a Complete 12-Bit 6 MSPS CCD/CIS Signal Processor AD9816 PRODUCT DESCRIPTION FEATURES 12-Bit 6 MSPS A/D Converter No Missing Codes Guaranteed 3-Channel or 1-Channel Operation Correlated Double Sampling 8-Bit Programmable Gain 8-Bit Offset Adjustment PGA Output Monitor Input Clamp Circuitry Internal Voltage Reference 3-Wire Serial Interface +3.3 V/+5 V Digital Output Compatibility 44-Lead MQFP Package Low Power CMOS: 420 mW Typ The AD9816 is a complete analog signal processor for CCD and CIS applications. Included is all the necessary circuitry to perform three-channel conditioning and sampling for a variety of imaging applications. The signal chain consists of an input clamp, correlated double sampler (CDS), offset adjust DAC, programmable gain amplifier and a 12-bit A/D converter. The CDS and input clamp may be disabled for CIS applications. The internal registers are programmed using a 3-wire serial interface and provide adjustment of the gain, offset and operating mode. The AD9816 operates from a +5 V supply, typically consumes 420 mW of power and is packaged in a 44-lead MQFP. FUNCTIONAL BLOCK DIAGRAM AVDD VINR AVSS CLAMP/CDS CAPT CAPB 6100mV + CML PGAOUT VREF DVDD DVSS DRVDD DRVSS AD9816 1X–6X OEB PGA BANDGAP REFERENCE DAC VING CLAMP/CDS + MUX PGA PGA CLAMP/CDS 8 SCLK 8 OFFSET REGISTERS CDSCLK1 CDSCLK2 DOUT 11:0 CONFIGURATION REGISTER + DAC OFFSET 12 MUX REGISTER DAC VINB 12-BIT ADC R G B R G B GAIN REGISTERS DIGITAL CONTROL PORT SLOAD SDATA ADCCLK REV. A 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 which may result from its use. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781/329-4700 World Wide Web Site: http://www.analog.com Fax: 781/326-8703 © Analog Devices, Inc., 1998 AD9816–SPECIFICATIONS ANALOG SPECIFICATIONS (TMIN to TMAX with AVDD = +5.0 V, DVDD = +5.0 V, DRVDD = +5.0 V, CDS Mode, fADCCLK = 6 MHz, fCDSCLK1 = 2 MHz, fCDSCLK2 = 2 MHz, PGA Gain = 1, Input Range = 3 V p-p, Input Capacitor = 1200 pF, unless otherwise noted) Parameter AD9816 AD9816-80010 Units MAXIMUM CONVERSION RATE 3-Channel Mode with CDS 1-Channel Mode with CDS 6 6 6 6 MSPS min MSPS min 12 ± 0.4 ± 1.0 12 ± 1.5 ± 4.0 2.4 4.3 12 ± 0.75 ± 2.5 Bits min LSB typ LSB max Bits Guaranteed LSB typ LSB max % FSR max % FSR max 0 3 AVSS – 0.3 AVDD + 0.3 10 10 0 3 AVSS – 0.3 AVDD + 0.3 10 10 V min V max V min V max pF typ nA typ 1 5.98 256 –100 +100 256 1 5.98 256 –100 +100 256 V/V min V/V max Steps mV min mV max Steps ACCURACY (Includes Entire Signal Path) ADC Resolution Differential Nonlinearity (DNL) No Missing Codes Integral Nonlinearity (INL) Offset Error Gain Error1 ANALOG INPUTS Input Voltage Range2 Input Limits3 Input Capacitance Input Current AMPLIFIERS PGA Gain Range PGA Gain Resolution Offset Range Offset Resolution NOISE AND CROSSTALK Total Output Noise at Min PGA Gain4 Total Output Noise at Max PGA Gain4 Channel-to-Channel Crosstalk5 0.5 0.8 1 LSB rms typ LSB rms typ LSB max POWER SUPPLY REJECTION (AVDD = +5 V/± 0.25 V) 0.28 % FSR max VOLTAGE REFERENCE 0.75 V Reference Tolerance (@ +25°C) 1.5 V Reference Tolerance (@ +25°C) ± 20 ± 34 mV max mV max TEMPERATURE RANGE Operating POWER SUPPLIES Operating Voltages AVDD, DVDD DRVDD Operating Current POWER CONSUMPTION 0 +70 0 +70 °C min °C max +4.75 +5.25 +3.3 +5.25 84 +4.75 +5.25 +3.3 +5.25 84 V min V max V min V max mA typ 420 500 420 500 mW typ mW max NOTES 1 Includes internal voltage reference error. 2 Input voltage range is the linear region over which the input signal can be processed by the input stage of the AD9816. 