2 R3 R4 R5 R6 R7 R8 R9
AX01
JP19
AX02
AX03
R46
100Ω
03152-089
AX13
R27
100Ω
R30
100Ω
RCOM
AX14
R26
100Ω
RCOM
AX15
Figure 89. Digital Data Port A Input Terminations, Rev. F Evaluation Board
Rev. B | Page 48 of 56
AD9786
R60
100Ω
BX13
R64
100Ω
3
4
5
6
7
8
9
1
2
BX13
3
BX12
4
BX11
5
BX10
6
BX09
7
BX08
8
BX07
1
BX06
2
BX05
3
BX04
4
BX03
5
BX02
6
BX01
7
BX00
8
6
5
8
7
10
9
12
11
14
13
16
15
18
17
20
19
22
21
24
23
26
25
28
27
30
29
32
31
34
33
SDO
36
35
CLK
38
37
SDI
40
39
CSB
1
RIBBON
J2
2
3
4
5
6
7
8
9 10
RP3 22
RP3 22
RP3 22
RP3 22
RP3 22
RP3 22
RP3 22
RP3 22
RP4 22
RP4 22
RP4 22
RP4 22
RP4 22
RP4 22
RP4 22
RP4 22
RP11
DNP
R1 R2 R3 R4 R5 R6 R7 R8 R9
BX00
BX07
R55
100Ω
BX04
2
RP12
10 DNP
BX14
3
BX05
R1 R2 R3 R4 R5 R6 R7 R8 R9
BX15
4
BX06
JP31
BX11
R54
100Ω
R53
100Ω
R56
100Ω
R47
100Ω
2
3
4
5
6
7
8
9
10
RP9
DNP
16
BD15
15
BD14
14
BD13
13
BD12
12
BD11
11
BD10
10
BD09
9
BD08
16
BD07
15
BD06
14
BD05
13
BD04
12
BD03
11
BD02
10
BD01
9
BD00
1
2
3
4
5
6
7
8
9 10
RP10
DNP
R1 R2 R3 R4 R5 R6 R7 R8 R9
JP25
R51
100Ω
R49
100Ω
1
R1 R2 R3 R4 R5 R6 R7 R8 R9
BX01
JP24
BX02
BX03
R52
100Ω
03152-090
1
RCOM
2
JP26
BX10
R63
100Ω
1
DATA-B
BX09
R59
100Ω
RCOM
BX12
BX08
R58
100Ω
RCOM
R61
100Ω
BX14
R57
100Ω
RCOM
R62
100Ω
BX15
Figure 90. Digital Data Port B Input Terminations, Rev. F Evaluation Board
Rev. B | Page 49 of 56
AD9786
DVDDS
OPCLK_3
+ C52
4.7μF
6.3V
C53
0.1μF
10
PRE
11
9
J
Q
13
CLK
12
7
K
Q_
CLR
14
74LCX112
DGND;8
U7
DVDDS;16
2
SPCSB
U5
1
12
4
U5
10
3
SPSDI
U5
8
5
1
U6
2
13
74AC14
R21
10kΩ
R20
10kΩ
3
R48
9kΩ
U5
9
R45
9kΩ
U6
11
4
U6
6
12
U6
DVDDS
10
74AC14
9
74AC14
U6
74AC14
74AC14
5
11
74AC14
74AC14
SPSDO
U5
74AC14
74AC14
6
13
SPI PORT
P1
1
2
3
4
5
6
74AC14
74AC14
SPCLK
U5
R50
9kΩ
U6
8
+ C43
4.7μF
6.3V
C50
0.1μF
74AC14
Figure 91. SPI and One-Port Clock Circuitry, Rev. F Evaluation Board
Rev. B | Page 50 of 56
+ C44
4.7μF
6.3V
C51
0.1μF
03152-091
OPCLK
4
PRE
3
5
J
Q
1
CLK
2
6
K
Q_
CLR
15
DGND;8
74LCX112
DVDDS;16
U7
03152-092
AD9786
03152-093
Figure 92. PCB Assembly, Primary Side, Rev. F Evaluation Board
Figure 93. PCB Assembly, Secondary Side, Rev. F Evaluation Board
Rev. B | Page 51 of 56
03152-094
AD9786
03152-095
Figure 94. PCB Assembly, Layer 1 Metal, Rev. F Evaluation Board
Figure 95. PCB Assembly, Layer 6 Metal, Rev. F Evaluation Board
Rev. B | Page 52 of 56
03152-096
AD9786
03152-097
Figure 96. PCB Assembly, Layer 2 Metal (Ground Plane),Rev. F Evaluation Board
Figure 97. PCB Assembly, Layer 3 Metal (Power Plane),Rev. F Evaluation Board
Rev. B | Page 53 of 56
03152-098
AD9786
03152-099
Figure 98. PCB Assembly, Layer 4 Metal (Power Plane), Rev. F Evaluation Board
Figure 99. PCB Assembly, Layer 5 Metal (Ground Plane), Rev. F Evaluation Board
Rev. B | Page 54 of 56
AD9786
OUTLINE DIMENSIONS
14.20
14.00 SQ
13.80
0.75
0.60
0.45
1.20
MAX
12.20
12.00 SQ
11.80
80
61
61
1
60
80
1
60
PIN 1
EXPOSED
PAD
TOP VIEW
(PINS DOWN)
BOTTOM VIEW
0° MIN
1.05
1.00
0.95
0.15
0.05
SEATING
PLANE
6.00
BSC SQ
0.20
0.09
7°
3.5°
0°
0.08 MAX
COPLANARITY
(PINS UP)
20
41
40
21
VIEW A
41
20
21
40
0.50 BSC
LEAD PITCH
0.27
0.22
0.17
VIEW A
ROTATED 90° CCW
COMPLIANT TO JEDEC STANDARDS MS-026-ADD-HD
Figure 100. 80-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
(SV-80-1)
Dimensions shown in millimeters
ORDERING GUIDE
Model
AD9786BSV
AD9786BSVRL
AD9786BSVZ1
AD9786BSVZRL1
AD9786-EB
1
Temperature Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
80-Lead TQFP_EP
80-Lead TQFP_EP
80-Lead TQFP_EP
80-Lead TQFP_EP
Evaluation Board
Z = Pb-free part.
Rev. B | Page 55 of 56
Package Option
SV-80-1
SV-80-1
SV-80-1
SV-80-1
AD9786
NOTES
© 2005 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D03152-0-10/05(B)
Rev. B | Page 56 of 56
工商网监
湘ICP备2023018690号
