14-Bit, 500 MSPS LVDS,
Dual Analog-to-Digital Converter
AD9684
Data Sheet
FEATURES
FUNCTIONAL BLOCK DIAGRAM
APPLICATIONS
Communications
Diversity multiband, multimode digital receivers
3G/4G, TD-SCDMA, W-CDMA, MC-GSM, LTE
General-purpose software radios
Ultrawideband satellite receiver
Instrumentation (spectrum analyzers, network analyzers,
integrated RF test solutions)
Radar
Digital oscilloscopes
High speed data acquisition systems
DOCSIS CMTS upstream receiver paths
HFC digital reverse path receivers
Rev. 0
AVDD1
(1.25V)
AVDD2
(2.5V)
AVDD3
(3.3V)
DRVDD
(1.25V)
DVDD
(1.25V)
SPIVDD
(1.8V TO 3.4V)
BUFFER
VIN+A
SIGNAL
MONITOR
FD_B
DIGITAL
DOWNCONVERTER
BUFFER
VIN+B
VIN–B
ADC
CORE
D0±
D1±
D2±
D3±
D4±
D5±
D6±
D7±
D8±
D9±
D10±
D11±
D12±
D13±
DCO±
STATUS±
14
CONTROL
REGISTERS
FAST
DETECT
V_1P0
SIGNAL MONITOR
CLOCK
GENERATION
CLK+
CLK–
16
LVDS/SYNC
CONTROL
FD_A
LVDS OUTPUT
STAGING
DIGITAL
DOWNCONVERTER
LVDS
OUTPUTS
14
ADC
CORE
VIN–A
FAST
DETECT
SYNC+
SYNC–
SPI CONTROL
÷2
÷4
AD9684
PDWN/
STBY
÷8
AGND
DRGND
DGND
SDIO SCLK CSB
13015-001
Parallel LVDS (DDR) outputs
1.1 W total power per channel at 500 MSPS (default settings)
SFDR = 85 dBFS at 170 MHz fIN (500 MSPS)
SNR = 68.6 dBFS at 170 MHz fIN (500 MSPS)
ENOB = 10.9 bits at 170 MHz fIN
DNL = ±0.5 LSB
INL = ±2.5 LSB
Noise density = −153 dBFS/Hz at 500 MSPS
1.25 V, 2.50 V, and 3.3 V supply operation
No missing codes
Internal analog-to-digital converter (ADC) voltage reference
Flexible input range and termination impedance
1.46 V p-p to 2.06 V p-p (2.06 V p-p nominal)
400 Ω, 200 Ω, 100 Ω, and 50 Ω differential
SYNC± input allows multichip synchronization
DDR LVDS (ANSI-644 levels) outputs
2 GHz usable analog input full power bandwidth
>96 dB channel isolation/crosstalk
Amplitude detect bits for efficient AGC implementation
Two integrated wideband digital processors per channel
12-bit numerically controlled oscillator (NCO)
3 cascaded half-band filters
Differential clock inputs
Serial port control
Integer clock divide by 2, 4, or 8
Small signal dither
Figure 1.
GENERAL DESCRIPTION
The AD9684 is a dual, 14-bit, 500 MSPS ADC. The device has
an on-chip buffer and a sample-and-hold circuit designed for
low power, small size, and ease of use. This product is designed
for sampling wide bandwidth analog signals. The AD9684 is
optimized for wide input bandwidth, a high sampling rate,
excellent linearity, and low power in a small package.
The dual ADC cores feature a multistage, differential pipelined
architecture with integrated output error correction logic. Each
ADC features wide bandwidth buffered inputs, supporting a
variety of user selectable input ranges. An integrated voltage
reference eases design considerations. Each ADC data output is
internally connected to an optional decimate by 2 block.
The analog input and clock signals are differential inputs. Each
ADC data output is internally connected to two digital
downconverters (DDCs). Each DDC consists of four cascaded
signal processing stages: a 12-bit frequency translator (NCO),
and three half-band decimation filters supporting a divide by
factor of two, four, and eight.
Document Feedback
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
©2015 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
AD9684
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1
DDC I/Q Output Selection ....................................................... 31
Applications ....................................................................................... 1
DDC General Description ........................................................ 31
Functional Block Diagram .............................................................. 1
Frequency Translation ................................................................... 37
General Description ......................................................................... 1
General Description ................................................................... 37
Revision History ............................................................................... 2
DDC NCO Plus Mixer Loss and SFDR ................................... 38
Product Highlights ........................................................................... 3
Numerically Controlled Oscillator .......................................... 38
Specifications..................................................................................... 4
FIR Filters ........................................................................................ 40
DC Specifications ......................................................................... 4
General Description ................................................................... 40
AC Specifications.......................................................................... 5
Half-Band Filters ........................................................................ 41
Digital Specifications ................................................................... 6
DDC Gain Stage ......................................................................... 42
Switching Specifications .............................................................. 7
DDC Complex to Real Conversion Block............................... 42
Timing Specifications .................................................................. 8
DDC Example Configurations ................................................. 43
Absolute Maximum Ratings .......................................................... 16
Digital Outputs ............................................................................... 47
Thermal Characteristics ............................................................ 16
Digital Outputs ........................................................................... 47
ESD Caution ................................................................................ 16
ADC Overrange .......................................................................... 47
Pin Configuration and Function Descriptions ........................... 17
Multichip Synchronization............................................................ 48
Typical Performance Characteristics ........................................... 19
SYNC± Setup and Hold Window Monitor ............................. 49
Equivalent Circuits ......................................................................... 22
Test Modes ....................................................................................... 51
Theory of Operation ...................................................................... 24
ADC Test Modes ........................................................................ 51
ADC Architecture ...................................................................... 24
Serial Port Interface (SPI) .............................................................. 52
Analog Input Considerations.................................................... 24
Configuration Using the SPI ..................................................... 52
Voltage Reference ....................................................................... 26
Hardware Interface ..................................................................... 52
Clock Input Considerations ...................................................... 27
SPI Accessible Features .............................................................. 52
Power-Down/Standby Mode..................................................... 28
Memory Map .................................................................................. 53
Temperature Diode .................................................................... 28
Reading the Memory Map Register Table............................... 53
ADC Overrange and Fast Detect .................................................. 29
Memory Map Register Table ..................................................... 54
ADC Overrange .......................................................................... 29
Applications Information .............................................................. 63
Fast Threshold Detection (FD_A and FD_B) ........................ 29
Power Supply Recommendations............................................. 63
Signal Monitor ................................................................................ 30
Outline Dimensions ....................................................................... 64
Digital Downconverters (DDCs).................................................. 31
Ordering Guide .......................................................................... 64
DDC I/Q Input Selection .......................................................... 31
REVISION HISTORY
5/15—Revision 0: Initial Version
Rev. 0 | Page 2 of 64
Data Sheet
AD9684
The AD9684 has several functions that simplify the automatic
gain control (AGC) function in a communications receiver. The
programmable threshold detector allows monitoring of the
incoming signal power using the fast detect output bits of the
ADC. If the input signal level exceeds the programmable
threshold, the fast detect indicator goes high. Because this
threshold indicator has low latency, the user can quickly reduce
the system gain to avoid an overrange condition at the ADC
input. In addition to the fast detect outputs, the AD9684 also
offers signal monitoring capability. The signal monitoring block
provides additional information about the signal that the ADC
digitized.
The dual ADC output data is routed directly to the one external,
14-bit LVDS output port, supporting double data rate (DDR)
formatting. An external data clock and status bit are offered for
data capture flexibility.
The LVDS outputs have several configurations, depending on
the acceptable rate of the receiving logic device and the sampling
rate of the ADC. Multiple device synchronization is supported
through the SYNC± input pins.
The AD9684 has flexible power-down options that allow
significant power savings when desired. All of these features can
be programmed using a 1.8 V to 3.4 V capable 3-wire serial port
interface (SPI).
The AD9684 is available in a Pb-free, 196-ball ball grid array
(BGA) and is specified over the −40°C to +85°C industrial
temperature range. This product is protected by a U.S. patent.
PRODUCT HIGHLIGHTS
1.
2.
3.
4.
5.
6.
7.
Rev. 0 | Page 3 of 64
Wide full power bandwidth supports intermediate
frequency (IF) sampling of signals up to 2 GHz.
Buffered inputs with programmable input termination ease
filter design and implementation.
Four integrated wideband decimation filters and NCO
blocks supporting multiband receivers.
Flexible SPI controls various product features and functions
to meet specific system requirements.
Programmable fast overrange detection and signal
monitoring.
SYNC± input allows synchronization of multiple devices.
12 mm × 12 mm, 196-ball BGA_ED.
AD9684
Data Sheet
SPECIFICATIONS
DC SPECIFICATIONS
AVDD1 = 1.25 V, AVDD2 = 2.5 V, AVDD3 = 3.3 V, DVDD = 1.25 V, DRVDD = 1.25 V, SPIVDD = 1.8 V, specified maximum sampling
rate (500 MSPS), 1.7 V p-p full-scale differential input, 1.0 V internal reference, AIN = −1.0 dBFS, default SPI settings, TA = 25°C, unless
otherwise noted.
Table 1.
Parameter
RESOLUTION
ACCURACY
No Missing Codes
Offset Error
Offset Matching
Gain Error
Gain Matching
Differential Nonlinearity (DNL)
Integral Nonlinearity (INL)
TEMPERATURE DRIFT
Offset Error
Gain Error
INTERNAL VOLTAGE REFERENCE
INPUT-REFERRED NOISE
VREF = 1.0 V
ANALOG INPUTS
Differential Input Voltage Range (Programmable)
Common-Mode Voltage (VCM)
Differential Input Capacitance 1
Analog Input Full Power Bandwidth
POWER SUPPLY
AVDD1
AVDD2
AVDD3
DVDD
DRVDD
SPIVDD
IAVDD1
IAVDD2
IAVDD3
IDVDD
IDRVDD
ISPIVDD
POWER CONSUMPTION
Total Power Dissipation 2
Power-Down Dissipation
Standby
1
2
Temperature
Full
Full
Full
Full
Full
Full
Full
Full
Min
14
−0.3
−6.5
−0.6
−4.5
Typ
Max
Unit
Bits
Guaranteed
0
0
0
0
±0.5
±2.5
+0.3
+0.3
+6.5
+5.0
+0.7
+5.0
% FSR
% FSR
% FSR
% FSR
LSB
LSB
25°C
25°C
Full
±3
−39
1.0
ppm/°C
ppm/°C
V
25°C
2.63
LSB rms
Full
25°C
25°C
25°C
1.46
2.06
2.05
1.5
2
2.06
V p-p
V
pF
GHz
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
1.22
2.44
3.2
1.22
1.22
1.22
1.25
2.50
3.3
1.25
1.25
1.8
448
396
103
108
106
2
1.28
2.56
3.4
1.28
1.28
3.4
503
455
124
127
119
6
V
V
V
V
V
V
mA
mA
mA
mA
mA
mA
Full
Full
Full
Differential capacitance is measured between the VIN+x and VIN−x pins (x = A or B).
Parallel interleaved LVDS mode. The power dissipation on DRVDD changes with the output data mode used.
Rev. 0 | Page 4 of 64
2.2
710
1.0
W
mW
W
Data Sheet
AD9684
AC SPECIFICATIONS
AVDD1 = 1.25 V, AVDD2 = 2.5 V, AVDD3 = 3.3 V, DVDD = 1.25 V, DRVDD = 1.25 V, SPIVDD = 1.8 V, specified maximum sampling
rate (500 MSPS), 1.7 V p-p full-scale differential input, 1.0 V internal reference, AIN = −1.0 dBFS, default SPI settings, TA = 25°C, unless
otherwise noted.
Table 2.
Parameter 1
ANALOG INPUT FULL SCALE
NOISE DENSITY 2
SIGNAL-TO-NOISE RATIO (SNR) 3
fIN = 10 MHz
fIN = 170 MHz
fIN = 340 MHz
fIN = 450 MHz
fIN = 765 MHz
fIN = 985 MHz
fIN = 1950 MHz
SIGNAL-TO-NOISE RATIO AND DISTORTION RATIO (SINAD)3
fIN = 10 MHz
fIN = 170 MHz
fIN = 340 MHz
fIN = 450 MHz
fIN = 765 MHz
fIN = 985 MHz
fIN = 1950 MHz
EFFECTIVE NUMBER OF BITS (ENOB)
fIN = 10 MHz
fIN = 170 MHz
fIN = 340 MHz
fIN = 450 MHz
fIN = 765 MHz
fIN = 985 MHz
fIN = 1950 MHz
SPURIOUS-FREE DYNAMIC RANGE (SFDR)3
fIN = 10 MHz
fIN = 170 MHz
fIN = 340 MHz
fIN = 450 MHz
fIN = 765 MHz
fIN = 985 MHz
fIN = 1950 MHz
WORST HARMONIC, SECOND OR THIRD3
fIN = 10 MHz
fIN = 170 MHz
fIN = 340 MHz
fIN = 450 MHz
fIN = 765 MHz
fIN = 985 MHz
fIN = 1950 MHz
Temperature
Full
Full
25°C
Full
25°C
25°C
25°C
25°C
25°C
25°C
Full
25°C
25°C
25°C
25°C
25°C
25°C
Full
25°C
25°C
25°C
25°C
25°C
25°C
Full
25°C
25°C
25°C
25°C
25°C
25°C
Full
25°C
25°C
25°C
25°C
25°C
Rev. 0 | Page 5 of 64
Min
67.5
67
10.8
76
Typ
2.06
−153
Max
Unit
V p-p
dBFS/Hz
69.2
68.6
68.4
68.0
64.4
63.8
60.5
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
68.7
68.5
67.6
67.2
63.8
62.5
58.3
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
11.1
10.9
10.8
10.8
10.3
10.1
9.5
Bits
Bits
Bits
Bits
Bits
Bits
Bits
83
85
82
86
81
76
69
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
−83
−85
−82
−86
−81
−76
−69
−76
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
AD9684
Data Sheet
Parameter 1
WORST OTHER, EXCLUDING SECOND OR THIRD HARMONIC3
fIN = 10 MHz
fIN = 170 MHz
fIN = 340 MHz
fIN = 450 MHz
fIN = 765 MHz
fIN = 985 MHz
fIN = 1950 MHz
TWO-TONE INTERMODULATION DISTORTION (IMD), AIN1 AND AIN2 = −7 dBFS
fIN1 = 185 MHz, fIN2 = 188 MHz
fIN1 = 338 MHz, fIN2 = 341 MHz
CROSSTALK 4
FULL POWER BANDWIDTH
Temperature
Min
Typ
25°C
Full
25°C
25°C
25°C
25°C
25°C
−93
−92
−90
−92
−89
−89
−85
25°C
25°C
25°C
25°C
−88
−87
96
2
Max
−76
Unit
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dBFS
dB
GHz
See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.