3 The input limits are defined as the maximum tolerable input voltage into the AD9816. This is not intended to be the linear input range of the device. Signals beyond the input limits will turn on the overvoltage protection diodes. 4 The total output noise is measured with the inputs of the AD9816 grounded. 5 The channel-to-channel crosstalk is measured with one input grounded, and the other two inputs at full scale. Specifications subject to change without notice. –2– REV. A AD9816 DIGITAL SPECIFICATIONS (TMIN to TMAX with AVDD = +5.0 V, DVDD = +5.0 V, DRVDD = +5.0 V, fADCCLK = 6 MHz, fCDSCLK1 = 2 MHz, fCDSCLK2 = 2 MHz, CL = 10 pF unless otherwise noted) Parameter Symbol Min LOGIC INPUTS High Level Input Voltage Low Level Input Voltage High Level Input Current Low Level Input Current Input Capacitance VIH VIL IIH IIL CIN 3.5 LOGIC OUTPUTS High Level Output Voltage Low Level Output Voltage High Level Output Current Low Level Output Current VOH VOL IOH IOL 4.5 Typ Max Units V V µA µA pF 1.0 10 10 10 V V µA µA 0.1 50 50 Specifications subject to change without notice. TIMING SPECIFICATIONS (T MIN to TMAX with DVDD = +5.0 V, DRVDD = +5.0 V) Parameter Symbol Min CLOCK PARAMETERS 3-Channel Conversion Rate 1-Channel Conversion Rate ADCCLK Pulsewidth CDSCLK1 Pulsewidth CDSCLK2 Pulsewidth CDSCLK1 Falling to CDSCLK2 Rising ADCCLK Falling to CDSCLK2 Rising CDSCLK2 Falling to ADCCLK Falling CDSCLK2 Falling to CDSCLK1 Rising Aperture Delay for CDS Clocks tCRA tCRB tADCLK tC1 tC2 tC1C2 tADC2 tC2AD tC2C1 tAD 500 160 80 20 60 5 0 30 10 10 SERIAL INTERFACE Maximum SCLK Frequency SLOAD to SCLK Set-Up Time SCLK to SLOAD Hold Time SDATA to SCLK Rising Set-Up Time SCLK Rising to SDATA Hold Time SCLK Falling to SDATA Valid fSCLK tLS tLH tDS tDH tRDV 10 10 10 10 10 10 DATA OUTPUT Output Delay 3-State to Data Valid Output Enable High to 3-State Latency (Pipeline Delay) REV. A tOD tDV tHZ Typ 2 tADCLK – 30 Units ns ns ns ns ns ns ns ns ns ns MHz ns ns ns ns ns 13 15 5 3 (Fixed) –3– Max ns ns ns ADCCLK Cycles AD9816 PIXEL (n+1) PIXEL n (R, G, B) ANALOG INPUTS PIXEL (n+m) PIXEL (n+2) tAD tAD tCRA tC2C1 tC1 CDSCLK1 tC1C2 tC2 CDSCLK2 tADCLK tADC2 tC2AD ADCCLK OUTPUT DATA D11:D0 tADCLK R(n–2) tOD G(n–2) B(n–2) R(n–1) G(n–1) R(n) G(n) B(n–1) R(n) G(n) B(n) R(n+1) G(n+1) R(n+1) PGAOUT_T G(n–1) B(n–1) B(n) R(n+2) B(n+1) PGAOUT_C Figure 1. 3-Channel CDS Mode Timing PIXEL n (R, G, B) PIXEL (n+1) PIXEL (n+m) PIXEL (n+2) tAD ANALOG INPUTS tC2 tCRA CDSCLK2 tADC2 tADCLK tC2AD ADCCLK OUTPUT DATA D11:D0 tADCLK R(n–2) tOD G(n–2) B(n–2) R(n–1) G(n–1) R(n) G(n) B(n–1) R(n) G(n) R(n+1) G(n+1) B(n) R(n+1) PGAOUT_T B(n–1) G(n–1) B(n) B(n+1) R(n+2) PGAOUT_C Figure 2. 3-Channel SHA Mode Timing ANALOG INPUTS PIXEL n tAD PIXEL (n+1) PIXEL (n+m) PIXEL (n+2) tAD tC1 tCRB tC2C1 CDSCLK1 tC2 tC1C2 CDSCLK2 tADC2 tC2AD ADCCLK OUTPUT DATA D11:D0 tADCLK PIXEL (n–4) tOD tADCLK PIXEL (n–3) PIXEL (n–2) PIXEL (n–1) PGAOUT_T PIXEL (n–1) PIXEL n PIXEL (n+1) PIXEL (n+2) PGAOUT_C Figure 3. 1-Channel CDS Mode Timing –4– REV. A AD9816 PIXEL (n+1) PIXEL n PIXEL (n+2) PIXEL (n+m) tAD ANALOG INPUTS tC2 tCRB CDSCLK2 tADC2 tC2AD ADCCLK OUTPUT DATA ,D11:D0. tADCLK tOD tADCLK PIXEL (n–4) PIXEL (n–3) PIXEL (n–2) PIXEL (n–1) PGAOUT_T PIXEL (n–1) PIXEL n PIXEL (n+1) PIXEL (n+2) PGAOUT_C Figure 4. 