AD AD9992_07

12-Bit CCD Signal Processor with
Precision Timing Generator
AD9992
FEATURES
GENERAL DESCRIPTION
1.8 V AFETG core
Internal LDO regulator and charge pump circuitry
Compatibility with 3 V or 1.8 V systems
24 programmable vertical clock outputs
Correlated double sampler (CDS) with −3 dB, 0 dB,
+3 dB, and +6 dB gain
6 dB to 42 dB, 10-bit variable gain amplifier (VGA)
12-bit, 40 MHz ADC
Black level clamp with variable level control
Complete on-chip timing generator
Precision Timing core with 400 ps resolution
On-chip 3 V horizontal and RG drivers
General-purpose outputs (GPOs) for shutter and
system support
On-chip driver for external crystal
On-chip sync generator with external sync input
105-lead CSP_BGA package, 8 mm × 8 mm, 0.65 mm pitch
The AD9992 is a highly integrated CCD signal processor for
digital still camera applications. It includes a complete analog
front end with analog to digital conversion combined with
a full-function programmable timing generator. The timing
generator is capable of supporting up to 24 vertical clock signals
to control advanced CCDs. A Precision Timing™ core allows
adjustment of high speed clocks with approximately 400 ps
resolution at 40 MHz operation. The AD9992 also contains
eight general-purpose inputs/outputs that can be used for
shutter and system functions.
The AD9992 is specified at pixel rates of up to 40 MHz. The
analog front end includes black level clamping, CDS, VGA, and
a 12-bit analog-to-digital converter (ADC). The timing generator
provides all the necessary CCD clocks: RG, H-clocks, V-clocks,
sensor gate pulses, substrate clock, and substrate bias control.
Operation is programmed using a 3-wire serial interface.
The AD9992 is specified over an operating temperature range
of −25°C to +85°C.
APPLICATIONS
Digital still cameras
FUNCTIONAL BLOCK DIAGRAM
REFT REFB
AD9992
6dB TO 42dB
CCDIN
CDS
VREF
–3dB, 0dB, +3dB, +6dB
3V INPUT
12
12-BIT
ADC
VGA
DOUT
CLAMP
LDO
REG
1.8V OUTPUT
1.8V INPUT
CHARGE
PUMP
3V OUTPUT
INTERNAL CLOCKS
RG
HL
8
HORIZONTAL
DRIVERS
PRECISION
TIMING
GENERATOR
SL
INTERNAL
REGISTERS
H1 TO H8
SDATA
24
XV1 TO XV24
XSUBCK
SCK
VERTICAL
TIMING
CONTROL
SYNC
GENERATOR
GPO1 TO GPO8
HD
VD SYNC CLI
CLO
05891-001
8
Figure 1.
Rev. C
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 ©2006–2007 Analog Devices, Inc. All rights reserved.
AD9992
TABLE OF CONTENTS
Features .............................................................................................. 1
Shutter Timing Control ............................................................. 47
Applications....................................................................................... 1
Substrate Clock Operation (SUBCK) ...................................... 47
General Description ......................................................................... 1
Field Counters............................................................................. 50
Functional Block Diagram .............................................................. 1
General-Purpose Outputs (GPOs) .......................................... 51
Revision History ............................................................................... 2
GP Look-Up Tables (LUT)........................................................ 55
Specifications..................................................................................... 3
Digital Specifications ................................................................... 4
Complete Exposure/Readout Operation Using Primary
Counter and GPO Signals ......................................................... 56
Analog Specifications................................................................... 5
Manual Shutter Operation Using Enhanced SYNC Modes.. 58
Timing Specifications .................................................................. 6
Analog Front End Description and Operation ...................... 62
Absolute Maximum Ratings............................................................ 7
Power-Up Sequence for Master Mode..................................... 64
Thermal Resistance ...................................................................... 7
Standby Mode Operation .......................................................... 67
ESD Caution.................................................................................. 7
CLI Frequency Change.............................................................. 67
Pin Configuration and Function Descriptions............................. 8
Circuit Layout Information........................................................... 69
Typical Performance Characteristics ........................................... 11
Typical 3 V System ..................................................................... 69
Equivalent Circuits ......................................................................... 12
Typical 1.8 V System .................................................................. 69
Terminology .................................................................................... 13
External Crystal Application .................................................... 69
System Overview ............................................................................ 14
Serial Interface Timing .............................................................. 72
High Speed Precision Timing Core........................................... 15
Layout of Internal Registers ...................................................... 73
Horizontal Clamping and Blanking......................................... 19
Updating New Register Values ................................................. 74
Horizontal Timing Sequence Example.................................... 25
Complete Register Listing ............................................................. 75
Vertical Timing Generation ...................................................... 26
Outline Dimensions ....................................................................... 92
Vertical Sequences (VSEQ) ....................................................... 29
Ordering Guide .......................................................................... 92
Vertical Timing Example........................................................... 45
REVISION HISTORY
10/07—Rev. B to Rev. C
Changes to Vertical Timing Generation Section........................ 26
Changes to Vertical Sequences (VSEQ) Section......................... 29
Changes to Vertical Timing Example Section ............................ 45
Changes to Power-Up Sequence for Master Mode Section ...... 64
Changes to Figure 80...................................................................... 70
Changes to Figure 81...................................................................... 71
9/07—Rev. A to Rev. B
Added Figure 2.................................................................................. 4
Deleted Endnote in Table 3 ............................................................. 5
Added Address 0x17 Bit 17 Information to Table 30................. 75
7/07—Rev. 0 to Rev. A
Changes to Table 3 and Related Endnote.......................................5
Added Slave Mode and SHP/SHD Information to Table 4..........6
Changes to Table 5.............................................................................7
Changes to Table 7.............................................................................8
Changes to Figure 18...................................................................... 17
Changes to Figure 75...................................................................... 66
Changes to Figure 81...................................................................... 71
1/06—Revision 0: Initial Version
Rev. C | Page 2 of 92
AD9992
SPECIFICATIONS
Table 1.
Parameter
TEMPERATURE RANGE
Operating
Storage
POWER SUPPLY VOLTAGE INPUTS
AVDD (AFE Analog Supply)
TCVDD (Timing Core Supply)
CLIVDD (CLI Input Supply)
RGVDD (RG, HL Driver)
HVDD1/HVDD2 (H1 to H8 Drivers) 1
DVDD (Digital Logic)
DRVDD (Parallel Data Output Drivers)
IOVDD (Digital I/O)
XVVDD (Vertical Output Drivers)
CP1P8 (CP Supply Input)
LDOIN (LDO Supply Input)
POWER SUPPLY CURRENTS—40 MHz OPERATION
AVDD (1.8 V)
TCVDD (1.8 V)
CLIVDD (3 V)
RGVDD (3.3 V, 20 pF RG Load, 20 pF HL Load)
HVDD1/HVDD2 (3.3 V, 480 pF Total Load on H1 to H8)1
DVDD (1.8 V)
DRVDD (3 V, 10 pF Load on Each DOUT Pin)
IOVDD (3 V, Depends on Load and Output Frequency of Digital I/O)
XVVDD (3 V, Depends on Load and Output Frequency of XV Signals)
POWER SUPPLY CURRENTS—STANDBY MODE OPERATION
Standby1 Mode
Standby2 Mode
Standby3 Mode
MAXIMUM CLOCK RATE (CLI)
1
Min
Typ
−25
−65
1.6
1.6
1.6
2.7
2.7
1.6
1.6
1.6
1.6
1.6
2.25
1.8
1.8
3.0
3.0
3.0
1.8
3.0
3.0
3.0
1.8
3.0
Unit
+85
+150
°C
°C
2.0
2.0
3.6
3.6
3.6
2.0
3.6
3.6
3.6
2.0
3.6
V
V
V
V
V
V
V
V
V
V
V
27
5
1.5
10
59
9.5
6
2
2
mA
mA
mA
mA
mA
mA
mA
mA
mA
12
5
1.5
mA
mA
mA
MHz
40
The total power dissipated by the HVDD (or RGVDD) supply can be approximated using the equation
Total HVDD Power = [CL × HVDD × Pixel Frequency] × HVDD
Reducing the capacitive load and/or reducing the HVDD supply reduces the power dissipation. CL is the total capacitance seen by all H-outputs.
Rev. C | Page 3 of 92
Max
AD9992
DIGITAL SPECIFICATIONS
IOVDD = 1.6 V to 3.6 V, RGVDD = HVDD = 2.7 V to 3.6 V, CL = 20 pF, TMIN to TMAX, unless otherwise noted.
Table 2.
Parameter
LOGIC INPUTS (IOVDD)
High Level Input Voltage
Low Level Input Voltage
High Level Input Current
Low Level Input Current
Input Capacitance
LOGIC OUTPUTS (IOVDD, XVDD)
High Level Output Voltage @ IOH = 2 mA
Low Level Output Voltage @ IOL = 2 mA
RG and H-DRIVER OUTPUTS (HVDD, RGVDD)
High Level Output Voltage @ Maximum Current
Low Level Output Voltage @ Maximum Current
Maximum Output Current (Programmable)
Maximum Load Capacitance (for Each Output)
Symbol
Min
VIH
VIL
IIH
IIL
CIN
VDD − 0.6
VOH
VOL
VDD − 0.5
VOH
VOL
VDD − 0.5
Typ
Max
0.6
10
10
10
0.5
0.5
18
60
MAXIMUM INPUT LIMIT = LESSER OF
2.2V or AVDD + 0.2V
+1.8V TYP (AVDD)
800mV
MAXIMUM
+1.3V TYP (AVDD – 0.5V)
DC RESTORE VOLTAGE
200mV MAX
OPTICAL BLACK PIXEL
1V MAXIMUM INPUT SIGNAL RANGE
(0dB CDS GAIN)
0V (AVSS)
MINIMUM INPUT LIMIT
(AVSS – 0.2V)
Figure 2. Input Signal Characteristics
(See Allowable OB Pixel Amplitude in Table 3)
Rev. C | Page 4 of 92
05891-091
500mV TYP
RESET TRANSIENT
Unit
V
V
μA
μA
pF
V
V
V
V
mA
pF
AD9992
ANALOG SPECIFICATIONS
AVDD = 1.8 V, fCLI = 40 MHz, typical timing specifications, TMIN to TMAX, unless otherwise noted.
Table 3.
Parameter
CDS
Allowable CCD Reset Transient
CDS Gain Accuracy
−3.0 dB CDS Gain
0 dB CDS Gain
+3 dB CDS Gain
+6 dB CDS Gain
Maximum Input Range Before Saturation
−3 dB CDS Gain
0 dB CDS Gain
+3 dB CDS Gain
+6 dB CDS Gain
Allowable OB Pixel Amplitude (See Figure 2)
0 dB CDS Gain (Default)
+6 dB CDS Gain
VARIABLE GAIN AMPLIFIER (VGA)
Gain Control Resolution
Gain Monotonicity
Gain Range
Low Gain (VGA Code 15, Default)
Maximum Gain (VGA Code 1023)
BLACK LEVEL CLAMP
Clamp Level Resolution
Clamp Level
Minimum Clamp Level (Code 0)
Maximum Clamp Level (Code 1023)
ADC
Resolution
Differential Nonlinearity (DNL)
No Missing Codes
Integral Nonlinearity (INL)
Full-Scale Input Voltage
VOLTAGE REFERENCE
Reference Top Voltage (REFT)
Reference Bottom Voltage (REFB)
SYSTEM PERFORMANCE
Gain Accuracy
Low Gain (VGA Code 15)
Maximum Gain (VGA Code 1023)
Peak Nonlinearity, 1.0 V Input Signal
Total Output Noise
Power Supply Rejection (PSR)
Min
−3.3
−0.5
2.4
5.0
Typ
Max
Unit
Test Conditions/Comments
0.5
0.8
V
The limit is the lower of AVDD + 0.2 V or 2.2 V.
VGA gain = 6.3 dB (Code 15, default value).
−2.8
0
2.9
5.5
−2.3
+0.5
3.4
6.0
dB
dB
dB
dB
VGA gain = 6.3 dB (Code 15, default value).
1.4
1.0
0.7
0.5
−100
−50
V p-p
V p-p
V p-p
V p-p
+200
+100
1024
Guaranteed
mV
mV
Steps
6.3
42.4
dB
dB
1024
Steps
0
255
LSB
LSB
Measured at ADC output.
12
−1.0
±0.5
+1.0
Guaranteed
1
4
2.0
1.4
0.4
5.8
41.9
6.3
42.4
0.1
0.5
50
Bits
LSB
LSB
V
V
V
6.8
42.9
0.2
Rev. C | Page 5 of 92
dB
dB
%
LSB rms
dB
Includes entire signal chain.
0 dB CDS gain.
Gain = (0.0358 × Code) + 5.76 dB.
6 dB VGA gain, 0 dB CDS gain applied.
AC-grounded input, 6 dB VGA gain applied.
Measured with step change on supply.
AD9992
TIMING SPECIFICATIONS
CL = 20 pF, AVDD = DVDD = TCVDD = 1.8 V, DRVDD = 3.0 V, fCLI = 40 MHz, unless otherwise noted.
Table 4.
Parameter
MASTER CLOCK (See Figure 15)
CLI Clock Period
CLI High/Low Pulse Width
Delay from CLI Rising Edge to Internal Pixel Position 0
SLAVE MODE SPECIFICATIONS (See Figure 76)
VD Falling Edge to HD Falling Edge in Slave Mode
2
Min
tCONV
25
10
tCLIDLY
HD Edge to CLI Rising Edge (Only Valid if OSC_RSTB = LO)
HD Edge to CLO Rising Edge (Only Valid if OSC_RSTB = HI)
Inhibit Region for SHP Edge Location
AFE CLPOB PULSE WIDTH (See Figure 22 and Figure 32) 1, 2
AFE SAMPLE LOCATION (See Figure 16 and Figure 19)1
SHP Sample Edge to SHD Sample Edge
DATA OUTPUTS (See Figure 20 and Figure 21)
Output Delay from DCLK Rising Edge
Inhibited Area for DOUTPHASE Edge Location
Pipeline Delay from SHP/SHD Sampling to DOUT
SERIAL INTERFACE (See Figure 83)
Maximum SCK Frequency (Must Not Exceed CLI Frequency)
SL to SCK Setup Time
SCK to SL Hold Time
SDATA Valid to SCK Rising Edge Setup
SCK Falling Edge to SDATA Valid Hold
INHIBIT REGION FOR SHP AND SHD WITH RESPECT TO H-CLOCK
EDGE PLACEMENT (see Figure 19) for H*POL = 1
RETIME = 0, MASK = 0
RETIME = 0, MASK = 1
RETIME = 1, MASK = 0
RETIME = 1, MASK = 1
1
Symbol
Typ
Max
Unit
12.5
6
15
ns
ns
ns
tVDHD
0
tHDCLI
tHDCLO
tSHPINH
3
3
48
2
20
ns
ns
Edge location
Pixels
tS1
11
12.5
ns
tOD
tDOUTINH
SHDLOC + 1
VD period − 5 ×
tCONV
tCONV − 2
tCONV − 2
63
1
SHDLOC + 15
16
fSCLK
tLS
tLH
tDS
tDH
40
10
10
10
10
tSHDINH
tSHDINH
tSHPINH
tSHPINH
H*NEGLOC – 15
H*POSLOC – 15
H*NEGLOC – 15
H*POSLOC – 15
ns
Edge location
Cycles
MHz
ns
ns
ns
ns
H*NEGLOC – 0
H*POSLOC – 0
H*NEGLOC – 0
H*POSLOC – 0
Parameter is programmable.
Minimum CLPOB pulse width is for functional operation only. Wider typical pulses are recommended to achieve good clamp performance.
Rev. C | Page 6 of 92
ns
Edge location
Edge location
Edge location
Edge location
AD9992
ABSOLUTE MAXIMUM RATINGS
Table 5.
Parameter
AVDD
TCVDD
HVDD1/HVDD2
RGVDD
DVDD
DRVDD
IOVDD
XVVDD
CLIVDD
CP1P8
RG Output
H1 to H8, HL Output
Digital Outputs
Digital Inputs
SCK, SL, SDATA
REFT, REFB, CCDIN
Junction Temperature
Lead Temperature,
10 sec
With
Respect To
AVSS
TCVSS
HVSS1/HVSS2
RGVSS
DVSS
DRVSS
DVSS
DVSS
TCVSS
CPVSS
RGVSS
HVSS1/HVSS2
DVSS
DVSS
DVSS
AVSS
Rating
−0.2 V to +2.2 V
−0.2 V to +2.2 V
−0.3 V to +3.9 V
−0.3 V to +3.9 V
−0.2 V to +2.2 V
−0.3 V to +3.9 V
−0.3 V to +3.9 V
−0.3 V to +3.9 V
−0.3 V to +3.9 V
−0.2 V to +2.2 V
−0.3 V to RGVDD + 0.3 V
−0.3 V to HVDD + 0.3 V
−0.3 V to IOVDD + 0.3 V
−0.3 V to IOVDD + 0.3 V
−0.3 V to IOVDD + 0.3 V
−0.2 V to AVDD + 0.2 V
150°C
350°C
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.
THERMAL RESISTANCE
θJA is specified for the worst-case conditions, that is, a device
soldered in a circuit board for surface-mount packages.
Table 6. Thermal Resistance
Package Type
106-Lead CSP_BGA
ESD CAUTION
Rev. C | Page 7 of 92
θJA
40.3
Unit
°C/W
AD9992
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
AD9992
A1 CORNER
INDEX AREA
1
TOP VIEW
(Not to Scale)
2
3
4
5
6
7
8
9 10 11
A
B
C
D
E
F
G
H
L
05891-003
J
K
Figure 3. Pin Configuration
Table 7. Pin Function Descriptions
Pin No.
A1
B2
C2
B1
B4
C1
D2
C3
D3
E2
D1
E7
E6
E5
E3
E1
F2
F3
F7
G3
F5
F6
G2
F1
G1
G5
H2
H1
G6
G7
J2
J1
L1
L2
Mnemonic
GPO8
GPO7
GPO6
GPO5
GPO4
GPO3
GPO2
GPO1
SYNC
VD
HD
RSTB
IOVDD
IOVSS
XVVDD
XSUBCK
XV1
XV2
XV3
XV4
XV5
XV6
XV7
XV8
XV9
XV10
XV11
XV12
XV13
XV14
XV15
XV16
XV17
XV18
Type 1
DIO
DIO
DIO
DIO
DIO
DIO
DIO
DIO
DI
DIO
DIO
DI
P
P
P
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
Description
General-Purpose Output 8
General-Purpose Output 7
General-Purpose Output 6
General-Purpose Output 5
General-Purpose Output 4
General-Purpose Output 3
General-Purpose Output 2
General-Purpose Output 1
External System Sync Input
Vertical Sync Pulse (input for slave mode, output for master mode)
Horizontal Sync Pulse (input for slave mode, output for master mode)
External Reset Input (active low pulse to reset, internal pull-up resistor)
Digital I/O Supply: 1.8 V, 3.0 V (GPO, SUBCK, HD/VD, SL, SCK, SDATA, SYNC, RSTB)
Digital I/O Ground
XV Output Supply: 1.8 V, 3.0 V
CCD Substrate Clock
CCD Vertical Clock 1
CCD Vertical Clock 2
CCD Vertical Clock 3
CCD Vertical Clock 4
CCD Vertical Clock 5
CCD Vertical Clock 6
CCD Vertical Clock 7
CCD Vertical Clock 8
CCD Vertical Clock 9
CCD Vertical Clock 10
CCD Vertical Clock 11
CCD Vertical Clock 12
CCD Vertical Clock 13
CCD Vertical Clock 14
CCD Vertical Clock 15
CCD Vertical Clock 16
CCD Vertical Clock 17
CCD Vertical Clock 18
Rev. C | Page 8 of 92
AD9992
Pin No.
L3
K1
K2
K3
J3
H3
L4
K4
L6
K6
J6
L7
K7
J7
L8
K8
J8
L9
K9
J9
L10
K10
L11
K11
J10
J11
H10
H11
H9
G11
G9
F10
F11
E11
E10
F9
E9
D11
C11
D10
C10
D9
C9
B11
A11
B10
A10
B9
B8
A8
C8
B7
A7
Mnemonic
XV19
XV20
XV21
XV22
XV23
XV24
DVDD
DVSS
D0
D1
D2
D3
D4
D5
D6
D7
D8
D9
D10
D11
DCLK
DRVSS
DRVDD
CP3P3
CPFCT
CPFCB
CPVSS
CP1P8
CPCLI
LDO3P2EN
LDOVSS
LDO1P8EN
SENSE
LDOOUT
LDOIN
H1
H2
HVSS1
HVDD1
H3
H4
H5
H6
HVSS2
HVDD2
H7
H8
HL
RGVSS
RGVDD
RG
TCVSS
TCVDD
Type 1
DO
DO
DO
DO
DO
DO
P
P
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
DO
P
P
P
AO
AO
P
P
DI
DI
P
DI
AI
AO
P
DO
DO
P
P
DO
DO
DO
DO
P
P
DO
DO
DO
P
P
DO
P
P
Description
CCD Vertical Clock 19
CCD Vertical Clock 20
CCD Vertical Clock 21
CCD Vertical Clock 22
CCD Vertical Clock 23
CCD Vertical Clock 24
Digital Logic Supply: 1.8 V
Digital Logic Ground
Data Output 0 (LSB)
Data Output 1
Data Output 2
Data Output 3
Data Output 4
Data Output 5
Data Output 6
Data Output 7
Data Output 8
Data Output 9
Data Output 10
Data Output 11 (MSB)
Data Clock Output
Data Driver Ground
Data Driver Supply: 1.8 V, 3.0 V
Charge Pump 3.3 V Output
Charge Pump Flying Capacitor Top
Charge Pump Flying Capacitor Bottom
Charge Pump Ground
Charge Pump 1.8 V Input
Charge Pump Clock Input
LDO 3.2 V Output Enable
LDO Ground
LDO 1.8 V Output Enable
LDO Output Sense
LDO Output Voltage
LDO 3.3 V Input
CCD Horizontal Clock 1
CCD Horizontal Clock 2
H-Driver Ground 1
H-Driver Supply 1: 3.3 V
CCD Horizontal Clock 3
CCD Horizontal Clock 4
CCD Horizontal Clock 5
CCD Horizontal Clock 6
H-Driver Ground 2
H-Driver Supply 2: 3.3 V
CCD Horizontal Clock 7
CCD Horizontal Clock 8
CCD Last Horizontal Clock
RG Driver Ground
RG Driver Supply: 3.3 V
CCD Reset Gate Clock
Analog Ground for Timing Core
Timing Core Supply: 1.8 V
Rev. C | Page 9 of 92
AD9992
Pin No.
C7
C6
C5
B6
A6
B5, A5
A4
A3
C4
A2
B3
A9, G10, K5, J4, J5, L5
1
Mnemonic
CLIVDD
CLO
CLI
AVDD
CCDIN
AVSS
REFT
REFB
SL
SDATA
SCK
NC
Type 1
P
DO
DI
P
AI
P
AO
AO
DI
DI
DI
Description
CLI Input Supply: 3.0 V
Clock Output for Crystal
Reference Clock Input
AFE Supply: 1.8 V
CCD Signal Input
Analog Supply Ground
Voltage Reference Top Bypass
Voltage Reference Bottom Bypass
3-Wire Serial Load Pulse (internal pull-up resistor)
3-Wire Serial Data Input
3-Wire Serial Clock
Not Internally Connected
DIO = digital input/output, DI = digital input, P = power, DO = digital output, AI = analog input, AO = analog output.
Rev. C | Page 10 of 92
AD9992
TYPICAL PERFORMANCE CHARACTERISTICS
500
150
+3dB CDS
450
400
POWER (mW)
RMS OUTPUT NOISE (LSB)
3.3V/1.8V
350
3.0V/1.8V
300
250
2.7V/1.8V
200
150
100
0dB CDS
–3dB CDS
100
50
20
25
30
FREQUENCY (MHz)
35
40
0
05891-004
0
15
0
Figure 4. Power vs. Frequency
(AVDD = TCVDD = DVDD = 1.8 V, All Other Supplies at 2.7 V, 3.0 V, or 3.3 V)
5
10
15
20
25
30
CDS + VGA GAIN (dB)
35
40
45
05891-006
50
Figure 6. Output Noise vs. Total Gain (CDS + VGA)
1.0
5
0.8
4
0.6
3
INL (LSB)
0.2
0
–0.2
–0.4
2
1
0
–0.6
–1.0
0
500
1000
1500
2000 2500
CODE
3000
3500
4000
Figure 5. Typical Differential Nonlinearity (DNL) Performance
–2
0
500
1000
1500
2000 2500
CODE
3000
3500
4000
Figure 7. Typical Integral Nonlinearity (INL) Performance
Rev. C | Page 11 of 92
05891-090
–1
–0.8
05891-089
DNL (LSB)
0.4
AD9992
EQUIVALENT CIRCUITS
AVDD
IOVDD
R
AVSS
AVSS
330Ω
05891-008
DIGITAL INPUT
IOVSS
Figure 10. Digital Inputs
Figure 8. CCDIN
DVDD
05891-010
CCDIN
HVDD OR
RGVDD
DRVDD
DATA
RG, H1 TO H8
DOUT
DRVSS
05891-009
DVSS
OUTPUT
THREE-STATE
HVSS OR
RGVSS
Figure 11. H1 to H8, HL, RG Drivers
Figure 9. Digital Data Outputs
Rev. C | Page 12 of 92
05891-011
THREESTATE
AD9992
TERMINOLOGY
Differential Nonlinearity (DNL)
An ideal ADC exhibits code transitions that are exactly 1 LSB
apart. DNL is the deviation from this ideal value. Therefore,
every code must have a finite width. No missing codes
guaranteed to 12-bit resolution indicates that all 4096 codes
must be present over all operating conditions.
Peak Nonlinearity
Peak nonlinearity, a full signal chain specification, refers to the
peak deviation of the output of the AD9992 from a true straight
line. The point used as zero scale occurs 0.5 LSB before the first
code transition. Positive full scale is defined as a level 1 LSB and
0.5 LSB beyond the last code transition. The deviation is measured
from the middle of each particular output code to the true straight
line. The error is then expressed as a percentage of the 2 V ADC
full-scale signal. The input signal is always appropriately amplified
to fill the full-scale range of the ADC.
Total Output Noise
The rms output noise is measured using histogram techniques.
The standard deviation of the ADC output codes is calculated
in LSB and represents the rms noise level of the total signal
chain at the specified gain setting. The output noise can be converted to an equivalent voltage using the relationship
1 LSB = (ADC Full Scale/2n Codes)
where n is the bit resolution of the ADC. For the AD9992, 1 LSB
is 0.488 mV.
Power Supply Rejection (PSR)
The PSR is measured with a step change applied to the supply
pins. The PSR specification is calculated from the change in
data outputs for a given step change in the supply voltage.
Rev. C | Page 13 of 92
AD9992
SYSTEM OVERVIEW
Figure 12 shows the typical system block diagram for the AD9992
in master mode. The CCD output is processed by AD9992 AFE
circuitry, which consists of a CDS, VGA, black level clamp, and
ADC. The digitized pixel information is sent to the digital
image processor chip, which performs the postprocessing and
compression. To operate the CCD, all CCD timing parameters
are programmed into the AD9992 from the system microprocessor
through the 3-wire serial interface. The AD9992 generates the
CCD’s horizontal and vertical clocks and internal AFE clocks
from the master clock, CLI, which is provided by the image
processor or external crystal,. External synchronization is provided
by a sync pulse from the microprocessor, which resets the
internal counters and resyncs the VD and HD outputs.
XV1 TO XV24, XSUBCK
V-DRIVER
H1 TO H8, HL,
RG
DOUT
CCDIN
CCD
AD9992
AFETG
GPO1 TO GPO8
DCLK
HD, VD
CLI
Alternatively, the AD9992 can operate in slave mode. In slave
mode, the VD and HD are provided externally from the image
processor, and all AD9992 timing synchronizes with VD and HD.
H-drivers for H1 to H8, HL, and RG are included in the AD9992,
allowing these clocks to be directly connected to the CCD.
An H-driver voltage of up to 3.3 V is supported. An external
V-driver is required for the vertical transfer clocks, the sensor
gate pulses, and the substrate clock.
The AD9992 includes programmable general-purpose outputs
(GPO), which can trigger mechanical shutter and strobe (flash)
circuitry.
Figure 13 and Figure 14 show the maximum horizontal and
vertical counter dimensions for the AD9992. All internal
horizontal and vertical clocking is controlled by these counters,
which specify line and pixel locations. Maximum HD length
is 8192 pixels per line; maximum VD length is 8192 lines per field.
MAXIMUM COUNTER DIMENSIONS
DIGITAL
IMAGE
PROCESSING
ASIC
13-BIT HORIZONTAL = 8192 PIXELS MAX
SERIAL
INTERFACE
MICROPROCESSOR
05891-012
SYNC
13-BIT VERTICAL = 8192 LINES MAX
05891-013
Figure 12. Typical System Block Diagram, Master Mode
Figure 13. Vertical and Horizontal Counters
MAX VD LENGTH IS 8192 LINES
VD
MAX HD LENGTH IS 8192 PIXELS
HD
05891-014
CLI
Figure 14. Maximum VD/HD Dimensions
Rev. C | Page 14 of 92
AD9992
clock is not available, it is possible to use a 2× reference clock by
programming the CLIDIVIDE register (AFE Register
Address 0x0D). The AD9992 then internally divides the CLI
frequency by 2.
HIGH SPEED PRECISION TIMING CORE
The AD9992 generates high speed timing signals using the flexible
Precision Timing core. This core is the foundation for generating
timing used for both the CCD and the AFE; it includes the reset
gate RG, horizontal drivers H1 to H8, HL, and SHP/SHD sample
clocks. A unique architecture makes it routine for the system
designer to optimize image quality by providing precise control
over the horizontal CCD readout and the AFE correlated
double sampling.
The AD9992 includes a master clock output (CLO) which
is the inverse of CLI. This output should be used as a crystal
driver. A crystal can be placed between the CLI and CLO pins
to generate the master clock for the AD9992.
High Speed Clock Programmability
The high speed timing of the AD9992 operates the same way in
either master or slave mode configuration. For more information on synchronization and pipeline delays, see the Power-Up
Sequence for Master Mode section.
Figure 16 shows when the high speed clocks RG, H1 to H8,
SHP, and SHD are generated. The RG pulse has programmable
rising and falling edges and can be inverted using the polarity
control. Horizontal Clock H1 has programmable rising and
falling edges and polarity control. In HCLK Mode 1, H3, H5,
and H7 are equal to H1. H2, H4, H6, and H8 are always inverses
of H1.
Timing Resolution
The Precision Timing core uses a 1× master clock input (CLI) as
a reference. This clock should be the same as the CCD pixel
clock frequency. Figure 15 illustrates how the internal timing
core divides the master clock period into 64 steps or edge
positions. Using a 40 MHz CLI frequency, the edge resolution of
the Precision Timing core is approximately 0.4 ns. If a 1× system
POSITION
P[0]
P[16]
The edge location registers are each six bits wide, allowing
selection of all 64 edge locations. Figure 19 shows the default
timing locations for all high speed clock signals.