Noise density is measured at a low analog input frequency (30 MHz).
See Table 9 for the recommended settings for full-scale voltage and buffer current control.
4
Crosstalk is measured at 170 MHz with a −1.0 dBFS analog input on one channel and no input on the adjacent channel.
1
2
3
DIGITAL SPECIFICATIONS
AVDD1 = 1.25 V, AVDD2 = 2.5 V, AVDD3 = 3.3 V, DVDD = 1.25 V, DRVDD = 1.25 V, SPIVDD = 1.8 V, specified maximum sampling
rate (500 MSPS), 1.7 V p-p full-scale differential input, 1.0 V internal reference, AIN = −1.0 dBFS, default SPI settings, TA = 25°C, unless
otherwise noted.
Table 3.
Parameter
CLOCK INPUTS (CLK+, CLK−)
Logic Compliance
Differential Input Voltage
Input Common-Mode Voltage
Input Resistance (Differential)
Input Capacitance
SYNC INPUTS (SYNC+, SYNC−)
Logic Compliance
Differential Input Voltage
Input Common-Mode Voltage
Input Resistance (Differential)
Input Capacitance (Differential)
LOGIC INPUTS (SDIO, SCLK, CSB, PDWN/STBY)
Logic Compliance
Logic 1 Voltage
Logic 0 Voltage
Input Resistance
LOGIC OUTPUT (SDIO)
Logic Compliance
Logic 1 Voltage (IOH = 800 µA)
Logic 0 Voltage (IOL = 50 µA)
LOGIC OUTPUTS (FD_A, FD_B)
Logic Compliance
Logic 1 Voltage
Logic 0 Voltage
Input Resistance
Temperature
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Min
600
Rev. 0 | Page 6 of 64
LVDS/LVPECL
1200
0.85
35
Max
Unit
1800
mV p-p
V
kΩ
pF
2.5
400
0.6
LVDS/LVPECL
1200
0.85
35
1800
2.0
2.5
0
Full
Full
Full
Full
Full
Full
Full
Typ
0.8
0
mV p-p
V
kΩ
pF
CMOS
0.8 × SPIVDD
0.2 × SPIVDD
30
V
V
kΩ
CMOS
0.8 × SPIVDD
0.2 × SPIVDD
V
V
CMOS
SPIVDD
0
30
V
V
kΩ
Data Sheet
Parameter
DIGITAL OUTPUTS (Dx±, 1 DCO±, STATUS±)
Logic Compliance
Differential Output Voltage
Output Common-Mode Voltage (VCM)
AC-Coupled
Short-Circuit Current (IDSHORT)
Differential Return Loss (RLDIFF) 2
Common-Mode Return Loss (RLCM) 2
Differential Termination Impedance
1
2
AD9684
Temperature
Min
Full
Full
25°C
25°C
25°C
25°C
Full
Typ
Max
Unit
230
430
mV p-p
0
−100
8
6
80
1.8
+100
V
mA
dB
dB
Ω
LVDS
100
120
Where x = 0 to 13.
Differential and common-mode return loss is measured from 100 MHz to 0.75 MHz × baud rate.
SWITCHING SPECIFICATIONS
AVDD1 = 1.25 V, AVDD2 = 2.5 V, AVDD3 = 3.3 V, DVDD = 1.25 V, DRVDD = 1.25 V, SPIVDD = 1.8 V, specified maximum sampling
rate, 1.7 V p-p full-scale differential input, 1.0 V internal reference, AIN = −1.0 dBFS, default SPI settings, TA = 25°C, unless otherwise
noted.
Table 4.
Parameter
CLOCK
Clock Rate (at CLK+/CLK− Pins)
Maximum Sample Rate 1
Minimum Sample Rate 2
Clock Pulse Width
High
Low
LVDS DATA OUTPUT PARAMETERS
Data Propagation Delay (tPD) 3
DCO± Propagation Delay (tDCO)3
DCO± to Data Skew
Rising Edge Data (tSKEWR)3
Falling Edge Data (tSKEWF)3
STATUS± Propagation Delay (tSTATUS) 4
DCO± to STATUS± Skew (tFRAME)4
Data Propagation Delay (tPD)3
DCO± Propagation Delay (tDCO)3
LATENCY 5
Pipeline Latency
Fast Detect Latency
HB1 Filter Latency3
HB1 + HB2 Filter Latency3
HB1 + HB2 + HB3 Filter Latency3
HB1 + HB2 + HB3 + HB4 Filter Latency3
Fast Detect Latency
Wake-Up Time 6
Standby
Power-Down
Temperature
Min
Full
Full
Full
0.25
500
250
Full
Full
1000
1000
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
25°C
25°C
Rev. 0 | Page 7 of 64
Typ
Max
Unit
4
GHz
MSPS
MSPS
ps
ps
2.225
2.2
−150
850
−150
−25
1.025
2.2
−25
2.225
2.2
ns
ns
+100
1100
+100
35
28
50
101
217
433
28
1
4
ps
ps
ns
ps
ns
ns
Clock cycles
Clock cycles
Clock cycles
Clock cycles
Clock cycles
Clock cycles
Clock cycles
ms
ms
AD9684
Data Sheet
Parameter
APERTURE
Aperture Delay (tA)
Aperture Uncertainty (Jitter, tj)
Out of Range Recovery Time
Temperature
Min
Typ
Full
Full
Full
Max
Unit
530
55
1
ps
fs rms
Clock Cycles
The maximum sample rate is the clock rate after the divider.
The minimum sample rate operates at 300 MSPS.
This specification is valid for parallel interleaved, channel multiplexed, and byte mode output modes.
4
This specification is valid for byte mode output mode only.
5
No DDCs used.
6
Wake-up time is defined as the time required to return to normal operation from power-down mode or standby mode.
1
2
3
TIMING SPECIFICATIONS
Table 5.
Parameter
CLK± to SYNC± TIMING REQUIREMENTS
tSU_SR
tH_SR
SPI TIMING REQUIREMENTS
tDS
tDH
tCLK
tS
tH
tHIGH
tLOW
tEN_SDIO
Description
See Figure 2
Device clock to SYNC± setup time
Device clock to SYNC± hold time
See Figure 3
Setup time between the data and the rising edge of SCLK
Hold time between the data and the rising edge of SCLK
Period of the SCLK
Setup time between CSB and SCLK
Hold time between CSB and SCLK
Minimum period that SCLK must be in a logic high state
Minimum period that SCLK must be in a logic low state
Time required for the SDIO pin to switch from an input to an
output relative to the SCLK falling edge (not shown in Figure 3)
Time required for the SDIO pin to switch from an output to an
input relative to the SCLK rising edge (not shown in Figure 3)
tDIS_SDIO
Min
Typ
Max
117
−96
Unit
ps
ps
2
2
40
2
2
10
10
10
ns
ns
ns
ns
ns
ns
ns
ns
10
ns
Timing Diagrams
CLK–
CLK+
tSU_SR
tH_SR
13015-002
SYNC–
SYNC+
Figure 2. SYNC± Setup and Hold Timing
tHIGH
tDS
tS
tCLK
tDH
tH
tLOW
CSB
SCLK DON’T CARE
R/W
A14
A13
A12
A11
A10
A9
A8
A7
D5
Figure 3. Serial Port Interface Timing Diagram
Rev. 0 | Page 8 of 64
D4
D3
D2
D1
D0
DON’T CARE
13015-003
SDIO DON’T CARE
DON’T CARE
Data Sheet
AD9684
APERTURE DELAY
N + 35
N
VIN±x
N + 39
N + 36
N + 40
N–1
N+x
N + 37
N+y
N + 38
N + 41
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF
THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
SYNC+
SYNC–
CLK+
CLK–
tCLK
FIXED DELAY FROM SYNC EVENT TO DCO KNOWN PHASE
CONSTANT LATENCY = X CLK CYCLES
tDCO
tPD
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
90° PHASE ADJUST 1
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
270° PHASE ADJUST 2
tSKEWF
tSKEWR
CONVERTER 0 CONVERTER 0 CONVERTER 0 CONVERTER 0 CONVERTER 0
SAMPLE
SAMPLE
SAMPLE
SAMPLE
SAMPLE
[N]
[N + 1]
[N + 2]
[N + 3]
[N + 4]
STATUS+
(OVERRANGE/STATUS BIT)
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
D13±
D13
D13
D13
D13
D13
D13
D13
D0±
D0
D0
D0
D0
D0
D0
D0
190° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
2270° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
Figure 4. Parallel Interleaved Mode—One Converter, ≤14-Bit Data
Rev. 0 | Page 9 of 64
13015-004
STATUS–
AD9684
Data Sheet
APERTURE DELAY
N + 36
N
VIN±x
N + 38
N+x
N + 37
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF
THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
SYNC+
SYNC–
CLK+
CLK–
CONSTANT LATENCY = X CLK CYCLES
tDCO
tCLK
tPD
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
tSKEWR
STATUS BIT SELECTED BY
REGISTER 0x559, BITS[2:0]
IN THE REGISTER MAP
STATUS+
(OVERRANGE/STATUS BIT)
tSKEWF
CONVERTER 0
SAMPLE
[N]
CONVERTER 1
SAMPLE
[N]
CONVERTER 0
SAMPLE
[N + 1]
CONVERTER 1
SAMPLE
[N + 1]
CONVERTER 0
SAMPLE
[N + 2]
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
D13±
D13
D13
D13
D13
D13
D13
D13
D13
D0±
D0
D0
D0
D0
D0
D0
D0
D0
Figure 5. Parallel Interleaved Mode—Two Converters, ≤14-Bit Data, Output Sample Rate < 625 MSPS
Rev. 0 | Page 10 of 64
13015-005
STATUS–
Data Sheet
AD9684
APERTURE DELAY
N + 36
N
VIN±x
N + 38
N+x
N + 37
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF
THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
SYNC+
SYNC–
CLK+
CLK–
CONSTANT LATENCY = X CLK CYCLES
tDCO
tPD
tCLK
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
tSKEWR
STATUS BIT SELECTED BY
REGISTER 0x559, BITS[2:0]
IN THE REGISTER MAP
STATUS+
(OVERRANGE/STAUS BIT)
STATUS–
STATUS
STATUS
S[N – y]
(ODD BITS)
S[N – x]
(EVEN BITS)
STATUS
tSKEWF
CONVERTERS
SAMPLE
[N]
CONVERTERS
SAMPLE
[N]
CONVERTERS
SAMPLE
[N + 1]
CONVERTERS
SAMPLE
[N + 1]
CONVERTERS
SAMPLE
[N + 2]
STATUS
STATUS
STATUS
STATUS
STATUS
S[N]
(ODD BITS)
S[N + 1]
(EVEN BITS)
S[N + 1]
(ODD BITS)
S[N + 2]
(EVEN BITS)
S[N]
S[N – 1]
(ODD BITS) (EVEN BITS)
CHANNEL A D0±/D1±
Figure 6. Channel Multiplexed (Even/Odd) Mode—One Converter, ≤14-Bit Data
Rev. 0 | Page 11 of 64
13015-006
CHANNEL A D12±/D13±
AD9684
Data Sheet
APERTURE DELAY
N + 36
N
VIN±x
N + 38
N+x
N + 37
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF
THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
SYNC+
SYNC–
CLK+
CLK–
tCLK
CONSTANT LATENCY = X CLK CYCLES
tDCO
tPD
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
tSKEWR
STATUS BIT SELECTED BY
REGISTER 0x559, BITS[2:0]
IN THE REGISTER MAP
STATUS+
(OVERRANGE/STATUS BIT)
STATUS–
STATUS
tSKEWF
CONVERTERS
SAMPLE
[N]
CONVERTERS
SAMPLE
[N]
CONVERTERS
SAMPLE
[N + 1]
CONVERTERS
SAMPLE
[N + 1]
CONVERTERS
SAMPLE
[N + 2]
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
STATUS
S[N – y]
(ODD BITS)
S[N – x]
(EVEN BITS)
S[N]
S[N – 1]
(ODD BITS) (EVEN BITS)
S[N]
(ODD BITS)
S[N + 1]
(EVEN BITS)
S[N + 1]
(ODD BITS)
S[N + 2]
(EVEN BITS)
S[N – y]
(ODD BITS)
S[N – x]
(EVEN BITS)
S[N]
S[N – 1]
(ODD BITS) (EVEN BITS)
S[N]
(ODD BITS)
S[N + 1]
(EVEN BITS)
S[N + 1]
(ODD BITS)
S[N + 2]
(EVEN BITS)
CHANNEL A D12±/D13±
CHANNEL B D12±/D13±
CHANNEL B D0±/D1±
Figure 7. Channel Multiplexed (Even/Odd) Mode—Two Converters, ≤14-Bit Data, Output Sample Rate < 625 MSPS
Rev. 0 | Page 12 of 64
13015-007
CHANNEL A D0±/D1±
Data Sheet
AD9684
APERTURE DELAY
N+z
N
VIN±x
N–1
N + 39
N + 36
N + 40
N + 42
N+x
N + 37
N + 38
N+y
N + 41
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF
THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
SYNC+
SYNC–
CLK+
CLK–
tCLK
FIXED DELAY FROM SYNC EVENT
TO DCO KNOWN PHASE
CONSTANT LATENCY = X CLK CYCLES
tDCO
tPD
tSTATUS
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
90° PHASE ADJUST 1
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
270° PHASE ADJUST 2
tFRAME
STATUS–
(FRAME CLOCK OUTPUT)3
FRAME 0
STAUS+
STATUS+
(OVERRANGE STATUS BIT)
FRAME 1
tSKEWF
tSKEWR
I0[N]
EVEN
I0[N]
ODD
Q 0[N]
EVEN
Q 0[N]
ODD
I0[N + 1]
EVEN
I0[N + 1] Q 0[N + 1] Q 0[N + 1]
ODD
EVEN
ODD
PAR 4
PAR
STATUS
PAR
STATUS
PAR
STATUS
PAR
STATUS
PAR
D7±
D15
D15
D14
D15
D14
D15
D14
D15
D14
D15
D0±
D1
D1
D0
D1
D0
D1
D0
D1
D0
D1
STATUS–
1) ENABLED (ALWAYS ON).