1-Channel SHA Mode Timing EFFECTIVE PIXELS OPTICAL BLACK OR DUMMY PIXELS ANALOG INPUTS CDSCLK1 CDSCLK2 ADCCLK Figure 5. Line Clamp Timing for 3-Channel CDS Mode ADCCLK tOD OUTPUT DATA ,D11:D0. tHZ OEB Figure 6. Output Enable Timing REV. A –5– tDV AD9816 ABSOLUTE MAXIMUM RATINGS* Parameter VIN, VREF PGA Outputs Clock Inputs AVDD DVDD DRVDD AVSS Digital Outputs Digital Inputs Junction Temperature Storage Temperature Lead Temperature (10 sec) With Respect To Min Max Units AVSS AVSS DVSS AVSS DVSS DRVSS DVSS DRVSS DVSS –0.3 –0.3 –0.3 –0.5 –0.5 –0.5 –0.3 –0.3 –0.3 AVDD + 0.3 AVDD + 0.3 DVDD + 0.3 +6.5 +6.5 +6.5 +0.3 DRVDD + 0.3 DVDD + 0.3 +150 +150 V V V V V V V V V °C °C +300 °C –65 *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 other conditions above those indicated in the operational sections of this specification is not implied. Exposure to absolute maximum ratings for extended periods may affect device reliability. ORDERING GUIDE Model AD9816JS AD9816JS-80010 AD9816-EB Temperature Range Package Description Package Option 0°C to +70°C 0°C to +70°C 44-Lead MQFP (Metric) Plastic Quad Flatpack 44-Lead MQFP (Metric) Plastic Quad Flatpack Evaluation Board S-44 S-44 CAUTION ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on the human body and test equipment and can discharge without detection. Although the AD9816 features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance degradation or loss of functionality. –6– WARNING! ESD SENSITIVE DEVICE REV. A AD9816 ADCCLK DVSS DVDD NC SCLK SDATA SLOAD DVSS DB0 DB1–DB5 DRVSS DRVDD DB6–DB10 DB11 OEB DI P P 42 NC No Connect. 43 PGAOUT_C AO PGA Output, Negative. This pin should be left unconnected except during evaluation. 44 PGAOUT_T AO PGA Output, Positive. This pin should be left unconnected except during evaluation. DI DIO DI P DO DO P P DO DO DI 32 DB5 31 DB4 CAPT 3 CAPT 4 30 DB3 CAPB 5 CAPB 6 VREF 7 29 DB2 AD9816 28 DB1 TOP VIEW (Not to Scale) 27 DB0 (LSB) 26 DVSS CML 8 VINR 9 25 SLOAD AVSS 10 24 SDATA 23 SCLK VING 11 12 13 14 15 16 17 18 19 20 21 22 NC = NO CONNECT NOTES See Applications Information for circuit configurations. TYPE: AI = Analog Input, AO = Analog Output, DI = Digital Input, DO = Digital Output, DIO = Digital Input/Output, P = Power. REV. A DRVDD 19 20 21 22 23 24 25 26 27 28–32 33 34 35–39 40 41 DB6 DI 33 DRVSS –7– NC CDSCLK2 DB7 18 DB8 DI PIN 1 IDENTIFIER AVDD 1 AVSS 2 DVSS DVDD CDSCLK1 44 43 42 41 40 39 38 37 36 35 34 ADCCLK 17 +5 V Analog Supply. Analog Ground. Reference Decoupling. Reference Decoupling. Internal Reference Output. Internal Bias Level. Analog Input, Red Channel. Analog Ground. Analog Input, Green Channel. Analog Ground. Analog Input, Blue Channel. Analog Ground. +5 V Analog Supply. Clamp bias level in CDS mode. Offset adjustment input in SHA mode. CDS Reset Level Sampling Clock. CDS Data Level Sampling Clock. A/D Converter Sampling Clock. Digital Ground. +5 V Digital Supply. No Connect. Clock Input for Serial Interface. Serial Data Input-Output. Load Pulse for Serial Interface. Digital Ground. Data Output (LSB). Data Outputs. Digital Driver Ground. Digital Driver Supply. Data Outputs. Data Output (MSB). Output Enable, Active Low. DB10 DB9 P P AO AO AO AO AI P AI P AI P P AI CDSCLK1 CDSCLK2 AVDD AVSS CAPT CAPB VREF CML VINR AVSS VING AVSS VINB AVSS AVDD OFFSET AVDD 1 2 3, 4 5, 6 7 8 9 10 11 12 13 14 15 16 OFFSET Description PGAOUT_C NC OEB DB11(MSB) Type VINB AVSS Pin Name AVSS Pin PIN CONFIGURATION PGAOUT_T PIN FUNCTION DESCRIPTIONS AD9816 DEFINITIONS OF SPECIFICATIONS FUNCTIONAL DESCRIPTION INTEGRAL NONLINEARITY (INL) The AD9816 can be operated in several different modes: 3-channel CDS mode, 3-channel SHA mode, 1-channel CDS mode, and 1-channel SHA mode. Each mode is selected by programming the Configuration Register through the serial interface. For more detail on CDS or SHA mode