P[32]
P[48]
P[64] = P[0]
CLI
tCLIDLY
ONE PIXEL
PERIOD
05891-015
NOTES
1. THE PIXEL CLOCK PERIOD IS DIVIDED INTO 64 POSITIONS, PROVIDING FINE EDGE RESOLUTION FOR HIGH SPEED CLOCKS.
2. THERE IS A FIXED DELAY FROM THE CLI INPUT TO THE INTERNAL PIXEL PERIOD POSITION (tCLIDLY).
Figure 15. High Speed Clock Resolution from CLI, Master Clock Input
1
2
CCD
SIGNAL
3
4
RG
5
6
7
8
H1, H3, H5, H7
H2, H4, H6, H8
HL
PROGRAMMABLE CLOCK POSITIONS:
05891-016
1SHP SAMPLE LOCATION.
2SHD SAMPLE LOCATION.
3RG RISING EDGE.
4RG FALLING EDGE.
5H1 RISING EDGE.
6H1 FALLING EDGE.
7HL RISING EDGE.
8HL FALLING EDGE.
Figure 16. High Speed Clock Programmable Locations (HCLKMODE = 001)
Rev. C | Page 15 of 92
AD9992
H-Driver and RG Outputs
In addition to the programmable timing positions, the AD9992
features on-chip output drivers for the RG, HL, and H1 to H8
outputs. These drivers are powerful enough to drive the CCD
inputs directly. The H-driver and RG current can be adjusted
for optimum rise/fall time for a particular load by using the drive
strength control registers (Address 0x35 and Address 0x36).
The 3-bit drive setting for each output is adjustable in 4.3 mA
increments: 0 = three-state; 1 = 4.3 mA; 2 = 8.6 mA; 3 = 12.9 mA;
and 4, 5, 6, 7 = 17.2 mA.
As shown in Figure 16, when HCLK Mode 1 is used, the H2,
H4, H6, and H8 outputs are inverses of the H1, H3, H5, and H7
outputs, respectively. Using the HCLKMODE register (Address
0x23, Bits [9:7]), it is possible to select a different configuration.
Table 9 shows a comparison of the different programmable
settings for each HCLK mode. Figure 17 and Figure 18 show
the settings for HCLK Mode 2 and HCLK Mode 3, respectively.
It is recommended that all H1 to H8 outputs on the AD9992
be used together for maximum flexibility in drive strength
settings. A typical CCD with H1 and H2 inputs should only
have the AD9992 H1, H3, H5, and H7 outputs connected
together to drive the CCD’s H1, and the H2, H4, H6, and H8
outputs connected together to drive the CCD’s H2. Similarly,
a CCD with H1, H2, H3, and H4 inputs should have
•
•
•
•
H1 and H3 connected to the CCD’s H1.
H2 and H4 connected to the CCD’s H2.
H5 and H7 connected to the CCD’s H3.
H6 and H8 connected to the CCD’s H4.
Table 8. Timing Core Register Parameters for H1, H2, HL, RG, SHP, SHD
Parameter
Polarity
Positive Edge
Negative Edge
Sampling Location
Drive Strength
Length
1b
6b
6b
6b
3b
Range
High/low
0 to 63 edge location
0 to 63 edge location
0 to 63 edge location
0 to 4 current steps
Description
Polarity control for H1, H2, HL, and RG (0 = inversion, 1 = no inversion)
Positive edge location for H1, H2, HL, and RG
Negative edge location for H1, H2, HL, and RG
Sampling location for internal SHP and SHD signals
Drive current for H1 to H8 , HL, and RG outputs (4.3 mA per step)
Table 9. HCLK Modes, Selected by Address 0x23, Bits[9:7]
HCLKMODE
Mode 1
Mode 2
Register Value
001
010
Mode 3
100
Invalid Selection
000, 011, 101, 110, 111
1
Description
H1 edges are programmable, with H3 = H5 = H7 = H1, H2 = H4 = H6 = H8 = inverse of H1
H1 edges are programmable, with H3 = H5 = H7 = H1
H2 edges are programmable, with H4 = H6 = H8 = H2
H1 edges are programmable, with H3 = H1 and H2 = H4 = inverse of H1
H5 edges are programmable, with H7 = H5 and H6 = H8 = inverse of H5
Invalid register settings
2
H1, H3, H5, H7
4
3
H2, H4, H6, H8
05891-017
H1 TO H8 PROGRAMMABLE LOCATIONS:
1H1 RISING EDGE.
2H1 FALLING EDGE.
3H2 RISING EDGE.
4H2 FALLING EDGE.
Figure 17. HCLK Mode 2 Operation
Rev. C | Page 16 of 92
AD9992
1
2
H1, H3
H2, H4
3
4
H5, H7
H6, H8
05891-018
H1 TO H8 PROGRAMMABLE EDGES:
1H1 RISING EDGE.
2H1 FALLING EDGE.
3H5 RISING EDGE.
4H5 FALLING EDGE.
Figure 18. HCLK Mode 3 Operation
POSITION
P[0]
P[16]
RGr[0]
RGf[16]
P[32]
P[48]
P[64] = P[0]
CLI
RG
H1r[0]
H1f[32]
48
tSHDINH
H1
63
tSHDINH
H2
tS2
tS1
CCD
SIGNAL
SHPLOC[32]
tSHPINH
SHP
SHDLOC[0]
SHD
1
12
tDOUTINH
NOTES:
1. ALL SIGNAL EDGES ARE FULLY PROGRAMMABLE TO ANY OF THE 64 POSITIONS WITHIN ONE PIXEL PERIOD.
TYPICAL POSITIONS FOR EACH SIGNAL ARE SHOWN. HCLK MODE 1 IS SHOWN.
2. CERTAIN POSITIONS SHOULD BE AVOIDED FOR EACH SIGNAL, SHOWN ABOVE AS INHIBIT REGIONS.
3. IF A SETTING IN THE INHIBIT REGION IS USED, AN UNSTABLE PIXEL SHIFT CAN OCCUR IN THE HBLK LOCATION OR AFE PIPELINE.
4. THE tSHPINH AREA FROM 50 TO 62 ONLY APPLIES IN SLAVE MODE.
5. THE tSHDINH AREA WLL APPLY TO EITHER H1 RISING OR FALLING EDGE, DEPENDING ON THE VALUE OF THE
H1HBLK MASKING POLARITY.
6. THE tSHDINH AREA CAN ALSO BE CHANGED TO A tSHPINH AREA IF THE H1HBLKRETIME BIT = 1.
05891-093
DOUTPHASEP
Figure 19. High Speed Timing Default Locations
Digital Data Outputs
The AD9992 data output and DCLK phase are programmable
using the DOUTPHASE registers (Address 0x38, Bits [11:0]).
DOUTPHASEP (Bits [5:0]) selects any edge location from 0 to
63, as shown in Figure 20. DOUTPHASEN (Bits [11:6]) does
not actually program the phase of the data outputs but is used
internally and should always be programmed to a value of
DOUTPHASEP plus 32 edges. For example, if DOUTPHASEP
is set to 0, DOUTPHASEN should be set to 32 (0x20).
Normally, the DOUT and DCLK signals track in phase, based
on the contents of the DOUTPHASE registers. The DCLK output
phase can also be held fixed with respect to the data outputs by
changing the DCLKMODE register high (Address 0x38, Bit 12). In
this mode, the DCLK output remains at a fixed phase equal to a
delayed version of CLI while the data output phase is still
programmable.
The pipeline delay through the AD9992 is shown in Figure 21.
After the CCD input is sampled by SHD, there is a 16-cycle
delay until the data is available.
Rev. C | Page 17 of 92
AD9992
P[0]
P[16]
P[48]
P[32]
P[64] = P[0]
PIXEL
PERIOD
DCLK
tOD
DOUT
05891-020
NOTES
1. DATA OUTPUT (DOUT) AND DCLK PHASE ARE ADJUSTABLE WITH RESPECT TO THE PIXEL PERIOD.
2. WITHIN ONE CLOCK PERIOD, THE DATA TRANSITION CAN BE PROGRAMMED TO 64 DIFFERENT LOCATIONS.
3. DCLK CAN BE INVERTED WITH RESPECT TO DOUT BY USING THE DCLKINV REGISTER.
Figure 20. Digital Output Phase Adjustment Using DOUTPHASEP Register
CLI
tCLIDLY
N
N+1
N+2
N+3
N+4
N – 14
N – 13
N+5
N+6
N+7
N – 11
N – 10
N+8
N+9
N + 10
N + 11
N + 12
N + 13
N + 14
N + 15
N + 16
N–7
N–6
N–5
N–4
N–3
N–2
N–1
N + 17
CCDIN
SAMPLE PIXEL N
SHD
(INTERNAL)
ADC DOUT
(INTERNAL)
N – 17
N – 16
N – 15
N – 12
N–9
N–8
N
N+1
tDOUTINH
DCLK
PIPELINE LATENCY = 16 CYCLES
N – 17
N – 16
N – 15
N – 14
N – 13
N – 12
N – 11
N – 10
N–9
N–8
N–7
N–6
N–5
N–4
NOTES
1. TIMING VALUES SHOWN ARE SHDLOC = 0, WITH DCLKMODE = 0.
2. HIGHER VALUES OF SHD AND/OR DOUT PHASE SHIFTS DOUT TRANSITION TO THE RIGHT, WITH RESPECT TO CLI LOCATION.
3. RECOMMENDED VALUE FOR DOUT PHASE IS TO USE SHPLOC OR UP TO 15 EDGES FOLLOWING SHPLOC.
Figure 21. Digital Data Output Pipeline Delay
Rev. C | Page 18 of 92
N–3
N–2
N–1
N
N+1
05891-021
DOUT
AD9992
HORIZONTAL CLAMPING AND BLANKING
CLPOB and PBLK Masking Areas
The horizontal clamping and blanking pulses of the AD9992
are fully programmable to suit a variety of applications.
Individual control is provided for CLPOB, PBLK, and HBLK in
the different regions of each field. This allows the dark pixel
clamping and blanking patterns to be changed at each stage of
the readout to accommodate different image transfer timing
and high speed line shifts.
Additionally, the AD9992 allows the CLPOB and PBLK signals to
be disabled in certain lines in the field without changing any of
the existing CLPOB pattern settings.
Individual CLPOB and PBLK Patterns
The AFE horizontal timing consists of CLPOB and PBLK, as
shown in Figure 22. These two signals are programmed
independently using the registers listed in Table 10. The start
polarity for the CLPOB (and PBLK) signal is CLPOBPOL
(PBLKPOL), and the first and second toggle positions of the pulse
are CLPOBTOG1 (PBLKTOG1) and CLPOBTOG2 (PBLKTOG2).
Both signals are active low and should be programmed
accordingly.
A separate pattern for CLPOB and PBLK can be programmed
for each vertical sequence. As described in the Vertical Timing
Generation section, several V-sequences can be created, each
containing a unique pulse pattern for CLPOB and PBLK. Figure 48
shows how the sequence change positions divide the readout field
into different regions. By assigning a different V-sequence to
each region, the CLPOB and PBLK signals can change with
each change in the vertical timing.
To use CLPOB (or PBLK) masking, the CLPMASKSTART
(PBLKMASKSTART) and CLPMASKEND (PBLKMASKEND)
registers are programmed to specify the start and end lines in
the field where the CLPOB (PBLK) patterns are ignored. The three
sets of start and end registers allow up to three CLPOB (PBLK)
masking areas to be created.
The CLPOB and PBLK masking registers are not specific to
a certain V-sequence; they are always active for any existing field of
timing. During operation, to disable the CLPOB masking
feature, these registers must be set to the maximum value of
0x1FFF or a value greater than the programmed VD length.
Note that to disable CLPOB (and PBLK) masking during power-up,
it is recommended to set CLPMASKSTART (PBLKMASKSTART)
to 8191 and CLPMASKEND (PBLKMASKEND) to 0. This
prevents any accidental masking caused by register update events.
Table 10. CLPOB and PBLK Pattern Registers
Register
CLPOBPOL
PBLKPOL
CLPOBTOG1
CLPOBTOG2
PBLKTOG1
PBLKTOG2
CLPMASKSTART
CLPMASKEND
PBLKMASKSTART
PBLKMASKEND
Length
1b
1b
13b
13b
13b
13b
13b
13b
13b
13b
Range
High/low
High/low
0 to 8191 pixel locations
0 to 8191 pixel locations
0 to 8191 pixel locations
0 to 8191 pixel locations
0 to 8191 line locations
0 to 8191 line locations
0 to 8191 line locations
0 to 8191 line locations
Description
Starting polarity of CLPOB for each V-sequence.
Starting polarity of PBLK for each V-sequence.
First CLPOB toggle position within line for each V-sequence.
Second CLPOB toggle position within line for each V-sequence.
First PBLK toggle position within line for each V-sequence.
Second PBLK toggle position within line for each V-sequence.
CLPOB masking area—starting line within field (maximum of three areas).
CLPOB masking area—ending line within field (maximum of three areas).
PBLK masking area—starting line within field (maximum of three areas).
PBLK masking area—ending line within field (maximum of three areas).
Rev. C | Page 19 of 92
AD9992
HD
2
CLPOB 1
PBLK
3
ACTIVE
ACTIVE
PROGRAMMABLE SETTINGS:
05891-022
1START POLARITY (CLAMP AND BLANK REGION ARE ACTIVE LOW).
2FIRST TOGGLE POSITION.
3SECOND TOGGLE POSITION.
Figure 22. Clamp and Preblank Pulse Placement
NO CLPOB SIGNAL
FOR LINE 600
NO CLPOB SIGNAL
FOR LINES 6 TO 8
VD
0
1
2
597 598
HD
CLPMASKSTART1 = 6
CLPMASKEND1 = 8
CLPMASKSTART2 = CLPMASKEND2 = 600
05891-023
CLPOB
Figure 23. CLPOB Masking Example
Individual HBLK Patterns
The HBLK programmable timing shown in Figure 24 is similar
to CLPOB and PBLK; however, there is no start polarity control.
Only the toggle positions are used to designate the start and
stop positions of the blanking period. Additionally, there are
separate masking polarity controls for H1, H2, and HL that
designate the polarity of the horizontal clock signals during
the blanking period. Setting HBLKMASK_H1 high sets H1, and
therefore H3, H5, and H7, low during the blanking, as shown in
Figure 25. As with the CLPOB and PBLK signals, HBLK
registers are available in each V-sequence, allowing different
blanking signals to be used with different vertical timing
sequences.
The AD9992 supports three modes of HBLK operation. HBLK
Mode 0 supports basic operation and some support for special
HBLK patterns. HBLK Mode 1 supports pixel mixing HBLK
operation. HBLK Mode 2 supports advanced HBLK operation.
The following sections describe each mode in detail. Register
parameters are described in detail in Table 11.
HBLK Mode 0 Operation
There are six toggle positions available for HBLK. Normally,
only two of the toggle positions are used to generate the standard
HBLK interval. However, the additional toggle positions can be
used to generate special HBLK patterns, as shown in Figure 26.
The pattern in this example uses all six toggle positions to
generate two extra groups of pulses during the HBLK interval.
By changing the toggle positions, different patterns can be
created.
Separate toggle positions are available for even and odd lines. If
alternation is not needed, the same values should be loaded into
the registers for even (HBLKTOGE) and odd (HBLKTOGO) lines.
Rev. C | Page 20 of 92
AD9992
HD
HBLKTOGE2
BLANK
HBLK
BLANK
05891-024
HBLKTOGE1
BASIC HBLK PULSE IS GENERATED USING HBLKTOGE1 AND HBLKTOGE2 REGISTERS (HBLKALT = 0)
Figure 24. Typical Horizontal Blanking Pulse Placement (HBLKMODE = 0)
HD
HBLK
H1/H3/H5/H7
THE POLARITY OF H1/H3/H5/H7 DURING BLANKING IS PROGRAMMABLE
(H2/H4/H6/H8 AND HL ARE SEPARATELY PROGRAMMABLE)
05891-025
H1/H3/H5/H7
H2/H4/H6/H8
Figure 25. HBLK Masking Polarity Control
HBLKTOGE2
HBLKTOGE1
HBLKTOGE4
HBLKTOGE3
HBLKTOGE6
HBLKTOGE5
HBLK
H1/H3
SPECIAL H-BLANK PATTERN IS CREATED USING MULTIPLE HBLK TOGGLE POSITIONS (HBLKALT = 0)
Figure 26. Using Multiple Toggle Positions for HBLK (HBLKMODE = 0)
Rev. C | Page 21 of 92
05891-026
H2/H4
AD9992
Table 11. HBLK Pattern Registers
Register
HBLKMODE
Length
2b
Range
0 to 2 HBLK modes
HBLKSTART
HBLKEND
HBLKLEN
HBLKREP
HBLKMASK_H1
HBLKMASK_H2
HBLKMASK_HL
HBLKTOGO1
HBLKTOGO2
HBLKTOGO3
HBLKTOGO4
HBLKTOGO5
HBLKTOGO6
HBLKTOGE1
HBLKTOGE2
HBLKTOGE3
HBLKTOGE4
HBLKTOGE5
HBLKTOGE6
RA0H1REPA/B/C
13b
13b
13b
13b
1b
1b
1b
13b
13b
13b
13b
13b
13b
13b
13b
13b
13b
13b
13b
12b
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixels
0 to 8191 repetitions
High/low
High/low
High/low
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 15 HCLK pulses for
each A, B, and C
RA1H1REPA/B/C
RA2H1REPA/B/C
RA3H1REPA/B/C
RA4H1REPA/B/C
RA5H1REPA/B/C
RA0H2REPA/B/C
12b
12b
12b
12b
12b
12b
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses for
each A, B, and C
RA1H2REPA/B/C
RA2H2REPA/B/C
RA3H2REPA/B/C
RA4H2REPA/B/C
RA5H2REPA/B/C
HBLKSTARTA
HBLKSTARTB
HBLKSTARTC
HBLKALT_PAT1
12b
12b
12b
12b
12b
13b
13b
13b
3b
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 15 HCLK pulses
0 to 8191 pixel location
0 to 8191 pixel location
0 to 8191 pixel location
0 to 5 even repeat area
Description
Enables different HBLK toggle position operation.
0: Normal mode. Six toggle positions available for even and odd lines.
If even/odd alternation is not needed, set toggles for even/odd the same.
1: Pixel mixing mode. In addition to the six toggle positions, the HBLKSTART,
HBLKEND, HBLKLEN, and HBLKREP registers can be used to generate HBLK
patterns. If even/odd alternation is not need, set toggles for even/odd the same.
2: Advanced HBLK mode. Divides HBLK interval into six repeat areas. Uses
HBLKSTARTA/B/C and RA*H*REPA/B/C registers.
3: Test mode only. Do not access.
Start location for HBLK in HBLK Mode 1 and HBLK Mode 2.
End location for HBLK in HBLK Mode 1 and HBLK Mode 2.
HBLK length in HBLK Mode 1 and HBLK Mode 2.
Number of HBLK repetitions in HBLK Mode 1 and HBLK Mode 2.
Masking polarity for H1, H3, H5, H7 during HBLK.
Masking polarity for H2, H4, H6, H8 during HBLK.
Masking polarity for HL during HBLK.
First HBLK toggle position for odd lines in HBLK Mode 0 and HBLK Mode 1.
Second HBLK toggle position for odd lines in HBLK Mode 0 and HBLK Mode 1.
Third HBLK toggle position for odd lines in HBLK Mode 0 and HBLK Mode 1.
Fourth HBLK toggle position for odd lines in HBLK Mode 0 and HBLK Mode 1.
Fifth HBLK toggle position for odd lines in HBLK Mode 0 and HBLK Mode 1.
Sixth HBLK toggle position for odd lines in HBLK Mode 0 and HBLK Mode 1.
First HBLK toggle position for even lines in HBLK Mode 0 and HBLK Mode 1.
Second HBLK toggle position for even lines in HBLK Mode 0 and HBLK Mode 1.
Third HBLK toggle position for even lines in HBLK Mode 0 and HBLK Mode 1.
Fourth HBLK toggle position for even lines in HBLK Mode 0 and HBLK Mode 1.
Fifth HBLK toggle position for even lines in HBLK Mode 0 and HBLK Mode 1.
Sixth HBLK toggle position for even lines in HBLK Mode 0 and HBLK Mode 1.
HBLK Repeat Area 0. Number of H1 repetitions for HBLKSTARTA/B/C in
HBLK Mode 2 for even lines; odd lines defined using HBLKALT_PAT.
[3:0] RA0H1REPA. Number of H1 pulses following HBLKSTARTA.
[7:4] RA0H1REPB. Number of H1 pulses following HBLKSTARTB.
[11:8] RA0H1REPC. Number of H1 pulses following HBLKSTARTC.
HBLK Repeat Area 1. Number of H1 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 2. Number of H1 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 3. Number of H1 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 4. Number of H1 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 5. Number of H1 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 0. Number of H2 repetitions for HBLKSTARTA/B/C in
HBLK Mode 2 for even lines; odd lines defined using HBLKALT_PAT.
[3:0] RA0H2REPA. Number of H2 pulses following HBLKSTARTA.
[7:4] RA0H2REPB. Number of H2 pulses following HBLKSTARTB.
[11:8] RA0H2REPC. Number of H2 pulses following HBLKSTARTC.
HBLK Repeat Area 1. Number of H2 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 2. Number of H2 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 3. Number of H2 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 4. Number of H2 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area 5. Number of H2 repetitions for HBLKSTARTA/B/C.
HBLK Repeat Area Start Position A for HBLK Mode 2. Set to 8191 if not used.
HBLK Repeat Area Start Position B for HBLK Mode 2. Set to 8191 if not used.
HBLK Repeat Area Start Position C for HBLK Mode 2. Set to 8191 if not used.
HBLK Mode 2, Odd Field Repeat Area 0 pattern, selected from even field
repeat areas previously defined.
Rev. C | Page 22 of 92
AD9992
Register
HBLKALT_PAT2
HBLKALT_PAT3
HBLKALT_PAT4
HBLKALT_PAT5
HBLKALT_PAT6
Length
3b
3b
3b
3b
3b
Range
0 to 5 even repeat area
0 to 5 even repeat area
0 to 5 even repeat area
0 to 5 even repeat area
0 to 5 even repeat area
HBLKTOGE2
HBLKSTART
HBLKTOGE1
Description
HBLK Mode 2, Odd Field Repeat Area 1 pattern.
HBLK Mode 2, Odd Field Repeat Area 2 pattern.
HBLK Mode 2, Odd Field Repeat Area 3 pattern.
HBLK Mode 2, Odd Field Repeat Area 4 pattern.
HBLK Mode 2, Odd Field Repeat Area 5 pattern.
HBLKTOGE4
HBLKTOGE3
HBLKEND
HBLK
HBLKLEN
HBLKREP = 3
H1/H3
HBLKREP NUMBER 1
HBLKREP NUMBER 2
05891-027
H2/H4
HBLKREP NUMBER 3
H-BLANK REPEATING PATTERN IS CREATED USING HBLKLEN AND HBLKREP REGISTERS
Figure 27. HBLK Repeating Pattern Using HBLKMODE = 1
HBLK Mode 1 Operation
Bits [7:4]) is set to a value between 1 and 15. When this register
is set to 0, the wide HCLK feature is disabled. The reduced
frequency occurs only for H1 to H8 pulses that are located
within the HBLK area.
Multiple repeats of the HBLK signal are enabled by setting
HBLKMODE to 1. In this mode, the HBLK pattern can be
generated using a different set of registers: HBLKSTART,
HBLKEND, HBLKLEN, and HBLKREP, along with the six
toggle positions (see Figure 27).
Separate toggle positions are available for even and odd lines. If
alternation is not needed, the same values should be loaded into
the registers for even (HBLKTOGE) and odd (HBLKTOGO) lines.
Generating HBLK Line Alternation
HBLK Mode 0 and HBLK Mode 1 provide the ability to
alternate different HBLK toggle positions on even and odd
lines. HBLK line alternation can be used in conjunction with
V-pattern odd/even alternation or on its own. Separate toggle
positions are available for even and odd lines. If even/odd line
alternation is not required, the same values should be loaded into
the registers for even (HBLKTOGE) and odd (HBLKTOGO) lines.
The HCLK_WIDTH register is generally used in conjunction
with special HBLK patterns to generate vertical and horizontal
mixing in the CCD.
Note that the wide HCLK feature is available only in HBLK
Mode 0 and HBLK Mode 1. HBLK Mode 2 does not support
wide HCLKs.
Table 12. HCLK Width Register
Register
HCLK_WIDTH
Increasing H-Clock Width During HBLK
HBLK Mode 0 and HBLK Mode 1 allow the H1 to H8 pulse
widths to be increased during the HBLK interval. As shown in
Figure 28, the H-clock frequency can be reduced by a factor of
1/2, 1/4, 1/6, 1/8, 1/10, 1/12, and so on, up to 1/30. To enable
this feature, the HCLK_WIDTH register (Address 0x34,
Rev. C | Page 23 of 92
Length
4b
Description
Controls H1 to H8 pulse widths
during HBLK as a fraction of pixel rate
0: Same frequency as pixel rate
1: 1/2 pixel frequency, that is, doubles
the HCLK pulse width
2: 1/4 pixel frequency
3: 1/6 pixel frequency
4: 1/8 pixel frequency
5: 1/10 pixel frequency
…
15: 1/30 pixel frequency
AD9992
HBLK
H1/H3
1/FPIX
2 × (1/FPIX)
05891-028
H2/H4
H-CLOCK FREQUENCY CAN BE REDUCED DURING HBLK BY 1/2 (AS SHOWN),
1/4, 1/6, 1/8, 1/10, 1/12, AND SO ON, UP TO 1/30 USING HBLKWIDTH REGISTER
Figure 28. Generating Wide H-Clock Pulses During HBLK Interval
HD
CREATE UP TO 3 GROUPS OF TOGGLES
A, B, C COMMON IN ALL REPEAT AREAS
A
CHANGE NUMBER OF A, B, C PULSES IN ANY
REPEAT AREA USING RA*H*REP* REGISTERS
MASK A, B, C PULSES IN ANY REPEAT
AREA BY SETTING RA*H*REP* = 0
B
C
H1
REPEAT AREA 0
REPEAT AREA 1 REPEAT AREA 2
REPEAT AREA 3
REPEAT AREA 4 REPEAT AREA 5
HBLKSTART
HBLKEND
05891-029
H2
Figure 29. HBLK Mode 2 Operation
HD
HBLKLEN
HBLK
HBLKSTARTA
ALL RA*H*REPA/B/C REGISTERS = 2 TO CREATE TWO HCLK PULSES
HBLKSTARTB
HBLKSTARTC
H1
RA0H1REPA RA0H1REPB
RA0H1REPC
RA1H1REPA RA1H1REPB
RA1H1REPC
RA1H2REPA RA1H2REPB
RA1H2REPC
H2
RA0H2REPA RA0H2REPB
RA0H2REPC
REPEAT AREA 1
REPEAT AREA 0
HBLKREP = 2
TO CREATE TWO REPEAT AREAS
Figure 30. HBLK Mode 2 Registers
Rev. C | Page 24 of 92
HBLKEND
05891-030
HBLKSTART
AD9992
HBLK Mode 2 Operation
HORIZONTAL TIMING SEQUENCE EXAMPLE
HBLK Mode 2 allows more advanced HBLK pattern operation.
If multiple areas of HCLK pulses that are unevenly spaced apart
from one another are needed, HBLK Mode 2 can be used. Using
a separate set of registers, HBLK Mode 2 can divide the HBLK
region into up to six repeat areas (see Table 11). As shown in
Figure 30, each repeat area shares a common group of toggle
positions, HBLKSTARTA, HBLKSTARTB, and HBLKSTARTC.
However, the number of toggles following each start position
can be unique in each repeat area by using the RA*H1REP* and
RA*H2REP* registers. As shown in Figure 29, setting the
RA*H1REPA/RA*H1REPB/RA*H1REPC or RA*H2REPA/
RA*H2REPB/RA*H2REPC registers to 0 masks HCLK groups
from appearing in a particular repeat area. Figure 30 shows only
two repeat areas being used, although six are available. It is possible
to program a separate number of repeat area repetitions for H1
and H2, but generally the same value is used for both H1 and
H2. Figure 30 shows an example of RA0H1REPA/RA0H1REPB/
RA0H1REPC = RA0H2REPA/RA0H2REPB/RA0H2REPC =
RA1H1REPA/RA1H1REPB/RA1H1REPC = RA1H2REPA/
RA1H2REPB/RA1H2REPC = 2.
Figure 31 shows an example CCD layout. The horizontal
register contains 28 dummy pixels, which occur on each line
clocked from the CCD. In the vertical direction, there are
10 optical black (OB) lines at the front of the readout and two at
the back of the readout. The horizontal direction has four OB
pixels in the front and 48 in the back.
Figure 32 shows the basic sequence layout to be used during
the effective pixel readout. The 48 OB pixels at the end of each
line are used for the CLPOB signals. PBLK is optional and is
often used to blank the digital outputs during the HBLK time.
HBLK is used during the vertical shift interval.
Because PBLK is used to isolate the CDS input (see the Analog
Preblanking section), the PBLK signal should not be used
during CLPOB operation. The change in the offset behavior
that occurs during PBLK impacts the accuracy of the CLPOB
circuitry.
The HBLK, CLPOB, and PBLK parameters are programmed in
the V-sequence registers. More elaborate clamping schemes,
such as adding in a separate sequence to clamp in the entire
shield OB lines, can be used. This requires configuring a separate
V-sequence for clocking out the OB lines.
Furthermore, HBLK Mode 2 allows a different HBLK pattern
on even and odd lines. The HBLKSTARTA, HBLKSTARTB, and
HBLKSTARTC registers, as well as the RA*H1REPA/RA*H1REPB
/RA*H1REPC and RA*H2REPA/ RA*H2REPB/RA*H2REPC
registers, define operation for the even lines. For separate
control of the odd lines, the HBLKALT_PAT registers specify
up to six repeat areas on the odd lines by reordering the repeat
areas used for the even lines. New patterns are not available, but
the order of the previously defined repeat areas on the even lines
can be changed for the odd lines to accommodate advanced
CCD operation.
The CLPMASK registers are also useful for disabling the CLPOB
on a few lines without affecting the setup of the clamping
sequences. It is important that CLPOB be used only during valid
OB pixels. During other portions on the frame timing, such as
vertical blanking or SG line timing, the CCD does not output
valid OB pixels. Any CLPOB pulse that occurs during this time
causes errors in clamping operation and changes in the black
level of the image.