2) DISABLED (ALWAYS OFF).
3) GAPPED PERIODIC (CONDITIONALLY ENABLED BASED ON PSEUDO-RANDOM BIT).
4STATUS BIT SELECTED BY REGISTER 0x559, BITS[2:0] IN THE REGISTER MAP.
Figure 8. LVDS Byte Mode—Two Virtual Converters, One DDC, I/Q Data Decimate by 4
Rev. 0 | Page 13 of 64
13015-008
190° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
2270° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
3FRAME CLOCK OUTPUT SUPPORTS 3 MODES OF OPERATION:
Rev. 0 | Page 14 of 64
D1
D0±
N+y
N+z
N + 36
N + 37
N + 38
N + 39
N + 40
N + 41
N + 42
tPD
tSTATUS
Figure 9. LVDS Byte Mode—Four Virtual Converters, Two DDCs, ≤16-Bit Data, I/Q Data Decimate by 8
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
Q0[N]
EVEN
tSKEWF
I0[N]
ODD
D1
D15
PAR
Q0[N]
ODD
D0
D14
STATUS
I1[N + 1]
EVEN
FRAME 0
FRAME 0
D1
D15
PAR
I1[N]
ODD
D0
D14
STATUS
Q1[N]
EVEN
FIXED DELAY FROM SYNC EVENT TO DCO KNOWN PHASE
I0[N]
EVEN
1) ENABLED (ALWAYS ON).
2) DISABLED (ALWAYS OFF).
3) GAPPED PERIODIC (CONDITIONALLY ENABLED BASED ON PSEUDO-RANDOM BIT).
4STATUS BIT SELECTED BY REGISTER 0x559, BITS[2:0] IN THE REGISTER MAP.
D1
D15
PAR
tSKEWR
tFRAME
CONSTANT LATENCY =
X CLK CYCLES
tDCO
D1
D15
PAR
Q1[N]
ODD
D0
D14
STATUS
I0[N+1]
EVEN
D1
D15
PAR
I0[N+1]
ODD
D0
D14
STATUS
Q0[N+1]
EVEN
D1
D15
PAR
Q0[N+1]
ODD
D0
D14
STATUS
I1[N + 1]
EVEN
D1
D15
PAR
I1[N+1]
ODD
FRAME 1
FRAME 1
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
N+x
APERTURE DELAY
190° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
2270° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
3FRAME CLOCK OUTPUT SUPPORTS 3 MODES OF OPERATION:
D15
PAR4
tCLK
N–1
D7±
STATUS–
STATUS+
(OVERRANGE STATUS BIT)
STATUS+
STATUS–
(FRAME CLOCK OUTPUT)3
DCO± (DATA CLOCK OUTPUT)
270° PHASE ADJUST 2
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
90° PHASE ADJUST 1
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
CLK–
CLK+
SYNC–
SYNC+
VIN±x
N
D0
D14
STATUS
Q1[N+1]
EVEN
D1
D15
PAR
Q1[N+1]
ODD
AD9684
Data Sheet
13015-010
Rev. 0 | Page 15 of 64
D1
D0±
N+2
N+3
N+4
N+5
N+6
N+7
N+8
N + 10
D1
D15
D14
D0
PAR
STATUS
CONSTANT LATENCY =
X CLK CYCLES
D0
D14
STATUS
I0[N]
EVEN
D1
D15
PAR
I0[N]
ODD
D0
D14
STATUS
Q0[N]
EVEN
D1
D15
PAR
D0
D14
STATUS
Q0[N] I1[N]
ODD EVEN
D1
D15
PAR
I1[N]
ODD
FIXED DELAY FROM SYNC EVENT
TO DCO KNOWN PHASE
D0
D14
STATUS
Q1[N]
EVEN
CLOCK OUTPUT SUPPORTS 3 MODES OF OPERATION:
1) ENABLED (ALWAYS ON).
2) DISABLED (ALWAYS OFF).
3) GAPPED PERIODIC (CONDITIONALLY ENABLED BASED ON PSEUDO-RANDOM BIT).
4STATUS BIT SELECTED BY REGISTER 0x559, BITS[2:0] IN THE REGISTER MAP.
3FRAME
N+9
D1
D15
PAR
D0
D14
STATUS
Q1[N] I2[N]
ODD EVEN
FRAME 0
D1
D15
PAR
I2[N]
ODD
D0
D14
STATUS
Q2[N]
EVEN
D1
D15
PAR
D0
D14
STATUS
Q2[N] I3[N]
ODD EVEN
D1
D15
PAR
I0[N]
ODD
D0
D14
STATUS
Q3[N]
EVEN
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
D0
D14
STATUS
D1
D15
PAR
Q3[N] I0[N+1] I0[N+1] Q0[N+1]Q0[N+1] I1[N+1] I1[N+1] Q1[N+1]Q1[N+1] I2[N+1] I2[N+1] Q2[N+1]Q2[N+1] I3[N+1] I3[N+1] Q3[N+1]Q3[N+1]
ODD EVEN ODD EVEN ODD EVEN ODD EVEN ODD EVEN ODD EVEN ODD EVEN ODD EVEN ODD
FRAME 1
SYNCHRONOUS LOW TO HIGH TRANSITIONS OF THE SYNC SIGNAL CAPTURED ON THE RISING EDGE OF THE CLK SIGNAL CAUSES THE DCO INTERNAL DIVIDER TO BE RESET
N+1
APERTURE DELAY
190° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
2270° PHASE ADJUST IS GENERATED USING THE FALLING EDGE OF CLK±.
D15
PAR4
D7±
STATUS+
(STATUS BIT)
STATUS–
STATUS+
(FRAME CLOCK OUTPUT)3
DCO± (DATA CLOCK OUTPUT)
270° PHASE ADJUST2
DCO± (DATA CLOCK OUTPUT)
180° PHASE ADJUST
DCO± (DATA CLOCK OUTPUT)
90° PHASE ADJUST1
DCO± (DATA CLOCK OUTPUT)
0° PHASE ADJUST
CLK+
(ENCODE CLOCK)
SYNC+
(SYSTEM REFERENCE)
VIN±x
N
13015-011
N–1
Data Sheet
AD9684
Figure 10. LVDS Byte Mode—Eight Virtual Converters, Four DDCs, ≤16-Bit Data, I/Q Data Decimate by 16
AD9684
Data Sheet
ABSOLUTE MAXIMUM RATINGS
THERMAL CHARACTERISTICS
Table 6.
Parameter
Electrical
AVDD1 to AGND
AVDD2 to AGND
AVDD3 to AGND
DVDD to DGND
DRVDD to DRGND
SPIVDD to AGND
AGND to DRGND
VIN±x to AGND
SCLK, SDIO, CSB to AGND
VIN±x Maximum Swing
PDWN/STBY to AGND
Environmental
Operating Temperature Range
(TCASE)
Maximum Junction Temperature
Storage Temperature Range
(Ambient)
Rating
Typical θJA, θJB, and θJC are specified vs. the number of printed
circuit board (PCB) layers in different airflow velocities (in
m/sec). Airflow increases heat dissipation effectively reducing
θJA and θJB. The use of appropriate thermal management
techniques is recommended to ensure that the maximum
junction temperature does not exceed the limits shown in Table 7.
1.32 V
2.75 V
3.63 V
1.32 V
1.32 V
3.63 V
−0.3 V to +0.3 V
3.2 V
−0.3 V to SPIVDD + 0.3 V
4.3 V p-p
−0.3 V to SPIVDD + 0.3 V
PCB
Type
JEDEC
2s2p
Board
−40°C to +85°C
10-Layer
PCB
125°C
−65°C to +150°C
Table 7. Simulated Thermal Data
Airflow
Velocity
(m/sec)
0.0
1.0
2.5
0.0
1.0
2.5
θJB
6.31, 3
5.91, 3
5.71, 3
4.6
4.6
4.6
θJC_TOP
4.71, 5
N/A4
N/A4
4.7
N/A4
N/A4
θJC_BOT
1.21, 5
N/A4
N/A4
1.2
N/A4
N/A4
Per JEDEC 51-7, plus JEDEC 51-5 2s2p test board.
Per JEDEC JESD51-2 (still air) or JEDEC JESD51-6 (moving air).
3
Per JEDEC JESD51-8 (still air).
4
N/A means not applicable.
5
Per MIL-STD 883, Method 1012.1.
1
2
Stresses at or above those listed under Absolute Maximum
Ratings may cause permanent damage to the product. This is a
stress rating only; functional operation of the product at these
or any other conditions above those indicated in the operational
section of this specification is not implied. Operation beyond
the maximum operating conditions for extended periods may
affect product reliability.
θJA
17.81, 2
15.61, 2
15.01, 2
13.8
12.7
12.0
ESD CAUTION
Rev. 0 | Page 16 of 64
Unit
°C/W
°C/W
°C/W
°C/W
°C/W
°C/W
Data Sheet
AD9684
1
2
3
4
5
6
7
8
9
10
11
12
13
14
A
AGND
AGND
AGND
AVDD2
AVDD1
AGND
CLK+
CLK–
AGND
AVDD1
AVDD2
AGND
AGND
AGND
A
B
AVDD3
AGND
AGND
AVDD2
AVDD1
AGND
AGND
AGND
AGND
AVDD1
AVDD2
AGND
AGND
AVDD3
B
C
AVDD3
AGND
AGND
AVDD2
AVDD1
AGND
SYNC+
SYNC–
AGND
AVDD1
AVDD2
AGND
AGND
AVDD3
C
D
AGND
AGND
AGND
AVDD2
AVDD1
AGND
AVDD1
AGND
AGND
AVDD1
AVDD2
AGND
AGND
AGND
D
E
VIN–B
AGND
AGND
AVDD2
AVDD1
AGND
AGND
AGND
AGND
AVDD1
AVDD2
AGND
AGND
VIN–A
E
F
VIN+B
AGND
AGND
AVDD2
AGND
AGND
AGND
AGND
AGND
AGND
AVDD2
AGND
AGND
VIN+A
F
G
AGND
AGND
AGND
AGND
AGND
AGND
AGND
AGND
AGND
AGND
AVDD2
AGND
AGND
AGND
G
H
AGND
AGND
AGND
CSB
AGND
AGND
AGND
AGND
AGND
V_1P0
AGND
AGND
AGND
AGND
H
J
FD_B
AGND
AGND
SCLK
SPIVDD
AGND
AGND
AGND
AGND
AVDD2
SPIVDD
AGND
PDWN/STBY
FD_A
J
K
DGND
DGND
AGND
SDIO
AGND
AGND
AGND
AGND
AGND
AGND
AGND
AGND
DCO–
DCO+
K
L
DVDD
DVDD
DGND
DGND
AGND
AGND
AGND
AGND
AGND
AGND
AGND
AGND
STATUS–
STATUS+
L
M
D1+
D1–
DVDD
DVDD
DRVDD
DRVDD
DRVDD
DRGND
DRGND
DRGND
DRGND
DRGND
D13–
D13+
M
N
D2–
D3–
D4–
D5–
D6–
D0–
DRVDD
DRGND
D7–
D8–
D9–
D10–
D11–
D12–
N
P
D2+
D3+
D4+
D5+
D6+
D0+
DRVDD
DRGND
D7+
D8+
D9+
D10+
D11+
D12+
P
1
2
3
4
5
7
8
9
10
11
12
13
14
6
13015-012
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
Figure 11. Pin Configuration (Top View)
Table 8. Pin Function Descriptions
Pin No.