operation, see Circuit Descriptions section. Integral nonlinearity error refers to the deviation of each individual code from a line drawn from “zero scale” through “positive full scale.” The point used as “zero scale” occurs 1/2 LSB before the first code transition. “Positive full scale” is defined as a level 1 1/2 LSB beyond the last code transition. The deviation is measured from the middle of each particular code to the true straight line. 3-Channel CDS Mode In 3-channel CDS mode, the AD9816 simultaneously samples the red, green and blue input voltages from the CCD outputs. The sampling points for each Correlated Double Sampler (CDS) are controlled by CDSCLK1 and CDSCLK2. CDSCLK1’s falling edge clamps the reference level of the CCD waveform at the analog inputs of the AD9816. CDSCLK2’s falling edge samples the data level of the CCD waveform. Each CDS amplifier outputs the difference between the CCD reference and data levels. Next, the output voltage of each CDS amplifier is level-shifted by an Offset DAC. The voltages are then scaled by the three Programmable Gain Amplifiers before being multiplexed to the common 12-bit ADC. The ADC sequentially samples the PGA outputs on the falling edges of ADCCLK. DIFFERENTIAL NONLINEARITY (DNL) An ideal ADC exhibits code transitions which are exactly 1 LSB apart. DNL is the deviation from this ideal value. Thus every code must have a finite width. No missing codes guaranteed to 12-bit resolution indicates that all 4096 codes, respectively, must be present over all operating ranges. OFFSET ERROR The first ADC code transition should occur at a level 1/2 LSB above the nominal zero scale voltage. The offset error is the deviation of the actual first code transition level from the ideal level. Timing for this mode is shown in Figure 1, using a 2× master clock. Although it is not required, it is recommended that the falling edge of CDSCLK2 be aligned with the rising edge of ADCCLK. The rising edge of CDSCLK2 should not occur before the previous falling edge of ADCCLK, as shown by tADC2. The maximum allowable width of CDSCLK2 will be dependent on the ADCCLK period, and equal to one ADCCLK period minus 30 ns. The output data latency is three clock cycles. GAIN ERROR The last code transition should occur for an analog value 1 1/2 LSB below the nominal full scale voltage. Gain error is the deviation of the actual difference between first and last code transitions and the ideal difference between the first and last code transitions. The offset and gain values for the red, green, and blue channels are programmed using the serial interface. The order in which the channels are switched through the multiplexer is selected by programming the MUX register. The rising edge of CDSCLK2 always resets the multiplexer. TOTAL OUTPUT NOISE An ideal ADC outputs only one code value for a dc input voltage. A real converter has noise sources that will cause a spread of codes at the output for a dc input voltage. The total output noise is measured with a grounded input and is equal to the standard deviation of the histogram of output codes. 