2 VERTICAL
OB LINES
V
EFFECTIVE IMAGE AREA
10 VERTICAL
OB LINES
H
48 OB PIXELS
4 OB PIXELS
05891-031
HORIZONTAL CCD REGISTER
28 DUMMY PIXELS
Figure 31. Example CCD Configuration
Rev. C | Page 25 of 92
AD9992
OPTICAL BLACK
OPTICAL BLACK
HD
CCD OUTPUT
VERTICAL SHIFT
DUMMY
EFFECTIVE PIXELS
OPTICAL BLACK
VERT. SHIFT
SHP
SHD
H1/H3/H5/H7
H2/H4/H6/H8
HBLK
PBLK
05891-032
CLPOB
NOTES
1. PBLK ACTIVE (LOW) SHOULD NOT BE USED DURING CLPOB ACTIVE (LOW).
Figure 32. Horizontal Sequence Example
3.
VERTICAL TIMING GENERATION
The AD9992 provides a flexible solution for generating vertical
CCD timing and can support multiple CCDs and different
system architectures. The vertical transfer clocks are used to shift
each line of pixels into the horizontal output register of the CCD.
The AD9992 allows these outputs to be individually programmed
into various readout configurations by using a 4-step process.
Figure 33 shows an overview of how the vertical timing is
generated in four steps.
1.
2.
The readout for an entire field is constructed by dividing
the field into different regions and then assigning a
sequence to each region.
Each field can contain up to nine different regions to
accommodate different steps of the readout, such as high
speed line shifts and unique vertical line transfers. The
total number of V-patterns, V-sequences, and fields is
programmable but limited by the number of registers.
4.
The individual pulse patterns for XV1 to XV24 are created
by using the vertical pattern group registers.
The V-pattern groups are used to build the sequences,
which is when additional information is added.
Rev. C | Page 26 of 92
The MODE registers allow the different fields to be
combined in any order for various readout configurations.
AD9992
CREATE THE VERTICAL PATTERN GROUPS,
UP TO FOUR TOGGLE POSITIONS FOR EACH OUTPUT.
1
BUILD THE V-SEQUENCES BY ADDING START POLARITY,
LINE START POSITION, NUMBER OF REPEATS, ALTERNATION,
GROUP A/B/C/D INFORMATION, AND HBLK/CLPOB PULSES.
2
XV1
XV1
XV2
VPAT0
XV2
XV3
XV3
V-SEQUENCE 0
(VPAT0, 1 REP)
XV23
XV23
XV24
XV24
XV1
XV2
XV1
VPAT1
XV3
V-SEQUENCE 1
(VPAT1, 2 REP)
XV2
XV3
XV23
XV24
XV23
XV24
XV1
XV2
XV3
V-SEQUENCE 2
(VPAT1, N REP)
XV23
XV24
USE THE MODE REGISTERS TO CONTROL WHICH FIELDS
ARE USED, AND IN WHAT ORDER (MAXIMUM OF SEVEN
FIELDS CAN BE COMBINED IN ANY ORDER).
3
BUILD EACH FIELD BY DIVIDING INTO DIFFERENT REGIONS
AND ASSIGNING A DIFFERENT V-SEQUENCE TO EACH
(MAXIMUM OF NINE REGIONS IN EACH FIELD).
FIELD 0
FIELD0
FIELD1
FIELD2
REGION 0: USE V-SEQUENCE 2
REGION 0: USE V-SEQUENCE 3
REGION 1: USE V-SEQUENCE 0
REGION 0: USE V-SEQUENCE 3
REGION
USE V-SEQUENCE
2
REGION
2: USE1:V-SEQUENCE
3
REGION 1: USE V-SEQUENCE 2
FIELD3
FIELD4
REGION 3: USE V-SEQUENCE 0
REGION 2: USE V-SEQUENCE 1
FIELD5
FIELD1
FIELD4
FIELD2
REGION 2: USE V-SEQUENCE 1
REGION 4: USE V-SEQUENCE 2
FIELD1
FIELD2
Figure 33. Summary of Vertical Timing Generation
Rev. C | Page 27 of 92
05891-033
4
AD9992
Vertical Pattern Groups (VPAT)
The vertical pattern groups define the individual pulse patterns for
each XV1 to XV24 output signal. Table 13 summarizes the registers
available for generating each of the V-pattern groups. The first,
second, third, and fourth toggle positions (XVTOG1, XVTOG2,
XVTOG3, and XVTOG4) are the pixel locations within the line
where the pulse transitions. All toggle positions are 13-bit values,
allowing their placement anywhere in the horizontal line.
More registers are included in the vertical sequence registers to
specify the output pulses. VPOL specifies the start polarity for
each signal; VSTART specifies the start position of the V-pattern
group within the line; VLEN designates the total length of
the V-pattern group, which determines the number of pixels
between each of the pattern repetitions when repetitions are used.
The VSTART position is actually an offset value for each toggle
position. The actual pixel location for each toggle, measured
from the HD falling edge (Pixel 0), is equal to the VSTART value
plus the toggle position.
When the selected V-output is designated as a VSG pulse, either
the XVTOG1/XVTOG2 or XVTOG3/XVTOG4 pair is selected
using V-Sequence Address 0x02, VSGPATSEL. All four toggle
positions are not simultaneously available for VSG pulses.
Unused V-channels must have their toggle positions programmed
to either 0 or maximum value. This prevents unpredictable
behavior because the default values of the V-pattern group registers
are unknown.
Table 13. Vertical Pattern Group Registers
Register
XVTOG1
XVTOG2
XVTOG3
XVTOG4
Length
13b
13b
13b
13b
Description
First toggle position within line for each XV1 to XV24 output, relative to VSTART value.
Second toggle position, relative to VSTART value
Third toggle position, relative to VSTART value
Fourth toggle position, relative to VSTART value
START POSITION OF VERTICAL PATTERN GROUP IS PROGRAMMABLE IN VERTICAL SEQUENCE REGISTERS.
HD
4
1
2
XV2
3
1
2
XV24
3
1
2
3
PROGRAMMABLE SETTINGS:
1START POLARITY (LOCATED IN V-SEQUENCE REGISTERS).
2FIRST TOGGLE POSITION.
3SECOND TOGGLE POSITION (THIRD AND FOURTH TOGGLE POSITIONS ALSO AVAILABLE FOR MORE COMPLEX PATTERNS).
4TOTAL PATTERN LENGTH FOR ALL VERTICAL OUTPUTS (LOCATED IN VERTICAL SEQUENCE REGISTERS).
Figure 34. Vertical Pattern Group Programmability
Rev. C | Page 28 of 92
05891-034
XV1
AD9992
into both registers. If a different number of repetitions is required
on odd and even lines, separate values can be used for each
register (see the Generating Line Alternation for V-Sequences
and HBLK section). The VSTARTA and VSTARTB registers
specify where in the line the V-pattern group starts. The
VMASK_EN register is used with the FREEZE/RESUME
registers to enable optional masking of the V-outputs. Either or
both of the FREEZE1/RESUME1 and FREEZE2/RESUME2
registers can be enabled.
VERTICAL SEQUENCES (VSEQ)
The vertical sequences are created by selecting one of the V-pattern
groups and adding repeats, start position, horizontal clamping,
and blanking information. The V-sequences are programmed
using the registers shown in Table 14. Figure 35 shows how
the different registers are used to generate each V-sequence.
The VPATSELA, VPATSELB, VPATSELC, and VPATSELD
registers select which V-pattern is used in a given V-sequence.
Having four groups available allows different vertical outputs to
be mapped to different V-patterns. The selected V-pattern group
can have repetitions added for high speed line shifts or for line
binning by using the VREP registers for odd and even lines.
Generally, the same number of repetitions is programmed
The line length (in pixels) is programmable using the HDLEN
registers. Each V-sequence can have a different line length to
accommodate various image readout techniques. The maximum
number of pixels per line is 8192. The last line of the field is
programmed separately using the HDLASTLEN register, which
is located in the field register section.
1
HD
2
XV1 TO XV24
HBLK
4
4
VREP 2
VREP 3
5
6
PROGRAMMABLE SETTINGS FOR EACH VERTICAL SEQUENCE:
1START POSITION IN THE LINE OF SELECTED V-PATTERN GROUP.
2HD LINE LENGTH.
3V-PATTERN SELECT (VPATSEL) TO SELECT ANY V-PATTERN GROUP.
4NUMBER OF REPETITIONS OF THE V-PATTERN GROUP (IF NEEDED).
5START POLARITY AND TOGGLE POSITIONS FOR CLPOB AND PBLK SIGNALS.
6MASKING POLARITY AND TOGGLE POSITIONS FOR HBLK SIGNAL.
Figure 35. V-Sequence Programmability
Rev. C | Page 29 of 92
05891-035
CLPOB
3
V-PATTERN GROUP
AD9992
Table 14. Summary of V-Sequence Registers (see Table 10 and Table 11 for the CLPOB, PBLK, and HBLK Pattern Registers)
Register
HOLD
Length
4b
VMASK_EN
4b
CONCAT_GRP
4b
VREP_MODE
2b
LASTREPLEN_EN
4b
LASTTOG_EN
4b
HDLENE
HDLENO
VPOL_A
VPOL_B
VPOL_C
VPOL_D
GROUPSEL_0
13b
13b
24b
24b
24b
24b
24b
GROUPSEL_1
24b
VPATSELA
VPATSELB
VPATSELC
VPATSELD
VSTARTA
VSTARTB
VSTARTC
VSTARTD
VLENA
VLENB
VLENC
VLEND
VREPA_1
VREPA_2
5b
5b
5b
5b
13b
13b
13b
13b
13b
13b
13b
13b
13b
13b
Description
Use in conjunction with VMASK_EN.
1: HOLD function instead of FREEZE/RESUME function.
Enables the masking of XV1 to XV24 outputs at the locations specified by the FREEZE/RESUME registers.
1: Enable masking for all groups. One bit for each set of Freeze and Resume Positions 1 to 4.
Combines toggle positions of Groups A/B/C/D when enabled. Only Group A settings for start, polarity, length,
and repetition are used when this mode is selected.
0: Disable.
1: Enable the addition of all toggle positions from VPATSELA/B/C/D.
2: Test mode only. Do not use.
…
15: Test mode only. Do not use.
Selects line alternation for V-output repetitions. Note separate controls for Group A and Groups B/C/D.
0: Disable alternation. Group A uses VREPA_1, Groups B/C/D use VREP _EVEN for all lines.
1: 2-line. Group A alternates VREPA_1 and VREPA_2. Groups B/C/D alternate VREP_EVEN and VREP_ODD.
2: 3-line. Group A alternates VREPA_1, VREPA_2, and VREPA_3. Groups B/C/D follow a VREP_EVEN, VREP_ODD,
VREP_ODD, VREP_EVEN, VREP_ODD, VREP_ODD pattern.
3: 4-line. Group A alternates VREPA_1, VREPA_2, VREPA_3, VREPA_4. Groups B/C/D follow 2-line alternation.
Enables a separate pattern length to be used during the last repetition of the V-sequence. One bit for
each group (A, B, C, and D). Set bit high to enable. Group A is the LSB. Recommended value is enabled.
Enables a final toggle position to be added at the end of the V-sequence. The toggle position is shared
by all V-outputs in the same group. One bit for each group. Set bit high to enable. Group A is the LSB.
HD line length for even lines in the V-sequence.
HD line length for odd lines in the V-sequence.
Group A start polarity bits for each XV1 to XV24 output.
Group B start polarity bits for each XV1 to XV24 output.
Group C start polarity bits for each XV1 to XV24 output.
Group D start polarity bits for each XV1 to XV24 output.
Assigns each XV1 to XV12 output to either Group A/B/C/D. Two bits for each signal. Bits [1:0] are for XV1, Bits
[3:2] are for XV2 … Bits [23:22] are for XV12.
0: Assign to Group A
1: Assign to Group B
2: Assign to Group C
3: Assign to Group D
Assigns each XV13 to XV24 output to either Group A/B/C/D. Two bits for each signal. Bits [1:0] are for XV13,
Bits [3:2] are for XV14 … Bits [23:22] are for XV24.
0: Assign to Group A.
1: Assign to Group B.
2: Assign to Group C.
3: Assign to Group D.
Selected V-pattern for Group A.
Selected V-pattern for Group B.
Selected V-pattern for Group C.
Selected V-pattern for Group D.
Start position for the selected V-Pattern Group A.
Start position for the selected V-Pattern Group B.
Start position for the selected V-Pattern Group C.
Start position for the selected V-Pattern Group D.
Length of selected V-Pattern Group A.
Length of selected V-Pattern Group B.
Length of selected V-Pattern Group C.
Length of selected V-Pattern Group D.
Number of repetitions for the V-Pattern Group A for first lines (even).
Number of repetitions for the V-Pattern Group A for second lines (odd).
Rev. C | Page 30 of 92
AD9992
Register
VREPA_3
VREPA_4
VREPB_ODD
VREPC_ODD
VREPD_ODD
VREPB_EVEN
VREPC_EVEN
VREPD_EVEN
FREEZE1
Length
13b
13b
13b
13b
13b
13b
13b
13b
13b
FREEZE2
13b
FREEZE3
13b
FREEZE4
13b
RESUME1
13b
RESUME2
13b
RESUME3
13b
RESUME4
13b
LASTREPLEN_A
13b
LASTREPLEN_B
13b
LASTREPLEN_C
13b
LASTREPLEN_D
13b
LASTTOG_A
13b
LASTTOG_B
13b
LASTTOG_C
13b
LASTTOG_D
13b
VSEQALT_EN
VALT_MAP
1b
1b
VALTSEL0_EVEN
18b
VALTSEL1_EVEN
18b
VALTSEL0_ODD
18b
VALTSEL1_ODD
18b
SPC_PAT_EN
1b
Description
Number of repetitions for the V-Pattern Group A for third lines.
Number of repetitions for the V-Pattern Group A for fourth lines.
Number of repetitions for the V-Pattern Group B for odd lines.
Number of repetitions for the V-Pattern Group C for odd lines.
Number of repetitions for the V-Pattern Group D for odd lines.
Number of repetitions for the V-Pattern Group B for even lines.
Number of repetitions for the V-Pattern Group C for even lines.
Number of repetitions for the V-Pattern Group D for even lines.
Pixel location where the V-outputs freeze or hold (see VMASK_EN). Also used as VALTSEL0_EVEN [12:0] register
when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs freeze or hold (see VMASK_EN). Also used as VALTSEL1_EVEN [12:0] register
when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs freeze or hold (see VMASK_EN). Also used as VALTSEL0_ODD [12:0] register
when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs freeze or hold (see VMASK_EN). Also used as VALTSEL1_ODD [12:0] register
when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs resume operation (see VMASK_EN). Also used as VALTSEL0_EVEN [17:13]
register when S within line for each XV1 to XV24 output when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs resume operation (see VMASK_EN). Also used as VALTSEL1_EVEN [17:13]
register when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs resume operation (see VMASK_EN). Also used as VALTSEL0_ODD [17:13]
register when special VSEQALT_EN mode is enabled.
Pixel location where the V-outputs resume operation (see VMASK_EN). Also used as VALTSEL1_ODD [17:13]
register when special VSEQALT_EN mode is enabled.
Separate length for last repetition of vertical pulses. Must be enabled using LASTREPLEN_EN.
Should be programmed to a value equal to the VLENA register.
Separate length for last repetition of vertical pulses. Must be enabled using LASTREPLEN_EN.
Should be programmed to a value equal to the VLENB register.
Separate length for last repetition of vertical pulses. Must be enabled using LASTREPLEN_EN.
Should be programmed to a value equal to the VLENC register.
Separate length for last repetition of vertical pulses. Must be enabled using LASTREPLEN_EN.
Should be programmed to a value equal to the VLEND register.
Optional fifth toggle position for the vertical signals. Must be enabled using LASTTOG_EN.
Note that the toggle position is common for all vertical signals.
Optional fifth toggle position for the vertical signals. Must be enabled using LASTTOG_EN.
Note that the toggle position is common for all vertical signals.
Optional fifth toggle position for the vertical signals. Must be enabled using LASTTOG_EN.
Note that the toggle position is common for all vertical signals.
Optional fifth toggle position for the vertical signals. Must be enabled using LASTTOG_EN.
Note that the toggle position is common for all vertical signals.
Special V-sequence alternation mode is enabled when this register is programmed high.
Enables the use of FREEZE/RESUME register locations to specify the VALTSEL0 and VALTSEL1 registers. Must be
enabled if VSEQALT mode is enabled.
Select lines for special V-sequence alternation mode for even lines. Used to concatenate VPAT Groups A/B/C/D into
unique merged patterns. Setting is used to specify one segment, with up to a maximum of 18 segments.
Select lines for special V-sequence alternation mode for even lines. Used to concatenate VPAT Groups A/B/C/D into
unique merged patterns. Setting is used to specify one segment, with up to a maximum of 18 segments.
Select lines for special V-sequence alternation mode for odd lines. Used to concatenate VPAT Groups A/B/C/D into
unique merged patterns. Setting is used to specify one segment, with up to a maximum of 18 segments.
Select lines for special V-sequence alternation mode for odd lines. Used to concatenate VPAT Groups A/B/C/D into
unique merged patterns. Setting is used to specify one segment, with up to a maximum of 18 segments.
Enable special V-pattern to be inserted into one repetition of a VPATA series.
SPC_PAT_EN [0]: Set to 1 to enable VPATB to be used as special pattern insertion.
SPC_PAT_EN [1]: Set to 1 to enable VPATC to be used as special pattern insertion.
SPC_PAT_EN [2]: Set to 1 to enable VPATD to be used as special pattern insertion.
Rev. C | Page 31 of 92
AD9992
HD
XV1 TO XV8 USE
V-PATTERN GROUP A
XV1
XV8
XV9, XV10 USE
V-PATTERN GROUP B
05891-036
XV9
XV10
Figure 36. Using Separate Group A and Group B V-Patterns
HD
V-PATTERN GROUP A
V-PATTERN GROUP B
V-PATTERN GROUP C
V-PATTERN GROUP D
05891-037
XV1
XV24
Figure 37. Combining Multiple V-Patterns Using CONCAT_GRP = 1
HD
V-PATTERN GROUP A
V-PATTERN GROUP B
XV1
GROUP A REP 1
GROUP A REP 2
GROUP A REP 3
05891-038
XV10
Figure 38. Combining Group A and Group B V-Patterns with Repetition
Group A/Group B/Group C/Group D Selection
The AD9992 has the flexibility to use four different V-pattern
groups in a vertical sequence. In general, the vertical outputs
use the same V-pattern group during a particular sequence. It is
possible to assign some of the outputs to a different V-pattern
group, which can be useful in certain CCD readout modes.
The GROUPSEL registers are used to select Group A, Group B,
Group C, or Group D for each V-output. In general, only a single
V-pattern group is needed for the vertical outputs; therefore,
Group A should be selected for all outputs by default
(GROUPSEL_0, GROUPSEL_1 = 0x00). In this configuration,
all outputs use the V-pattern group specified by the VPATSELA
register.
If additional flexibility is needed, some outputs can be set to
Group B, Group C, or Group D in the GROUPSEL registers.
In this case, those selected outputs use the V-pattern group
specified by the VPATSELB, VPATSELC, or VPATSELD
registers. Figure 36 shows an example where outputs XV9 and
XV10 are using a separate V-Pattern Group B to perform
special CCD timing.
Another application of the Group A, Group B, Group C, and
Group D registers is to combine up to four different V-pattern
groups together for more complex patterns. This is accomplished by setting the CONCAT_GRP register (Address 0x00,
Bits [13:10]) equal to 0x01. This setting combines the toggle
positions from the V-pattern groups specified by the VPATSELA,
VPATSELB, VPATSELC, and VPATSELD registers for a maximum
of up to 16 toggle positions. Example timing for the CONCAT_
GRP = 1 feature is shown in Figure 37.
If only two groups are needed (up to eight toggle positions) for
the specified timing, the VPATSELB, VPATSELC, and
VPATSELD registers can be programmed to the same value.
If only three groups are needed, VPATSELC and VPATSELD
can be programmed to the same value. Following this approach
Rev. C | Page 32 of 92
AD9992
conserves register memory if the four separate V-patterns are
not needed.
Note that when CONCAT_GRP is enabled, Group A settings
are used only for start position, polarity, length, and repetitions.
All toggle positions for Group A, Group B, Group C, and Group D
are combined together and applied using the settings in the
VSTARTA, VPOL_A, VLENA, and VREPA registers.
Special Vertical Sequence Alternation (SVSA) Mode
The AD9992 has additional flexibility for combining four
different V-pattern groups in a random sequence that can be
programmed for specific CCD requirements. This mode of
operation allows custom vertical sequences for CCDs that
require more complex vertical timing patterns. For example,
using the special vertical sequence alternation mode, it is
possible to support random pattern concatenation, with
additional support for odd/even line alternation.
Figure 39 illustrates four common and repetitive vertical pattern
segments, A through D, that are derived from the complete vertical
pattern. Figure 40 illustrates how each group can be concatenated
in an arbitrary order.
To enable the SVSA mode, write the VSEQALT_EN bit,
Address 0x20 Bit 13, equal to 0x01. The location of the VALTSEL
registers is shared with the VPAT registers for XV24. When
SVSA mode is enabled, the VALTSEL register function is
selected.
To create SVSA timing, divide the complete vertical timing
pattern into four common and repetitive segments. Identify
the related segments as VPATA, VPATB, VPATC, or VPATD.
Up to four toggle positions for each segment can be programmed
using the V-pattern registers.
Table 15 shows how the segments are specified using a 2-bit
representation. Each bit from VALTSEL0 and VALTSEL1 is
combined to produce four values, corresponding to Pattern A,
Pattern B, Pattern C, and Pattern D.
Table 15. VALTSEL Bit Settings for Even and Odd Lines
Parameter
VALTSEL0_EVEN
VALTSEL1_EVEN
VALTSEL0_ODD
VALTSEL1_ODD
Resulting pattern for even lines
Resulting pattern for odd lines
VALTSEL Bit Settings
0
0
1
1
0
1
0
1
0
0
1
1
0
1
0
1
A
B
C
D
A
B
C
D
When the entire pattern is divided, program VALTSEL0 (even
and odd) [17:0] and VALTSEL1 (even and odd) [17:0] so that
the segments are concatenated in the desired order. If separate
odd and even lines are not required, set the odd and even registers
to the same value. Figure 41 illustrates the process of using six
vertical pattern segments that have been concatenated into a small,
merged pattern.
Program the register VREPA_1 to specify the number of segments
to concatenate into each merged pattern. The maximum number of
segments that can be concatenated to create a merged pattern is
18. Program VLENA, VLENB, VLENC, VLEND to be of equal
length. Finally, program HBLK to generate the proper H-clock
timing using the procedure for HBLK Mode 2 described in the
HBLK Mode 2 Operation section.
It is important to note that because the FREEZE/RESUME registers
are used to specify the VALTSEL registers, the VALT_MAP
register must be enabled when using the special VALT mode.
Table 16. VALTSEL Register Locations1
Register Function
When VSEQALT_EN = 1
VALTSEL0_EVEN [12:0]
VALTSEL0_EVEN [17:13]
VALTSEL1_EVEN [12:0]
VALTSEL1_EVEN [17:13]
VALTSEL0_ODD [12:0]
VALTSEL0_ODD [17:13]
VALTSEL1_ODD [12:0]
VALTSEL1_ODD [17:13]
1
Register Location
VSEQ register FREEZE1 [12:0]
VSEQ register RESUME1 [17:13]
VSEQ register FREEZE2 [12:0]
VSEQ register RESUME2 [17:13]
VSEQ register FREEZE3 [12:0]
VSEQ register RESUME3 [17:13]
VSEQ register FREEZE4 [12:0]
VSEQ register RESUME4 [17:13]
The VALT_MAP register must be set to 1 to enable the use of VALTSEL
registers.
Rev. C | Page 33 of 92
AD9992
V-PATTERN A
V-PATTERN B
VLENA
VLENB
V-PATTERN C
V-PATTERN D
XV1
XV2
XV3
XV23
VLENC
VLEND
05891-039
NOTES
1. EACH SEGMENT MUST BE THE SAME LENGTH.
VLENA = VLENB = VLENC = VLEND.
Figure 39. Vertical Timing Divided Into Four Segments: VPATA, VPATB, VPATC, and VPATD
HD
COMBINED
V-PATTERN
A B B D A C C
B C B D A B A
A
A
05891-040
NOTES
1. ABLE TO CONCATENATE PATTERNS TOGETHER ARBITRARILY.
2. EACH PATTERN CAN HAVE UP TO FOUR TOGGLES PROGRAMMED.
3. MAY CONCATENATE UP TO 18 PATTERNS INTO A MERGED PATTERN.
4. ODD AND EVEN LINES CAN HAVE A DIFFERENT PATTERN CONCATENATION
SPECIFIED BY VALTSEL EVEN AND ODD REGISTERS.
Figure 40. Concatenating Each VPAT Group in Arbitrary Order
HD
A
XV1 TO XV23
SEGMENT 1
C
B
D
SEGMENT 2
D
A
SEGMENT 3
SEGMENT4
VPATB
VPATD
1
0
1
1
SEGMENT 5
SEGMENT 6
XV1
XV2
XV3
XV23
0
0
VPATC
0
1
VPATD
1
1
VPATA
0
0
NOTES
1. SIX V-PATTERN SEGMENTS CONCATENATED INTO A MERGED PATTERN.
2. COMMON AND REPETITIVE VTP SEGMENTS DERIVED FROM THE COMPLETE VTP PATTERN.
3. VALTSEL REGISTERS SPECIFY SEGMENT ORDER TO CREATE THE CONCATENATED MERGED PATTERN.
Figure 41. Special V-Sequence Alternation Mode Using VALTSEL Registers to Specify Segment Order
Rev. C | Page 34 of 92
05891-041
VPATA
VALTSEL0_EVEN
VALTSEL1_EVEN
AD9992
Using the LASTREPLEN_EN
The LASTREPLEN_EN register (Address 0x00, Bits [19:16] in
the sequence registers) is used to enable a separate pattern
length to be used in the final repetition of several pulse
repetitions. It is recommended that the LASTREPLEN_EN
register bits be set high (enabled) and the LASTREPLEN_A,
LASTREPLEN_B, LASTREPLEN_C, and LASTREPLEN_D
registers be set to a value equal to the VLENA, VLENB,
VLENC, and VLEND register values, respectively.
Generating Line Alternation for V-Sequences and HBLK
During low resolution readout, some CCDs require a different
number of vertical clocks on alternate lines. The AD9992 can
support this by using the VREP registers. This allows a different
number of V-pattern group repetitions to be programmed on
odd and even lines. Only the number of repeats can be different
in odd and even lines, while the V-pattern group remains the same.
There are separate controls for the assigned Group A, Group B,
Group C, and Group D patterns. All groups can support odd and
even line alternation. Group A uses the VREPA_1 register and
the VREPA_2 register; Group B, Group C, and Group D use
the corresponding VREP_ODD register and VREP_EVEN
register. With the additional VREPA_3 register and VREPA_4
register, Group A can also support 3-line and 4-line alternation.
As discussed in the Generating HBLK Line Alternation section,
the HBLK signal can be alternated for odd and even lines. Figure 42
shows an example of V-pattern group repetition alternation
and HBLK Mode 0 alternation used together.
HD
VREPA_1 = 2
(OR VREPB/C/D_EVEN = 2)
VREPA_2 = 5
(OR VREPB/C/D_ODD = 5)
VREPA_1 = 2
(OR VREPB/C/D_EVEN = 2)
XV1
XV2
XV24
XVTOGE1
XVTOGE2
XVTOGO1
XVTOGO2
XVTOGE1
XVTOGE2
NOTES
1. THE NUMBER OF REPEATS FOR V-PATTERN GROUPS A/B/C/D CAN BE ALTERNATED ON ODD AND EVEN LINES.
2. GROUP A ALSO SUPPORTS 3- AND 4-LINE ALTERNATION USING THE ADDITIONAL VREPA_3 AND VREPA_4 REGISTERS.
3. THE HBLK TOGGLE POSITIONS CAN BE ALTERNATED BETWEEN ODD AND EVEN LINES TO GENERATE DIFFERENT HBLK PATTERNS.
Figure 42. Odd/Even Line Alternation of V-Pattern Group Repetitions and HBLK Toggle Positions
Rev. C | Page 35 of 92
05891-042
HBLK
AD9992
Vertical Masking Using FREEZE/RESUME Registers
As shown in Figure 43 and Figure 44, the FREEZE/RESUME
registers are used to temporarily mask the V-outputs. The pixel
locations to begin the masking (FREEZE) and end the masking
(RESUME) create an area in which the vertical toggle positions
are ignored. At the pixel location specified in the FREEZE register,
the V-outputs are held static at their current dc state, high or low.
The V-outputs are held until the pixel location that is specified
by the RESUME register is reached, at which point the signals
continue with any remaining toggle positions, if any exist.
HD
Four sets of FREEZE/RESUME registers are provided, allowing the
vertical outputs to be interrupted up to four times in the same line.
The FREEZE and RESUME Position 1 to Position 4 are enabled
independently and applied to all groups (Group A, Group B, Group
C, and Group D) using the VMASK_EN register.
Note that when masking is enabled, Group A, Group B, Group C,
and Group D use the same FREEZE/RESUME positions.
Note that the FREEZE/RESUME registers are also used as the
VALTSEL0 and VALTSEL1 registers during special vertical
alternation mode.
NO MASKING AREA
05891-043
XV1
XV24
Figure 43. No FREEZE/RESUME
HD
V-MASKING AREA
FREEZE
RESUME
XV1
NOTES
1. ALL TOGGLE POSITIONS WITHIN THE FREEZE/RESUME MASKING AREA ARE IGNORED. H-COUNTER CONTINUES TO COUNT DURING MASKING.
2. FOUR SEPARATE MASKING AREAS ARE AVAILABLE, USING FREEZE1/RESUME1, FREEZE2/RESUME2, FREEZE3/RESUME3, AND
FREEZE4/RESUME4 REGISTERS.
Figure 44. Using FREEZE/RESUME
Rev. C | Page 36 of 92
05891-044
XV24
AD9992
Hold Area Using FREEZE/RESUME Registers
The FREEZE/RESUME registers can also be used to create a
hold area in which the V-outputs are temporarily held and later
continued, starting at the point where they were held. As shown
in Figure 45, this is different from the VMASK_EN register
HD
FREEZE
because the V-outputs continue from where they stopped rather
than from where they would have been. The hold area
temporarily stops the pixel counter for the V-outputs, while the
V-masking allows the counter to continue in the masking area.
HOLD AREA
FOR GROUP A
RESUME
XV1
XV8
XV9
NOTES
1. WHEN HOLD = 1 FOR ANY V-SEQUENCE GROUP, THE FREEZE AND RESUME REGISTERS ARE USED TO SPECIFY THE HOLD AREA.
2. ABOVE EXAMPLE: XV1 TO XV10 ARE ASSIGNED TO GROUP A. HOLD BIT FOR GROUP A = 1.
3. H-COUNTER FOR GROUP A (XV1 TO XV10) STOPS DURING HOLD AREA.
Figure 45. Hold Area for Group A
Rev. C | Page 37 of 92
05891-045
XV10
AD9992
Special Pattern Insertion
been added into the middle of the sequence. Figure 47 shows more
detail on how to set the registers to achieve the desired timing.
