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