Power Supplies
A5, A10, B5, B10, C5, C10, D5, D7,
D10, E5, E10
A4, A11, B4, B11, C4, C11, D4,
D11, E4, E11, F4, F11, G11, J10
B1, B14, C1, C14
L1, L2, M3, M4
M5, M6, M7, N7, P7
J5, J11
K1, K2, L3, L4
M8 to M12, N8, P8
A1, A2, A3, A6, A9, A12, A13, A14,
B2, B3, B6, B7, B8, B9, B12, B13,
C2, C3, C6, C9, C12, C13, D1,
D2, D3, D6, D8, D9, D12, D13,
D14, E2, E3, E6 to E9, E12, E13,
F2, F3, F5 to F10, F12, F13, G1
to G10, G12, G13, G14, H1, H2,
H3, H5 to H9, H11 to H14, J2,
J3, J6 to J9, J12, K3, K5 to K12,
L5 to L12
Analog
E14, F14
E1, F1
H10
A7, A8
Mnemonic
Type
Description
AVDD1
Supply
Analog Power Supply (1.25 V Nominal).
AVDD2
Supply
Analog Power Supply (2.50 V Nominal).
AVDD3
DVDD
DRVDD
SPIVDD
DGND
DRGND
AGND
Supply
Supply
Supply
Supply
Ground
Ground
Ground
Analog Power Supply (3.3 V Nominal)
Digital Power Supply (1.25 V Nominal).
Digital Driver Power Supply (1.25 V Nominal).
Digital Power Supply for SPI (1.8 V to 3.4 V).
Ground Reference for DVDD.
Ground Reference for DRVDD.
Ground Reference for AVDD.
VIN−A, VIN+A
VIN−B, VIN+B
V_1P0
Input
Input
Input/DNC
CLK+, CLK−
Input
ADC A Analog Input Complement/True.
ADC B Analog Input Complement/True.
1.0 V Reference Voltage Input/Do Not Connect. This pin is
configurable through the SPI as a no connect or as an input.
Do not connect this pin if using the internal reference. This pin
requires a 1.0 V reference voltage input if using an external
voltage reference source.
Clock Input True/Complement.
Rev. 0 | Page 17 of 64
AD9684
Pin No.
CMOS Outputs
J14, J1
Digital Inputs
C7, C8
Data Outputs
N6, P6
M1, M2
N1, P1
N2, P2
N3, P3
N4, P4
N5, P5
N9, P9
N10, P10
N11, P11
N12, P12
N13, P13
N14, P14
M13, M14
L13, L14
K13, K14
SPI Controls
K4
J4
H4
J13
Data Sheet
Mnemonic
Type
Description
FD_A, FD_B
Output
Fast Detect Outputs for Channel A and Channel B.
SYNC+, SYNC−
Input
Active High LVDS SYNC Input—True/Complement.
D0−, D0+
D1+, D1−
D2−, D2+
D3−, D3+
D4−, D4+
D5−, D5+
D6−, D6+
D7−, D7+
D8−, D8+
D9−, D9+
D10−, D10+
D11−, D11+
D12−, D12+
D13−, D13+
STATUS−, STATUS+
DCO−, DCO+
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
Output
LVDS Lane 0 Output Data—Complement/True.
LVDS Lane 1 Output Data—True/Complement.
LVDS Lane 2 Output Data—Complement/True.
LVDS Lane 3 Output Data—Complement/True.
LVDS Lane 4 Output Data—Complement/True.
LVDS Lane 5 Output Data—Complement/True.
LVDS Lane 6 Output Data—Complement/True.
LVDS Lane 7 Output Data—Complement/True.
LVDS Lane 8 Output Data—Complement/True.
LVDS Lane 9 Output Data—Complement/True.
LVDS Lane 10 Output Data—Complement/True.
LVDS Lane 11 Output Data—Complement/True.
LVDS Lane 12 Output Data—Complement/True.
LVDS Lane 13 Output Data—Complement/True.
LVDS Status Output Data—Complement/True.
LVDS Digital Clock Output Data—Complement/True.
SDIO
SCLK
CSB
PDWN/STBY
Input/output
Input
Input
Input
SPI Serial Data Input/Output.
SPI Serial Clock.
SPI Chip Select (Active Low).
Power-Down Input (Active High). The operation of this pin
depends on the SPI mode and can be configured as powerdown or standby.
Rev. 0 | Page 18 of 64
Data Sheet
AD9684
TYPICAL PERFORMANCE CHARACTERISTICS
AVDD1 = 1.2 V, AVDD2= 2.5 V, AVDD3 = 3.3 V, DVDD = 1.2 V, DRVDD = 1.2 V, SPIVDD = 1.8 V, sampling rate = 500 MHz, 1.6 V p-p
full-scale differential input, AIN = −1.0 dBFS, default SPI settings, TA = 25°C, 256k FFT sample, unless otherwise noted.
0
0
AIN = −1dBFS
SNR = 68.9dBFS
ENOB = 10.9 BITS
SFDR = 83dBFS
BUFFER CONTROL 1 = 2.0×
–20
–40
–40
AMPLITUDE (dBFS)
–60
–80
–100
–120
–60
–80
–100
0
25
50
75
100
125
150
175
200
225
250
FREQUENCY (MHz)
–140
13015-014
–140
0
50
75
100
125
150
175
200
225
250
FREQUENCY (MHz)
Figure 12. Single Tone FFT with fIN = 10.3 MHz
Figure 15. Single Tone FFT with fIN = 450.3 MHz
0
0
AIN = −1dBFS
SNR = 68.7dBFS
ENOB = 10.9 BITS
SFDR = 84dBFS
BUFFER CONTROL 1 = 2.0×
–20
AIN = −1dBFS
SNR = 63.9dBFS
ENOB = 10.3 BITS
SFDR = 81dBFS
BUFFER CONTROL 1 = 5.0×
–20
–40
–40
AMPLITUDE (dBFS)
AMPLITUDE (dBFS)
25
13015-017
–120
–60
–80
–100
–120
–60
–80
–100
–120
0
25
50
75
100
125
150
175
200
225
250
FREQUENCY (MHz)
–140
13015-015
–140
0
25
50
75
100
125
150
175
200
225
250
225
250
FREQUENCY (MHz)
Figure 13. Single Tone FFT with fIN = 170.3 MHz
13015-018
AMPLITUDE (dBFS)
AIN = −1dBFS
SNR = 67.3dBFS
ENOB = 10.8 BITS
SFDR = 86dBFS
BUFFER CONTROL 1 = 4.5×
–20
Figure 16. Single Tone FFT with fIN = 765.3 MHz
0
AIN = −1dBFS
SNR = 62.8dBFS
ENOB = 10.1 BITS
SFDR = 76dBFS
BUFFER CONTROL 1 = 5.0×
0
AIN = −1dBFS
SNR = 67.8dBFS
ENOB = 10.8 BITS
SFDR = 82dBFS
BUFFER CONTROL 1 = 4.5×
–40
AMPLITUDE (dBFS)
–20
–20
–60
–80
–60
–80
–100
–100
–120
–140
–140
0
25
50
75
100
125
150
175
200
225
FREQUENCY (MHz)
250
Figure 14. Single Tone FFT with fIN = 340.3 MHz
Rev. 0 | Page 19 of 64
0
25
50
75
100
125
150
175
200
FREQUENCY (MHz)
Figure 17. Single-Tone FFT with fIN = 985.3 MHz
13015-019
–120
13015-016
AMPLITUDE (dBFS)
–40
AD9684
Data Sheet
95
0
AIN = −1dBFS
SNR = 61.7dBFS
ENOB = 9.9 BITS
SFDR = 70dBFS
BUFFER CONTROL 1 = 8.0×
–20
90
SFDR
85
SNR/SFDR (dBFS)
–40
AMPLITUDE (dBFS)
2.0×
2.0×
3.0×
3.0×
4.0×
4.0×
–60
–80
80
75
–100
70
–120
65
0
25
50
75
100
125
150
175
200
225
250
FREQUENCY (MHz)
60
13015-020
–140
Figure 18. Single Tone FFT with fIN = 1205.3 MHz
0
AIN = −1dBFS
SNR = 60.1dBFS
ENOB = 9.7 BITS
SFDR = 71dBFS
BUFFER CONTROL 1 = 8.0×
AIN1 AND AIN2 = –7dBFS
SFDR = 88dBFS
IMD2 = 95dBFS
IMD3 = 88dBFS
BUFFER CONTROL 1 = 2.0×
–20
AMPLITUDE (dBFS)
–60
–80
–100
–120
–40
–60
–80
0
25
50
75
100
125
150
175
200
225
250
FREQUENCY (MHz)
–120
0
100
150
200
250
FREQUENCY (MHz)
Figure 19. Single Tone FFT with fIN = 1630.3 MHz
Figure 22. Two-Tone FFT with fIN1 = 184 MHz and fIN2 = 187 MHz
0
0
AIN = −1dBFS
SNR = 59.0dBFS
ENOB = 9.5 BITS
SFDR = 69dBFS
BUFFER CONTROL 1 = 8.0×
–20
50
13015-025
–140
13015-021
–100
AIN1 AND AIN2 = –7dBFS
SFDR = 87dBFS
IMD2 = 94dBFS
IMD3 = 87dBFS
BUFFER CONTROL 1 = 2.0×
–20
AMPLITUDE (dBFS)
–40
–60
–80
–100
–120
–40
–60
–80
–140
0
25
50
75
100
125
150
175
200
225
FREQUENCY (MHz)
250
13015-022
–100
Figure 20. Single Tone FFT with fIN = 985.3 MHz
–120
0
50
100
150
200
FREQUENCY (MHz)
Figure 23. Two-Tone FFT; fIN1 = 338 MHz, fIN2 = 341 MHz
Rev. 0 | Page 20 of 64
250
13015-026
AMPLITUDE (dBFS)
–40
AMPLITUDE (dBFS)
500
Figure 21. SNR/SFDR vs. Analog Input Frequency (fIN);
fIN < 500 MHz; Buffer Control 1 Setting = 2.0×, 3.0×, and 4.0×
0
–20
400
100
200
300
ANALOG INPUT FREQUENCY (MHz)
0
13015-023
SNR
Data Sheet
AD9684
90
0
SFDR
–20
85
SNR/SFDR (dBFS)
–40
IMD3 (dBc)
–60
–80
SFDR (dBFS)
80
75
–100
70
IMD3 (dBFS)
–120
–140
–90 –84 –78 –72 –66 –60 –54 –48 –42 –36 –30 –24 –18 –12 –6
INPUT AMPLITUDE (dBFS)
65
–40
Figure 24. Two-Tone SFDR/IMD3 vs. Input Amplitude (AIN) with fIN1 = 184 MHz
and fIN2 = 187 MHz
–25
–10
0
25
15
TEMPERATURE (°C)
40
70
55
85
13015-030
SNR
13015-027
SFDR/IMD3 (dBc AND dBFS)
SFDR (dBc)
Figure 27. SNR/SFDR vs. Temperature, fIN = 170.3 MHz
0
2.30
2.25
–20
2.20
POWER DISSIPATION (W)
SFDR/IMD3 (dBc AND dBFS)
SFDR (dBc)
–40
IMD3 (dBc)
–60
–80
SFDR (dBFS)
–100
2.15
2.10
2.05
2.00
1.95
1.90
IMD3 (dBFS)
Figure 25. Two-Tone SFDR/IMD3 vs. Input Amplitude (AIN) with fIN1 = 338 MHz
and fIN2 = 341 MHz
100
SFDR (dBc)
80
SFDR (dBFS)
40
SNR (dBc)
20
0
–20
INPUT AMPLITUDE (dBFS)
0
13015-029
–6
–12
–18
–24
–30
–36
–42
–48
–60
–54
–66
–72
–78
–84
–40
–90
SNR/SFDR (dBc AND dBFS)
SNR (dBc)
60
Figure 26. SNR/SFDR vs. Input Amplitude, fIN = 170.3 MHz
Rev. 0 | Page 21 of 64
500
480
460
440
420
400
380
360
340
320
1.80
SAMPLE RATE (MSPS)
Figure 28. Power Dissipation vs. Sample Rate (fS) (Default SPI)
13015-031
INPUT AMPLITUDE (dBFS)
13015-028
1.85
–140
–90 –84 –78 –72 –66 –60 –54 –48 –42 –36 –30 –24 –18 –12 –6
300
–120
AD9684
Data Sheet
EQUIVALENT CIRCUITS
AVDD3
AVDD3
AVDD3
3pF 1.5pF
200Ω
67Ω
200Ω
28Ω
VIN+x
200Ω
67Ω
28Ω
200Ω
400Ω
10pF
VCM
BUFFER
SPIVDD
ESD
PROTECTED
AVDD3
AVDD3
SPIVDD
1kΩ
SCLK
VIN–x
ESD
PROTECTED
13015-036
3pF 1.5pF
13015-032
30kΩ
AIN
CONTROL
(SPI)
Figure 29. Analog Inputs
Figure 33. SCLK Inputs
AVDD1
SPIVDD
25Ω
CLK+
ESD
PROTECTED
30kΩ
1kΩ
CSB
AVDD1
20kΩ
VCM = 0.85V
13015-033
20kΩ
13015-037
ESD
PROTECTED
25Ω
Figure 30. Clock Inputs
Figure 34. CSB Input
AVDD1
SYNC+
SPIVDD
1kΩ
ESD
PROTECTED
SDO
20kΩ
AVDD1
SDI
30kΩ
VCM = 0.85V
ESD
PROTECTED
20kΩ
13015-034
1kΩ
Figure 31. SYNC± Inputs
Figure 35. SDIO
SPIVDD
SWING CONTROL
(SPI)
ESD
PROTECTED
DRVDD
DATA+
OUTPUT
DRIVER
DATA–
DRGND
ESD
PROTECTED
DRVDD
Dx±
DRGND
FD
FD_A/FD_B
Dx±
TEMPERATURE DIODE
(FD_A ONLY)
FD_x PIN CONTROL (SPI)
13015-035
SYNC–
1kΩ
SDIO
13015-038
LEVEL
TRANSLATOR
SPIVDD
Figure 32. LVDS Digital Outputs, STATUS±, DCO±
Figure 36. FD_A/FD_B Outputs
Rev. 0 | Page 22 of 64
13015-039
CLK–
Data Sheet
AD9684
AVDD2
SPIVDD
ESD
PROTECTED
1kΩ
PDWN/STBY
CONTROL (SPI)
V_1P0
ESD
PROTECTED
13015-040
PDWN/
STBY
ESD
PROTECTED
30kΩ
V_1P0 PIN
CONTROL (SPI)
Figure 38. V_1P0 Input/Output
Figure 37. PDWN/STBY Input
Rev. 0 | Page 23 of 64
13015-041
ESD
PROTECTED
AD9684
Data Sheet
THEORY OF OPERATION
The AD9684 has several functions that simplify the AGC
function in a communications receiver. The programmable
threshold detector allows monitoring of the incoming signal
power using the fast detect output bits of the ADC. If the input
signal level exceeds the programmable threshold, the fast detect
indicator goes high. Because this threshold indicator has low
latency, the user can quickly reduce the system gain to avoid an
overrange condition at the ADC input.