3-Channel SHA Mode CHANNEL-TO-CHANNEL CROSSTALK In an ideal three-channel system, the signal in one channel will not influence the signal level of another channel. The channelto-channel crosstalk specification is a measure of the change that occurs in one channel as the other two channels are varied. In the AD9816, one channel is grounded and the other two channels are exercised with full-scale input signals. The change in the output codes from the first channel is measured and compared with the result when all three channels are grounded. The difference is the channel-to-channel crosstalk, stated in LSBs. APERTURE DELAY The aperture delay is the time delay that occurs from when a sampling edge is applied to the AD9816 until the actual sample of the input signal is held. For CDSCLK1, the aperture delay represents the amount of time it takes for the clamp switch to open after CDSCLK1 transitions from high to low. For CDSCLK2, the aperture delay is the amount of time after the CDSCLK2 falling edge that the input signal is sampled. In 3-channel SHA mode, the AD9816 simultaneously samples the red, green, and blue input voltages. The sample-and-hold amplifier’s sampling point is controlled by CDSCLK2. CDSCLK2’s falling edge samples the input waveforms on each channel. The output voltages from the three SHAs are modified by the offset DACs and then scaled by the three PGAs. The outputs of the PGAs are then multiplexed through the 12-bit ADC. The ADC sequentially samples the PGA outputs on the falling edges of ADCCLK. The input signal is sampled with respect to the voltage applied to the OFFSET pin. With the OFFSET pin grounded, a zero volt input corresponds to the ADC’s zero scale output. The input clamp is disabled in this mode. However, the OFFSET pin may be used as a coarse offset adjust pin. A voltage applied to this pin will be subtracted from the voltages applied to the red, green and blue inputs in the first amplifier stage of the AD9816. For more information, see the Circuit Descriptions section. Timing for this mode is shown in Figure 2, using a 1× master clock. CDSCLK1 should be grounded in this mode. Although it is not required, it is recommended that the falling edge of CDSCLK2 be aligned with the rising edge of ADCCLK. The rising edge of CDSCLK2 should not occur before the previous falling edge of ADCCLK, as shown by tADC2. The maximum allowable width of CDSCLK2 will be dependent on the ADCCLK POWER SUPPLY REJECTION Power supply rejection specifies the maximum full-scale change that occurs from the initial value when the supplies are varied over the specified limits. –8– REV. A AD9816 period, and equal to one ADCCLK period minus 30 ns. The output data latency is three ADCCLK cycles. REGISTER OVERVIEW The serial interface is used to program the eight internal registers of the AD9816. The address bits A2–A0 determine the register in the AD9816 where serial data D7–D0 is written to or read from. The offset and gain values for the red, green and blue channels are programmed using the serial interface. The order in which the channels are switched through the multiplexer is selected by programming the MUX register. The rising edge of CDSCLK2 always resets the multiplexer. The Configuration Register controls the operating mode of the AD9816. Bits 7 (MSB), 6 and 0 are test mode bits and should always be set to zero. Bit 5 is set high to enable the CDS mode. Setting this bit low enables the SHA mode. Set Bit 4 high to enable the 3 V input span. Set Bit 3 high to enable the 1.5 V span. Bits 2 and 1 set the channel mode. Bit 2 enables 3-channel simultaneous sampling. Bit 1 enables single channel mode, with the appropriate channel set in the MUX Register. At power-on, this register defaults to 3-channel CDS mode with a 3 V input span, as shown in Table I. 1-Channel CDS Mode This mode operates in the same way as the 3-channel CDS mode. The difference is that the multiplexer remains fixed in this mode, so only the channel specified in the MUX register is processed. Because the AD9816 is still sampling all three channels, the unused inputs should be grounded through 1200 pF capacitors. Timing for this mode is shown in Figure 3, using a 3× master clock. Although it is not required, it is recommended that the falling edge of CDSCLK2 be aligned with the rising edge of ADCCLK. 