Additional flexibility is available using the SPC_PAT_EN registers,
which allows a Group B, Group C, or Group D pattern to be
inserted into a series of Group A repetitions. This feature is
useful when a different pattern is needed at the start, end, or
middle of a sequence.
Note that VREPB is used to specify which repetition number
has the special pattern inserted instead of VPATA. VPATB
always has priority over VPATC or VPATD if more than one
SPC_PAT_EN bit is enabled (SPC_PAT_EN [0] has priority).
Figure 46 shows an example of a sweep region using VPATA
with multiple repetitions where a single repetition of VPATB has
VD
HD
SCP1
LINE 0
LINE 1
SCP2
LINE 2
LINE 24
LINE 25
XV1 TO XV24
REGION 1: SWEEP REGION
REGION 2
05891-046
REGION 0
PATTERN B INSERTED DURING PATTERN A REPETITIONS
Figure 46. Example of Special Pattern Insertion
HD
REP 1
REP 2
REP 3
REP 4
REP 5
REP N
XV1
REGISTER SETTINGS:
SPC_PAT_EN[0] = 1
VREPA = N
VREPB = 4
V-PATTERN B
V-PATTERN A
DESCRIPTION:
V-PATTERN B IS USED AS SPECIAL PATTERN
TOTAL NUMBER OF REPS USED FOR SEQUENCE (N REPS)
REP 4 USES V-PATTERN B INSTEAD OF V-PATTERN A
NOTES
1. VSTARTB MUST BE SET EQUAL TO VSTARTA.
Figure 47. Example of Special Pattern Insertion, Detail
Rev. C | Page 38 of 92
05891-047
V-PATTERN A
AD9992
Complete Field: Combining V-Sequences
After the V-sequences are created, they are combined to create
different readout fields. A field consists of up to nine regions,
and within each region, a different V-sequence can be selected.
Figure 48 shows how the sequence change positions (SCP)
designate the line boundary for each region and how the SEQ
registers then select which V-sequence is used in each region.
Registers to control the VSG outputs are also included in the field
registers. Table 17 summarizes the registers used to create the
fields.
The SEQ registers, one for each region, select which V-sequences
are active in each region. The MULT_SWEEP registers, one for
each region, are used to enable sweep mode and/or multiplier
mode in any region. The SCP registers create the line boundaries
for each region. The VDLEN register specifies the total number of
lines in the field. The HDLEN register specifies the total number of
pixels per line, and the HDLASTLEN register specifies the
number of pixels in the last line of the field.
The VPATSECOND register is used to add a second V-pattern
group to the XV1 to XV10 outputs in the vertical sensor gate
(VSG) line. The SGMASK register is used to enable or disable each
individual VSG output. There are two bits for each VSG output
to enable separate masking in SGACTLINE1 and SGACTLINE2.
Setting a masking bit high masks the output; setting it low enables
the output. The VSGPATSEL register assigns one of the eight
SG patterns to each VSG output. Individual SG patterns are created
separately using the SG pattern registers. The SGACTLINE1
register specifies which line in the field contains the VSG outputs.
The optional SGACTLINE2 register allows the same VSG pulses to
be repeated on a different line. Separate masking is not available
for SGACTLINE1 and SGACTLINE2.
Table 17. Field Registers (CLPOB, PBLK Masking Shown in Table 10)
Register
SEQx
MULT_SWEEP
Length
5b
2b
Range
0 to 31 V-sequence number
0 to 3
SCP
VDLEN
HDLASTLEN
VSGPATSEL
13b
13b
13b
24b
0 to 8191 line number
0 to 8191 lines
0 to 8191 pixels
High/low
SGMASK
24b
High/low, each VSG
SGACTLINE1
SGACTLINE2
13b
13b
0 to 8191 line number
0 to 8191 line number
Description
Selected V-sequence for each region in the field.
Enables multiplier mode and/or sweep mode for each region.
0: Multiplier off, sweep off.
1: Multiplier off, sweep on.
2: Multiplier on, sweep off.
3: Multiplier on, sweep on.
Sequence change position for each region.
Total number of lines in each field.
Length in pixels of the last HD line in each field.
VSGPATSEL selects which V-pattern toggle positions are used. When set to 0,
Toggle 1 and Toggle 2 are used. When set to 1, Toggle 3 and Toggle 4 are used.
[0]: XV1 selection (0 = use XVTOG1, XVTOG2; 1 = use XVTOG3, XVTOG4).
…
[23]: XV24 selection.
Set high to mask each individual VSG output.
[0]: XV1 mask.
…
[23]: XV24 mask.
Selects the line in the field where the VSG signals are active.
Selects a second line in the field to repeat the VSG signals. If not used,
set this equal to SGACTLINE1 or to the maximum value.
Rev. C | Page 39 of 92
AD9992
SCP1
SCP0
SCP2
SCP4
SCP3
SCP5
SCP8
VD
REGION 0
REGION 1
REGION 2
REGION 3
REGION 4
REGION 8
SEQ0
SEQ1
SEQ2
SEQ3
SEQ4
SEQ8
HD
XV1 TO XVN
SGACTLINE1
VSG
05891-048
FIELD SETTINGS:
1. SEQUENCE CHANGE POSITIONS (SCP0 TO SCP8) DEFINE EACH OF THE NINE AVAILABLE REGIONS IN THE FIELD.
2. SEQ0 TO SEQ8 SELECT THE DESIRED V-SEQUENCE FOR EACH REGION.
3. SGACTLINE1 REGISTER SELECTS WHICH HD LINE IN THE FIELD CONTAINS THE SENSOR GATE PULSE(S).
Figure 48. Complete Field Divided into Regions
VD
HD
SCP1
LINE 0
LINE 1
SCP2
LINE 2
LINE 24
LINE 25
REGION 0
REGION 1: SWEEP REGION
REGION 2
05891-049
XV1 TO XVN
Figure 49. Example of Sweep Region for High Speed Vertical Shift
Sweep Mode Operation
The AD9992 contains an additional mode of vertical timing
operation called sweep mode. This mode is used to generate
a large number of repetitive pulses that span across multiple HD
lines. An example of where this mode is needed is at the start of
the CCD readout operation. At the end of the image exposure
before the image is transferred by the sensor gate pulses, the vertical
interline CCD registers should be free of all charge. This can be
accomplished by quickly shifting out any charge using a long
series of pulses from the vertical outputs. Depending on the vertical
resolution of the CCD, up to 3000 clock cycles may be needed
to shift the charge out of each vertical CCD line. This operation
spans across multiple HD line lengths. Normally, the AD9992
vertical timing must be contained within one HD line length,
but when sweep mode is enabled, the HD boundaries are
ignored until the region is finished. To enable sweep mode within
any region, program the appropriate SWEEP register to high.
Figure 49 shows an example of the sweep mode operation. The
number of vertical pulses needed depends on the vertical resolution of the CCD. The toggle positions for the XV1 to XV24
signals are generated using the V-pattern registers (shown in
Table 13). A single pulse is created using the polarity and toggle
position registers. The number of repetitions is then programmed
to match the number of vertical shifts required by the CCD.
Repetitions are programmed into the V-sequence registers
(shown in Table 14) by using the VREP registers. This produces
a pulse train of the appropriate length. Normally, the pulse train
is truncated at the end of the HD line length, but when sweep
mode is enabled for this region, the HD boundaries are ignored.
In Figure 49, the sweep region occupies 23 HD lines. After the
sweep mode region is complete, normal sequence operation
resumes in the next region. When using sweep mode, be sure to
set the region boundaries (using the sequence change positions)
to the appropriate lines to prevent the sweep operation from
overlapping the next V-sequence.
Rev. C | Page 40 of 92
AD9992
Multiplier Mode
Because the XVTOG register is multiplied by VLEN, the
resolution of the toggle position placement is reduced. If VLEN =
4, the toggle position precision is reduced to 4-pixel increments
instead of to single-pixel increments. Table 18 summarizes how
the V-pattern group registers are used in multiplier mode
operation. In multiplier mode, the VREP registers must always be
programmed to the same value as the highest toggle position.
To generate very wide vertical timing pulses, a vertical region
can be configured into a multiplier region. This mode uses
the V-pattern registers in a slightly different manner. Multiplier
mode can be used to support unusual CCD timing requirements,
such as vertical pulses that are wider than the 13-bit V-pattern
toggle position counter. In general, the 13-bit toggle position
counter can be used with the sweep mode feature to support very
wide pulses; however, multiplier mode can be used to generate
even wider pulses.
Figure 50 illustrates this operation. The first toggle position is 2,
and the second toggle position is 9. In nonmultiplier mode, this
causes the V-sequence to toggle at Pixel 2 and then at Pixel 9 within
a single HD line. However, in multiplier mode, toggle positions are
multiplied by the value of VLEN (in this case, 4); therefore, the first
toggle occurs at Pixel 8, and the second toggle occurs at Pixel 36.
Sweep mode has also been enabled to allow the toggle positions
to cross the HD line boundaries.
The start polarity and toggle positions are still used in the same
manner as the standard V-pattern group programming, but
VLEN is used differently. Instead of using the pixel counter
(HD counter) to specify the toggle position locations (XVTOG1,
XVTOG2, XVTOG3, and XVTOG4) of the V-pattern group,
the VLEN is multiplied with the XVVTOG position to allow
very long pulses to be generated. To calculate the exact toggle
position, which is counted in pixels after the start position, use the
following equation:
Multiplier Mode Toggle Position = XVTOG × VLEN
Table 18. Multiplier Mode Register Parameters
Register
MULTI
VPOL
XVTOG
VLEN
VREP
Length
1b
1b
13b
13b
13b
Range
High/low
High/low
0 to 8191 pixel location
0 to 8191 pixels
0 to 8191 pixel location
Description
High enables multiplier mode.
Starting polarity of XV1 to XV10 signals in each V-pattern group.
Toggle positions for XV1 to XV10 signals in each V-pattern group.
Used as multiplier factor for toggle position counter.
VREP_EVEN/VREP_ODD must be set to the same value as the highest XVTOG value.
START POSITION OF VPAT GROUP IS STILL PROGRAMMED IN THE V-SEQUENCE REGISTERS
HD
5
3
5
VLEN
1
2
3
4
1
2
3
4
1
PIXEL
NUMBER
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
4
2
3
4
4
XV1 TO XV10
2
2
Figure 50. Example of Multiplier Region for Wide Vertical Pulse Timing
Rev. C | Page 41 of 92
05891-050
1
MULTIPLIER MODE V-PATTERN GROUP PROPERTIES:
1START POLARITY (STARTPOL = 0).
2FIRST, SECOND, AND THIRD TOGGLE POSITIONS (XVTOG1 = 2, XVTOG2 = 9).
3LENGTH OF VPAT COUNTER (VLEN = 4); THIS IS THE MINIMUM RESOLUTION FOR TOGGLE POSITION CHANGES.
4TOGGLE POSITIONS OCCUR AT LOCATION EQUAL TO (XVTOG × VLEN).
5IF SWEEP REGION IS ENABLED, THE V-PULSES MAY ALSO CROSS THE HD BOUNDRIES, AS SHOWN ABOVE.
AD9992
Vertical Sensor Gate (Shift Gate) Patterns
In an interline CCD, the vertical sensor gate (VSG) pulses are
used to transfer the pixel charges from the light-sensitive image
area into light-shielded vertical registers. From the lightshielded vertical registers, the image is clocked out line-by-line
using the vertical transfer pulses (XV signals) in conjunction
with the high speed horizontal clocks. The AD9992 has 24
vertical signals, and each signal can be assigned as a VSG pulse
instead of an XV pulse.
Table 19 summarizes the VSG control registers, which are mostly
located in the field registers space (see Table 17). The VSGSELECT
register (Address 0x1C in the fixed address space) determines
which vertical outputs are assigned as VSG pulses. When a signal
is selected to be a VSG pulse, only the starting polarity and two of
the V-pattern toggle positions are used. The VSGPATSEL register
in the sequence registers is used to assign either XVTOG1 and
XVTOG2 or XVTOG3 and XVTOG4 to the VSG signal.
Note that only two of the four V-pattern toggle positions are
available when a vertical signal is selected to be a VSG pulse.
The SGACTLINE1 and SGACTLINE2 registers are used to
select which line in the field is the VSG line. The VSG active
line location is used to reference when the substrate clocking
(SUBCK) signal begins to operate in each field. For more
information, see the Substrate Clock Operation (SUBCK)
section.
Also located in the field registers, the SGMASK register selects
which individual VSG pulses are active in a given field. Therefore,
all SG patterns to be preprogrammed into the V-pattern registers
and the appropriate pulses for the different fields can be enabled
separately.
Table 19. VSG Control Registers (Also see Field Registers in Table 17)
Register
VSGSELECT
(Located in Fixed
Address Space, 0x1C)
Length
24b
Range
High/low
VSGPATSEL
24b
High/low
SGMASK
24b
High/low, each VSG
SGACTLINE1
SGACTLINE2
13b
13b
0 to 8191 line number
0 to 8191 line number
Description
Selection of VSG signals from XV signals. Set to 1 to make signal a VSG.
[0]: XV1 selection (0 = XV pulse; 1 = VSG pulse).
[1]: XV2 selection.
…
[23]: XV24 selection.
When VSG signal is selected using the VSGSELECT register, VSGPATSEL
selects which V-pattern toggle positions are used. When set to 0, Toggle 1
and Toggle 2 are used. When set to 1, Toggle 3 and Toggle 4 are used.
[0]: XV1 selection (0 = use XVTOG1, XVTOG2; 1 = use XVTOG3, XVTOG4).
[1]: XV2 selection.
…
[23]: XV24 selection.
Set high to mask each individual VSG output.
[0]: XV1 mask.
…
[23]: XV24 mask.
Selects the line in the field where the VSG signals are active.
Selects a second line in the field to repeat the VSG signals. If not used,
set this equal to SGACTLINE1 or to the maximum value.
VD
4
HD
1
2
VSG PATTERN
Figure 51. Vertical Sensor Gate Pulse Placement
Rev. C | Page 42 of 92
05891-051
3
PROGRAMMABLE SETTINGS FOR EACH PATTERN:
1START POLARITY OF PULSE (FROM VPOL IN SEQUENCE REGISTERS).
2FIRST TOGGLE POSITION (FROM V-PATTERN REGISTERS).
3SECOND TOGGLE POSITION (FROM V-PATTERN REGISTERS).
4ACTIVE LINE FOR VSG PULSES WITHIN THE FIELD (FROM FIELD REGISTERS).
AD9992
MODE Registers
The MODE registers are used to select the field timing of the
AD9992. Typically, all of the field, V-sequence, and V-pattern
information is programmed into the AD9992 at startup. During
operation, the MODE registers allow the user to select any combination of field timing to meet the requirements of the system.
The advantage of using the MODE registers in conjunction with
preprogrammed timing is that it greatly reduces the system programming requirements during camera operation. Only a few
register writes are required when the camera operating mode is
changed, rather than having to program all of the vertical timing
information with each camera mode change.
A basic still camera application can require six fields of vertical
timing—one for draft mode operation, one for autofocusing,
and four for still image readout. All of the register timing
information for the six fields is loaded at startup. Then, during
camera operation, the MODE registers select which field timing is
active, depending on how the camera is being used.
Table 20 shows how the MODE registers are used. The MODE
register (Address 0x2A) specifies how many total fields are
used. Any value from 1 to 7 can be selected using these three
bits. The other two registers (0x2B and 0x2C) are used to select
which of the programmed fields are used and in which order.
Up to seven fields can be used in a single MODE write. The
AD9992 starts with the field timing specified by FIELD0 and,
on the next VD, switches to the timing specified by FIELD1 and
so on. After completing the total number of fields specified by
MODE, the AD9992 repeats by starting at the first field. This
continues until a new write to the MODE register occurs. Figure 54
shows example MODE register settings for different field configurations.
Note that only a write to Address 0x2C properly resets the field
counter. Therefore, when changing the values in any of the mode
registers, it is recommended that all three registers be updated
together in the same field (VD period).
Caution
The MODE registers are SCK updated by default. If they are
configured as VD-updated registers by writing Address 0xB4 =
0x03FF and Address 0xB5 = 0xFC00, the new MODE information
is updated on the second VD falling edge after the write occurs,
rather than on the first VD falling edge (see Figure 53).
Table 20. MODE Registers—VD Updated
Address
2A
2B
2C
Name
MODE
FIELD0
FIELD1
FIELD2
FIELD3
FIELD4
FIELD5
FIELD6
Length
3b
5b
5b
5b
5b
5b
5b
5b
Description
Total number of fields to cycle through. Set from 1 to 7.
Selected FIELD (from FIELD registers in configurable memory) for the first field to cycle through.
Selected FIELD (from FIELD registers in configurable memory) for the second field to cycle through.
Selected FIELD (from FIELD registers in configurable memory) for the third field to cycle through.
Selected FIELD (from FIELD registers in configurable memory) for the fourth field to cycle through.
Selected FIELD (from FIELD registers in configurable memory) for the fifth field to cycle through.
Selected FIELD (from FIELD registers in configurable memory) for the sixth field to cycle through.
Selected FIELD (from FIELD registers in configurable memory) for the seventh field to cycle through.
Rev. C | Page 43 of 92
AD9992
VD
MODE WRITE
MODE UPDATE
A
REGISTER WRITE
MODE FIELD NUMBER
4 (DRAFT)
4 (DRAFT)
0 (STILL 1ST FIELD)
1 (STILL 2ND FIELD)
2
05891-052
EXAMPLE MODE REGISTER CHANGE:
REGISTER WRITE A––WRITE TO MODE REGISTERS 0x2A, 0x2B, 0x2C TO SPECIFY
CHANGE FROM DRAFT MODE (FIELD4) TO STILL MODE (FIELD0/1/2/3).
ALSO WRITE TO VGA GAIN OR ANY NEW REGISTER VALUES NEEDED
FOR STILL FRAME OPERATION, SUCH AS NEW FIELD INFORMATION.
Figure 52. Update of MODE Register, SCK Updated (Default Setting)
VD
MODE WRITE
MODE UPDATE
A
REGISTER WRITE
MODE FIELD NUMBER
B
4 (DRAFT)
4 (DRAFT)
0 (STILL 1ST FIELD)
1 (STILL 2ND FIELD)
2
NOTES
1. NEW MODE INFORMATION IS UPDATED AT SECOND VD FALLING EDGE AFTER
SERIAL WRITE A.
05891-053
EXAMPLE MODE REGISTER CHANGE:
REGISTER WRITE A––WRITE TO MODE REGISTERS 0x2A, 0x2B, 0x2C TO SPECIFY
CHANGE FROM DRAFT MODE (FIELD4) TO STILL MODE (FIELD0/1/2/3).
REGISTER WRITE B––WRITE TO VGA GAIN OR ANY NEW REGISTER VALUES NEEDED
FOR STILL FRAME OPERATION, SUCH AS NEW FIELD INFORMATION.
Figure 53. Update of MODE Register if Changed to VD-Updated Register
EXAMPLE 1:
TOTAL FIELDS = 3, FIRST FIELD = FIELD0, SECOND FIELD = FIELD1, THIRD FIELD = FIELD2
MODE SETTINGS:
0x2A = 0x3
0x2B = 0x820
0x2C = 0x0
FIELD0
FIELD1
FIELD2
EXAMPLE 2:
TOTAL FIELDS = 1, FIRST FIELD = FIELD3
MODE SETTINGS:
0x2A = 0x1
0x2B = 0x3
0x2C = 0x0
FIELD3
EXAMPLE 3:
TOTAL FIELDS = 4, FIRST FIELD = FIELD5, SECOND FIELD = FIELD1, THIRD FIELD = FIELD4, FOURTH FIELD = FIELD2
MODE SETTINGS:
0x2A = 0x4
0x2B = 0x11025
0x2C = 0x0
FIELD1
FIELD4
FIELD2
05891-054
FIELD5
Figure 54. Using the MODE Registers to Select Field Timing
Rev. C | Page 44 of 92
AD9992
VERTICAL TIMING EXAMPLE
To better understand how AD9992 vertical timing generation
is used, consider the example CCD timing chart in Figure 55.
This example illustrates a CCD using a general 3-field readout
technique. As described in the Complete Field: Combining VSequences section, each readout field must be divided into
separate regions to perform each step of the readout. The sequence
change positions (SCP) determine the line boundaries for each
region, and the SEQx registers assign a particular V-sequence to
each region. The V-sequences contain the specific timing
information required in each region: XV1 to XV6 pulses (using
V-pattern groups), HBLK/CLPOB timing, and VSG patterns for
the SG active lines.
This timing example requires four regions for each of the three
fields, labeled Region 0, Region 1, Region 2, and Region 3.
Because the AD9992 allows many individual fields to be programmed, FIELD0, FIELD1, and FIELD2 can be used to meet
the requirements of this timing example. The four regions for
each field are very similar in this example, but the individual
registers for each field allow flexibility to accommodate other
timing charts.
Region 0 is a high speed, vertical shift region. Sweep mode can
be used to generate this timing operation with the desired
number of high speed vertical pulses needed to clear any charge
from the CCD vertical registers.
Region 1 consists of only two lines and uses standard singleline, vertical shift timing. The timing of this region is the same
as the timing in Region 3.
Region 2 is the sensor gate line where the VSG pulses transfer
the image into the vertical CCD registers. This region may
require the use of the second V-pattern group for the SG
active line.
Region 3 also uses the standard single-line, vertical shift timing,
the same timing as Region 1. Four regions are required in each
of the three fields.
The timing for Region 1 and Region 3 is essentially the same,
reducing the complexity of the register programming. Other
registers need to be used during the actual readout operation.
These include the MODE registers, shutter control registers
(PRIMARY_ACTION, SUBCK, GPO for MSHUT, and VSUB
control), and AFE gain register.
Important Note Regarding Signal Polarities
When programming the AD9992 to generate the XV1 to XV24
and SUBCK signals, the external V-driver circuit usually inverts
these signals. Carefully check the required timing signals needed
at the input and the output of the V-driver circuit being used and
adjust the polarities of the AD9992 outputs accordingly.
Rev. C | Page 45 of 92
Rev. C | Page 46 of 92
Figure 55. CCD Timing Example—Dividing Each Field into Regions
05891-055
CCD
OUT
VSUB
MSHUT
SUBCK
XV6
XV5
XV4
XV3
XV2
XV1
HD
VD
OPEN
REGION 0
N–5
N–2
REGION 2
REGION 3
1
4
7
10
13
16
FIELD 0
REGION 1
FIRST FIELD READOUT
CLOSED
EXPOSURE (tEXP)
REGION 0
N–4
N–1
REGION 2
REGION 3
2
5
8
11
14
17
20
FIELD 1
REGION 1
SECOND FIELD READOUT
REGION 0
N–3
N
REGION 2
REGION 3
3
6
9
12
15
18
21
FIELD 2
REGION 1
THIRD FIELD READOUT
OPEN
AD9992
AD9992
SHUTTER TIMING CONTROL
SUBCK: Low Speed Operation
The AD9992 supports the generation of electronic shuttering
(SUBCK) and also features flexible general-purpose outputs
(GPO) to control mechanical shuttering, CCD substrate bias
switching, and strobe circuitry. In the following documentation,
the terms sense gate (SG) and vertical sense gate (VSG) are used
interchangeably.
Normal and high precision shutter operations are used when
the exposure time is less than 1 field. For exposure times greater
than 1 field, the low speed (LS) shutter features can be used.
The AD9992 includes a field counter (primary field counter) to
regulate long exposure times. The primary field counter must
be activated (Address 0x70) to serve as the trigger for the LS
operation. The durations of the LS exposure and read are
specified by the SGMASK_NUM and SUBCKMASK_NUM
register (Address 0x74), respectively. As shown in Figure 58,
this mode suppresses the SUBCK and VSG outputs for up to
8192 fields (VD periods).
SUBSTRATE CLOCK OPERATION (SUBCK)
The CCD image exposure time is controlled by the substrate
clock signal (SUBCK), which pulses the CCD substrate to clear
out accumulated charge. The AD9992 supports three types of
electronic shuttering: normal, high precision, and low speed.
Along with the SUBCK pulse placement, the AD9992 can
accommodate different readout configurations to further
suppress the SUBCK pulses during multiple field readouts.
The SUBCK signal is a programmable string of pulses, each
occupying a line following the primary sense gate active line,
SGACTLINE1 (registers are shown in Table 21). The SUBCK
signal has programmable pulse width, line placement, and
number of pulses to accurately control the exposure time.
SUBCK: Normal Operation
By default, the AD9992 operates in the normal SUBCK
configuration, in which the SUBCK signal is pulsing in every
VD field (see Figure 56). The SUBCK pulse occurs once per
line, and the total number of repetitions within the field
determines the length of the exposure time. The SUBCK pulse
polarity and toggle positions within a line are programmable using
the SUBCK_POL and SUBCK_TOG1 registers (see Table 21).
The number of SUBCK pulses per field is programmed in the
SUBCKNUM register (Address 0x75).
As shown in Figure 56, the SUBCK pulses always begin in the line
following the SG-active line, which is specified in the SGACTLINE
registers for each field. The SUBCK_POL, SUBCK_TOG1,
SUBCK_TOG2, SUBCKNUM, and SUBCKSTARTLINE
registers are updated at the start of the line after the sensor gate
line, as described in the Updating New Register Values section.
SUBCK: High Precision Operation
High precision shuttering is used in the same manner as normal
shuttering but uses an additional register to control the last
SUBCK pulse. In this mode, the SUBCK still pulses once per
line, but the last SUBCK in the field has an additional SUBCK
pulse, whose location is determined by the SUBCKHP_TOG
registers, as shown in Figure 57. Finer resolution of the exposure
time is possible using this mode. Leaving the SUBCKHP_TOG
registers set to its maximum value (0xFFFFFF) disables the last
SUBCK pulse (default setting).
To activate an LS shutter operation, trigger the start of the exposure
by writing to the PRIMARY_ACTION register bits according to
the desired effect. When the primary counter is activated, the next
VD period becomes the first active period of the exposure for
which the VSG and SUBCK masks are applied.
Optionally, if the SUBCKMASK_SKIP1 register is enabled, the
AD9992 ignores the first VSG and SUBCK masks in subsequent
fields. This is generally desired so that the exposure time begins
in the field after the exposure operation is initiated. Figure 58
shows operation with SUBCKMASK_SKIP1 = 1.
If the PRIMARY_ACTION register is used while the
SUBCKMASK_NUM and SGMASK_NUM registers are set to 0,
the behavior of the SUBCK and VSG signals are not different
from the normal shutter or high precision shutter operations.
Therefore, the primary field counter can be used for other tasks
(described in the General-Purpose Outputs (GPOs) section)
without disrupting the normal activity. In addition, there exists
a secondary field counter that has no effect on the SUBCK and
VSG signals. These counters are described in detail in the Field
Counters section.
SUBCK Start Line
By default, the SUBCK pulses begin in the line following
SGACTLINE1. For applications where the SUBCK pulse should
be suppressed for one or more lines following the VSG line, the
SUBCKSTARTLINE register can be programmed. This register
setting delays the start of the SUBCK pulses until the specified
number of lines following SGACTLINE1.
Caution
A value of 1 should not be used in the SUBCKSTARTLINE
register. A value of 0 is used to specify the SUBCK pulses to
begin in the next line after the SG line. A value of 2 is used to
specify the SUBCK pulses to begin two lines after the SG line,
and so on.
Rev. C | Page 47 of 92
AD9992
Read After Exposure
To read the CCD data after exposure, the SG should resume
normal activity while the SUBCK remains null. By default,
the AD9992 generates the VSG pulses in every field. When only
a single exposure and a single frame read are desired, such as is
the case in the preview mode, the VSG and SUBCK pulses can
operate in every field.
Other applications require that more frames are read, in which
case SUBCK must be masked until the readout is finished.
The SUBCKMASK_NUM register specifies the total number of
fields (exposure and read) to mask SUBCK. A 2-field CCD
frame read mode typically requires two additional fields of
SUBCK masking (SUBCKMASK_NUM = 2). A 3-field, 6-phase
CCD requires three additional fields of SUBCK masking after
the read begins (SUBCKMASK_NUM = 3).
Note that the SUBCKMASK_SKIP1 register setting allows
SUBCK pulses at the beginning of the field of exposure.
Table 21. SUBCK and Exposure/Read Register Parameters
Register
SGMASK_NUM
SUBCKMASK_NUM
SUBCKMASK_SKIP1
SUBCKSTARTLINE 1
Length
13b
13b
1b
13b
Range
0 to 8191 fields
0 to 8191 fields
On/off
0, 2 to 8191 line locations
SUBCKNUM1
SG_SUPPRESS1
SUBCK_TOG1
SUBCK_TOG2
SUBCK_POL
SUBCKHP_TOG1
SUBCKHP_TOG2
13b
1b
13b
13b
1b
13b
13b
1 to 8191 pulses
On/off
0 to 8191 pixel locations
0 to 8191 pixel locations
Low/high
0 to 8191 pixel locations
0 to 8191 pixel locations
1
Description
Exposure duration (number of fields to suppress VSG) for LS operation.
Exposure plus readout duration (number of fields to suppress SUBCK) for LS.
Suppress SG/SUBCK masks for one field (default = 0). Typically set to 1.
Line location to start the SUBCK pulses, relative to SGLINE location.
A value of 1 is invalid. See the SUBCK Start Line section.
Total number of SUBCKs per field, at 1 pulse per line. Must be <VDLEN.
Suppress the SG and allow SUBCK to finish at SUBCKNUM.
SUBCK Toggle Position 1.
SUBCK Toggle Position 2.
SUBCK start polarity.
High precision SUBCK Toggle Position 1. Selectable as SG or VD updated.
High precision SUBCK Toggle Position 2. Selectable as SG or VD updated.
Register is not VD updated but is updated at the start of the line after the sensor gate line.