The LVDS outputs can be configured depending on the
decimation ratio. Multiple device synchronization is supported
through the SYNC± input pins.
ADC ARCHITECTURE
The architecture of the AD9684 consists of an input buffered pipelined ADC. The input buffer provides a termination impedance to
the analog input signal. This termination impedance can be
changed using the SPI to meet the termination needs of the driver/
amplifier. The default termination value is set to 400 Ω. The input
buffer is optimized for high linearity, low noise, and low power.
The input buffer provides a linear high input impedance (for
ease of drive) and reduces kickback from the ADC. The buffer
is optimized for high linearity, low noise, and low power. The
quantized outputs from each stage are combined into a final
14-bit result in the digital correction logic. The pipelined
architecture permits the first stage to operate with a new input
sample, whereas the remaining stages operate with the preceding
samples. Sampling occurs on the rising edge of the clock.
ANALOG INPUT CONSIDERATIONS
The analog input to the AD9684 is a differential buffer. The
internal common-mode voltage of the buffer is 2.05 V. The
clock signal alternately switches the input circuit between
sample mode and hold mode. When the input circuit is switched
into sample mode, the signal source must be capable of charging
the sample capacitors and settling within one-half of a clock cycle.
A small resistor, in series with each input, helps reduce the peak
transient current injected from the output stage of the driving
source. In addition, low Q inductors or ferrite beads can be placed
on each leg of the input to reduce high differential capacitance
at the analog inputs and, thus, achieve the maximum bandwidth
of the ADC. Such use of low Q inductors or ferrite beads is
required when driving the converter front end at high IF
For best dynamic performance, the source impedances driving
VIN+x and VIN−x must be matched such that common-mode
settling errors are symmetrical. These errors are reduced by the
common-mode rejection of the ADC. An internal reference buffer
creates a differential reference that defines the span of the ADC core.
Maximum SNR performance is achieved by setting the ADC to
the largest span in a differential configuration. In the case of the
AD9684, the available span is 2.06 V p-p differential.
Differential Input Configurations
There are several ways to drive the AD9684, either actively or
passively. However, optimum performance is achieved by
driving the analog input differentially.
For applications in which SNR and SFDR are key parameters,
differential transformer coupling is the recommended input
configuration because the noise performance of most amplifiers
is not adequate to achieve the true performance of the AD9684.
For low to midrange frequencies, a double balun or double
transformer network is recommended for optimum performance
of the AD9684 (see Figure 39). For higher frequencies in the
second and third Nyquist zones, it is better to remove some of
the front-end passive components to ensure wideband
operation (see Figure 40).
ETC1-11-13/
MABA007159
1:1Z
10Ω
10Ω
0.1µF
25Ω
4pF
0.1µF
25Ω
10Ω
2pF
ADC
10Ω
0.1µF
13015-042
The dual ADC cores feature a multistage, differential pipelined
architecture with integrated output error correction logic. Each
ADC features wide bandwidth inputs that support a variety of
user selectable input ranges. An integrated voltage reference
eases design considerations.
frequencies. Place either a differential capacitor or two singleended capacitors on the inputs to provide a matching passive
network. This ultimately creates a low-pass filter at the input, which
limits unwanted broadband noise. For more information, see the
AN-742 Application Note, the AN-827 Application Note, and the
Analog Dialogue article “Transformer-Coupled Front-End for
Wideband A/D Converters” (Volume 39, April 2005). In general,
the precise values depend on the application.
4pF
Figure 39. Differential Transformer-Coupled Configuration for First and
Second Nyquist Frequencies
25Ω
MARKI
BAL-0006
OR
BAL-0006SMG
25Ω
25Ω
0.1µF
0.1µF
25Ω
0.1µF
ADC
13015-043
The AD9684 has two analog input channels and 14 LVDS
output lane pairs. The ADC is designed to sample wide
bandwidth analog signals of up to 2 GHz. The AD9684 is
optimized for wide input bandwidth, a high sampling rate,
excellent linearity, and low power in a small package.
Figure 40. Differential Transformer-Coupled Configuration for Second and
Third Nyquist Frequencies
Rev. 0 | Page 24 of 64
Data Sheet
AD9684
Input Common Mode
95
Analog Input Controls and SFDR Optimization
4.5×
85
3.0×
75
SFDR (dBFS)
The analog inputs of the AD9684 are internally biased to the
common mode as shown in Figure 41. The common-mode
buffer has a limited range in that the performance suffers greatly
if the common-mode voltage drops by more than 100 mV.
Therefore, in dc-coupled applications, set the common-mode
voltage to 2.05 V ± 100 mV to ensure proper ADC operation.
2.0×
65
1.5×
55
The AD9684 offers flexible controls for the analog inputs, such
as input termination and buffer current. All of the available
controls are shown in Figure 41.
1.0×
35
50
AVDD3
150
100
200
250
300
350
400
450
500
INPUT FREQUENCY (MHz)
13015-046
45
Figure 43. Buffer Current Sweeps (SFDR vs. Input Frequency and IBUFF),
10 MHz < fIN < 500 MHz
VIN+x
3pF
200Ω
67Ω
200Ω
28Ω
90
AVDD3
VCM
BUFFER
SFDR (dBFS)
200Ω
67Ω
28Ω
200Ω
400Ω
10pF
4.5×
5.0×
6.0×
7.0×
8.0×
85
AVDD3
VIN–x
80
75
3pF
70
65
500
Figure 41. Analog Input Controls (Should the AIN
Using Register 0x018, the buffer currents on each channel can
be scaled to optimize the SFDR over various input frequencies
and bandwidths of interest. As the input buffer currents are set,
the amount of current required by the AVDD3 supply changes.
For a complete list of buffer current settings, see Table 29.
550
600
650
700
750
800
850
900
950
1000
INPUT FREQUENCY (MHz)
13015-047
13015-044
AIN CONTROL
(SPI) REGISTERS
(REG 0x008, REG 0x015,
REG 0x016, REG 0x018,
REG 0x025)
Figure 44. Buffer Current Sweeps (SFDR vs. Input Frequency and IBUFF),
500 MHz < fIN < 1000 MHz
80
75
250
70
210
IAVDD3 (mA)
190
170
60
55
150
130
50
110
4.5×
5.0×
6.0×
7.0×
8.0×
8.0×
45
1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
90
INPUT FREQUENCY (MHz)
2.5×
3.5×
4.5×
5.5×
6.5×
7.5×
8.5×
BUFFER CURRENT SETTING
Figure 42. AVDD3 Power (IAVDD3) vs. Buffer Current Control Setting in
Register 0x018
13015-045
70
50
1.5×
65
13015-048
SFDR (dBFS)
230
Figure 45. Buffer Current Sweeps (SFDR vs. Input Frequency and IBUFF),
1 GHz < fIN < 2 GHz, Front-End Network Shown in Figure 40
Figure 43, Figure 44, and Figure 45 show how the SFDR can be
optimized using the buffer current setting in Register 0x018 for
different Nyquist zones. At frequencies greater than 1 GHz, it is
better to run the ADC at input amplitudes less than −1 dBFS
(−3 dBFS, for example). This greatly improves the linearity of
the converted signal without sacrificing SNR performance.
Rev. 0 | Page 25 of 64
AD9684
Data Sheet
The use of an external reference may be necessary, in some
applications, to enhance the gain accuracy of the ADC or
improve thermal drift characteristics. Figure 47 shows the
typical drift characteristics of the internal 1.0 V reference.
Table 9 shows the recommended buffer current and full-scale
voltage settings for the different analog input frequency ranges.
Absolute Maximum Input Swing
The absolute maximum input swing allowed at the inputs of the
AD9684 is 4.3 V p-p differential. Signals operating near or at
this level can cause permanent damage to the ADC.
1.0010
1.0009
1.0008
VOLTAGE REFERENCE
V_1P0 VOLTAGE (V)
1.0007
A stable and accurate 1.0 V voltage reference is built into the
AD9684. This internal 1.0 V reference sets the full-scale input
range of the ADC. For more information on adjusting the input
swing, see Table 29. Figure 46 shows the block diagram of the
internal 1.0 V reference controls.
1.0006
1.0005
1.0004
1.0003
1.0002
1.0000
VIN–A/
VIN–B
0.9999
0.9998
–50
INTERNAL
V_1P0
GENERATOR
0
ADC
CORE
FULL-SCALE
VOLTAGE
ADJUST
25
90
TEMPERATURE (°C)
Figure 47. Typical V_1P0 Drift
The external reference must be a stable 1.0 V reference. The
ADR130 is a good option for providing the 1.0 V reference.
Figure 48 shows how the ADR130 can be used to provide the
external 1.0 V reference to the AD9684. The grayed out areas
show unused blocks within the AD9684 while using the
ADR130 to provide the external reference.
INPUT FULL-SCALE
RANGE ADJUST
SPI REGISTER
(REGISTER 0x025)
V_1P0
13015-049
V_1P0 PIN
CONTROL SPI
REGISTER
(REGISTER 0x024)
Figure 46. Internal Reference Configuration and Controls
Register 0x024 enables the user either to use this internal 1.0 V
reference, or to provide an external 1.0 V reference. When using
an external voltage reference, provide a 1.0 V reference. The
full-scale adjustment is made using the SPI, irrespective of the
reference voltage. For more information on adjusting the fullscale level of the AD9684, see the Memory Map Register Table
section.
Table 9. SFDR Optimization for Input Frequencies
Frequency
DC to 250 MHz
250 MHz to 500
MHz
500 MHz to
1 GHz
1 GHz to 2 GHz
Buffer Control 1
(Register 0x018)
0x20 (2.0×)
0x70 (4.5×)
Input Full-Scale Range
(Register 0x025)
0x0C (2.06 V p-p)
0x0C (2.06 V p-p)
Input Full-Scale Control
(Register 0x030)
0x04
0x04
Input Termination
(Register 0x016) 1
0x0C/0x1C/0x6C
0x0C/0x1C/0x6C
0x80 (5.0×)
0x08 (1.46 V p-p)
0x18
0x0C/0x1C/0x6C
0xF0 (8.5×)
0x08 (1.46 V p-p)
0x18
0x0C/0x1C/0x6C
The input termination can be changed to accommodate the application with little or no impact to ac performance.
INTERNAL
V_1P0
GENERATOR
ADR130
INPUT
1
NC
2
GND SET 5
3
VIN
0.1µF
FULL-SCALE
VOLTAGE
ADJUST
NC 6
VOUT 4
V_1P0
0.1µF
FULL-SCALE
CONTROL
Figure 48. External Reference Using the ADR130
Rev. 0 | Page 26 of 64
13015-054
1
13015-050
1.0001
VIN+A/
VIN+B
Data Sheet
AD9684
CLOCK INPUT CONSIDERATIONS
Input Clock Divider
For optimum performance, drive the AD9684 sample clock
inputs (CLK+ and CLK−) with a differential signal. This signal
is typically ac-coupled to the CLK+ and CLK− pins via a
transformer or clock drivers. These pins are biased internally
and require no additional biasing.
The AD9684 contains an input clock divider with the ability to
divide the Nyquist input clock by 1, 2, 4, and 8. The divider
ratios can be selected using Register 0x10B. This is shown in
Figure 52.
Figure 49 shows a preferred method for clocking the AD9684. The
low jitter clock source is converted from a single-ended signal to
a differential signal using an RF transformer.
0.1µF
CLK+
100Ω
CLK+
ADC
CLK–
0.1µF
CLK–
÷2
÷4
Figure 49. Transformer Coupled Differential Clock
÷8
Another option is to ac couple a differential CML or LVDS
signal to the sample clock input pins, as shown in Figure 50 and
Figure 51.
3.3V
71Ω
33Ω
0.1µF
ADC
Z0 = 50Ω
0.1µF
13015-052
CLK+
CLK–
Input Clock Divider ½ Period Delay Adjustment
Figure 50. Differential CML Sample Clock
CLK+
LVDS
DRIVER
CLK+
100Ω
CLK–
CLOCK INPUT
50Ω1
50Ω1
Clock Fine Delay Adjustment
ADC
CLK–
0.1µF
150Ω RESISTORS ARE OPTIONAL.
13015-053
0.1µF
The input clock divider inside the AD9684 provides phase delay
in increments of ½ the input clock cycle. Program Register 0x10C
to enable this delay independently for each channel.
0.1µF
0.1µF
CLOCK INPUT
Figure 52. Clock Divider Circuit
The AD9684 clock divider can be synchronized using the external
SYNC± input. A valid SYNC± input causes the clock divider to
reset to a programmable state. This feature is enabled by setting
Bit 7 of Register 0x10D. This synchronization feature allows
multiple devices to have their clock dividers aligned to guarantee
simultaneous input sampling.