7 6 5 4 3 2 1 0 TEST MODE (LSB) 1-CHANNEL MODE 1-Channel SHA Mode 3-CHANNEL MODE This mode operates the same way as the 3-channel SHA mode, except that the multiplexer remains stationary. Only the channel specified in the MUX register is processed. Because the AD9816 is still sampling all three channels, the unused inputs should be grounded. 1.5 V INPUT SPAN 3 V INPUT SPAN CDS ENABLE TEST MODE TEST MODE (MSB) Figure 7. Configuration Register The input signal is sampled with respect to the voltage applied to the OFFSET pin. With the OFFSET pin grounded, a zero volt input corresponds to the ADC’s zero scale output. The input clamp is disabled in this mode. However, the OFFSET pin may be used as a coarse offset adjust pin. A voltage applied to this pin will be subtracted from the voltages applied to the red, green and blue inputs in the first amplifier stage of the AD9816. For more information, see the Circuit Descriptions section. The MUX Register determines the order of channels that the multiplexer will switch to in the different modes of operation. Bit 7 and Bit 1 are test modes and should be set to zero. Bit 0 is a test mode bit and should be set high. In 3-channel mode, Table II shows how to set the order in which the channels are converted. The multiplexer is always reset on the rising edge of CDSCLK2. In 1-channel mode, the multiplexer is stationary, and only converts the channel selected in Table III. At poweron, this register defaults to 3-channel RGB mode. Timing for this mode is shown in Figure 4, using a 1× master clock. CDSCLK1 should be grounded in this mode of operation. Although it is not required, it is recommended that the falling edge of CDSCLK2 be aligned with the rising edge of ADCCLK. 7 6 5 4 3 2 1 0 TEST MODE (LSB) TEST MODE 1-CHANNEL RED 1-CHANNEL GREEN 1-CHANNEL BLUE 3-CHANNEL BIT 0 3-CHANNEL BIT 1 TEST MODE (MSB) Figure 8. MUX Register Table I. Register Map REV. A A2 A1 A0 Register Power-On Default Value 0 0 0 0 1 1 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 Configuration Register MUX Register Red PGA Register Green PGA Register Blue PGA Register Red Offset Register Green Offset Register Blue Offset Register 0 0 1 1 0 1 0 0 (LSB) 0 0 1 0 0 0 0 1 (LSB) Undetermined Undetermined Undetermined Undetermined Undetermined Undetermined –9– AD9816 gain. The gain of the PGA increases linearly as the gain word increases, and can be calculated by the following equation: Table II. 3-Channel Selection MUX Register Bits 6 5 Channel Sequence 0 1 Red, Green, Blue Blue, Green, Red 1 0 PGA Gain = 1 + (Gain Code/51.2) where Gain Code varies from 0 to 255. For more information, refer to the Circuit Descriptions section. 7 6 5 4 3 2 1 0 Table III. 1-Channel Selection D0 (LSB) MUX Register Bits 4 3 2 Channel 0 0 1 Red Green Blue D1 D2 D3 0 1 0 1 0 0 D4 D5 D6 D7 (MSB) The offset is variable from –100 mV to +100 mV, and is applied at the output of the CDS, before the PGA. The resolution is 8 bits, and a sign magnitude coding scheme is used. Table IV shows the offset voltage that corresponds to the register value. 7 6 5 4 3 2 1 0 D0 (LSB) D1 D2 D3 D4 D5 D6 D7 (MSB) Figure 9. Offset Registers for Red, Green and Blue Channels Table IV. Offset Adjustment Offset Register Offset Voltage 0111 1111 (LSB) . . . 