Rev. C | Page 48 of 92
AD9992
VD
HD
VSG
tEXP
tEXP
SUBCK PROGRAMMABLE SETTINGS:
1. PULSE POLARITY USING THE SUBCK_POL REGISTER.
2. NUMBER OF PULSES WITHIN THE FIELD USING THE SUBCKNUM REGISTER (SUBCKNUM = 3 IN THE ABOVE EXAMPLE).
3. PIXEL LOCATION OF PULSE WITHIN THE LINE AND PULSE WIDTH PROGRAMMED USING THE SUBCK1 TOGGLE POSITION REGISTERS.
05891-056
SUBCK
Figure 56. Normal SUBCK Operation
VD
HD
VSG
tEXP
tEXP
NOTES
1. SECOND SUBCK PULSE IS ADDED IN THE LAST SUBCK LINE.
2. LOCATION OF SECOND PULSE IS FULLY PROGRAMMABLE USING THE SUBCKHP TOGGLE POSITION REGISTERS.
05891-057
SUBCK
Figure 57. High Precision SUBCK Operation
TRIGGER
EXPOSURE
(0x70)
VD
VSG
tEXP
NOTES
1. SUBCK CAN BE SUPPRESSED FOR MULTIPLE FIELDS BY PROGRAMMING THE EXPOSURE REGISTER TO BE GREATER THAN 0.
2. ABOVE EXAMPLE USES EXPOSURE = 1.
3. TRIGGER REGISTER MUST ALSO BE USED TO START THE LOW SPEED EXPOSURE.
4. VD/HD OUTPUTS CAN ALSO BE SUPPRESSED USING THE VDHDOFF REGISTER = 1.
Figure 58. Low Speed SUBCK Operation
Rev. C | Page 49 of 92
05891-058
SUBCK
AD9992
FIELD COUNTERS
The AD9992 contains three field counters (primary, secondary,
and mode). When these counters are active, they increment
with each VD cycle. The mode counter is the field counter used
with the mode register to control the vertical timing signals (see
the MODE Registers section. The primary and secondary
counters are more flexible and are generally used for shuttering
signal applications. Both the primary and secondary counters
have several modes of operation that are dictated by Address
0x70, including
•
•
•
Normal (single count)
RapidShot (repeating count)
ShotDelay (delayed count)
•
•
•
•
ShotDelay with RapidShot
Manual exposure
Manual readout
Force to idle
The primary counter regulates the expose and read actions by
regulating the SUBCK and VSG signals. If the RapidShot
feature is used with the primary counter, the SUBCK and VSG
masking automatically repeats as necessary for multiple expose/
read cycles. The secondary counter has no effect on the SUBCK
or VSG signal. Both counters can be used to regulate the generalpurpose signals described in the General-Purpose Outputs
(GPOs) section.
Table 22. Primary/Secondary Field Counter Registers (Address 0x70, Address 0x71, and Address 0x72)
Register
PRIMARY_ACTION
SECOND_ACTION
Length
3b
3b
PRIMARY_MAX
SECOND_MAX
VDHD_MASK
PRIMARY_DELAY
13b
12b
3b
13b
PRIMARY_SKIP
SECOND_DELAY
1b
13b
SECOND_SKIP
1b
Description
0: Idle, no counter action. GPO signals can still be controlled using polarity or GP_PROTOCOL = 1.
1: Activate counter. Single cycle of counter from 1 to counter maximum value, and then returns to idle state.
2: RapidShot. After reaching maximum counter value, counter wraps and repeats until reset.
3: ShotTimer. Active single cycle of counter after added delay of N fields (use the corresponding DELAY
register).
4: ShotTimer with RapidShot. Same as 2, with added delay of N fields between each repetition.
5: Manual exposure. Primary counter stays in exposure until manual readout or reset to idle.
This mode keeps the SUBCK and VSG pulses masked indefinitely.
6: Manual readout. Primary counter switches to readout (VSG pulses becomes active).
7: Force to idle.
Primary counter maximum value.
Secondary counter maximum value.
Mask VD/HD during counter operation.
ShotTimer. Number of fields to delay before the next primary count (exposure) starts. If using ShotTimer
with RapidShot, delay value is used between each repeat.
When using ShotTimer with RapidShot, use primary delay value only before first count (exposure).
ShotTimer. Number of fields to delay before the next secondary count starts. If using ShotTimer with
RapidShot, delay value is used between each repeat.
When using ShotTimer with RapidShot, use secondary delay value only before first count.
Rev. C | Page 50 of 92
AD9992
GENERAL-PURPOSE OUTPUTS (GPOS)
For Protocol 1 (no counter association), skip Step 3 and Step 4.
The AD9992 provides programmable outputs to control a
mechanical shutter, strobe/flash, the CCD bias select signal, or
any other external component with general-purpose (GP)
signals. Eight GP signals, with up to four toggles each, are
available that can be programmed and assigned to special GPO
pins. These pins are bidirectional and allow visibility (as an output)
and external control (as an input) of HBLK, PBLK, CLPOB, and
OUTCONTROL. The registers introduced in this section are
described in Table 23.
With these four steps, the GP signals can be programmed to
accomplish many common tasks. Careful protocol selection and
application of the field counters yields efficient results to allow
the GP signals smooth integration with concurrent operations.
GP Toggles
When configured as an output, each GPO1 to GPO8 output can
deliver a signal that is the result of programmable toggle positions.
The GP signals are independent and can be linked to either
a specific VD period or over a range of VD periods via the primary
or secondary field counters through the GP protocol registers
(Address 0x73). As a result of their associations with the field
counters, the GP toggles inherit the characteristics of the field
counters, such as RapidShot and ShotDelay.
To use the GP toggles,
1.
2.
3.
4.
Program the toggle positions (Address 0x7A to
Address 0xA9).
Program the protocol (Address 0x73).
Program the counter parameters (Address 0x71 to
Address 0x72).
Activate the counter (Address 0x70).
Note that the SUBCK and VSG masks are linked to the primary
counter; however, if their parameters are 0, the GPO can use
the primary counter without expose/read activity.
The secondary counter is independent and can be used
simultaneously with the primary counter. Some applications
may require the use of both primary and secondary field
counters with different GPO protocols, start times, and
durations. Such operations are easily handled by the AD9992.
Several simple examples of GPO applications using only one
GPO and one field counter follow. These examples can be used
as building blocks for more complex GPO activity. In addition,
specific GPO signals can be passed through a 4-input LUT to
realize combinational logic between them. For example, GP1
and GP2 can be sent through an XOR look-up table, and the result
can be delivered on GP1, GP2, or both. Also, either GP1 or GP2
can deliver their original toggles.
Rev. C | Page 51 of 92
AD9992
Table 23. GPO Registers
Register
GP1_PROTOCOL
GP2_PROTOCOL
GP3_PROTOCOL
GP4_PROTOCOL
GP5_PROTOCOL
GP6_PROTOCOL
GP7_PROTOCOL
GP8_PROTOCOL
MANUAL_TRIG
GP<1:8>_POL
SEL_GP<1:8>
Length
3b
3b
3b
3b
3b
3b
3b
8b
8b
8b
8b
Range
0 to 7
0 to 7
0 to 7
0 to 7
0 to 7
0 to 7
0 to 7
0 to 7
Off/on
Low/high
Off/on
GPO_OUTPUT_EN
8b
Off/on
GP*_USE_LUT
LUT_FOR_GP12
LUT_FOR_GP34
LUT_FOR_GP56
LUT_FOR_GP78
8b
4b
4b
4b
4b
Off/on
Logic setting
Logic setting
Logic setting
Logic setting
GP*_TOG*_FD
GP*_TOG*_LN
GP*_TOG*_PX
GPO_INT_EN
13b
13b
13b
1b
0 to 8191 field
0 to 8191 line
0 to 8191 pixel
Off/on
Description
0: Idle.
1: No counter association, use MANUAL_TRIG bits to enable each GP signal.
2: Link to primary counter.
3: Link to secondary counter.
4: Link to mode counter (from vertical timing generation).
5: Primary repeat (allows GP signals to repeat with RapidShot).
6: Secondary repeat (allows GP signals to repeat with RapidShot).
7: Keep on.
Manual trigger for each GP signal, for use with Protocol 1.
Starting polarity for GP signals, only updated during PROTOCOL = 0.
1: Select GP toggles visible at GPO1 to GPO8 when output is enabled (default);
0: Select vertical signals visible at GPO4 to GPO8 when output is enabled.
GPO4: SUBCK.
GPO5: XV21.
GPO6: XV22.
GPO7: XV23.
GPO8: XV24.
1: Enable GPO1 to GPO8 outputs (one bit per output).
0: Disable GPO1 to GPO8 outputs, pins will be high-Z state (default).
Send GP signals through a programmable look-up table (LUT).
Desired logic to be realized on GP1 combined with GP2.
Desired logic to be realized on GP3 combined with GP4.
Desired logic to be realized on GP5 combined with GP6.
Desired logic to be realized on GP7 combined with GP8.
Example logic settings for LUT_FOR_GPxy:
0x6: GPy XOR GPx (see Figure 64).
0x7: GPy NAND GPx.
0x8: GPy AND GPx.
0xE: GPy OR GPx.
Field of activity, relative to primary and secondary counter for corresponding toggle.
Line of activity for corresponding toggle.
Pixel of activity for corresponding toggle.
When set to 1, internal signals are viewable on GPO5 to GPO8.
GPO5: OUTCONTROL.
GPO6: HBLK.
GPO7: CLPOB.
GPO8: PBLK.
Rev. C | Page 52 of 92
AD9992
Single-Field Toggles
Scheduled Toggles
Single-field toggles occur in the next field only. There can be up to
four toggles in the field. The mode is set with GP_PROTOCOL
equal to 1, and then the toggles are triggered in the next field by
writing to the MANUAL_TRIG register (0x70 [13:6]). In this
mode, the field toggle settings must be set to a value of 1. Two
consecutive fields do not have activity. If toggles are required to
repeat in the next field, the MANUAL_TRIG register can be
written to in consecutive fields.
Scheduled toggles are programmed to occur during any upcoming
field. For example, there can be one toggle in Field 1, two
toggles in Field 3, and a last toggle in Field 4. The mode is set
with GP_PROTOCOL = 2 or GP_PROTOCOL = 3. Mode 2
tells the GPO to obey the primary field counter, and Mode 3
tells the GPO to obey the secondary field counter.
Preparation
The GP toggle positions can be programmed any time prior to
use. For example,
0x7A
0x7B
0x7C
0x7D
0x7E
0x7F
Preparation
The GP toggle positions can be programmed any time prior to
use. For example,
0x7A
0x7B
0x7C
Å 0x000A001
Å 0x0002000
Å 0x000000F
Å 0x00C4002
Å 0x0004000
Å 0x00000B3
Å 0x00C4002
Å 0x0004000
Å 0x00000B3
Details
A) Field 0:
Details
0x70
0x73
Å 0x0000008
Å 0x0000003
VD
A) Field 0:
B) Field 1:
0x70
0x73
Å 0x0000040
Å 0x0000001
0x73
Å 0x0000000
VD
REGISTER WRITE
1
A
GP1_PROTOCOL 0
SECONDARY 0 (IDLE)
COUNT
1
2
3
1
2
0
GPO1
GP1_PROTOCOL 0
A
B
1
CAUTION! THE PRIMARY COUNTER REGULATES THE SUBCK
AND VSG ACTIVITY. LINK A GPO TO THE PRIMARY COUNTER
ONLY IF IT IS TO HAPPEN DURING EXPOSURE/READ.
0
Figure 60. Scheduled Toggles Using GP1_PROTOCOL = 3
NOTES
1. THE FIELD TOGGLE POSITION MUST BE SET TO 1 WHEN
GP PROTOCOL IS 1.
CAUTION! THE GP_PROTOCOL MUST BE RESET BEFORE
USING AGAIN.
05891-059
GPO1
Figure 59. Single-Field Toggles Using GP1_PROTOCOL = 1
Rev. C | Page 53 of 92
05891-060
REGISTER WRITE
AD9992
RapidShot Sequences
ShotDelay Sequences
RapidShot technology provides continuous repetition of
scheduled toggles.
ShotDelay technology provides internal delay of scheduled
toggles. The delay is in terms of fields.
Preparation
Preparation
The GP toggle positions can be programmed any time prior
to use. For example,
The GP toggle positions can be programmed any time prior
to use. For example,
Å 0x0004000
Å 0x000A001
Å 0x0002000
Å 0x000000F
Å 0x00C4002
Å 0x0004000
Å 0x00000B3
Å 0x0000006
0x71
0x72
0x7A
0x7B
0x7C
0x73
Å 0x0004000
Å 0x000C000
Å 0x000A001
Å 0x0002000
Å 0x000000F
Å 0x0000003
Details
A) Field 0:
Details
A) Field 0:
0x70
Å 0x0000010
B) Field 2:
0x70
Å 0x0000007
VD
REGISTER WRITE
1
VD
2
3
4
A
A
3
B
SECONDARY 0 (IDLE)
COUNT
GP1_PROTOCOL 0
2
1
5
GP1_PROTOCOL 0
REGISTER WRITE
Å 0x0000018
0x70
1
2
3
1
2
0
6
GPO1
SECONDARY 0 (IDLE)
COUNT
1
2
1
2
1
0
Figure 62. ShotDelay Toggle Operation Using GP1_PROTOCOL = 3
TERMINATED
AT VD EDGE
NOTES
1. THE GP PROTOCOLS ARE THE SAME AS THE SCHEDULED
TOGGLES, EXCEPT THE TOGGLES CAN BE EXCLUDED FROM
REPETITION BY CHOOSING GP PROTOCOL 2 OR 3.
CAUTION! THE FIELD COUNTER MUST BE FORCED INTO IDLE
STATE TO TERMINATE REPETITIONS.
05891-062
GPO1
Figure 61. RapidShot Toggle Operation Using GP1_PROTOCOL = 6
Rev. C | Page 54 of 92
05891-063
0x71
0x7A
0x7B
0x7C
0x7D
0x7E
0x7F
0x73
AD9992
GP LOOK-UP TABLES (LUT)
Table 24. LUT Results Based on GP1 and GP2 Values
The AD9992 is equipped with a look-up table for each pair of
consecutive GP signals when configured as outputs. GP1 is
always combined with GP2, GP3 is always combined with GP4,
GP5 is always combined with GP6, and GP7 is always combined
with GP8. The external GPO outputs from each pair can output
the result of the LUT or the original GP internal signal.
GP2
0
0
1
1
GP1
0
1
0
1
LUT: XOR
0
1
1
0
LUT: NAND
1
1
1
0
LUT: AND
0
0
0
1
LUT: OR
0
1
1
1
LUT_FOR_GP12[11:8] = 0x06
GP2_USE_LUT = 1 GP1_USE_LUT = 0
GP1_USE_LUT
GP1
GP1
GP2
GPO1
GPO2
1
GPO1
LUT
GP2
1
NOTES
1. LOGIC COMBINATION (XOR) OF PROGRAMMED TOGGLES
GP1 AND GP2.
GPO2
Figure 64. LUT Example for GP1 XOR GP2
05891-064
0
GP2_USE_LUT
05891-065
0
Field Counter and GPO Limitations
Figure 63. Internal LUT for GP1 and GP2 Signals
The following is a summary of the known limitations of the field
counters and GPO signals that dictate usability:
Address 0x79 dictates the behavior of the LUT and which
signals receive the result. Each 4-bit LUT_FOR_GPxy register
can realize any logic combination of GPx and GPy. For example,
Table 24 shows how the register values of LUT_FOR_GP12 [11:8]
are determined. XOR, NAND, AND, and OR results are shown,
but any 4-bit combination is possible. A simple example of XOR
gating is shown in Figure 64.
• The field counter trigger (PRIMARY_ACTION and
SECONDARY_ACTION registers, Address 0x70) is self-reset
at the start of every VD period. Therefore, there must be one
VD period between sequential programming to that address.
• If GP*_PROTOCOL = 1, it must be manually reset to
GP*_PROTOCOL = 0 one VD period before it can be used
again. If manual toggles are desired in sequential fields, the
MANUAL_TRIG register should be used in conjunction with
GP*_PROTOCOL = 1.
Rev. C | Page 55 of 92
AD9992
Write to the MODE registers to configure the next five
fields. The first two fields during exposure are the same as
the current draft mode fields, and the following three fields
are the still-frame readout fields. The register settings for
the draft mode field and the three readout fields are
previously programmed. Note that if the MODE registers
are changed to VD updated, only one field of exposure
should be included (the second one) because the MODE
settings will be delayed an extra field.
COMPLETE EXPOSURE/READOUT OPERATION
USING PRIMARY COUNTER AND GPO SIGNALS
Figure 65 demonstrates a typical expose/read cycle while exercising
the GPO signals. Using a 3-field CCD with an exposure time
that is greater than one field but less than two fields in duration
requires a total of five fields for the entire exposure/readout
operation. Other exposure times and other CCD field configurations require modification of these example settings.
Note that, if the MODE registers are changed to VD updated, as
shown in the MODE Registers section and in Figure 53, the
MODE update is delayed by one additional field. This should be
accounted for in selecting the number of fields to cycle and the VD
location to write to the MODE registers.
1.
The primary counter is used to control the masking of VSG
and SUBCK during exposure/readout. The PRIMARY_MAX
register should be set equal to the total number of fields
used for exposure and readout. In this example,
PRIMARY_MAX = 5.
The SUBCK masking should not occur immediately at
the next VD edge (Step 2) because this would define an
exposure time that begins in the previous field. Write to
the PRIMARY_DELAY register to delay the masking of
VSG and SUBCK pulses in the first exposure field. In this
example, MASKDELAY = 1.
Write to the SUBCKMASK_NUM register (Address 0x74)
to specify the number of fields to mask SUBCK while the
CCD data is read. In this example, SUBCKMASK_NUM = 4.
Write to the SGMASK_NUM register (Address 0x74) to
specify the number of fields to mask VSG outputs during
exposure. In this example, SGMASK_NUM = 1.
Write to the PRIMARY_ACTION register (Address 0x70)
to trigger the GP1 (STROBE), GP2 (MSHUT), and GP3
(VSUB) signals and to start the expose/read operation.
2.
VD/HD falling edge updates the serial writes from 1.
3.
GP3 (VSUB) output turns on at the field/line/pixel specified.
VSUB Example 1 and Example 2 use GP3TOG1_FD = 1.
4.
GP1 (STROBE) output turns on and off at the location
specified.
5.
GP2 (MSHUT) output turns off at the location specified.
6.
The next VD falling edge automatically starts the first
read field.
7.
The next VD falling edge automatically starts the second
read field.
8.
The next VD falling edge automatically starts the third
read field.
9.
Write to the MODE register to reconfigure the single draft
mode field timing. Note that if the MODE registers are
changed to VD updated, this write should occur one field
earlier.
10. VD/HD falling edge updates the serial writes from Step 9.
VSG outputs return to draft mode timing. SUBCK output
resumes operation. GP2 (MSHUT) output returns to the on
position (active or open). GP3 (VSUB) output returns to
the off position (inactive)
Rev. C | Page 56 of 92
Rev. C | Page 57 of 92
05891-066
CCD
OUT
VSUB
(GPO3)
MECHANICAL
SHUTTER
MSHUT
(GPO2)
STROBE
(GPO1)
SUBCK
VSG
VD
PRIMARY
COUNT
SERIAL
WRITES
DRAFT IMAGE
0 (IDLE)
1
2
3
tEXP
2
EXAMPLE 1
4
DRAFT IMAGE
1
CLOSED
EXAMPLE 2
OPEN
5
6
3
STILL IMAGE
FIRST FIELD
7
STILL IMAGE
SECOND FIELD
STILL IMAGE READOUT
4
8
5
9
STILL IMAGE
THIRD FIELD
10
10
10
10
0
OPEN
0
DRAFT IMAGE
AD9992
Figure 65. Complete Exposure/Readout Operation Using Primary Counter and GPO Signals
AD9992
MANUAL SHUTTER OPERATION USING
ENHANCED SYNC MODES
Shutter Operation in SLR Mode
The AD9992 also supports an external signal to control exposure,
using the SYNC input. Generally, the SYNC input is used as an
asynchronous reset signal during master mode operation. When
the enhanced SYNC mode is enabled, the SYNC input provides
additional control of the exposure operation.
1.
To turn on VSUB, write to the appropriate GP registers to
start VSUB and manual exposure [PRIMARY_ACTION = 5].
This change takes effect after the next VD. SUBCK is suppressed during the exposure and readout phases.
2.
To turn on MSHUT during the interval between the next
VD and SYNC, write to the appropriate GP register. When
MSHUT is in the on position, it has line and pixel control.
This change takes effect on the SYNC falling edge because
there is an internal VD.
3.
If the MODE register is programmed to cycle through
multiple fields (5, 7, 3, 5, 7, 3, …, in this example), the internal
field designator increments. If the MODE register is not
required to increment, set up the MODE register such that it
outputs only one field. This prevents the MODE counter
from incrementing during the SYNC interval.
4.
Write to the manual readout trigger to begin the manual
readout [PRIMARY_ACTION = 6]. Write to the appropriate
GP registers to trigger MSHUT to toggle low at the end of
the exposure. This change takes effect on the SYNC rising
edge during readout. Because VD register update is
disabled, the trigger takes effect on the SYNC rising edge.
The MSHUT falling edge is aligned to the SYNC rising
edge. Because the MSHUT falling edge is aligned with VD,
it may be necessary to insert a dummy VD to delay the
readout.
Referring to Figure 71,
Normal SYNC Mode (Mode 1)
By default, the SYNC input is used in master mode for synchronizing the internal counters of the AD9992 with external
timing. The SYNC During Master Mode Operation section
describes how horizontal, vertical, and field designator signals are
reset by the rising edge of the SYNC pulse. Figure 66 also shows
how this mode operates, highlighting the behavior of the mode
field designator.
Enhanced SYNC Modes (Mode 2 and Mode 3)
The enhanced SYNC modes can be used to accommodate unique
synchronization requirements during exposure operations. In
SYNC Mode 2, the V and VSG outputs are suspended and the
VD output is masked. The V-outputs are held at the dc value
established by the Sequence 0 start polarities. There is no SCP
operation, but the H-counter is still enabled. Finally, the AFE
sampling clocks HD, H/RG, CLPOB, and HBLK are operational
and use Sequence 0 behavior. See Figure 67 for more details.
Set Register ENH_ SYNC_EN (Address 0x13 Bit 3) to 1 to
enable the enhanced SYNC modes
Mode 3 uses all of these features, but the V-outputs are continuous
through the SYNC pulse interval. VD control pulses are masked
during the SYNC interval, and the HD pulse can also be masked
if required (see Figure 68).
It is important to note that in both of these enhanced modes,
the SYNC pulse resets the counters at both the falling edge and
the rising edge of the SYNC pulse.
Register Update and Field Designator
When using special SYNC Mode 2 or SYNC Mode 3, the VDupdated registers, such as PRIMARY_ACTION, are not
updated during the SYNC interval, and the SCP0 function is
ignored and held at 0 (see Figure 69).
When using SYNC Mode 2 or SYNC Mode 3, both the rising and
falling edges increment the internal field designator; therefore,
the new register data takes effect and VTP information is
updated to new SEQ0 data. However, this does not occur if the
MODE register creates an output of one field. In that case,
the region, sequence, and group information does not change
(see Figure 70).
Note that because the internal exposure counter (PRIMARY
counter) is not used during manual SYNC mode operation and
the VD register update is disabled, control is lost on the fine
placement of the GP signals for VSUB, MSHUT, and STROBE
edges while SYNC is low.
New Serial Registers
SYNC Mode 2 and SYNC Mode 3 are controlled using the registers
listed in Table 25. Note that registers for enhanced SYNC modes
are located at Address 0x13 Bits [6:3].
Table 25. Registers for Enhanced SYNC Modes
Register
ENH_SYNC_EN
Length
1b
SYNC_MASK_V
1b
SYNC_MASK_VD
1b
SYNC_MASK_HD
1b
Rev. C | Page 58 of 92
Description
HI active to enable
(default LO)
HI active to enable masking
(default LO)
HI active to enable masking
(default HI)
HI active to enable masking
(default HI)
AD9992
SYNC
VD
FIELD
DESIGNATOR
7
3
5
SUSPEND
HD
NOTES
1. THE SYNC RISING EDGE RESETS VD/HD AND COUNTERS TO 0.
2. SYNC POLARITY IS PROGRAMMABLE USING SYNCPOL REGISTER (ADDR 0x13).
3. DURING SYNC LOW, ALL INTERNAL COUNTERS ARE RESET AND VD/HD CAN BE SUSPENDED USING THE SYNCSUSPEND REGISTER (ADDR 0x13).
4. THE SYNC RISING EDGE CAUSES THE INTERNAL FIELD DESIGNATOR TO INCREMENT.
5. IF SYNCSUSPEND = 1, VERTICAL CLOCKS, H1 TO H4, AND RG ARE HELD AT THE SAME POLARITY SPECIFIED BY OUTCONTROL = LOW.
6. IF SYNCSUSPEND = 0, ALL CLOCK OUTPUTS CONTINUE TO OPERATE NORMALLY UNTIL SYNC RESET EDGE.
05891-083
H1 TO H4, RG,
XV1 TO XV24
VSG, SUBCK
Figure 66. Default Mode 1
2
1
SYNC
3
VD
VDLEN
HD
4
SCP
5
1FALLING EDGE RESYNCS THE CIRCUIT TO THE LINE/PIXEL NUMBER 0. VD AND HD INTERNALLY RESYNC.
2RISING EDGE RESETS COUNTERS.
3VD IS DISABLED DURING SYNC. THE REGISTER IS PROGRAMMABLE.
4SCP, HBLK, AND CLPOB ARE HELD AT SEQ0 VALUE.
5XV1 TO XV24 SIGNALS ARE HELD AT THE V-OUTPUT START POLARITY.
Figure 67. Enhanced SYNC Mode 2 with Vertical Signals Held at VTP Start Value
Rev. C | Page 59 of 92
05891-084
XV1 TO XV24
AD9992
SYNC
1
VD
VDLEN
2
HD
SCP
3
XV1 TO XV24
05891-085
1SYNC_MASK_VD IS A NEW REGISTER. HI WILL MASK VD. DEFAULT = HI.
2SYNC_MASK_HD IS A NEW REGISTER. HI WILL MASK HD. DEFAULT = LO.
3V-OUTPUT PULSES CONTINUE IN SEQUENCE.
Figure 68. Enhanced SYNC Mode 3
SYNC
VD
1
1VD
1
1
1
1
REGISTERS ARE UPDATED HERE.
05891-086
1
NOTES
1. VD-UPDATED REGISTERS (FOR EXAMPLE, PRIMARY_ACTION) ARE DISABLED DURING THE SYNC INTERVAL.
Figure 69. Register Update Behavior
SYNC
VD
5
7
5
3
1
1
1FIELD
DESIGNATOR IS INCREMENTED ON BOTH SYNC EDGES.
Figure 70. Special SYNC Mode Effect on Field Designator
Rev. C | Page 60 of 92
7
05891-087
FIELD
DESIGNATOR
AD9992
SYNC
1
4
2
VD
3
FIELD
DESIGNATOR
3
5
7
3
5
7
3
5
7
V-OUTPUTS
MSHUT
5
DRAFT
EXPOSURE
5
DUMMY
FIELD
READOUT
ODD
READOUT
EVEN
1SEE THE SHUTTER OPERATION IN SLR MODE SECTION.
2SEE THE SHUTTER OPERATION IN SLR MODE SECTION.
3SEE THE SHUTTER OPERATION IN SLR MODE SECTION.
4SEE THE SHUTTER OPERATION IN SLR MODE SECTION.
5SUBCK OUTPUT IS SUPPRESSED DURING EXPOSURE AND READOUT WHEN EXPOSURE TRIGGER IS USED.
Figure 71. Enhanced SYNC Mode—Manual Shutter Operation, SLR Mode
Rev. C | Page 61 of 92
DRAFT
05891-088
VSUB
AD9992
ANALOG FRONT END DESCRIPTION AND OPERATION
0.1µF 0.1µF
REFB REFT
0.4V
DC RESTORE
SHP
SHD
0.1µF
FIXED
DELAY
CLI
SHP
PBLK (WHEN DCBYP = 1)
1.2V
1.4V
INTERNAL
VREF
DCLK
0
DOUTPHASE
DCLK
MODE
2V FULL SCALE
6dB ~ 42dB
AD9992
1
DCLKINV
S11
CCDIN
CDS
–3dB, 0dB,
+3dB, +6dB
S21
PBLK
DAC
VGA GAIN
REGISTER
CDS GAIN
REGISTER
OUTPUT
DATA
LATCH
12-BIT
ADC
VGA
CLPOB PBLK
BLANK TO
ZERO OR
CLAMP LEVEL
CLAMP LEVEL
REGISTER
PBLK
VD
V-H
TIMING
GENERATION
HD
05891-067
PRECISION
TIMING
GENERATION
CLI
CLPOB
DOUT
OPTICAL BLACK
CLAMP
DIGITAL
FILTER
DOUTPHASE
SHP SHD
12
1S1 IS NORMALLY CLOSED; S2 IS NORMALLY OPEN.
Figure 72. Analog Front-End Functional Block Diagram
The AD9992 signal processing chain is shown in Figure 72.
Each processing step is essential for achieving a high quality
image from the raw CCD pixel data.
Note that, because the CDS input is shorted during PBLK, the
CLPOB pulse should not be used during the same active time as
the PBLK pulse.
DC Restore
Correlated Double Sampler (CDS)
To reduce the large dc offset of the CCD output signal, a dc
restore circuit is used with an external 0.1 μF series coupling
capacitor. This restores the dc level of the CCD signal to
approximately 1.2 V, making it compatible with the 1.8 V core
supply voltage of the AD9992. The dc restore switch is active
during the SHP sample pulse time.
The CDS circuit samples each CCD pixel twice to extract the video
information and to reject low frequency noise. The timing
shown in Figure 19 illustrates how the two internally generated
CDS clocks, SHP and SHD, are used to sample the reference
level and data level of the CCD signal, respectively. The placement
of the SHP and SHD sampling edges is determined by the setting
of the SHPLOC and SHDLOC registers located at Address 0x37.
Placement of these two clock signals is critical for achieving the
best performance from the CCD.
The dc restore circuit can be disabled when the optional PBLK
signal is used to isolate large-signal swings from the CCD input
(see the Analog Preblanking section). Bit 6 of AFE Register
Address 0x00 controls whether the dc restore is active during the
PBLK interval.
Analog Preblanking
During certain CCD blanking or substrate clocking intervals,
the CCD input signal to the AD9992 can increase in amplitude
beyond the recommended input range. The PBLK signal can be
used to isolate the CDS input from large-signal swings. While
PBLK is active (low), the CDS input is internally shorted to ground.
The CDS gain is variable in three steps by using the AFE
Address 0x04: −3 dB, 0 dB (default), and +3 dB. Improved noise
performance results from using the +3 dB setting, but the input
range is reduced (see the Analog Specifications section).
Rev. C | Page 62 of 92
AD9992
Variable Gain Amplifier
The VGA stage provides a gain range of approximately 6 dB to
42 dB, programmable with 10-bit resolution through the serial
digital interface. A gain of 6 dB is needed to match a 1 V input
signal with the ADC full-scale range of 2 V. When compared to
1 V full-scale systems, the equivalent gain range is 0 dB to 36 dB.
The VGA gain curve follows a linear-in-dB characteristic. The
exact VGA gain is calculated for any gain register value by
Gain (dB) = (0.0358 × Code) + 5.75
The AD9992 digital output data is latched using the rising edge
of the DOUTPHASE register value, as shown in Figure 72.
Output data timing is shown in Figure 20 and Figure 21. It is
also possible to leave the output latches transparent so that the data
outputs are valid immediately from the ADC. Programming the
AFE Register Address 0x01, Bit D1, to 1 sets the output latches to
transparent. The data outputs can also be disabled (three-stated) by
setting the AFE Register Address 0x01, Bit D0, to 1.