10pF
33Ω
Z0 = 50Ω
REG 0x10B
13015-055
1:1Z
50Ω
13015-051
CLOCK
INPUT
The maximum frequency at the CLK± inputs is 4 GHz. This is
the limit of the divider. In applications where the clock input is
a multiple of the sample clock, the appropriate divider ratio
must be programmed into the clock divider before applying the
clock signal. This ensures that the current transients during
device startup are controlled.
Figure 51. Differential LVDS Sample Clock
Clock Duty Cycle Considerations
Typical high speed ADCs use both clock edges to generate a
variety of internal timing signals. As a result, these ADCs may
be sensitive to the clock duty cycle. Commonly, a 5% tolerance
is required on the clock duty cycle to maintain dynamic performance
characteristics. In applications where the clock duty cycle cannot
be guaranteed to be 50%, a higher multiple frequency clock can be
supplied to the device. The AD9684 can be clocked at 2 GHz with
the internal clock divider set to 2. The output of the divider offers
a 50% duty cycle, high slew rate (fast edge) clock signal to the
internal ADC. See the Memory Map section for more details on
using this feature.
To adjust the AD9684 sampling edge instant, write to Register 0x117
and Register 0x118. Setting Bit 0 of Register 0x117 enables the fine
delay feature, and Register 0x118, Bits[7:0] set the value of the
delay. This value can be programmed individually for each channel.
The clock delay can be adjusted from −151.7 ps to +150 ps in
~1.7 ps increments. The clock delay adjust takes effect immediately
when it is enabled via SPI writes. Enabling the clock fine delay
adjustment in Register 0x117 causes a datapath reset.
Clock Jitter Considerations
High speed, high resolution ADCs are sensitive to the quality of
the clock input. The degradation in SNR at a given input
frequency (fA) due only to aperture jitter (tJ) can be calculated by
SNR = 20 × log 10 (2 × π × fA × tJ)
In this equation, the rms aperture jitter represents the root mean
square of all jitter sources, including the clock input, analog input
signal, and ADC aperture jitter specifications. IF undersampling
applications are particularly sensitive to jitter (see Figure 53).
Rev. 0 | Page 27 of 64
AD9684
POWER-DOWN/STANDBY MODE
RMS CLOCK JITTER REQUIREMENT
120
The AD9684 has a PDWN/STBY pin that configures the device
in power-down or standby mode. The default operation is the
power-down function. The PDWN/STBY pin is a logic high pin.
The power-down option can also be set via Register 0x03F and
Register 0x040.
SNR (dB)
110
100
16 BITS
90
14 BITS
80
12 BITS
TEMPERATURE DIODE
70
10 BITS
60
40
30
1
100
10
ANALOG INPUT FREQUENCY (MHz)
1000
13015-056
0.125ps
0.25ps
0.5ps
1.0ps
2.0ps
8 BITS
50
Figure 53. Ideal SNR vs. Analog Input Frequency and Jitter
Treat the clock input as an analog signal when aperture jitter
may affect the dynamic range of the AD9684. Separate the power
supplies for the clock drivers from the ADC output driver
supplies to avoid modulating the clock signal with digital noise. If
the clock is generated from another type of source (by gating,
dividing, or other methods), retime the clock by the original clock
at the last step. For more in-depth information about jitter
performance as it relates to ADCs, see the AN-501 Application
Note and the AN-756 Application Note.
Figure 54 shows the estimated SNR of the AD9684 across the
input frequency for different clock induced jitter values. Estimate
the SNR using the following equation:
− SNR JITTER
− SNR ADC
SNR (dBFS) = 10log10 10 + 10 10
The AD9684 contains a diode-based temperature sensor for
measuring the temperature of the die. This diode can output a
voltage and serve as a coarse temperature sensor to monitor the
internal die temperature.
The temperature diode voltage can be output to the FD_A pin
using the SPI. Use Register 0x028, Bit 0 to enable or disable the
diode. Register 0x028 is a local register. Channel A must be
selected in the device index register (Register 0x008) to enable
the temperature diode readout. Configure the FD_A pin to
output the diode voltage by programming Register 0x040,
Bits[2:0]. See Table 29 for more information.
The voltage response of the temperature diode (with SPIVDD =
1.8 V) is shown in Figure 55.
0.90
0.85
DIODE VOLTAGE (V)
130
Data Sheet
0.80
0.75
0.70
75
0.65
0.60
–55 –45 –35 –25 –15 –5
SNR (dBFS)
65
5
15 25 35 45 55 65 75 85 95 105 115 125
TEMPERATURE (°C)
25fs
50fs
75fs
100fs
125fs
150fs
175fs
200fs
60
55
Figure 55. Diode Voltage vs. Temperature
45
1M
10M
100M
INPUT FREQUENCY (Hz)
1G
10G
13015-057
50
Figure 54. Estimated SNR Degradation for the AD9684 vs. Input Frequency
and Clock Jitter
Rev. 0 | Page 28 of 64
13015-058
70
Data Sheet
AD9684
ADC OVERRANGE AND FAST DETECT
The operation of the upper threshold and lower threshold
registers, along with the dwell time registers, is shown in
Figure 56.
In receiver applications, it is desirable to have a mechanism to
reliably determine when the converter is about to be clipped.
The standard overrange pin outputs information on the state of
the analog input. It is also helpful to have a programmable
threshold below full scale that allows time to reduce the gain
before the clip actually occurs. In addition, because input
signals can have significant slew rates, the latency of this
function is of major concern. Highly pipelined converters can
have significant latency. The AD9684 contains fast detect
circuitry for individual channels to monitor the threshold and
assert the FD_A and FD_B pins.
The FD_x indicator is asserted if the input magnitude exceeds
the value programmed in the fast detect upper threshold
registers, in Register 0x247 and Register 0x248. The selected
threshold register is compared with the signal magnitude at the
output of the ADC. The fast upper threshold detection has a
latency of 28 clock cycles (maximum). The approximate upper
threshold magnitude is defined by
Upper Threshold Magnitude (dBFS) = 20log(Threshold
Magnitude/213)
ADC OVERRANGE
The ADC overrange indicator is asserted when an overrange is
detected on the input of the ADC. The overrange indicator can
be output on the STATUS± pins (when CSB > 0). The latency of
this overrange indicator matches the sample latency.
The FD indicators are not cleared until the signal drops below
the lower threshold for the programmed dwell time. The lower
threshold is programmed in the fast detect lower threshold
registers, in Register 0x249 and Register 0x24A. The fast detect
lower threshold register is a 13-bit register that is compared with
the signal magnitude at the output of the ADC. This
comparison is subject to the ADC pipeline latency, but is
accurate in terms of converter resolution. The lower threshold
magnitude is defined by
The AD9684 also records any overrange condition in any of the
four virtual converters. The overrange status of each virtual
converter is registered as a sticky bit in Register 0x563. The
contents of Register 0x563 can be cleared using Register 0x562,
by toggling the bits corresponding to the virtual converter to set
and reset the position.
Lower Threshold Magnitude (dBFS) = 20log(Threshold
Magnitude/213)
FAST THRESHOLD DETECTION (FD_A AND FD_B)
For example, to set an upper threshold of −6 dBFS, write 0xFFF
to Register 0x247 and Register 0x248. To set a lower threshold
of −10 dBFS, write 0xA1D to Register 0x249 and Register 0x24A.
The fast detect (FD) bit (enabled via the control bits in
Register 0x559) is immediately set whenever the absolute value
of the input signal exceeds the programmable upper threshold
level. The FD bit is cleared only when the absolute value of the
input signal drops below the lower threshold level for greater
than the programmable dwell time. This feature provides
hysteresis and prevents the FD bit from excessively toggling.
The dwell time can be programmed from 1 to 65,535 sample
clock cycles by placing the desired value in the fast detect dwell
time registers, in Register 0x24B and Register 0x24C. See the
Memory Map section (Register 0x040, and Register 0x245 to
Register 0x24C in Table 29) for more details.
UPPER THRESHOLD
DWELL TIME
TIMER RESET BY
RISE ABOVE
LOWER
THRESHOLD
DWELL TIME
FD_A OR FD_B
Figure 56. Threshold Settings for FD_A and FD_B Signals
Rev. 0 | Page 29 of 64
TIMER COMPLETES BEFORE
SIGNAL RISES ABOVE
LOWER THRESHOLD
13015-059
MIDSCALE
LOWER THRESHOLD
AD9684
Data Sheet
SIGNAL MONITOR
The signal monitor block provides additional information about
the signal being digitized by the ADC. The signal monitor
computes the peak magnitude of the digitized signal. This
information can be used to drive an AGC loop to optimize the
range of the ADC in the presence of real-world signals.
The results of the signal monitor block can be obtained by
reading back the internal values from the SPI port. A global, 24-bit
programmable period controls the duration of the measurement.
Figure 57 shows the simplified block diagram of the signal
monitor block.
The peak detector captures the largest signal within the
observation period. The detector only observes the magnitude
of the signal. The resolution of the peak detector is a 13-bit
value and the observation period is 24 bits and represents
converter output samples.
Derive the peak magnitude using the following equation:
Peak Magnitude (dBFS) = 20log(Peak Detector Value/213)
The magnitude of the input port signal is monitored over a
programmable time period, which is determined by the signal
SIGNAL MONITOR
PERIOD REGISTER
(SMPR)
REG 0x271, REG 0x272, REG 0x273
After enabling this mode, the value in the SMPR is loaded into a
monitor period timer that decrements at the decimated clock
rate. The magnitude of the input signal is compared with the
value in the internal magnitude storage register (not accessible
to the user), and the greater of the two is updated as the current
peak level. The initial value of the magnitude storage register is
set to the current ADC input signal magnitude. This comparison
continues until the monitor period timer reaches a count of 1.
When the monitor period timer reaches a count of 1, the 13-bit
peak level value is transferred to the signal monitor holding
register, which can be read through the memory map. The
monitor period timer is reloaded with the value in the SMPR,
and the countdown is restarted. In addition, the magnitude of
the first input sample is updated in the magnitude storage
register.
DOWN
COUNTER
IS
COUNT = 1?
LOAD
CLEAR
FROM
INPUT
MAGNITUDE
STORAGE
REGISTER
LOAD
LOAD
SIGNAL
MONITOR
HOLDING
REGISTER
COMPARE
A>B
Figure 57. Signal Monitor Block
Rev. 0 | Page 30 of 64
TO STATUS± PINS
AND MEMORY MAP
13015-060
FROM
MEMORY
MAP
monitor period register (SMPR). To enable the peak detector
function, set Bit 1 of Register 0x270 in the signal monitor
control register. The 24-bit SMPR must be programmed before
activating this mode.
Data Sheet
AD9684
DIGITAL DOWNCONVERTERS (DDCs)
The AD9684 includes four digital downconverters that provide
filtering and reduce the output data rate. This digital processing
section includes an NCO, a half-band decimating filter, a finite
impulse response (FIR) filter, a gain stage, and a complex to real
conversion stage. Each of these processing blocks has a control
line that allows the block to be independently enabled and
disabled to provide the desired processing function. The DDCs
can be configured to output either real data or complex output data.
DDC I/Q INPUT SELECTION
The AD9684 has two ADC channels and four DDC channels.
Each DDC channel has two input ports that can be paired to
support both real and complex inputs through the I/Q crossbar
mux. For real signals, both DDC input ports must select the
same ADC channel (that is, DDC Input Port I = ADC Channel A
and DDC Input Port Q = ADC Channel A). For complex
signals, each DDC input port must select different ADC
channels (that is, DDC Input Port I = ADC Channel A and
DDC Input Port Q = ADC Channel B).
The inputs to each DDC are controlled by the DDC input selection
registers (Register 0x311, Register 0x331, Register 0x351, and
Register 0x371). See Table 29 for information on how to
configure the DDCs.
DDC I/Q OUTPUT SELECTION
Each DDC channel has two output ports that can be paired to
support both real or complex outputs. For real output signals,
only the DDC Output Port I is used (the DDC Output Port Q is
invalid). For complex I/Q output signals, both DDC Output
Port I and DDC Output Port Q are used.
The I/Q outputs to each DDC channel are controlled by the
DDC complex to real enable bit in the DDC control registers
(Bit 3 in Register 0x310, Register 0x330, Register 0x350, and
Register 0x370).
The Chip I only bit in the chip application mode register
(Register 0x200, Bit 5) controls the chip output muxing of all
the DDC channels. When all DDC channels use real outputs,
set this bit high to ignore all DDC Q output ports. When any of
the DDC channels are set to use complex I/Q outputs, the user
must clear this bit to use both DDC Output Port I and DDC
Output Port Q.
DDC GENERAL DESCRIPTION
The four DDC blocks extract a portion of the full digital
spectrum captured by the ADCs. They are intended for IF
sampling or oversampled baseband radios requiring wide
bandwidth input signals.
Each DDC block contains the following signal processing stages:
•
•
•
•
Frequency translation stage (optional)
Filtering stage
Gain stage (optional)
Complex to real conversion stage (optional)
Frequency Translation Stage (Optional)
This stage consists of a 12-bit complex NCO and quadrature
mixers that can be used for frequency translation of both real or
complex input signals. This stage shifts a portion of the available
digital spectrum down to baseband.
Filtering Stage
After shifting down to baseband, this stage decimates the
frequency spectrum using a chain of up to four half-band, lowpass filters for rate conversion. The decimation process lowers
the output data rate, which, in turn, reduces the output interface
rate.
Gain Stage (Optional)
Due to losses associated with mixing a real input signal down to
baseband, this stage compensates by adding an additional 0 dB
or 6 dB of gain.
Complex to Real Conversion Stage (Optional)
When real outputs are necessary, this stage converts the complex
outputs back to real outputs by performing an fS/4 mixing
operation in addition to a filter to remove the complex
component of the signal.
Figure 58 shows the detailed block diagram of the DDCs
implemented in the AD9684.
Rev. 0 | Page 31 of 64
AD9684
Data Sheet
GAIN = 0dB
OR 6dB
COMPLEX TO REAL
CONVERSION
(OPTIONAL)
COMPLEX TO REAL
CONVERSION
(OPTIONAL)
COMPLEX TO REAL
CONVERSION
(OPTIONAL)
COMPLEX TO REAL
CONVERSION
(OPTIONAL)
HB1 FIR
DCM = 2
GAIN = 0dB
OR 6dB
ADC
SAMPLING
AT fS
GAIN = 0dB
OR 6dB
REAL/I
GAIN = 0dB
OR 6dB
REAL/Q Q
HB2 FIR
DCM = BYPASS OR 2
NCO
+
MIXER
(OPTIONAL)
HB3 FIR
DCM = BYPASS OR 2
I
HB4 FIR
DCM = BYPASS OR 2
DDC 0
REAL/I
REAL/I
CONVERTER 0
Q CONVERTER 1
SYNC±
Q CONVERTER 3
REAL/Q Q
ADC
SAMPLING
AT fS
HB1 FIR
DCM = 2
I
HB2 FIR
DCM = BYPASS OR 2
REAL/I
HB3 FIR
DCM = BYPASS OR 2
DDC 2
REAL/I
CONVERTER 4
OUTPUT INTERFACE
HB1 FIR
DCM = 2
REAL/I
CONVERTER 2
SYNC±
NCO
+
MIXER
(OPTIONAL)
REAL/I
HB2 FIR
DCM = BYPASS OR 2
REAL/Q Q
HB3 FIR
DCM = BYPASS OR 2
NCO
+
MIXER
(OPTIONAL)
HB4 FIR
DCM = BYPASS OR 2
I/Q CROSSBAR MUX
I
HB4 FIR
DCM = BYPASS OR 2
DDC 1
REAL/I
Q CONVERTER 5
SYNC±
SYNC±
SYNCHRONIZATION
CONTROL CIRCUITS
HB1 FIR
DCM = 2
REAL/I
CONVERTER 6
Q CONVERTER 7
13015-061
REAL/Q Q
HB2 FIR
DCM = BYPASS OR 2
NCO
+
MIXER
(OPTIONAL)
HB3 FIR
DCM = BYPASS OR 2
I
HB4 FIR
DCM = BYPASS OR 2
DDC 3
REAL/I
SYNC±
Figure 58. DDC Detailed Block Diagram
Figure 59 shows an example usage of one of the four DDC
blocks with a real input signal and four half-band filters (HB4 +
HB3 + HB2 + HB1). It shows both complex (decimate by 16)
and real (decimate by 8) output options.
When DDCs have different decimation ratios, the chip decimation
ratio (Register 0x201) must be set to the lowest decimation ratio
for all the DDC blocks. In this scenario, samples of higher decimation ratio DDCs are repeated to match the chip decimation ratio
sample rate. Whenever the NCO frequency is set or changed,
the DDC soft reset must be issued. If the DDC soft reset is not
issued, the output may potentially show amplitude variations.
Table 10 through Table 15 show the DDC samples when the
chip decimation ratio is set to 1, 2, 4, 8, or 16, respectively.
When DDCs have different decimation ratios, the chip
decimation ratio must be set to the lowest decimation ratio of
all the DDC channels. In this scenario, samples of higher
decimation ratio DDCs are repeated to match the chip
decimation ratio sample rate.
Rev. 0 | Page 32 of 64
Data Sheet
AD9684
ADC
ADC
SAMPLING
AT fS
REAL
REAL INPUT—SAMPLED AT fS
BANDWIDTH OF
INTEREST IMAGE
–fS/2
–fS/3
BANDWIDTH OF
INTEREST
fS/32
–fS/32
DC
–fS/16
fS/16
–fS/8
–fS/4
REAL
FREQUENCY TRANSLATION STAGE (OPTIONAL)
DIGITAL MIXER + NCO FOR fS/3 TUNING, THE FREQUENCY
TUNING WORD = ROUND ((fS/3)/fS × 4096) = +1365 (0x555)
fS/3
fS/4
fS/8
fS/2
I
NCO TUNES CENTER OF
BANDWIDTH OF INTEREST
TO BASEBAND
cos(wt)
REAL
12-BIT
NCO
90°
0°
–sin(wt)
Q
DIGITAL FILTER
RESPONSE
–fS/2
–fS/3
fS/32
–fS/32
DC
–fS/16
fS/16
–fS/8
–fS/4
BANDWIDTH OF
INTEREST IMAGE
(–6dB LOSS DUE TO
NCO + MIXER)
BANDWIDTH OF INTEREST
(–6dB LOSS DUE TO
NCO + MIXER)
fS/4
fS/8
fS/2
fS/3
FILTERING STAGE
I
HALFBAND
FILTER
Q
HALFBAND
FILTER
2
HALFBAND
FILTER
2
HALFBAND
FILTER
2
HALFBAND
FILTER
2
HALFBAND
FILTER
2
HALFBAND
FILTER
I
HB1 FIR
HB2 FIR
HB3 FIR
HB4 FIR
HB1 FIR
HB2 FIR
HB3 FIR
HB4 FIR
4 DIGITAL HALF-BAND FILTERS
(HB4 + HB3 + HB2 + HB1)
2
HALFBAND
FILTER
Q
6dB GAIN TO
COMPENSATE FOR
NCO + MIXER LOSS
COMPLEX (I/Q) OUTPUTS
GAIN STAGE (OPTIONAL)
DIGITAL FILTER
RESPONSE
I
GAIN STAGE (OPTIONAL)
Q
0dB OR 6dB GAIN
COMPLEX TO REAL
CONVERSION STAGE (OPTIONAL)
fS/4 MIXING + COMPLEX FILTER TO REMOVE Q
–fS/32
fS/32
DC
–fS/16
fS/16
–fS/8
I
REAL (I) OUTPUTS
+6dB
+6dB
fS/8
2
+6dB
2
+6dB
I
Q
–fS/32
fS/32
DC
–fS/16
fS/16
DOWNSAMPLE BY 2
I
DECIMATE BY 8
Q
DECIMATE BY 16
0dB OR 6dB GAIN
Q
COMPLEX REAL/I
TO
REAL
–fS/8
–fS/32
fS/32
DC
–fS/16
fS/16
fS/8
Figure 59. DDC Theory of Operation Example (Real Input, Decimate by 16)
Rev. 0 | Page 33 of 64
13015-062
6dB GAIN TO
COMPENSATE FOR
NCO + MIXER LOSS
AD9684
Data Sheet
Table 10. DDC Samples When the Chip Decimation Ratio = 1
HB1 FIR
(DCM 1 = 1)
N
N+1
N+2
N+3
N+4
N+5
N+6
N+7
N+8
N+9
N + 10
N + 11
N + 12
N + 13
N + 14
N + 15
N + 16
N + 17
N + 18
N + 19
N + 20
N + 21
N + 22
N + 23
N + 24
N + 25
N + 26
N + 27
N + 28
N + 29
N + 30
N + 31
1
Real (I) Output (Complex to Real Enabled)
HB2 FIR +
HB3 FIR + HB2
HB4 FIR + HB3 FIR
HB1 FIR
FIR + HB1 FIR
+ HB2 FIR + HB1
(DCM1 = 2)
(DCM1 = 4)
FIR (DCM1 = 8)
N
N
N
N+1
N+1
N+1
N
N
N
N+1
N+1
N+1
N+2
N
N
N+3
N+1
N+1
N+2
N
N
N+3
N+1
N+1
N+4
N+2
N
N+5
N+3
N+1
N+4
N+2
N
N+5
N+3
N+1
N+6
N+2
N
N+7
N+3
N+1
N+6
N+2
N
N+7
N+3
N+1
N+8
N+4
N+2
N+9
N+5
N+3
N+8
N+4
N+2
N+9
N+5
N+3
N + 10
N+4
N+2
N + 11
N+5
N+3
N + 10
N+4
N+2
N + 11
N+5
N+3
N + 12
N+6
N+2
N + 13
N+7
N+3
N + 12
N+6
N+2
N + 13
N+7
N+3
N + 14
N+6
N+2
N + 15
N+7
N+3
N + 14
N+6
N+2
N + 15
N+7
N+3
Complex (I/Q) Outputs (Complex to Real Disabled)
HB2 FIR +
HB3 FIR + HB2
HB4 FIR + HB3 FIR +
HB1 FIR
HB1 FIR
FIR + HB1 FIR
HB2 FIR + HB1 FIR
(DCM1 = 2)
(DCM1 = 4)
(DCM1 = 8)
(DCM1 = 16)
N
N
N
N
N+1
N+1
N+1
N+1
N
N
N
N
N+1
N+1
N+1
N+1
N+2
N
N
N
N+3
N+1
N+1
N+1
N+2
N
N
N
N+3
N+1
N+1
N+1
N+4
N+2
N
N
N+5
N+3
N+1
N+1
N+4
N+2
N
N
N+5
N+3
N+1
N+1
N+6
N+2
N
N
N+7
N+3
N+1
N+1
N+6
N+2
N
N
N+7
N+3
N+1
N+1
N+8
N+4
N+2
N
N+9
N+5
N+3
N+1
N+8
N+4
N+2
N
N+9
N+5
N+3
N+1
N + 10
N+4
N+2
N
N + 11
N+5
N+3
N+1
N + 10
N+4
N+2
N
N + 11
N+5
N+3
N+1
N + 12
N+6
N+2
N
N + 13
N+7
N+3
N+1
N + 12
N+6
N+2
N
N + 13
N+7
N+3
N+1
N + 14
N+6
N+2
N
N + 15
N+7
N+3
N+1
N + 14
N+6
N+2
N
N + 15
N+7
N+3
N+1
DCM means decimation.
Table 11. DDC Samples When the Chip Decimation Ratio = 2
Real (I) Output (Complex to Real Enabled)
HB4 FIR +
HB3 FIR +
HB3 FIR +
HB2 FIR +
HB2 FIR +
HB2 FIR +
HB1 FIR
HB1 FIR
HB1 FIR
(DCM1 = 4)
(DCM1 = 8)
(DCM 1 = 2)
N
N
N
N+1
N+1
N+1
N+2
N
N
N+3
N+1
N+1
N+4
N+2
N
N+5
N+3
N+1
N+6
N+2
N
N+7
N+3
N+1
N+8
N+4
N+2
N+9
N+5
N+3
Complex (I/Q) Outputs (Complex to Real Disabled)
HB4 FIR +
HB3 FIR +
HB3 FIR +
HB2 FIR +
HB2 FIR +
HB2 FIR +
HB1 FIR
HB1 FIR
HB1 FIR
HB1 FIR
(DCM1 = 2)
(DCM1 = 4)
(DCM1 = 8)
(DCM1 = 16)
N
N
N
N
N+1
N+1
N+1
N+1
N+2
N
N
N
N+3
N+1
N+1
N+1
N+4
N+2
N
N
N+5
N+3
N+1
N+1
N+6
N+2
N
N
N+7
N+3
N+1
N+1
N+8
N+4
N+2
N
N+9
N+5
N+3
N+1
Rev. 0 | Page 34 of 64
Data Sheet
Real (I) Output (Complex to Real Enabled)
HB4 FIR +
HB3 FIR +
HB3 FIR +
HB2 FIR +
HB2 FIR +
HB2 FIR +
HB1 FIR
HB1 FIR
HB1 FIR
(DCM1 = 4)
(DCM 1 = 2)
(DCM1 = 8)
N + 10
N+4
N+2
N + 11
N+5
N+3
N + 12
N+6
N+2
N + 13
N+7
N+3
N + 14
N+6
N+2
N + 15
N+7
N+3
1
AD9684
Complex (I/Q) Outputs (Complex to Real Disabled)
HB4 FIR +
HB3 FIR +
HB3 FIR +
HB2 FIR +
HB2 FIR +
HB2 FIR +
HB1 FIR
HB1 FIR
HB1 FIR
HB1 FIR
(DCM1 = 2)
(DCM1 = 4)
(DCM1 = 8)
(DCM1 = 16)
N + 10
N+4
N+2
N
N + 11
N+5
N+3
N+1
N + 12
N+6
N+2
N
N + 13
N+7
N+3
N+1
N + 14
N+6
N+2
N
N + 15
N+7
N+3
N+1
DCM means decimation.
Table 12. DDC Samples When the Chip Decimation Ratio = 4
Real (I) Output (Complex to Real Enabled)
HB3 FIR + HB2 FIR + HB1
HB4 FIR + HB3 FIR + HB2 FIR
FIR (DCM 1 = 4)
+ HB1 FIR (DCM1 = 8)
N
N
N+1
N+1
N+2
N
N+3
N+1
N+4
N+2
N+5
N+3
N+6
N+2
N+7
N+3
1
Complex (I/Q) Outputs (Complex to Real Disabled)
HB2 FIR + HB1 FIR
HB3 FIR + HB2 FIR +
HB4 FIR + HB3 FIR + HB2 FIR
(DCM1 = 4)
HB1 FIR (DCM1 = 8)
+ HB1 FIR (DCM1 = 16)
N
N
N
N+1
N+1
N+1
N+2
N
N
N+3
N+1
N+1
N+4
N+2
N
N+5
N+3
N+1
N+6
N+2
N
N+7
N+3
N+1
DCM means decimation.