42
VGA GAIN (dB)
36
30
24
18
127
255
383
511
639
767
VGA GAIN REGISTER CODE
895
1023
05891-068
12
0
The CLPOB pulse should be aligned with the CCD’s optical
black pixels. It is recommended that the CLPOB pulse duration
be at least 20 pixels wide. Shorter pulse widths can be used, but
the ability for the loop to track low frequency variations in the
black level will be reduced. See the Horizontal Clamping and
Blanking section for timing examples.
Digital Data Outputs
where Code is the range of 0 to 1023.
6
If the CLPOB loop is disabled, higher VGA gain settings reduce
the dynamic range because the uncorrected offset in the signal
path is amplified.
Figure 73. VGA Gain Curve
ADC
The AD9992 uses a high performance ADC architecture
optimized for high speed and low power. Differential nonlinearity (DNL) performance is typically better than 0.5 LSB.
The ADC uses a 2 V input range. See Figure 5 and Figure 7 for
typical linearity and noise performance plots for the AD9992.
Optical Black Clamp
The optical black clamp loop is used to remove residual offsets in
the signal chain and to track low frequency variations in the CCD’s
black level. During the optical black (shielded) pixel interval on
each line, the ADC output is compared with a fixed black level
reference, selected by the user in the clamp level register. The value
can be programmed between 0 LSB and 255 LSB in 1023 steps.
The resulting error signal is filtered to reduce noise; the correction
value is applied to the ADC input through a DAC. Normally,
the optical black clamp loop is turned on once per horizontal
line, but this loop can be updated more slowly to suit a particular
application. If external digital clamping is used during postprocessing, the AD9992 optical black clamping can be disabled
using Bit D2 in the AFE Register Address 0x00. When the loop is
disabled, the clamp level register can still be used to provide
fixed offset adjustment.
The DCLK output can be used for external latching of the data
outputs. By default, the DCLK output tracks the values of the
DOUTPHASE registers. By changing the DCLKMODE register,
the DCLK output can be held at a fixed phase, and DOUTPHASE
register values are ignored. The DCLK output can also be inverted
with respect to DOUT, using the DCLKINV register bit.
The switching of the data outputs can couple noise back into
the analog signal path. To minimize switching noise, it is
recommended that the DOUTPHASE registers be set to the same
edge as the SHP sampling location, or up to 15 edges after the SHP
sampling location. Other settings can produce good results, but
experimentation is necessary. It is recommended that the
DOUTPHASE location not occur between the SHD sampling
location and 15 edges after the SHD location. For example, if
SHDLOC = 0, DOUTPHASE should be set to an edge location
of 16 or greater. If adjustable phase is not required for the data
outputs, the output latch can be left transparent by using
Address 0x01, Bit D1.
The data output coding is normally straight binary, but the coding
can be changed to gray coding by setting the AFE Register
Address 0x01, Bit D2, to 1.
Rev. C | Page 63 of 92
AD9992
and enables the OB clamp (AFE Register Address 0x00,
Bit 2). If the CLO output is being used to drive a crystal, also
power up the CLO oscillator by writing 1 to Address 0x15.
By default, the internal timing core is held in a reset state,
with TGCORE_RSTB register = 0. Write 1 to the
TGCORE_RSTB register (Address 0x14) to start the internal
timing core operation. Note that, if a 2× clock is used for
the CLI input, the CLIDIVIDE register (0x0D) should be
set to 1 before resetting the timing core.
Configure the AD9992 for master mode timing by writing 1
to the MASTER register (Address 0x20).
Write 1 to the OUTCONTROL register (Address 0x11).
This allows the outputs to become active after the next
SYNC rising edge. Normally OUTCONTROL takes effect
after the next VD edge; however, because the part is just
being powered up, there is no VD edge until the rising
edge of the SYNC signal.
Generate a SYNC event. If SYNC is high at power-up,
bring the SYNC input low for a minimum of 100 ns, and
then bring SYNC high again. This causes the internal
counters to reset and starts VD/HD operation. The first
VD/HD edge allows VD-updated register updates to
occur, including OUTCONTROL to enable all outputs.
If a hardware SYNC is not available, the SWSYNC register
(Address 0x13, Bit 14) can be used to initiate a SYNC event.
POWER-UP SEQUENCE FOR MASTER MODE
When the AD9992 is powered up, the following sequence is
recommended (refer to Figure 74 for each step). A SYNC signal
is required for master mode operation. If an external SYNC
pulse is not available, it is possible to generate an internal SYNC
event by writing to the SWSYNC register.
2.
3.
4.
5.
6.
Turn on the power supplies for AD9992 and start the master
clock, CLI.
Reset the internal AD9992 registers by writing 1 to the
SW_RST register (Address 0x10).
By default, Vertical Output XV1 to Vertical Output XV24
are low. If necessary, write to the Standby3 output polarity
(Address 0x26) to set different polarities for the vertical
outputs in order to avoid damage to the V-driver and
CCD. Write to Address 0x1C to configure each V-output
as a vertical transfer clock (XV) or sensor pulse (VSG).
If using an external V-driver in conjunction with the
AD9992, power up the V-driver supplies, VH and VL,
anytime after Step 3 is complete to set the proper polarities.
Load the required registers to configure the necessary
vertical timing, horizontal timing, high speed timing, and
shutter timing. Set the recommended start-up address,
0xD8, to 0x888.
To place the part into normal power operation, write 0x04
to Register Address 0x00. This sets the STANDBY register
(AFE Register Address 0x00, Bits [1:0]) to normal operation
8.
9.
10.
VH SUPPLY FOR V-DRIVER (IF USING EXTERNAL V-DRIVER)
POWER 0V
SUPPLIES
4
VL SUPPLY FOR V-DRIVER (IF USING EXTERNAL V-DRIVER)
CLI
(INPUT)
2
3
5
6
7
8
9
SERIAL
WRITES
10
SYNC
(INPUT)
VD
(OUTPUT)
HD
(OUTPUT)
XV1 TO XV24
SUBCK
tSYNC
FIRST FIELD
1H
HI-Z BY
DEFAULT
LOW BY
DEFAULT
HI-Z BY
DEFAULT
H-CLOCKS
1V
HI-Z BY
DEFAULT
H2, H4, H6, H8
CLOCKS ACTIVE WHEN OUTCONTROL
REGISTER IS UPDATED AT VD/HD EDGE
H1, H3, H5, H7, RG
Figure 74. Recommended Power-Up Sequence and Synchronization, Master Mode
Rev. C | Page 64 of 92
05891-069
1.
7.
AD9992
Power-Up and Synchronization in Slave Mode
Table 26. Power-Up Register Write Sequence
Address
0x10
0x26
0x20 to
0xFFF
0xD8
0x00
0x15
0x14
0x20
0x11
0x13
Data
0x01
User-defined
User-defined
Description
Resets all registers to default values
Standby3 vertical output polarities
Horizontal, vertical, shutter timing
0x888
0x04
0x01
0x01
0x01
0x01
0x4XX1
Configures start-up register
Powers up the AFE, enables OB clamp
Starts CLO oscillator (if using crystal)
Starts internal timing core
Configures for master mode
Enables all outputs after SYNC
SWSYNC (if using software SYNC)
The power-up procedure for slave mode operation is the same
as the procedure for master mode operation with two exceptions:
•
Eliminate Step 8. Do not write the part into master mode.
•
No SYNC pulse is required in slave mode. Substitute
Step 10 with starting the external VD and HD signals.
This synchronizes the part, allows the register updates,
and starts the timing operation.
When the AD9992 is used in slave mode, the VD/HD inputs are
used to synchronize the internal counters. After a falling edge of
VD, there is a latency of 36 master clock CLI edges after the falling
edge of HD until the internal H-counter is reset. The reset
operation is shown in Figure 76.
Using the SWSYNC Register
Additional Restrictions in Slave Mode
If an external SYNC pulse is not available, it is possible to
generate an internal SYNC in the AD9992 by writing 1 to the
SWSYNC register (Address 0x13, Bit 14). If the software SYNC
option is used, the SYNC input (Pin D3) should be low (VSS)
during power-up. The SYNCENABLE register (Address 0x13,
Bit 0) should be set high.
When operating in slave mode, the following restrictions
should be noted:
•
The HD falling edge should be located in the same CLI
clock cycle as the VD falling edge or later than the VD
falling edge. The HD falling edge should not be located
within five cycles prior to the VD falling edge.
SYNC During Master Mode Operation
•
If possible, all start-up serial writes should be performed
with VD and HD disabled. This prevents unknown
behavior caused by partial updating of registers before all
information is loaded.
The hardware SYNC input can be used anytime during
operation to synchronize the AD9992 counters with external
timing, as shown in Figure 75. The operation of the digital
outputs can be suspended during the SYNC operation by setting the SYNCSUSPEND register (Address 0x13, Bit 2) to 1. If
SYNCSUSPEND = 1, the polarities of the outputs are held at the
same state as OUTCONTROL = low, as shown in Table 27.
SYNC
VD
SUSPEND
HD
NOTES
1. THE SYNC RISING EDGE RESETS VD/HD AND COUNTERS TO 0.
2. SYNC POLARITY IS PROGRAMMABLE USING SYNCPOL REGISTER (ADDR 0x13).
3. DURING SYNC LOW, ALL INTERNAL COUNTERS ARE RESET AND VD/HD CAN BE SUSPENDED USING THE SYNCSUSPEND REGISTER (ADDR 0x13).
4. IF SYNCSUSPEND = 1, VERTICAL CLOCKS, H1 TO H4, AND RG ARE HELD AT THE SAME POLARITY SPECIFIED BY OUTCONTROL = LOW.
5. IF SYNCSUSPEND = 0, ALL CLOCK OUTPUTS CONTINUE TO OPERATE NORMALLY UNTIL THE SYNC RESET EDGE.
Figure 75. SYNC Timing to Synchronize the AD9992 with External Timing
Rev. C | Page 65 of 92
05891-070
H1 TO H4, RG,
XV1 TO XV24,
VSG, SUBCK
AD9992
VD
tVDHD
HD
tHDCLI
CLI
tHDCLO
tCLIDLY
CLO
SHPLOC
INTERNAL
SHDLOC
INTERNAL
HD
INTERNAL
H-COUNTER
RESET
H-COUNTER
(PIXEL COUNTER)
X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 0
1
2
NOTES:
1. EXTERNAL HD FALLING EDGE IS LATCHED BY CLI RISING EDGE, THEN LATCHED AGAIN BY SHPLOC (INTERNAL SAMPLING EDGE).
2. INTERNAL H-COUNTER IS ALWAYS RESET 35.5 CLOCK CYCLES AFTER THE INTERNAL HD FALLING EDGE, AT SHDLOC (INTERNAL SAMPLING EDGE).
3. DEPENDING ON THE VALUE OF SHPLOC, H-COUNTER RESET CAN OCCUR 36 OR 37 CLI CLOCK EDGES AFTER THE EXTERNAL HD FALLING EDGE.
4. SHPLOC = 32, SHDLOC = 0 IS SHOWN IN ABOVE EXAMPLE. IN THIS CASE, THE H-COUNTER RESET OCCURS 36 CLI RISING EDGES AFTER HD FALLING EDGE.
5. HD FALLING EDGE SHOULD OCCUR COINCIDENT WITH VD FALLING EDGE (WITHIN SAME CLI CYCLE) OR AFTER VD FALLING EDGE. HD FALLING
EDGE SHOULD NOT OCCUR WITHIN 1 CYCLE IMMEDIATELY BEFORE VD FALLING EDGE.
05891-094
35.5 CYCLES
Figure 76. External VD/HD and Internal H-Counter Synchronization, Slave Mode
PIXEL NO.
0
60
100 103
112
HD
1
2
H1
3
1HBLKTOG1
2HBLKTOG2
3CLPOB_TOG1
4CLPOB_TOG2
MASTER MODE
SLAVE MODE
60
100
103
112
(60 – 36) = 24
(100 – 36) = 64
(103 – 36) = 67
(112 – 36) = 76
4
05891-072
CLPOB
Figure 77. Example of Slave Mode Register Setting to Obtain Desired Toggle Positions
Vertical Toggle Position Placement Near Counter Reset
An additional consideration during the reset of the internal
counters is the vertical toggle position placement. Before
the internal counters are reset, there is a region of 36 pixels
during which no toggle positions should be programmed.
As shown in Figure 78 for master mode, the last 36 pixels before
the HD falling edge must not be used for toggle position placement
of the V, VSG, SUBCK, HBLK, PBLK, or CLPOB pulses.
Figure 79 shows the same example for slave mode. The same
restriction applies: the last 36 pixels before the counters are
reset cannot be used. However, in slave mode, the counter reset
is delayed with respect to VD/HD placement, so the inhibited
area is different than it is in master mode.
It is recommended that Pixel Location 0 not be used for any of
the toggle positions for the VSG and SUBCK pulses.
Rev. C | Page 66 of 92
AD9992
VD
H-COUNTER
RESET
HD
NO TOGGLE POSITIONS ALLOWED IN THIS AREA
X
X
X
X
N – 35 N – 34 N – 33 N – 32
N – 13 N – 12 N – 11 N – 10 N – 9
N–8
N–7
N–6
N–5
N–4
N–3
N–2
N–1
N
0
1
2
3
4
1
2
05891-073
H-COUNTER
(PIXEL COUNTER)
NOTES
1. TOGGLE POSITIONS CANNOT BE PROGRAMMED WITHIN 36 PIXELS OF PIXEL 0 LOCATION.
Figure 78. Toggle Position Inhibited Area—Master Mode
VD
H-COUNTER
RESET
HD
H-COUNTER
(PIXEL COUNTER)
X
X
X
X
X
X
N – 35 N – 34 N – 33 N – 32
N – 13 N – 12 N – 11 N – 10
N–9
N–8
N–7
N–6
N–5
N–4
N–3
N–2
N–1
N
0
NOTES
1. TOGGLE POSITIONS CANNOT BE PROGRAMMED WITHIN 36 PIXELS OF PIXEL 0 LOCATION.
05891-074
NO TOGGLE POSITIONS ALLOWED IN THIS AREA
Figure 79. Toggle Position Inhibited Area—Slave Mode
STANDBY MODE OPERATION
The AD9992 contains three standby modes to optimize the overall
power dissipation in a particular application. Bits [1:0] of Address
0x00 control the power-down state of the device:
• STANDBY [1:0] = 0 = normal operation (full power)
• STANDBY [1:0] = 1 = Standby1 mode
• STANDBY [1:0] = 2 = Standby2 mode
• STANDBY [1:0] = 3 = Standby3 mode (lowest power)
Table 27 summarizes the operation of each power-down mode.
The OUTCONTROL register takes priority over the Standby1
and Standby2 modes in determining the digital output states,
but Standby3 mode takes priority over OUTCONTROL.
Standby3 has the lowest power consumption and even shuts
down the crystal oscillator circuit between CLI and CLO.
Therefore, if CLI and CLO are being used with a crystal to
generate the master clock, this circuit is powered down and
there is no clock signal. When returning from Standby3 mode
to normal operation, the timing core must be reset at least
500 μs after the STANDBY register is written to. This allows
sufficient time for the crystal circuit to settle.
The vertical outputs can also be programmed to hold a specific
value during the Standby3 mode by using Address 0x26. This
register is useful during power-up if different polarities are
required by the V-driver and CCD to prevent damage when VH
and VL areas are applied. The polarities for Standby1 mode and
Standby2 mode are also programmable, using Address 0x25.
OUTCONTROL = low also uses the same polarities programmed
for Standby1 and Standby2 modes in Address 0x25. The GPO
polarities are programmable using Address 0x27.
Note that the GPO outputs are high-Z by default at power-up
until Address 0x78 is used to select them as outputs.
CLI FREQUENCY CHANGE
If the input clock CLI is interrupted or changed to a different
frequency, the timing core must be reset for proper operation.
After the CLI clock settles to the new frequency, or the previous
frequency is resumed, write 0 and then 1 to the TGCORE_RSTB
register (Address 0x14). This guarantees that the timing core
operates properly.
Rev. C | Page 67 of 92
AD9992
Table 27. Standby Mode Operation (Standby Polarities for XV, XSUBCK, GPO Outputs Are Programmable)
I/O Block
AFE
Timing Core
CLO Oscillator
CLO
H1
H2
H3
H4
H5
H6
H7
H8
HL
RG
VD
HD
DCLK
DOUT
XV1 to XV24
XSUBCK
GPO1 to GPO8
Standby3 (Default) 1, 2
Off
Off
Off
Low
High-Z
High-Z
High-Z
High-Z
High-Z
High-Z
High-Z
High-Z
High-Z
High-Z
Low
Low
Low
Low
Low
Low
Low
OUTCONTROL = Low2
No change
No change
No change
No change
Low
High
Low
High
Low
High
Low
High
Low
Low
VDHDPOL value
VDHDPOL value
Running
Low
Low
Low
Low
1
Standby2 3, 4
Off
Off
Off
Low
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
Low (4.3 mA)
VDHDPOL value
VDHDPOL value
Low
Low
Low
Low
Low
Standby13, 4
Only REFT, REFB on
On
On
Running
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
High (4.3 mA)
Low (4.3 mA)
Low (4.3 mA)
Running
Running
Running
Low
Low
Low
Low
To exit Standby3, write 00 to STANDBY (Address 0x00, Bits [1:0]), and then reset the timing core after 500 μs to guarantee proper settling of the oscillator and external crystal.
Standby3 mode takes priority over OUTCONTROL for determining the output polarities.
3
These polarities assume OUTCONTROL = high because OUTCONTROL = low takes priority over Standby1 and Standby2.
4
Standby1 and Standby2 set H and RG drive strength to minimum value (4.3 mA).
2
Rev. C | Page 68 of 92
AD9992
CIRCUIT LAYOUT INFORMATION
The PCB layout is critical in achieving good image quality from
the AD9992. All of the supply pins, particularly the AVDD,
TCVDD, RGVDD, and HVDD supplies, must be decoupled to
ground with good quality high frequency chip capacitors. The
decoupling capacitors should be located as close as possible to
the supply pins and should have a very low impedance path to
a continuous ground plane. If possible, there should be a 4.7 μF
or larger value bypass capacitor for each main supply—AVDD,
HVDD, and DRVDD—although this is not necessary for each
individual pin. In most applications, the supply for RGVDD and
HVDD is shared, which can be done as long as the individual
supply pins are separately bypassed with 0.1 μF capacitors. A
separate 3 V supply can also be used for DRVDD, but this
supply pin should still be decoupled to the same ground plane
as the rest of the chip. A separate ground for DRVSS is not
recommended.
The analog bypass pins (REFT and REFB) should be carefully
decoupled to ground as close as possible to their respective pins.
The analog input (CCDIN) capacitor should be located close to
the CCDIN pin.
The H1 to H8, HL, and RG traces should be designed to have
low inductance to minimize distortion of the signals. The complementary signals, H1/H3/H5/H7 and H2/H4/H6/H8, should
be routed as close together and as symmetrically as possible to
minimize mutual inductance. Heavier PCB traces are recommended because of the large transient current demand on H1 to
H8 by the CCD. If possible, physically locating the AD9992
closer to the CCD reduces the inductance on these lines. As
usual, the routing path should be as direct as possible from the
AD9992 to the CCD.
TYPICAL 3 V SYSTEM
The AD9992 typical circuit connections for a 3 V system are
shown in Figure 80. This application uses an external 3.3 V
supply, which is connected to the AD9992 LDO input. The LDO is
configured to output 1.8 V for the AD9992 core supply by
connecting the LDO1P8EN pin to 3.3 V and the LDO3P2EN
pin to ground. The LDOOUT and SENSE pins are shorted
together and used to supply 1.8 V to the AVDD, TCVDD, and
DVDD pins.
TYPICAL 1.8 V SYSTEM
The internal LDO can be disabled by tying the LDO pins to
ground (LDOIN, LDO1P8EN, LDO3P2EN, LDOOUT, and
SENSE). In this case, an external 1.8 V regulator is required to
supply 1.8 V to the AVDD, TCVDD, and DVDD pins.
All of the AD9992 remaining supplies can be directly supplied
with 1.8 V. The internal charge pump (CP) can be used to
generate 3.3 V for the H and RG supplies.
The AD9992 typical circuit connections for a 1.8 V system are
shown in Figure 81.
EXTERNAL CRYSTAL APPLICATION
The AD9992 contains an on-chip oscillator for driving an
external crystal. Figure 82 shows an example application using
a typical 27 MHz crystal. For the exact values of the external
resistors and capacitors, it is best to consult the crystal
manufacturer’s data sheet.
Note that a 2× crystal is not recommended for use with the CLO
oscillator circuit. The crystal frequency should not exceed
40 MHz.
It is recommended that all H1 to H8 outputs on the AD9992 be
used together for maximum flexibility in drive strength settings.
A typical CCD with H1 and H2 inputs only should have the
AD9992 H1, H3, H5, and H7 outputs connected together to drive
the CCD’s H1, and H2, H4, H6, and H8 outputs connected
together to drive the CCD’s H2. Similarly, a CCD with H1, H2,
H3, and H4 inputs should have the following:
•
•
•
•
H1 and H3 connected to the CCD’s H1.
H2 and H4 connected to the CCD’s H2.
H5 and H7 connected to the CCD’s H3.
H6 and H8 connected to the CCD’s H4.
Rev. C | Page 69 of 92
AD9992
MASTER CLOCK INPUT
(3V LOGIC)
+3V CLI SUPPLY
0.1µF
0.1µF
ANALOG OUTPUT FROM CCD
+1.8V LDO OUT
0.1µF
0.1µF
+3V H, RG SUPPLY
0.1µF
0.1µF
0.1µF
3
SCK
SDATA
SL
REFB
REFT
AVSS
AVSS
CCDIN
AVDD
CLI
CLO
CLIVDD
TCVDD
TCVSS
RG
RGVDD
RGVSS
HL
NC
GENERAL-PURPOSE
OUTPUTS
8
AD9992BBCZ
NOT DRAWN TO SCALE
A10
B10
A11
B11
C9
D9
C10
D10
C11
D11
E9
F9
E10
E11
F11
F10
G9
G11
G10
H9
H11
H10
J11
J10
K11
L11
K10
H8
H7
HVDD2
HVSS2
H6
H5
H4
H3
HVDD1
HVSS1
H2
H1
LDOIN
LDOOUT
SENSE
LDO1P8EN
LDOVSS
LDO3P2EN
NC
CPCLI
CP1P8
CPVSS
CPFCB
CPFCT
CP3P3
DRVDD
DRVSS
H7, H8 TO CCD
+3V H, RG SUPPLY
0.1µF
H5, H6 TO CCD
H3, H4 TO CCD
+3V, H, RG SUPPLY
0.1µF
H1, H2 TO CCD
0.1µF
+3V LDOIN
+1.8V LDOOUT TO
AVDD, TCVDD, DVDD
0.1µF
+3V DIGITAL SUPPLY
0.1µF
XV17
XV18
XV19
XV20
XV21
XV22
XV23
XV24
DVDD
DVSS
NC
NC
NC
NC
(LSB) D0
D1
D2
D3
D4
D5
D6
D7
D8
D9
D10
(MSB) D11
DCLK
XSUBCK OUTPUT
(TO V-DRIVER)
A1
B2
C2
B1
B4
C1
D2
C3
E7
D3
E2
D1
E6
E5
E3
E1
F2
F3
F7
G3
F5
F6
G2
F1
G1
G5
H2
H1
G6
G7
J2
J1
L1
L2
L3
K1
K2
K3
J3
H3
L4
K4
J4
L5
K5
J5
L6
K6
J6
L7
K7
J7
L8
K8
J8
L9
K9
J9
L10
EXTERNAL RESET IN
EXTERNAL SYNC IN
VERTICAL SYNC IN/OUT
HORIZONTAL SYNC IN/OUT
+3V DIGITAL I/O
XV SUPPLY
0.1µF
GPO8
GPO7
GPO6
GPO5
GPO4
GPO3
GPO2
GPO1
RSTB
SYNC
VD
HD
IOVDD
IOVSS
XVVDD
XSUBCK
XV1
XV2
XV3
XV4
XV5
XV6
XV7
XV8
XV9
XV10
XV11
XV12
XV13
XV14
XV15
XV16
RG TO CCD
HL TO CCD
B3
A2
C4
A3
A4
A5
B5
A6
B6
C5
C6
C7
A7
B7
C8
A8
B8
B9
A9
SERIAL INTERFACE
(FROM ASIC/DSP)
VERTICAL OUTPUTS
(TO V-DRIVER)
DCLK OUTPUT
12
+1.8V LDOOUT
0.1µF
DATA OUTPUTS
05891-075
24
Figure 80. Typical 3 V Circuit Configuration
Rev. C | Page 70 of 92
AD9992
MASTER CLOCK INPUT
(1.8V LOGIC)
+1.8V SUPPLY
0.1µF
0.1µF
ANALOG OUTPUT FROM CCD
+1.8V SUPPLY
0.1µF
+3V CP OUTPUT
0.1µF
0.1µF
0.1µF
SERIAL INTERFACE
(FROM ASIC/DSP)
3
8
0.1µF
A10
B10
A11
B11
C9
D9
C10
D10
C11
D11
E9
F9
E10
E11
F11
F10
G9
G11
G10
H9
H11
H10
J11
J10
K11
L11
K10
AD9992BBCZ
NOT DRAWN TO SCALE
H8
H7
HVDD2
HVSS2
H6
H5
H4
H3
HVDD1
HVSS1
H2
H1
LDOIN
LDOOUT
SENSE
LDO1P8EN
LDOVSS
LDO3P2EN
NC
CPCLI
CP1P8
CPVSS
CPFCB
CPFCT
CP3P3
DRVDD
DRVSS
0.1µF
H7, H8 TO CCD
+3V CP OUTPUT
H5, H6 TO CCD
H3, H4 TO CCD
+3V CP OUTPUT
0.1µF
H1, H2 TO CCD
+1.8V CP INPUT
0.1µF
0.1µF
+3V CP OUTPUT
3.3µF
+1.8V DIGITAL SUPPLY
0.1µF
XV17
XV18
XV19
XV20
XV21
XV22
XV23
XV24
DVDD
DVSS
NC
NC
NC
NC
(LSB) D0
D1
D2
D3
D4
D5
D6
D7
D8
D9
D10
(MSB) D11
DCLK
XSUBCK OUTPUT
(TO V-DRIVER)
A1
B2
C2
B1
B4
C1
D2
C3
E7
D3
E2
D1
E6
E5
E3
E1
F2
F3
F7
G3
F5
F6
G2
F1
G1
G5
H2
H1
G6
G7
J2
J1
L1
L2
L3
K1
K2
K3
J3
H3
L4
K4
J4
L5
K5
J5
L6
K6
J6
L7
K7
J7
L8
K8
J8
L9
K9
J9
L10
EXTERNAL RESET IN
EXTERNAL SYNC IN
VERTICAL SYNC IN/OUT
HORIZONTAL SYNC IN/OUT
+1.8V DIGITAL I/O
XV SUPPLY
GPO8
GPO7
GPO6
GPO5
GPO4
GPO3
GPO2
GPO1
RSTB
SYNC
VD
HD
IOVDD
IOVSS
XVVDD
XSUBCK
XV1
XV2
XV3
XV4
XV5
XV6
XV7
XV8
XV9
XV10
XV11
XV12
XV13
XV14
XV15
XV16
B3
A2
C4
A3
A4
A5
B5
A6
B6
C5
C6
C7
A7
B7
C8
A8
B8
B9
A9
SCK
SDATA
SL
REFB
REFT
AVSS
AVSS
CCDIN
AVDD
CLI
CLO
CLIVDD
TCVDD
TCVSS
RG
RGVDD
RGVSS
HL
NC
GENERAL-PURPOSE
OUTPUTS
RG TO CCD
HL TO CCD
VERTICAL OUTPUTS
(TOV-DRIVER)
DCLK OUTPUT
12
+1.8V SUPPLY
DATA OUTPUTS
05891-076
0.1µF
Figure 81. Typical 1.8 V Circuit Configuration Using Charge Pump for HVDD and RGVDD
~7MΩ
AD9992
K7
K6
CLI
375Ω
CLO
USER DEFINED
5pF ~ 20pF
24MHz TO 40MHz
XTAL
5pF ~ 20pF
05891-092
24
Figure 82. Crystal Application Using CLI/CLO (Consult Crystal Data Sheet for Component Values)
Rev. C | Page 71 of 92
AD9992
Figure 84 shows a more efficient way to write to the registers,
using the AD9992 address autoincrement capability. Using this
method, the lowest desired address is written first, followed by
multiple 28-bit data-words. Each new 28-bit data-word is
automatically written to the next highest register address. By
eliminating the need to write each 12-bit address, faster register
loading is achieved. Continuous write operations can be used
starting with any register location.
SERIAL INTERFACE TIMING
The internal registers of the AD9992 are accessed through a
3-wire serial interface. Each register consists of a 12-bit address
and a 28-bit data-word. Both the 12-bit address and 28-bit dataword are written starting with the LSB. To write to each register,
a 40-bit operation is required, as shown in Figure 83. Although
many registers are fewer than 28 bits wide, all 28 bits must be
written for each register. For example, if the register is only 20 bits
wide, the upper eight bits are don’t cares and must be filled with
0s during the serial write operation. If fewer than 28 data bits are
written, the register is not updated with new data.
12-BIT ADDRESS
A0
SDATA
A1
A2
A3
A4
A5
A6
A7
tDS
SCK
1
2
3
4
5
A8
28-BIT DATA
A9
A10
A11
D0
D1
D2
D3
D25
D26
D27
tDH
6
7
8
9
10
11
12
13
14
15
16
38
tLS
39
40
tLH
SL
05891-078
NOTES
1. SDATA BITS ARE LATCHED ON SCK RISING EDGES. SCK CAN IDLE HIGH OR LOW BETWEEN WRITE OPERATIONS.
2. ALL 40 BITS MUST BE WRITTEN: 12 BITS FOR ADDRESS AND 28 BITS FOR DATA.
3. IF THE REGISTER LENGTH IS <28 BITS, 0s MUST BE USED TO COMPLETE THE 28-BIT DATA LENGTH.
4. NEW DATA VALUES ARE UPDATED IN THE SPECIFIED REGISTER LOCATION AT DIFFERENT TIMES, DEPENDING ON THE
PARTICULAR REGISTER WRITTEN TO. SEE THE UPDATING OF NEW REGISTER VALUES SECTION FOR MORE INFORMATION.