Table 13. DDC Samples When the Chip Decimation Ratio = 8
Real (I) Output (Complex to Real Enabled)
HB4 FIR + HB3 FIR + HB2 FIR + HB1 FIR (DCM 1 = 8)
N
N+1
N+2
N+3
N+4
N+5
N+6
N+7
1
Complex (I/Q) Outputs (Complex to Real Disabled)
HB4 FIR + HB3 FIR + HB2 FIR + HB1
HB3 FIR + HB2 FIR + HB1 FIR(DCM1 = 8)
FIR (DCM1 = 16)
N
N
N+1
N+1
N+2
N
N+3
N+1
N+4
N+2
N+5
N+3
N+6
N+2
N+7
N+3
DCM means decimation.
Table 14. DDC Samples When the Chip Decimation Ratio = 16
Real (I) Output (Complex to Real Enabled)
HB4 FIR + HB3 FIR + HB2 FIR + HB1 FIR (DCM 1 = 16)
Not applicable
Not applicable
Not applicable
Not applicable
1
Complex (I/Q) Outputs (Complex to Real Disabled)
HB4 FIR + HB3 FIR + HB2 FIR + HB1 FIR (DCM1 = 16)
N
N+1
N+2
N+3
DCM means decimation.
Rev. 0 | Page 35 of 64
AD9684
Data Sheet
For example, if the chip decimation ratio is set to decimate by 4,
DDC 0 is set to use HB2 + HB1 filters (complex outputs,
decimate by 4) and DDC 1 is set to use HB4 + HB3 + HB2 +
HB1 filters (real outputs, decimate by 8). DDC 1 repeats its
output data two times for every one DDC 0 output. The
resulting output samples are shown in Table 15.
Table 15. Chip Decimation Ratio = 4, DDC 0 Decimation = 4 (Complex), and DDC 1 Decimation = 8 (Real)
DDC Input Samples
N
N+1
N+2
N+3
N+4
N+5
N+6
N+7
N+8
N+9
N + 10
N + 11
N + 12
N + 13
N + 14
N + 15
Output Port I
I0 (N)
DDC 0
Output Port Q
Q0 (N)
Output Port I
I1 (N)
DDC 1
Output Port Q
Not applicable
I0 (N + 1)
Q0 (N + 1)
I1 (N + 1)
Not applicable
I0 (N + 2)
Q0 (N + 2)
I1 (N)
Not applicable
I0 (N + 3)
Q0 (N + 3)
I1 (N + 1)
Not applicable
Rev. 0 | Page 36 of 64
Data Sheet
AD9684
FREQUENCY TRANSLATION
GENERAL DESCRIPTION
Variable IF Mode
Frequency translation is accomplished using a 12-bit complex
NCO with a digital quadrature mixer. The frequency translation
translates either a real or complex input signal from an IF to a
baseband complex digital output (carrier frequency = 0 Hz).
The NCO and the mixers are enabled. The NCO output
frequency can be used to digitally tune the IF frequency.
0 Hz IF (ZIF) Mode
The mixers are bypassed and the NCO is disabled.
The frequency translation stage of each DDC can be controlled
individually and supports four different IF modes using Bits[5:4] of
the DDC control registers (Register 0x310, Register 0x330,
Register 0x350, and Register 0x370). These IF modes are
The mixers and the NCO are enabled in a special downmixing
by fS/4 mode to save power.
Test Mode
Variable IF mode
0 Hz IF, or zero IF (ZIF), mode
fS/4 Hz IF mode
Test mode
The input samples are forced to 0.999 × full scale to positive full
scale. The NCO is enabled. This test mode allows the NCOs to
drive the decimation filters directly.
Figure 60 and Figure 61 show examples of the frequency
translation stage for both real and complex inputs.
NCO FREQUENCY TUNING WORD (FTW) SELECTION
12-BIT NCO FTW = MIXING FREQUENCY/ADC SAMPLE RATE × 4096
I
ADC + DIGITAL MIXER + NCO
REAL INPUT—SAMPLED AT fS
REAL
ADC
SAMPLING
AT fS
cos(wt)
REAL
12-BIT
NCO
90°
0°
COMPLEX
–sin(wt)
Q
BANDWIDTH OF
INTEREST
BANDWIDTH OF
INTEREST IMAGE
–fS/2
–fS/3
–fS/4
–fS/8
–fS/32
fS/32
DC
fS/16
–fS/16
fS/8
fS/4
fS/3
fS/2
–6dB LOSS DUE TO
NCO + MIXER
12-BIT NCO FTW =
ROUND ((fS/3)/fS × 4096) = +1365 (0x555)
POSITIVE FTW VALUES
–fS/32
DC
fS/32
12-BIT NCO FTW =
ROUND ((fS/3)/fS × 4096) = –1365 (0xAAB)
–fS/32
NEGATIVE FTW VALUES
DC
fS/32
Figure 60. DDC NCO Frequency Tuning Word Selection—Real Inputs
Rev. 0 | Page 37 of 64
13015-063
•
•
•
•
fS/4 Hz IF Mode
AD9684
Data Sheet
NCO FREQUENCY TUNING WORD (FTW) SELECTION
12-BIT NCO FTW = MIXING FREQUENCY/ADC SAMPLE RATE × 4096
I
+
I
I
I
Q
Q
90°
PHASE
12-BIT
NCO
90°
0°
Q
Q
ADC
SAMPLING
AT fS
Q
Q
I
I
–
–sin(wt)
QUADRATURE ANALOG MIXER +
2 ADCs + QUADRATURE DIGITAL REAL
MIXER + NCO
COMPLEX INPUT—SAMPLED AT fS
QUADRATURE MIXER
ADC
SAMPLING
AT fS
I
+
COMPLEX
Q
+
BANDWIDTH OF
INTEREST
IMAGE DUE TO
ANALOG I/Q
MISMATCH
–fS/3
–fS/4
–fS/32
fS/32
–fS/16
fS/16
DC
–fS/8
fS/8
fS/4
fS/3
fS/2
12-BIT NCO FTW =
ROUND ((fS/3)/fS × 4096) = +1365 (0x555)
POSITIVE FTW VALUES
–fS/32
fS/32
13015-064
–fS/2
DC
Figure 61. DDC NCO Frequency Tuning Word Selection—Complex Inputs
DDC NCO PLUS MIXER LOSS AND SFDR
Setting Up the NCO FTW and POW
When mixing a real input signal down to baseband, 6 dB of loss
is introduced in the signal due to filtering of the negative image.
The NCO introduces an additional 0.05 dB of loss. The total
loss of a real input signal mixed down to baseband is 6.05 dB. For
this reason, it is recommended to compensate for this loss by
enabling the 6 dB of gain in the gain stage of the DDC to
recenter the dynamic range of the signal within the full scale of
the output bits.
The NCO frequency value is given by the 12-bit, twos
complement number entered in the NCO FTW. Frequencies
between ±fS/2 (+fS/2 excluded) are represented using the
following frequency words:
When mixing a complex input signal down to baseband, the
maximum value that each I/Q sample can reach is 1.414 × full
scale after it passes through the complex mixer. To avoid an
overrange of the I/Q samples and to keep the data bit-widths
aligned with real mixing, introduce 3.06 dB of loss (0.707 × fullscale) in the mixer for complex signals. The NCO introduces an
additional 0.05 dB of loss. The total loss of a complex input
signal mixed down to baseband is −3.11 dB.
The worst case spurious signal from the NCO is greater than
102 dBc SFDR for all output frequencies.
NUMERICALLY CONTROLLED OSCILLATOR
The AD9684 has a 12-bit NCO for each DDC that enables the
frequency translation process. The NCO allows the input
spectrum to be tuned to dc, where it can be effectively filtered
by the subsequent filter blocks to prevent aliasing. The NCO
can be set up by providing a frequency tuning word (FTW) and
a phase offset word (POW).
•
•
•
0x800 represents a frequency of −fS/2.
0x000 represents dc (frequency is 0 Hz).
0x7FF represents a frequency of +fS/2 − fS/212.
Calculate the NCO frequency tuning word using the following
equation:
mod ( f C , f S )
NCO_ FTW = round 212
fS
where:
NCO_FTW is a 12-bit, twos complement number representing
the NCO FTW.
fC is the desired carrier frequency in Hz.
fS is the AD9684 sampling frequency (clock rate) in Hz.
mod( ) is a remainder function. For example, mod(110,100) =
10, and for negative numbers, mod(−32, +10) = −2.
round( ) is a rounding function. For example, round(3.6) = 4,
and for negative numbers, round(−3.4) = −3.
Note that this equation applies to the aliasing of signals in the
digital domain (that is, aliasing introduced when digitizing
analog signals).
Rev. 0 | Page 38 of 64
Data Sheet
AD9684
For example, if the ADC sampling frequency (fS) is 1250 MSPS
and the carrier frequency (fC) is 416.667 MHz,
Use the following two methods to synchronize multiple PAWs
within the chip:
mod( 416.667,1250 )
NCO _ FTW = round 212
= 1365 MHz
1250
•
This, in turn, converts to 0x555 in the 12-bit, twos complement
representation for NCO_FTW. Calculate the actual carrier
frequency based on the following equation:
f C _ ACTUAL =
NCO _ FTW × f S
= 416.56 MHz
212
•
A 12-bit POW is available for each NCO to create a known
phase relationship between multiple AD9684 chips or
individual DDC channels inside one AD9684 chip.
The following procedure must be followed to update the FTW
and/or POW registers to ensure proper operation of the NCO:
1.
2.
3.
Write to the FTW registers for all the DDCs.
Write to the POW registers for all the DDCs.
Synchronize the NCOs either through the DDC soft reset bit,
accessible through the SPI, or through the assertion of the
SYNC± pins.
Note that the NCOs must be synchronized either through the
SPI or through the SYNC± pins after all writes to the FTW or
POW registers are complete. This synchronization is necessary
to ensure the proper operation of the NCO.
NCO Synchronization
Each NCO contains a separate phase accumulator word (PAW)
that determines the instantaneous phase of the NCO. The initial
reset value of each PAW is determined by the POW, described
in the Setting Up the NCO FTW and POW section. The phase
increment value of each PAW is determined by the FTW.
Using the SPI. Use the DDC NCO soft reset bit in the DDC
synchronization control register (Register 0x300, Bit 4) to
reset all the PAWs in the chip. This is accomplished by
toggling the DDC NCO soft reset bit. Note that this
method synchronizes DDC channels within the same
AD9684 chip only.
Using the SYNC± pins. When the SYNC± pins are enabled
in the SYNC± control registers (Register 0x120 and
Register 0x121), and the DDC synchronization is enabled
in Bits[1:0] in the DDC synchronization control register
(Register 0x300), any subsequent SYNC± event resets all
the PAWs in the chip. Note that this method synchronizes
DDC channels within the same AD9684 chip or DDC
channels within separate AD9684 chips.
Mixer
The NCO is accompanied by a mixer, which operates similarly
to an analog quadrature mixer. It performs the downconversion
of input signals (real or complex) by using the NCO frequency
as a local oscillator. For real input signals, this mixer performs a
real mixer operation with two multipliers. For complex input
signals, the mixer performs a complex mixer operation with
four multipliers and two adders. The mixer adjusts its operation
based on the input signal (real or complex) provided to each
individual channel. The selection of real or complex inputs can be
controlled individually for each DDC block using Bit 7 of the DDC
control registers (Register 0x310, Register 0x330, Register 0x350,
and Register 0x370).
Rev. 0 | Page 39 of 64
AD9684
Data Sheet
FIR FILTERS
GENERAL DESCRIPTION
Table 16 shows the different bandwidth options by including
different half-band filters. In all cases, the DDC filtering stage of
the AD9684 provides less than −0.001 dB of pass-band ripple
and greater than 100 dB of stop band alias rejection.
There are four sets of decimate by 2, low-pass, half-band, FIR
filters (labeled HB1 FIR, HB2 FIR, HB3 FIR, and HB4 FIR in
Figure 58) following the frequency translation stage. After the
carrier of interest is tuned down to dc (carrier frequency =
0 Hz), these filters efficiently lower the sample rate while
providing sufficient alias rejection from unwanted adjacent
carriers around the bandwidth of interest.
Table 17 shows the amount of stop band alias rejection for
multiple pass-band ripple/cutoff points. The decimation ratio of
the filtering stage of each DDC can be controlled individually
through Bits[1:0] of the DDC control registers (Register 0x310,
Register 0x330, Register 0x350, and Register 0x370).
HB1 FIR is always enabled and cannot be bypassed. The HB2,
HB3, and HB4 FIR filters are optional and can be bypassed for
higher output sample rates.
Table 16. DDC Filter Characteristics
ADC Sample
Rate (MSPS)
1000
1
DDC Decimation
Ratio
2 (HB1)
4 (HB1 + HB2)
8 (HB1 + HB2 + HB3)
16 (HB1 + HB2 +
HB3 + HB4)
Real Output
Sample Rate
(MSPS)
1000
500
250
125
Complex (I/Q)
Output Sample
Rate (MSPS)
500 (I) + 500 (Q)
250 (I) + 250 (Q)
125 (I) + 125 (Q)
62.5 (I) + 62.5 (Q)
Alias Protected
Bandwidth
(MHz)
385.0
192.5
96.3
48.1
Ideal SNR
Improvement 1
(dB)
1
4
7
10
Pass-Band
Ripple
(dB)
100
The ideal SNR improvement due to oversampling and filtering = 10log(bandwidth/(fS/2)).
Table 17. DDC Filter Alias Rejection
Alias Rejection (dB)
>100
90
85
63.3
25
19.3
10.7
1
Pass-Band Ripple/Cutoff Point (dB)