Figure 83. Serial Write Operation
DATA FOR STARTING
REGISTER ADDRESS
SDATA
SCK
A0
1
A1
2
A2
3
A3
4
A10
11
A11
12
D0
13
D1
14
D26
39
DATA FOR NEXT
REGISTER ADDRESS
D27
40
D0
D1
41
42
D26
D27
67
68
D0
69
D1
70
D2
71
NOTES
1. MULTIPLE SEQUENTIAL REGISTERS CAN BE LOADED CONTINUOUSLY.
2. THE FIRST (LOWEST) REGISTER ADDRESS IS WRITTEN, FOLLOWED BY MULTIPLE 28-BIT DATA-WORDS.
3. THE ADDRESS AUTOMATICALLY INCREMENTS WITH EACH 28-BIT DATA-WORD (ALL 28 BITS MUST BE WRITTEN).
4. SL IS HELD LOW UNTIL THE LAST DESIRED REGISTER HAS BEEN LOADED.
Figure 84. Continuous Serial Write Operation
Rev. C | Page 72 of 92
05891-079
SL
AD9992
LAYOUT OF INTERNAL REGISTERS
The AD9992 address space is divided into two register areas, as
illustrated in Figure 85. In the first address space, Address 0x00
to Address 0xFF contain the registers for the AFE, miscellaneous,
VD/HD, I/O and CP, timing core, shutter and GPO, mode, and
update control functions. The second address space, beginning
at Address 0x800, consists of the V-pattern groups, V-sequences,
and field registers. This is a configurable set of registers; the user
can decide how many V-pattern groups, V-sequences, and fields
are used in a particular design. Therefore, the addresses for
these registers vary, depending on the number of V-patterns
and V-sequences chosen.
Address 0x28 specifies the total number of V-pattern groups
and V-sequences used. The starting address for the V-pattern
groups is always 0x800. The starting address for the V-sequences
is based on the number of V-pattern groups used, with each
V-pattern group occupying 48 register addresses. The starting
address for the field registers depends on both the number of
V-pattern groups and the number of V-sequences. Each Vsequence occupies 40 register addresses, and each field occupies
16 register addresses.
ADDR 0x00
FIXED REGISTER AREA
The starting address for the V-sequences is equal to 0x800 plus
the number of V-pattern groups times 48. The starting address
for the fields is equal to the starting address of the V-sequences
plus the number of V-sequences times 40. The V-pattern,
V-sequence and field registers must always occupy a continuous
block of addresses.
Figure 86 shows an example in which three V-pattern groups,
four V-sequences, and two fields are used. The starting address
for the V-pattern groups is always 0x800. Because VPATNUM = 3,
the V-pattern groups occupies 144 address locations. The start of
the V-sequence registers is 0x890 (that is, 0x800 + 144). With
VSEQNUM = 4, the V-sequences occupy 160 address locations.
Therefore, the field registers begin at 0x930 (that is, 0x890 + 160).
The AD9992 address space contains many unused addresses.
Undefined addresses between Address 0x00 and Address 0xFF
should not be written to; otherwise, the AD9992 may operate
incorrectly. Continuous register writes should be performed
carefully so that undefined registers are not written to.
VPAT START 0x800
CONFIGURABLE REGISTER AREA
AFE REGISTERS
MISCELLANEOUS REGISTERS
V-PATTERN GROUPS
VD/HD REGISTERS
I/O AND CP REGISTERS
VSEQ START
MODE REGISTERS
TIMING CORE REGISTERS
V-SEQUENCES
TEST REGISTERS
TEST REGISTERS
FIELD START
SHUTTER AND GPO REGISTERS
05891-080
FIELDS
UPDATE CONTROL REGISTERS
INVALID DO NOT ACCESS
MAX 0xFFF
ADDR 0xFF
Figure 85. Layout of AD9992 Registers
ADDR 0x800
3 V-PATTERN GROUPS
(48 × 3 = 144 REGISTERS)
ADDR 0x890
4 V-SEQUENCES
(40 × 4 = 160 REGISTERS)
ADDR 0x930
2 FIELDS
(16 × 2 = 32 REGISTERS)
UNUSED MEMORY
MAX 0xFFF
Figure 86. Example Register Configuration
Rev. C | Page 73 of 92
05891-081
ADDR 0x950
AD9992
These registers control the SUBCK signal so that the SUBCK
output is not updated until the SG line occurs.
SCP Updated—At the next SCP where they are used,
the V-pattern group and V-sequence registers are updated.
For example, in Figure 87 this field has selected Region 1 to
use VSEQ3 for the vertical outputs. This means that a write
to any of the VSEQ3 registers, or any of the V-pattern
group registers, which are referenced by VSEQ3, updates at
SCP1. If multiple writes are made to the same register, the last
one done before SCP1 is the one that is updated. Likewise,
register writes to any VSEQ5 registers are updated at SCP2;
register writes to any VSEQ8 registers are updated at SCP3.
UPDATING NEW REGISTER VALUES
The AD9992 internal registers are updated at different times,
depending on the particular register. Table 28 summarizes the four
register update types: SCK, VD, SG-Line, and SCP. Tables in the
Complete Register Listing section also contain an Update Type
column that identifies when each register is updated.
•
•
•
•
SCK Updated—As soon as the 28th data bit (D27) is clocked
in, some registers are immediately updated. These registers
are used for functions that do not require gating with the next
VD boundary, such as power-up and reset functions.
VD Updated—More registers are updated at the next VD
falling edge. By updating these values at the next VD edge,
the current field is not corrupted and the new register
values are applied to the next field. The VD update can be
further delayed past the VD falling edge by using the
UPDATE register (Address 0x17). This delays the VDupdated register updates to any HD line in the field. Note
that the field registers are not affected by the UPDATE
register.
SG-Line Updated—A few of the shutter registers are updated
at the HD falling edge at the start of the SG active line.
Caution
It is recommended that the registers in the configurable address
area not be written within 36 pixels of any HD falling edge
where a sequence change position (SCP) occurs. See Figure 78
and Figure 79 for examples of what this inhibit area looks like in
master and slave modes. This restriction applies to the V-pattern,
V-sequence, and field registers. As shown in Figure 87, writing
to these registers before the VD falling edge typically avoids
loading these registers during SCP locations.
Table 28. Register Update Locations
Update Type
SCK
VD
Description
When the 28th data bit (D27) is clocked in, the register is immediately updated.
Register is updated at the VD falling edge. VD-updated registers can be delayed further by using the UPDATE register at
Address 0x17. FIELD registers are not affected by the UPDATE register.
Register is updated at the HD falling edge at the start of the SG-active line.
Register is updated at the next SCP when the register is used.
SG-Line
SCP
SCK
UPDATED
VD
UPDATED
SG
UPDATED
SCP
UPDATED
SERIAL
WRITE
VD
HD
SGLINE
VSG
USE VSEQ2
USE VSEQ3
USE VSEQ5
REGION 0
REGION 1
REGION 2
SCP0
SCP1
SCP2
USE VSEQ8
REGION 3
SCP3
Figure 87. Register Update Locations (See Table 28 for Definitions)
Rev. C | Page 74 of 92
SCP0
05891-082
XV1 TO XV24
AD9992
COMPLETE REGISTER LISTING
When an address contains fewer than 28 data bits, all remaining bits must be written as 0s.
Table 29. AFE Registers
Data
Bits
[1:0]
Default
Value
3
[2]
1
CLPENABLE
[3]
0
CLPSPEED
[4]
0
FASTUPDATE
[5]
0
PBLK_LVL
[6]
0
DCBYP
[0]
0
[1]
0
DOUTLATCH
0x02
0x03
0x04
[2]
[3]
[0]
[23:0]
[2:0]
0
1
0
FFFFFF
0
SCK
SCK
VD
GRAY_EN
TEST
TEST
TEST
CDSGAIN
0x05
0x06
0x0D
[9:0]
[9:0]
[0]
F
1EC
0
VD
VD
VD
VGAGAIN
CLAMPLEVEL
CLIDIVIDE
Address
0x00
0x01
Update
Type
SCK
SCK
Mnemonic
STANDBY
DOUTDISABLE
Description
Standby modes:
0: Normal operation.
1: Standby1 mode.
2: Standby2 mode.
3: Standby3 mode.
0: Disable OB clamp.
1: Enable OB clamp.
0: Select normal OB clamp settling.
1: Select fast OB clamp settling.
0: Ignore CDS gain.
1: Very fast clamping when CDS gain is updated.
0: Blank data outputs to 0 during PBLK.
1: Blank data outputs to programmed clamp level during PBLK.
0: Enable input dc restore circuit during PBLK.
1: Disable input dc restore circuit during PBLK.
0: Data outputs are driven.
1: Data outputs are three-stated.
0: Latch data outputs using the rising edge of DOUTPHASEP
(DOUTPHASEP register setting).
1: Output latch is transparent.
1: Enable gray encoding of the digital data outputs.
Set to 0.
Do not access, or set to 0.
Do not access, or set to 0xFFFFFF.
CDS gain setting:
0: −3 dB.
4: 0 dB.
6: +3 dB.
7: +6 dB.
All other values are invalid.
VGA gain, 6 dB to 42 dB (0.035 dB per step).
Optical black clamp level, 0 to 1023 LSB (1 LSB per step).
0: No division of CLI.
1: Divide CLI input frequency by 2.
Table 30. Miscellaneous Registers
Address
0x10
Data
Bits
[0]
Default
Value
0
Update
Type
SCK
Mnemonic
SW_RST
0x11
[0]
0
VD
OUTCONTROL
0x12
[0]
[4:1]
[0]
0
0
1
SCK
RSTB_EN
TEST
SYNCENABLE
[1]
0
0x13
SCK
SYNCPOL
Rev. C | Page 75 of 92
Description
Software reset. Bit self-clears to 0 when a reset occurs.
1: Reset Address 0x00 to Address 0xFF to default values.
0: Make all outputs dc inactive.
1: Enable outputs at next VD edge.
1: Configure SYNC pin as RSTB input signal.
Test mode only. Must be set to 0.
1: External synchronization enable (configures Pin D3 as an
input).
SYNC active polarity.
AD9992
Data
Bits
[2]
Default
Value
0
[3]
[4]
[5]
[6]
[7]
[12:8]
[13]
0
0
1
1
0
0
0
0x14
[14]
[0]
0
0
SCK
SWSYNC
TGCORE_RSTB
0x15
[0]
0
SCK
OSC_RSTB
0x16
0x17
[27:0]
[12:0]
0
0
SCK
SCK
TEST
UPDATE
[13]
0
PREVENTUP
[14]
0
SYNC_RST_SHUTEN
[15]
[16]
[17]
0
0
0
REG_RST_SHUT
GPO_RST_SYNC
GPO_LINE_COUNT_OFFSET
[27:0]
[27:0]
[27:0]
[27:0]
[23:0]
[23:0]
0
0
0
A
FF0000
0
[24]
0
[0]
[1]
1
1
Address
0x18
0x19
0x1A
0x1B
0x1C
0x1D
0x1F
Update
Type
Mnemonic
SYNCSUSPEND
Description
Suspend clocks during SYNC active pulse:
0: Do not suspend.
1: Suspend.
1: Enable enhanced sync/shutter operations.
1: Mask HD during SYNCSUSPEND.
1: Mask VD during SYNCSUSPEND.
1: Mask XV outputs during SYNCSUSPEND.
1: Enable use of shadow registers.
Test mode only. Must be set to 0.
1: Writes to shadow bits affect shadow registers, not
primary.
1: Initiate software SYNC event (self-clears to 0 after SYNC).
Timing core reset bar. 0: reset TG core;
1: resume operation.
CLO oscillator reset bar:
0: Oscillator in power-down state.
1: Resume oscillator operation.
Test mode only. Must be set to 0.
Serial update line.
Sets the line (HD) within the field to update the VD-updated
registers.
Prevents the update of the VD-updated registers:
0: Normal update.
1: Prevent update of VD-updated registers.
1: Enable reset of the shutter control after SYNC operation
occurs.
1: Forces shutter control to reset.
1: Reset shutter and GPO control at SYNC operation.
0: First line in a field is considered Line 1 for GPOs.
1: First line in a field is considered Line 0 for GPOs.
Test mode only. Must be set to 0.
Test mode only. Must be set to 0.
Test mode only. Must be set to 0.
Test mode only. Must be set to 0xA.
Each bit selects XV pulses for use as VSG pulses.
VSG masking. Overrides settings in field registers when
enabled.
0: Disable VSGMAK_CTL bits. VSG masking is controlled by
field registers.
1: Enable VSGMASK_CTL bits to control VSG masking.
1: Enable 14-bit H-counter.
1: Disable clamp operation if PBLK is active at the same time
as CLPOB.
ENH_SYNC_EN
SYNC_MASK_HD
SYNC_MASK_VD
SYNC_MASK_V
SHADOW_EN
TEST
UPDATE_SHADOW
SCK
SCK
SCK
SCK
SCK
SCK
TEST
TEST
TEST
TEST
VSGSELECT
VSGMASK_CTL
VSGMASK_CTL_EN
SCK
HCNT14_EN
PBLK_MASK_EN
Table 31. VD/HD Registers
Address
0x20
Data
Bits
[0]
Default
Value
0
Update
Type
SCK
Mnemonic
MASTER
0x21
[0]
0
VD
VDHDPOL
0x22
[12:0]
[25:13]
0
0
VD
HDRISE
VDRISE
Description
VD/HD master or slave mode:
0: Slave mode.
1: Master mode.
VD/HD active polarity:
0: Low.
1: High.
Rising edge location for HD. Minimum value is 36 pixels.
Rising edge location for VD.
Rev. C | Page 76 of 92
AD9992
Table 32. I/O and Charge Pump Registers
Data
Bits
[0]
Default
Value
0
[1]
0
XV_NVR
[2]
0
IO_NVR
[3]
0
DATA_NVR
[4]
[5]
[6]
[9:7]
0
0
0
1
TEST
TEST
LDO_32_EN
HCLKMODE
0x24
[0]
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
0
1
0
0
0
1
0
1
1
1
1
0
0
0
1
SCK
SEL_VCO
SEL_DIV
SEL_CLI
O31V
O32V
O33V
O34V
TEST
TEST
TEST
TEST
TEST
TEST
TEST
CP_PDN
0x25
[24:0]
0
SCK
VT_STBY12
0x26
[24:0]
0
SCK
VT_STBY3
0x27
[7:0]
0
SCK
GP_STDBY12
Address
0x23
[15:8]
Update
Type
SCK
Mnemonic
OSC_NVR
GP_STDBY3
Description
Oscillator normal voltage range. Set to match CLIVDD supply voltage.
0: 1.8 V.
1: 3.3 V.
XV output normal voltage range. Set to match XVVDD supply voltage.
0: 1.8 V.
1: 3.3 V.
I/O normal voltage range. Set the match IOVDD supply voltage.
0: 1.8 V.
1: 3.3 V.
Data pin normal voltage range. Set to match DRVDD supply voltage.
0: 1.8 V I/O.
1: 3.3 V I/O.
Test mode only. Set to 0.
Test mode only. Set to 0.
1: Internal regulator enable for 3.2 V output.
Selects HCLK output configuration. Should be written to desired value.
001: Mode 1.
010: Mode 2.
100: Mode 3.
All other values are invalid.
1: Internal CP clock select VCO.
1: Internal CP clock select divided-down version of CLI (default).
1: Internal CP clock select CLI.
1: CP output voltage is 3.1 V.
1: CP output voltage is 3.2 V.
1: CP output voltage is 3.3 V.
1: CP output voltage is 3.4 V.
Test mode only. Use default values only.
Test mode only. Use default values only.
Test mode only. Use default values only.
Test mode only. Use default values only.
Test mode only. Use default values only.
Test mode only. Use default values only.
Test mode only. Use default values only.
Charge pump power-down.
1: Power-down.
0: CP is running.
[23:0] Standby1 and Standby2 polarity for XV [23:0].
[24] Standby1 and Standby2 polarity for XSUBCK.
Settings also apply when OUTCONTROL = low.
[23:0] Standby3 polarity for XV [23:0].
[24] Standby3 polarity for XSUBCK.
Standby1 and Standby2 polarity for GPO [7:0].
Settings also apply when OUTCONTROL = low.
Standby3 polarity for GPO [7:0].
Rev. C | Page 77 of 92
AD9992
Table 33. Memory Configuration and MODE Registers
Address
0x28
0x2A
0x2B
0x2C
Data Bits
[4:0]
[9:5]
[2:0]
[4:0]
[9:5]
[14:10]
[19:15]
[24:20]
[4:0]
[9:5]
Default Value
0
0
0
0
0
0
0
0
0
0
Update Type
SCK
SCK
SCK
SCK
Mnemonic
VPATNUM
SEQNUM
MODE
FIELD0
FIELD1
FIELD2
FIELD3
FIELD4
FIELD5
FIELD6
Description
Total number of V-pattern groups.
Total number of V-sequences.
Total number of fields in MODE.
Selected first field in MODE.
Selected second field in MODE.
Selected third field in MODE.
Selected fourth field in MODE.
Selected fifth field in MODE.
Selected sixth field in MODE.
Selected seventh field in MODE.
Table 34. Timing Core Registers
Address
0x30
Data Bits
[5:0]
[13:8]
[16]
Default
Value
0
20
1
Update
Type
SCK
0x31
[5:0]
[13:8]
[16]
0
20
1
SCK
H2POSLOC
H2NEGLOC
H2POL
0x32
[5:0]
[13:8]
[16]
0
20
1
SCK
HLPOSLOC
HLNEGLOC
HLPOL
0x33
[5:0]
[13:8]
[16]
0
10
1
SCK
RGPOSLOC
RGNEGLOC
RGH2POL
0x34
[0]
0
SCK
H1HBLKRETIME
[1]
0
H2HBLKRETIME
[2]
[3]
0
0
HLHBLKRETIME
HL_HBLK_EN
[7:4]
4
HCLK_WIDTH
[2:0]
1
0x35
SCK
Mnemonic
H1POSLOC
H1NEGLOC
H1POL
H1DRV
Rev. C | Page 78 of 92
Description
H1 rising edge location.
H1 falling edge location.
H1 polarity control:
0: Inverse of Figure 19.
1: No inversion.
H2 rising edge location (H5 in HCLK Mode 3).
H2 falling edge location (H5 in HCLK Mode 3).
H2 polarity (H5 in HCLK Mode 3):
0: Inverse of Figure 19.
1: No inversion.
HL rising edge location.
HL falling edge location.
HL polarity control:
0: Inverse of Figure 19.
1: No inversion.
RG rising edge location.
RG falling edge location.
RG polarity control:
0: Inverse of Figure 19.
1: No inversion.
Retime H1, H2, HL HBLK to the internal clock:
0: No retime.
1: Retime.
Recommended setting is retime enabled (1). Setting to 1
adds one cycle delay to programmed HBLK positions.
Enable HBLK for HL output:
0: Disable.
1: Enable.
Enables wide H-clocks during HBLK interval. Set to 0 to
disable.
H1 drive strength:
0: Off.
1: 4.3 mA.
2: 8.6 mA.
3: 12.9 mA.
4: 4.3 mA.
5: 8.6 mA.
6: 12.9 mA.
7: 17.2 mA.
AD9992
Address
0x36
0x37
0x38
0x39
Data Bits
[6:4]
[10:8]
[14:12]
[18:16]
[22:20]
[2:0]
[6:4]
[10:8]
[14:12]
[5:0]
[11:6]
[17:12]
[5:0]
Default
Value
1
1
1
1
1
1
1
1
1
0
20
10
0
Update
Type
[11:6]
20
DOUTPHASEN
[12]
[14:13]
0
0
DCLKMODE
DOUTDELAY
[15]
0
DCLKINV
[2:0]
7
SCK
SCK
SCK
Mnemonic
H2DRV
H3DRV
H4DRV
HLDRV
RGDRV
H5DRV
H6DRV
H7DRV
H8DRV
SHDLOC
SHPLOC
SHPWIDTH
DOUTPHASEP
Description
H2 drive strength (same range as H1DRV).
H3 drive strength (same range as H1DRV).
H4 drive strength (same range as H1DRV).
HL drive strength (same range as H1DRV).
RG drive strength (same range as H1DRV).
H5 drive strength (same range as H1DRV).
H6 drive strength (same range as H1DRV).
H7 drive strength (same range as H1DRV).
H8 drive strength (same range as H1DRV).
SHD sampling edge location.
SHP sampling edge location.
SHP width (controls input dc restore switch active time).
DOUT phase control, positive edge. Specifies location of
DOUT.
DOUT phase control, negative edge. Always set to
DOUTPHASEP plus 32 edges to maintain 50% duty cycle of
internal DOUTPHASE clocking.
DCLK mode. 0: DCLK tracks DOUT; 1: DCLK phase is fixed.
Data output delay (tOD) with respect to DCLK rising edge:
0: No delay.
1: ~3 ns.
2: ~6 ns.
3: ~9 ns.
Invert DCLK output:
0: No inversion.
1: Inversion of DCLK.
Enable H-masking during CP operation.
SCK
CPHMASK
Update
Type
Mnemonic
Description
Test registers only. Do not access.
Update
Type
Mnemonic
Description
Test registers only. Do not access.
Table 35. Test Registers—Do Not Access
Address
0x3E to 0x4F
Data Bits
Default
Value
Table 36. Test Registers—Do Not Access
Address
0x50 to 0x6F
Data Bits
Default
Value
Table 37. Shutter and GPO Registers
Address
0x70
Data Bits
[2:0]
[5:3]
Default
Value
0
0
Update
Type
VD
Mnemonic
PRIMARY_ACTION
SECOND_ACTION
Rev. C | Page 79 of 92
Description
Selects action for primary and secondary counters.
0: Idle (do nothing) autoreset on VD.
1: Activate counter (primary: auto exposure/readout).
2: RapidShot, wrap/repeat counter.
3: ShotTimer, delay start of count.
4: ShotTimer with RapidShot.
5: SLR exposure (manual).
6: SLR read (manual).
7: Force to idle.
AD9992
Address
Data Bits
[13:6]
Default
Value
0
Update
Type
0x71
[12:0]
[24:13]
[27:25]
[12:0]
0
0
0
0
VD
[13]
0
PRIMARY_SKIP
[26:14]
[27]
[2:0]
[5:3]
[8:6]
[11:9]
[14:12]
[17:15]
[20:18]
[23:21]
0
0
0
0
0
0
0
0
0
0
SECOND_DELAY
SECOND_SKIP
GP1_PROTOCOL
GP2_PROTOCOL
GP3_PROTOCOL
GP4_PROTOCOL
GP5_PROTOCOL
GP6_PROTOCOL
GP7_PROTOCOL
GP8_PROTOCOL
[12:0]
[25:13]
[26]
0
0
1
[27]
0
[0]
[13:1]
0
0
[26:14]
0
SUBCKNUM
[27]
0
SG_SUPPRESS
[12:0]
[25:13]
[26]
[12:0]
[25:13]
[0]
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
1FFF
1FFF
0
1FFF
1FFF
0
0
0
0
0
0
0
0
1
1
0x72
0x73
0x74
0x75
0x76
0x77
0x78
VD
VD
VD
Mnemonic
MANUAL_TRIG
PRIMARY_MAX
SECOND_MAX
VDHD_MASK
PRIMARY_DELAY
SGMASK_NUM
SUBCKMASK_NUM
SUBCKTOG_UPDATE
SUBCKMASK_SKIP1
SG
VD
VD/SG
VD
TEST
SUBCKSTARTLINE
SUBCK_TOG1
SUBCK_TOG2
SUBCK_POL
SUBCKHP_TOG1
SUBCKHP_TOG2
GP1_POL
GP2_POL
GP3_POL
GP4_POL
GP5_POL
GP6_POL
GP7_POL
GP8_POL
SEL_GP1
SEL_GP2
Rev. C | Page 80 of 92
Description
1: Manual trigger for GP signals, when Protocol 1 is
selected.
Bit 6: GP1 manual trigger.
…
Bit 13: GP8 manual trigger.
Primary counter maximum value.
Secondary counter maximum value.
Mask VD/HD during counter operation.
Number of fields to delay before the next count
(exposure) starts.
ShotTimer with RapidShot, skip delay before first count
(exposure).
Number of fields to delay before the next count starts.
ShotTimer with RapidShot, skip delay before first count.
Selects protocol for each general-purpose signal.
Idle = 0.
No counter association = 1.
Link to primary = 2.
Link to secondary = 3.
Link to mode = 4.
Primary repeat = 5.
Secondary repeat = 6.
Keep on = 7.
Exposure: number of fields to mask SGs.
Exposure plus readout: number of fields to mask SUBCK.
0: SUBCK toggles (Register 0x77) updated on SG line.
1: SUBCK toggles (Register 0x77) updated on UPDATE line
(VD-updated).
Skip the SUBCK mask for the first exposure field only.
Typically set to 1.
Reserved for test purpose. Must be set to 0.
Line location after VSG line to begin SUBCK pulses.
Must not be set to 1.
Number of SUBCK pulses per field. Must be set less than
VDLEN.
Suppress the SG and allow SUBCK to finish at
SUBCKNUM.
SUBCK Toggle Position 1.
SUBCK Toggle Position 2.
SUBCK start polarity.
High precision SUBCK Toggle Position 1.
High precision SUBCK Toggle Position 2.
GP1 low/high start polarity.
GP2 low/high start polarity.
GP3 low/high start polarity.
GP4 low/high start polarity.
GP5 low/high start polarity.
GP6 low/high start polarity.
GP7 low/high start polarity.
GP8 low/high start polarity.
1: GP1 signal is selected for GPO1 output.
1: GP2 signal is selected for GPO2 output.
AD9992
Address
0x79
0x7A
0x7B
0x7C
0x7D
0x7E
0x7F
0x80
Data Bits
[10]
[11]
Default
Value
1
1
Update
Type
[12]
1
SEL_GP5
[13]
1
SEL_GP6
[14]
1
SEL_GP7
[15]
1
SEL_GP8
[23:16]
0
GPO_OUTPUT_EN
[24]
0
GPO5_OVERRIDE
[25]
0
GPO6_OVERRIDE
[26]
0
GPO7_OVERRIDE
[27]
0
GPO8_OVERRIDE
[7:0]
[11:8]
[15:12]
[19:16]
[23:20]
[12:0]
0
{0, 0, 0, 0}
{0, 0, 0, 0}
{0, 0, 0, 0}
{0, 0, 0, 0}
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
VD
VD
Mnemonic
SEL_GP3
SEL_GP4
GP*_USE_LUT
LUT_FOR_GP12
LUT_FOR_GP34
LUT_FOR_GP56
LUT_FOR_GP78
GP1_TOG1_FD
GP1_TOG1_LN
VD
GP1_TOG1_PX
GP1_TOG2_FD
VD
GP1_TOG2_LN
GP1_TOG2_PX
VD
GP1_TOG3_FD
GP1_TOG3_LN
VD
GP1_TOG3_PX
GP1_TOG4_FD
VD
GP1_TOG4_LN
GP1_TOG4_PX
VD
GP2_TOG1_FD
GP2_TOG1_LN
Rev. C | Page 81 of 92
Description
1: GP3 signal is selected for GPO3 output.
1: GP4 signal is selected for GPO4 output.
0: SUBCK is selected.
1: GP5 signal is selected for GPO5 output.
0: XV21 is selected.
1: GP6 signal is selected for GPO6 output.
0: XV22 is selected.
1: GP7 signal is selected for GPO7 output.
0: XV23 is selected.
1: GP8 signal is selected for GPO8 output.
0: XV24 is selected.
1: GPO outputs enabled.
0: GPO is input high-Z state (default).
1: When GPO5 configured as input, overrides internal
OUT_CONT.
1: When GPO6 configured as input, overrides internal
HBLK.
1: When GPO7 configured as input, overrides internal
CLPOB.
1: When GPO8 configured as input, overrides internal
PBLK.
Use result from LUT, or else GP* is unaltered.
Two-input look-up table results.
Examples: {LUT_FOR_GP12} Å [GP2:GP1].
{0, 1, 1, 0} = GP2 XOR GP1; {1, 1, 1, 0} = GP2 OR GP1.
{0, 1, 1, 1} = GP2 NAND GP1; {1, 0, 0, 0} = GP2 AND GP1.
General-Purpose Signal 1, first toggle position, field
location.
General-Purpose Signal 1, first toggle position, line
location.
General-Purpose Signal 1, first toggle position, pixel
location.
General-Purpose Signal 1, second toggle position, field
location.
General-Purpose Signal 1, second toggle position, line
location.
General-Purpose Signal 1, second toggle position, pixel
location.
General-Purpose Signal 1, third toggle position, field
location.
General-Purpose Signal 1, third toggle position, line
location.
General-Purpose Signal 1, third toggle position, pixel
location.
General-Purpose Signal 1, fourth toggle position, field
location.
General-Purpose Signal 1, fourth toggle position, line
location.
General-Purpose Signal 1, fourth toggle position, pixel
location.
General-Purpose Signal 2, first toggle position, field
location.
General-Purpose Signal 2, first toggle position, line
location.
AD9992
Address
0x81
0x82
0x83
0x84
0x85
0x86
0x87
0x88
0x89
0x8A
0x8B
0x8C
0x8D
0x8E
Data Bits
[12:0]
Default
Value
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
Update
Type
VD
Mnemonic
GP2_TOG1_PX
GP2_TOG2_FD
VD
GP2_TOG2_LN
GP2_TOG2_PX
VD
GP2_TOG3_FD
GP2_TOG3_LN
VD
GP2_TOG3_PX
GP2_TOG4_FD
VD
GP2_TOG4_LN
GP2_TOG4_PX
VD
GP3_TOG1_FD
GP3_TOG1_LN
VD
GP3_TOG1_PX
GP3_TOG2_FD
VD
GP3_TOG2_LN
GP3_TOG2_PX
VD
GP3_TOG3_FD
GP3_TOG3_LN
VD
GP3_TOG3_PX
GP3_TOG4_FD
VD
GP3_TOG4_LN
GP3_TOG4_PX
VD
GP4_TOG1_FD
GP4_TOG1_LN
VD
GP4_TOG1_PX
GP4_TOG2_FD
VD
GP4_TOG2_LN
GP4_TOG2_PX
Rev. C | Page 82 of 92
Description
General-Purpose Signal 2, first toggle position, pixel
location.
General-Purpose Signal 2, second toggle position, field
location.
General-Purpose Signal 2, second toggle position, line
location.
General-Purpose Signal 2, second toggle position, pixel
location.
General-Purpose Signal 2, third toggle position, field
location.
General-Purpose Signal 2, third toggle position, line
location.
General-Purpose Signal 2, third toggle position, pixel
location.
General-Purpose Signal 2, fourth toggle position, field
location.
General-Purpose Signal 2, fourth toggle position, line
location.
General-Purpose Signal 2, fourth toggle position, pixel
location.
General-Purpose Signal 3, first toggle position, field
location.
General-Purpose Signal 3, first toggle position, line
location.
General-Purpose Signal 3, first toggle position, pixel
location.
General-Purpose Signal 3, second toggle position, field
location.
General-Purpose Signal 3, second toggle position, line
location.
General-Purpose Signal 3, second toggle position, pixel
location.
General-Purpose Signal 3, third toggle position, field
location.
General-Purpose Signal 3, third toggle position, line
location.
General-Purpose Signal 3, third toggle position, pixel
location.
General-Purpose Signal 3, fourth toggle position, field
location.
General-Purpose Signal 4, fourth toggle position, line
location.
General-Purpose Signal 4, fourth toggle position, pixel
location.
General-Purpose Signal 4, first toggle position, field
location.
General-Purpose Signal 4, first toggle position, line
location.
General-Purpose Signal 4, first toggle position, pixel
location.
General-Purpose Signal 4, second toggle position, field
location.
General-Purpose Signal 4, second toggle position, line
location.
General-Purpose Signal 4, second toggle position, pixel
location.
AD9992
Address
0x8F
0x90
0x91
0x92
0x93
0x94
0x95
0x96
0x97
0x98
0x99
0x9A
0x9B
0x9C
Data Bits
[12:0]
Default
Value
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
Update
Type
VD
Mnemonic
GP4_TOG3_FD
GP4_TOG3_LN
VD
GP4_TOG3_PX
GP4_TOG4_FD
VD
GP4_TOG4_LN
GP4_TOG4_PX
VD
GP5_TOG1_FD
GP5_TOG1_LN
VD
GP5_TOG1_PX
GP5_TOG2_FD
VD
GP5_TOG2_LN
GP5_TOG2_PX
VD
GP5_TOG3_FD
GP5_TOG3_LN
VD
GP5_TOG3_PX
GP5_TOG4_FD
VD
GP5_TOG4_LN
GP5_TOG4_PX
VD
GP6_TOG1_FD
GP6_TOG1_LN
VD
GP6_TOG1_PX
GP6_TOG2_FD
VD
GP6_TOG2_LN
GP6_TOG2_PX
VD
GP6_TOG3_FD
GP6_TOG3_LN
VD
GP6_TOG3_PX
GP6_TOG4_FD
Rev. C | Page 83 of 92
Description
General-Purpose Signal 4, third toggle position, field
location.
General-Purpose Signal 4, third toggle position, line
location.
General-Purpose Signal 4, third toggle position, pixel
location.
General-Purpose Signal 4, fourth toggle position, field
location.
General-Purpose Signal 4, fourth toggle position, line
location.
General-Purpose Signal 4, fourth toggle position, pixel
location.
General-Purpose Signal 5, first toggle position, field
location.
General-Purpose Signal 5, first toggle position, line
location.
General-Purpose Signal 5, first toggle position, pixel
location.
General-Purpose Signal 5, second toggle position, field
location.
General-Purpose Signal 5, second toggle position, line
location.
General-Purpose Signal 5, second toggle position, pixel
location.
General-Purpose Signal 5, third toggle position, field
location.
General-Purpose Signal 5, third toggle position, line
location.
General-Purpose Signal 5, third toggle position, pixel
location.
General-Purpose Signal 5, fourth toggle position, field
location.
General-Purpose Signal 5, fourth toggle position, line
location.
General-Purpose Signal 5, fourth toggle position, pixel
location.
General-Purpose Signal 6, first toggle position, field
location.
General-Purpose Signal 6, first toggle position, line
location.
General-Purpose Signal 6, first toggle position, pixel
location.
General-Purpose Signal 6, second toggle position, field
location.
General-Purpose Signal 6, second toggle position, line
location.
General-Purpose Signal 6, second toggle position, pixel
location.
General-Purpose Signal 6, third toggle position, field
location.
General-Purpose Signal 6, third toggle position, line
location.
General-Purpose Signal 6, third toggle position, pixel
location.
General-Purpose Signal 6, fourth toggle position, field
location.
AD9992
Address
0x9D
0x9E
0x9F
0xA0
0xA1
0xA2
0xA3
0xA4
0xA5
0xA6
0xA7
0xA8
0xA9
0xAA
Data Bits
[12:0]
Default
Value
0
Update
Type
VD
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[12:0]
0
[25:13]
0
[0]
0
VD
SUBCK_TOG1_13
[1]
0
VD
SUBCK_TOG2_13
Mnemonic
GP6_TOG4_LN
GP6_TOG4_PX
VD
GP7_TOG1_FD
GP7_TOG1_LN
VD
GP7_TOG1_PX
GP7_TOG2_FD
VD
GP7_TOG2_LN
GP7_TOG2_PX
VD
GP7_TOG3_FD
GP7_TOG3_LN
VD
GP7_TOG3_PX
GP7_TOG4_FD
VD
GP7_TOG4_LN
GP7_TOG4_PX
VD
GP8_TOG1_FD
GP8_TOG1_LN
VD
GP8_TOG1_PX
GP8_TOG2_FD
VD
GP8_TOG2_LN
GP8_TOG2_PX
VD
GP8_TOG3_FD
GP8_TOG3_LN
VD
GP8_TOG3_PX
GP8_TOG4_FD
VD
GP8_TOG4_LN
GP8_TOG4_PX
Rev. C | Page 84 of 92
Description
General-Purpose Signal 6, fourth toggle position, line
location.
General-Purpose Signal 6, fourth toggle position, pixel
location.
General-Purpose Signal 7, first toggle position, field
location.
General-Purpose Signal 7, first toggle position, line
location.
General-Purpose Signal 7, first toggle position, pixel
location.
General-Purpose Signal 7, second toggle position, field
location.
General-Purpose Signal 7, second toggle position, line
location.
General-Purpose Signal 7, second toggle position, pixel
location.
General-Purpose Signal 7, third toggle position, field
location.
General-Purpose Signal 7, third toggle position, line
location.
General-Purpose Signal 7, third toggle position, pixel
location.
General-Purpose Signal 7, fourth toggle position, field
location.
General-Purpose Signal 7, fourth toggle position, line
location.
General-Purpose Signal 7, fourth toggle position, pixel
location.
General-Purpose Signal 8, first toggle position, field
location.
General-Purpose Signal 8, first toggle position, line
location.
General-Purpose Signal 8, first toggle position, pixel
location.
General-Purpose Signal 8, second toggle position, field
location.
General-Purpose Signal 8, second toggle position, line
location.
General-Purpose Signal 8, second toggle position, pixel
location.
General-Purpose Signal 8, third toggle position, field
location.
General-Purpose Signal 8, third toggle position, line
location.
General-Purpose Signal 8, third toggle position, pixel
location.
General-Purpose Signal 8, fourth toggle position, field
location.
General-Purpose Signal 8, fourth toggle position, line
location.
General-Purpose Signal 8, fourth toggle position, pixel
location.
Bit 13 for SUBCK Toggle Position 1. For 14-bit H-counter
mode.
Bit 13 for SUBCK Toggle Position 2. For 14-bit H-counter
mode.
AD9992
Address
Data Bits
[2]
[3]
Default
Value
0
0
Update
Type
VD/SG
VD/SG
Mnemonic
SUBCKHP_TOG1_13
SUBCKHP_TOG2_13
Description
Bit 13 for SUBCK HP Toggle 1. For 14-bit H-counter mode.
Bit 13 for SUBCK HP Toggle 2. For 14-bit H-counter mode.
Table 38. Update Control Registers
Address
0xB0
Data
Bits
[15:0]
Default
Value
1803
Update
SCK
Mnemonic
AFE_UPDT_SCK
0xB1
[15:0]
E7FC
SCK
AFE_UPDT_VD
0xB2
[15:0]
F8FD
SCK
MISC_UPDT_SCK
0xB3
[15:0]
0702
SCK
MISC_UPDT_VD
0xB4
0xB5
[15:0]
[15:0]
FFF9
0006
SCK
SCK
VDHD_UPDT_SCK
VDHD_UPDT_VD
Mnemonic
TEST
GPO_INT_EN
Description
Each bit corresponds to one address location.
AFE_UPDT_SCK [0] = 1, update Address 0x00 on SL rising edge.
AFE_UPDT_SCK [1] = 1, update Address 0x01 on SL rising edge.
…
AFE_UPDT_SCK [15] = 1, update Address 0x0F on SL rising edge.
Each bit corresponds to one address location.
AFE_UPDT_VD [0] = 1, update Address 0x00 on VD rising edge.
AFE_UPDT_VD [1] = 1, update Address 0x01 on VD rising edge.
…
AFE_UPDT_VD [15] = 1, update Address 0x0F on VD rising edge.
Enable SCK update of miscellaneous registers, Address 0x10 to
Address 0x1F.
Enable VD update of miscellaneous registers, Address 0x10 to
Address 0x1F.
Enable SCK update of VDHD registers, Address 0x20 to Address 0x2F.
Enable VD update of VDHD registers, Address 0x20 to Address 0x2F.
Table 39. Extra Registers
Address
0xD4
Data
Bits
[0]
[1]
Default
Value
0
0
Update
SCK
0xD7
[9:2]
[0]
[1]
0
0
0
SCK
TEST
TEST
XV24_SWAP
0xD8
[27:0]
0
SCK
START
Description
Test mode only. Set to 0.
Allow observation of internal signals at GPO5 to GPO8 outputs:
GPO5: OUTCONTROL.
GPO6: HBLK.
GPO7: CLPOB.
GPO8: PBLK.
Test mode only. Set to 0.
Test mode only. Set to 0.
Set to 1 to change the V-driver output configuration so that XV15 is
output on the XV24 output pin. Useful with special vertical sequence
alternation mode when the XV24 register is reserved for pattern selection.
Recommended start-up register. Should be set to 0x888.
Rev. C | Page 85 of 92
AD9992
Table 40. V-Pattern Group (VPAT) Register Map
Address
0x00
0x01
0x02
0x03
0x04
0x05
0x06
0x07
0x08
0x09
0x0A
0x0B
0x0C
0x0D
0x0E
0x0F
0x10
0x11
0x12
0x13
0x14
0x15
0x16
0x17
0x18
Data Bits
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
Default Value
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Update Type
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
Rev. C | Page 86 of 92
Mnemonic
XV1TOG1
XV1TOG2
XV1TOG3
XV1TOG4
XV2TOG1
XV2TOG2
XV2TOG3
XV2TOG4
XV3TOG1
XV3TOG2
XV3TOG3
XV3TOG4
XV4TOG1
XV4TOG2
XV4TOG3
XV4TOG4
XV5TOG1
XV5TOG2
XV5TOG3
XV5TOG4
XV6TOG1
XV6TOG2
XV6TOG3
XV6TOG4
XV7TOG1
XV7TOG2
XV7TOG3
XV7TOG4
XV8TOG1
XV8TOG2
XV8TOG3
XV8TOG4
XV9TOG1
XV9TOG2
XV9TOG3
XV9TOG4
XV10TOG1
XV10TOG2
XV10TOG3
XV10TOG4
XV11TOG1
XV11TOG2
XV11TOG3
XV11TOG4
XV12TOG1
XV12TOG2
XV12TOG3
XV12TOG4
XV13TOG1
XV13TOG2
Description
XV1 Toggle Position 1.
XV1 Toggle Position 2.
XV1 Toggle Position 3.
XV1 Toggle Position 4.
XV2 Toggle Position 1.
XV2 Toggle Position 2.
XV2 Toggle Position 3.
XV2 Toggle Position 4.
XV3 Toggle Position 1.
XV3 Toggle Position 2.
XV3 Toggle Position 3.
XV3 Toggle Position 4.
XV4 Toggle Position 1.
XV4 Toggle Position 2.
XV4 Toggle Position 3.
XV4 Toggle Position 4.
XV5 Toggle Position 1.
XV5 Toggle Position 2.
XV5 Toggle Position 3.
XV5 Toggle Position 4.
XV6 Toggle Position 1.
XV6 Toggle Position 2.
XV6 Toggle Position 3.
XV6 Toggle Position 4.
XV7 Toggle Position 1.
XV7 Toggle Position 2.
XV7 Toggle Position 3.
XV7 Toggle Position 4.
XV8 Toggle Position 1.
XV8 Toggle Position 2.
XV8 Toggle Position 3.
XV8 Toggle Position 4.
XV9 Toggle Position 1.
XV9 Toggle Position 2.
XV9 Toggle Position 3.
XV9 Toggle Position 4.
XV10 Toggle Position 1.
XV10 Toggle Position 2.
XV10 Toggle Position 3.
XV10 Toggle Position 4.
XV11 Toggle Position 1.
XV11 Toggle Position 2.
XV11 Toggle Position 3.
XV11 Toggle Position 4.
XV12 Toggle Position 1.
XV12 Toggle Position 2.
XV12 Toggle Position 3.
XV12 Toggle Position 4.
XV13 Toggle Position 1.
XV13 Toggle Position 2.
AD9992
Address
0x19
0x1A
0x1B
0x1C
0x1D
0x1E
0x1F
0x20
0x21
0x22
0x23
0x24
0x25
0x26
0x27
0x28
0x29
0x2A
0x2B
0x2C
0x2D
0x2E
0x2F
Data Bits
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
Default Value
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Update Type
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
Rev. C | Page 87 of 92
Mnemonic
XV13TOG3
XV13TOG4
XV14TOG1
XV14TOG2
XV14TOG3
XV14TOG4
XV15TOG1
XV15TOG2
XV15TOG3
XV15TOG4
XV16TOG1
XV16TOG2
XV16TOG3
XV16TOG4
XV17TOG1
XV17TOG2
XV17TOG3
XV17TOG4
XV18TOG1
XV18TOG2
XV18TOG3
XV18TOG4
XV19TOG1
XV19TOG2
XV19TOG3
XV19TOG4
XV20TOG1
XV20TOG2
XV20TOG3
XV20TOG4
XV21TOG1
XV21TOG2
XV21TOG3
XV21TOG4
XV22TOG1
XV22TOG2
XV22TOG3
XV22TOG4
XV23TOG1
XV23TOG2
XV23TOG3
XV23TOG4
XV24TOG1
XV24TOG2
XV24TOG3
XV24TOG4
Description
XV13 Toggle Position 3.
XV13 Toggle Position 4.
XV14 Toggle Position 1.
XV14 Toggle Position 2
XV14 Toggle Position 3.
XV14 Toggle Position 4.
XV15 Toggle Position 1.
XV15 Toggle Position 2.
XV15 Toggle Position 3.
XV15 Toggle Position 4.
XV16 Toggle Position 1.
XV16 Toggle Position 2.
XV16 Toggle Position 3.
XV16 Toggle Position 4.
XV17 Toggle Position 1.
XV17 Toggle Position 2.
XV17 Toggle Position 3.
XV17 Toggle Position 4.
XV18 Toggle Position 1.
XV18 Toggle Position 2.
XV18 Toggle Position 3.
XV18 Toggle Position 4.
XV19 Toggle Position 1.
XV19 Toggle Position 2.
XV19 Toggle Position 3.
XV19 Toggle Position 4.
XV20 Toggle Position 1.
XV20 Toggle Position 2.
XV20 Toggle Position 3.
XV20 Toggle Position 4.
XV21 Toggle Position 1.
XV21 Toggle Position 2.
XV21 Toggle Position 3.
XV21 Toggle Position 4.
XV22 Toggle Position 1.
XV22 Toggle Position 2.
XV22 Toggle Position 3.
XV22 Toggle Position 4.
XV23 Toggle Position 1.
XV23 Toggle Position 2.
XV23 Toggle Position 3.
XV23 Toggle Position 4.
XV24 Toggle Position 1.
XV24 Toggle Position 2.
XV24 Toggle Position 3.
XV24 Toggle Position 4.
AD9992
Table 41. V-Sequence (VSEQ) Registers
Data
Bits
[0]
[1]
[5:2]
[9:6]
[13:10]
Default
Value
X
X
X
X
X
[15:14]
X
VREP_MODE
[19:16]
[23:20]
[25:24]
X
X
X
LASTREPLEN_EN
LASTTOG_EN
HBLK_MODE
0x02
[12:0]
[25:13]
[23:0]
X
X
X
0x03
0x04
0x05
0x06
0x07
[24]
[25]
[23:0]
[23:0]
[23:0]
[23:0]
[23:0]
0x08
[23:0]
Address
0x00
0x01
Update
Type
SCP
SCP
Mnemonic
CLPOBPOL
PBLKPOL
HOLD
VMASK_EN
CONCAT_GRP
SCP
HDLENE
HDLENO
VSGPATSEL
X
X
X
X
X
SCP
SCP
SCP
SCP
SCP
HDLENE_13
HDLENO_13
VPOL_A
VPOL_B
VPOL_C
VPOL_D
GROUPSEL_0
X
SCP
GROUPSEL_1
Description
CLPOB start polarity.
PBLK start polarity.
1: Enable HOLD function for each VPAT group (A, B, C, D).
1: Enable FREEZE/RESUME for each VPAT group (A, B, C, D).
Combine multiple VPAT groups together in one sequence. Set register equal
to 0x01 to enable.
Defines V-alternation repetition mode.
00: Single pattern alternation for all groups.
01: Two pattern alternation for all groups.
10: Three-pattern alternation for Group A. Groups B, C, and D
follow pattern {0, 1, 1, 0, 1, 1…}.
11: Four-pattern alternation for Group A. Two-pattern alternation
for Groups B, C, and D.
Enable use of last repetition counter for last repetition length of each group.
Enable the fifth toggle position for all V-signals in each group.
Selection of HBLK modes:
00: HBLK Mode 0 (normal six-toggle operation).
01: HBLK Mode 1.
10: HBLK Mode 2. (Address 0x19 to Address 0x1E operate differently.)
11: Test only, do not access.
HD line length for even lines.
HD line length for odd lines.
Selects which two toggle positions are used by each V-output when they
are configured as VSG pulses (Miscellaneous Register Address 0x1C, fixed
register area):
0: Use Toggle 1, Toggle 2.
1: Use Toggle 3, Toggle 4.
HD length Bit 13 for even lines when 14-bit H-counter is enabled.
HD length Bit 13 for odd lines when 14-bit H-counter is enabled.
Starting polarities for each V-output signal (Group A).
Starting polarities for each V-output signal (Group B).
Starting polarities for each V-output signal (Group C).
Starting polarities for each V-output signal (Group D).
Select which group each XV1 to XV12 signal is assigned to:
00: Group A.
01: Group B.
10: Group C.
11: Group D.
[1:0]: XV1.
[3:2]: XV2.
…
[23:22]: XV12.
Select which group each XV13 to XV24 signal is assigned to:
00: Group A.
01: Group B.
10: Group C.
11: Group D.
[1:0]: XV13.
[3:2]: XV14.
…
[23:22]: XV24.
Rev. C | Page 88 of 92
AD9992
Address
0x09
0x0A
0x0B
0x0C
0x0D
0x0E
0x0F
0x10
0x11
0x12
0x13
0x14
0x15
0x16
0x17
0x18
0x19
0x1A
0x1B
0x1C
0x1D
0x1E
Data
Bits
[4:0]
[9:5]
[14:10]
[19:15]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[20:13]
[21]
[22]
[23]
[25:24]
[12:0]
Default
Value
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
X
X
X
X
X
X
X
X
X
X
X
Update
Type
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
SCP
Mnemonic
VPATSELA
VPATSELB
VPATSELC
VPATSELD
VSTARTA
VLENA
VREPA_1
VREPA_2
VREPA_3
VREPA_4
VSTARTB
VLENB
VREPB_ODD
VREPB_EVEN
VSTARTC
VLENC
VREPC_ODD
VREPC_EVEN
VSTARTD
VLEND
VREPD_ODD
VREPD_EVEN
FREEZE1
RESUME1
FREEZE2
RESUME2
FREEZE3
RESUME3
FREEZE4
RESUME4
HBLKSTART
HBLKEND
HBLKLEN
HBLKREP
HBLKMASK_H1
HBLKMASK_H2
HBLKMASK_HL
TEST
HBLKTOGO1
HBLKTOGO2
HBLKTOGO3
HBLKTOGO4
HBLKTOGO5
HBLKTOGO6
HBLKTOGE1
HBLKTOGE2
HBLKTOGE3
HBLKTOGE4
HBLKTOGE5
HBLKTOGE6
Description
Selected VPAT group for Group A, from VPAT Group 0 to Group 31.
Selected VPAT group for Group B, from VPAT Group 0 to Group 31.
Selected VPAT group for Group C, from VPAT Group 0 to Group 31.
Selected VPAT group for Group D, from VPAT Group 0 to Group 31.
Start position of selected V-Pattern Group A.
Length of selected V-Pattern Group A.
Number of repetitions for V-Pattern Group A for first lines.
Number of repetitions for V-Pattern Group A for second lines.
Number of repetitions for V-Pattern Group A for third lines.
Number of repetitions for V-Pattern Group A for fourth lines.
Start position of selected V-Pattern Group B.
Length of selected V-Pattern Group B.
Number of repetitions for V-Pattern Group B for odd lines.
Number of repetitions for V-Pattern Group B for even lines.
Start position of selected V-Pattern Group C.
Length of selected V-Pattern Group C.
Number of repetitions for V-Pattern Group C for odd lines.
Number of repetitions for V-Pattern Group C for even lines.
Start position of selected V-Pattern Group D.
Length of selected V-Pattern Group D.
Number of repetitions for V-Pattern Group D for odd lines.
Number of repetitions for V-Pattern Group D for even lines.
Holds the V-outputs at their current levels.
Resumes the operation of V-outputs to finish the pattern.
Holds the V-outputs at their current levels.
Resumes the operation of V-outputs to finish the pattern.
Holds the V-outputs at their current levels.
Resumes the operation of V-outputs to finish the pattern.
Holds the V-outputs at their current levels.
Resumes the operation of V-outputs to finish the pattern.
Start location for HBLK in HBLK Mode 1 and HBLK Mode 2.
End location for HBLK in HBLK Mode 1 and HBLK Mode 2.
HBLK length in HBLK Mode 1 and HBLK Mode 2.
Number of HBLK repetitions in HBLK Mode 1 and HBLK Mode 2.
Masking polarity for H1/H3/H5/H7 during HBLK.
Masking polarity for H2/H4/H6/H8 during HBLK.
Masking polarity for HL during HBLK.
Test mode only. Set to 0.
First HBLK toggle position for odd lines, or RA0H1REPA/B/C in HBLK Mode 2
(see the HBLK Mode 2 Operation section for more information).
Second HBLK toggle position for odd lines, or RA1H1REPA/B/C.
Third HBLK toggle position for odd lines, or RA2H1REPA/B/C.
Fourth HBLK toggle position for odd lines, or RA3H1REPA/B/C.
Fifth HBLK toggle position for odd lines, or RA4H1REPA/B/C.
Sixth HBLK toggle position for odd lines, or RA5H1REPA/B/C.
First HBLK toggle position for even lines, or RA0H2REPA/B/C.
Second HBLK toggle position for even lines, or RA1H2REPA/B/C.
Third HBLK toggle position for even lines, or RA2H2REPA/B/C.
Fourth HBLK toggle position for even lines, or RA3H2REPA/B/C.
Fifth HBLK toggle position for even lines, or RA4H2REPA/B/C.
Sixth HBLK toggle position for even lines, or RA5H2REPA/B/C.
Rev. C | Page 89 of 92
AD9992
Address
0x1F
0x20
0x21
0x22
0x23
0x24
0x25
0x26
0x27
Data
Bits
[12:0]
[25:13]
[12:0]
[13]
[14]
Default
Value
X
X
X
X
X
[17:15]
X
[2:0]
[6:4]
[10:8]
[14:12]
[18:16]
[22:20]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Update
Type
SCP
SCP
Mnemonic
HBLKSTARTA
HBLKSTARTB
HBLKSTARTC
VSEQALT_EN
VALT_MAP
SPC_PAT_EN
SCP
SCP
SCP
SCP
SCP
SCP
SCP
HBLKALT_PAT1
HBLKALT_PAT2
HBLKALT_PAT3
HBLKALT_PAT4
HBLKALT_PAT5
HBLKALT_PAT6
CLPOBTOG1
CLPOBTOG2
PBLKTOG1
PBLKTOG2
LASTREPLEN_A
LASTREPLEN_B
LASTREPLEN_C
LASTREPLEN_D
LASTTOG_A
LASTTOG_B
LASTTOG_C
LASTTOG_D
Description
HBLK Repeat Area Start Position A for HBLK Mode 2. Set to 8191 if not used.
HBLK Repeat Area Start Position B for HBLK Mode 2. Set to 8191 if not used.
HBLK Repeat Area Start Position C for HBLK Mode 2. Set to 8191 if not used.
Special V-sequence alternation enable.
1: Enables operation of VALTSEL0_EVEN/ODD, VALTSEL1_EVEN/ODD
registers in FREEZE/RESUME registers. Must be enabled if special VALT mode
is used.
1: Enables use of special vertical pattern insertion into VPATA sequence.
[0]: Use VPATB as the special pattern.
[1]: Use VPATC as the special pattern.
[2]: Use VPATD as the special pattern.
HBLK Mode 2, Repeat Area 0 pattern for odd lines.
HBLK Mode 2, Repeat Area 0 pattern for odd lines.
HBLK Mode 2, Repeat Area 0 pattern for odd lines.
HBLK Mode 2, Repeat Area 0 pattern for odd lines.
HBLK Mode 2, Repeat Area 0 pattern for odd lines.
HBLK Mode 2, Repeat Area 0 pattern for odd lines.
CLPOB Toggle Position 1.
CLPOB Toggle Position 2.
PBLK Toggle Position 1.
PBLK Toggle Position 2.
Last repetition length for Group A. Set equal to VLENA.
Last repetition length for Group B. Set equal to VLENB.
Last repetition length for Group C. Set equal to VLENC.
Last repetition length for Group D. Set equal to VLEND.
Optional fifth toggle position for Group A.
Optional fifth toggle position for Group B.
Optional fifth toggle position for Group C.
Optional fifth toggle position for Group D.
Table 42. Field Registers
Address
0x00
0x01
Data
Bits
[4:0]
[9:5]
[14:10]
[19:15]
[24:20]
[4:0]
[9:5]
[14:10]
[19:15]
[21:20]
[23:22]
[25:24]
Default
Value
X
X
X
X
X
X
X
X
X
Update
Type
VD
VD
Mnemonic
SEQ0
SEQ1
SEQ2
SEQ3
SEQ4
SEQ5
SEQ6
SEQ7
SEQ8
MULT_SWEEP0
MULT_SWEEP1
MULT_SWEEP2
Description
Selected V-sequence for first region in the field.
Selected V-sequence for second region in the field.
Selected V-sequence for third region in the field.
Selected V-sequence for fourth region in the field.
Selected V-sequence for fifth region in the field.
Selected V-sequence for sixth region in the field.
Selected V-sequence for seventh region in the field.
Selected V-sequence for eighth region in the field.
Selected V-sequence for ninth region in the field.
Enables multiplier mode and/or sweep mode for Region 0:
0: Multiplier off/sweep off.
1: Multiplier off/sweep on.
2: Multiplier on/sweep off.
3: Multiplier on/sweep on.
Enables multiplier mode and/or sweep mode for Region 2.
Enables multiplier mode and/or sweep mode for Region 1.
Rev. C | Page 90 of 92
AD9992
Address
0x02
0x03
0x04
0x05
0x06
0x07
0x08
0x09
0x0A
0x0B
0x0C
0x0D
0x0E
0x0F
Data
Bits
[12:0]
[14:13]
[16:15]
[18:17]
[20:19]
[22:21]
[24:23]
[25]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[23:0]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
[12:0]
[25:13]
Default
Value
X
X
X
X
X
X
X
Update
Type
VD
X
X
X
X
X
X
X
X
X
VD
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
VD
VD
VD
VD
VD
VD
VD
VD
VD
VD
VD
VD
Mnemonic
HDLASTLEN
MULT_SWEEP3
MULT_SWEEP4
MULT_SWEEP5
MULT_SWEEP6
MULT_SWEEP7
MULT_SWEEP8
HDLASTLEN_13
SCP0
SCP1
SCP2
SCP3
SCP4
SCP5
SCP6
SCP7
SCP8
VDLEN
SGACTLINE1
SGACTLINE2
SGMASK
CLPMASKSTART1
CLPMASKEND1
CLPMASKSTART2
CLPMASKEND2
CLPMASKSTART3
CLPMASKEND3
PBLKMASKSTART1
PBLKMASKEND1
PBLKMASKSTART2
PBLKMASKEND2
PBLKMASKSTART3
PBLKMASKEND3
Description
HD last line length. Line length of last line in the field.
Enables multiplier mode and/or sweep mode for Region 3.
Enables multiplier mode and/or sweep mode for Region 4.
Enables multiplier mode and/or sweep mode for Region 5.
Enables multiplier mode and/or sweep mode for Region 6.
Enables multiplier mode and/or sweep mode for Region 7.
Enables multiplier mode and/or sweep mode for Region 8.
HD last line length Bit 13 when 14-bit H-counter is enabled.
V-Sequence Change Position 0.
V-Sequence Change Position 1.
V-Sequence Change Position 2.
V-Sequence Change Position 3.
V-Sequence Change Position 4.
V-Sequence Change Position 5.
V-Sequence Change Position 6.
V-Sequence Change Position 7.
V-Sequence Change Position 8.
VD field length (number of lines in the field).
SG Active Line 1.
SG Active Line 2 (set to SG Active Line 1 or maximum if not used).
Masking of VSG outputs during SG active line.
CLPOB Mask Region 1 start position. Set to 8191 to disable.
CLPOB Mask Region 1 end position. Set to 0 to disable.
CLPOB Mask Region 2 start position. Set to 8191 to disable.
CLPOB Mask Region 2 end position. Set to 0 to disable.
CLPOB Mask Region 3 start position. Set to 8191 to disable.
CLPOB Mask Region 3 end position. Set to 0 to disable.
PBLK Mask Region 1 start position. Set to 8191 to disable.
PBLK Mask Region 1 end position. Set to 0 to disable.
PBLK Mask Region 2 start position. Set to 8191 to disable.
PBLK Mask Region 2 end position. Set to 0 to disable.
PBLK Mask Region 3 start position. Set to 8191 to disable.
PBLK Mask Region 3 end position. Set to 0 to disable.
Rev. C | Page 91 of 92
AD9992
OUTLINE DIMENSIONS
A1 BALL
CORNER
8.10
8.00 SQ
7.90
11 10 9 8 7 6 5 4
3 2 1
A
BALL A1
PAD CORNER
B
C
D
6.50
BSC SQ
E
F
G
H
J
K
L
0.65
BSC
BOTTOM VIEW
TOP VIEW
DETAIL A
DETAIL A
1.31
1.16
0.91 MIN
0.25 MIN
0.45
0.40
0.35
BALL DIAMETER
SEATING
PLANE
*COMPLIANT TO JEDEC STANDARDS MO-225
WITH THE EXCEPTION TO PACKAGE HEIGHT.
COPLANARITY
0.10
080807-A
*1.40
Figure 88. 105-Lead Chip Scale Package Ball Grid Array [CSP_BGA]
8 mm × 8 mm Body
(BC-105-1)
Dimensions shown in millimeters
ORDERING GUIDE
Model
AD9992BBCZ 1
AD9992BBCZRL1
1
Temperature Range
–25°C to +85°C
–25°C to +85°C
Package Description
105-Lead CSP_BGA
105-Lead CSP_BGA
Z = RoHS Compliant Part.
©2006–2007 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D05891-0-10/07(C)
Rev. C | Page 92 of 92
Package Option
BC-105-1
BC-105-1