0000 0001 0000 0000 1000 0000 1000 0001 . . . 1111 1111 +100 mV . . . +0.8 mV 0.0 mV 0.0 mV –0.8 mV . . . –100 mV Figure 10. PGA Registers for Red, Green and Blue Channels SERIAL TIMING The 3-wire serial interface timing is shown below. To write to the AD9816, SLOAD is first taken low. Next, a total of 16 bits are sent to SDATA, which get latched into the AD9816 on the rising edges of SCLK. Additional SCLK pulses will be ignored. The first bit, R/W, should be low to specify a write operation. The next three bits, A2–A0, are the address bits to specify the destination register for the data word D7–D0. After all 16 bits have been clocked, SLOAD is taken high, which internally latches the data to the appropriate register. The read operation also starts by taking SLOAD low. First, a one is written to R/W, to specify a read operation. Next, the three Address Bits A2–A0 are written to specify the register that will be read. On the 8th SCLK falling edge, SDATA will begin to output the information from the desired register. After all eight data bits have been read, SLOAD is taken back high. R/Wb A2 A1 A0 SDATA tDH D7 D6 D5 D4 D3 D2 D1 D0 tDS SCLK tLH tLS SLOAD Figure 11. Write Operation Timing R/Wb A2 SDATA tDH A1 A0 D7 D6 D5 D4 D3 D2 D1 D0 tDS tRDV SCLK The PGA is used for correcting color imbalance and for fine adjustment of the input span before the ADC. Gain is variable from 1× to 6× (0 dB to 15.5 dB) with 8-bit resolution. An all “zeros” word (00 . . . 0) corresponds to the minimum gain, and an all “ones” word (11 . . . 1) corresponds to the maximum tLS tLH SLOAD Figure 12. Read Operation Timing –10– REV. A AD9816 CIRCUIT DESCRIPTIONS Analog Input Configuration for CDS and SHA Mode CDS Mode Operation Figure 13 shows the equivalent input circuit for the CDS mode of operation. The CCD signal is connected to the AD9816’s analog inputs through a coupling capacitor CIN. The CCD reference level is clamped during the CDSCLK1 pulse, when the clamp switch closes and connects the externally-generated 3 V bias to the analog input. After the clamp switch opens (CDSCLK1 low), the CCD data level will be level shifted by the voltage held across CIN, and the SHA will sample the input signal when the CDSCLK2 pulse goes low (see Figures 1 and 3 for CDS mode timing). In this sampling technique, the CDS function is effectively performed across the input capacitor, CIN. This CDS method has two additional considerations. First, the CCD signal cannot be dc-coupled into the AD9816, because the input capacitor is required. Second, the input clamp of the AD9816 is operating as a pixel clamp, and must be asserted on every pixel for true CDS operation. If line clamp operation is desired, CDSCLK1 may be used at the start of each line to set the proper dc voltage on CIN. Then, during the effective pixels of each line, CDSCLK1 can be held low while CDSCLK2 samples the data levels of each pixel. Figure 5 shows the timing for line clamp operation. AD9816 CIN RS CCD SIGNAL VING IBIAS SHA 11 CSTRAY BUFFER CLAMP SWITCH +5V 1mF 0.1mF 1.5kV Input Capacitor CIN The recommended value for CIN is 1200 pF. This value has been selected to provide the best overall performance when considering three factors: input attenuation, linearity and signal droop. The value of CIN may be optimized for a particular application if these three factors are understood. 1. Attenuation (Gain Error) The input voltage will be attenuated by the interaction of CIN and CSTRAY. CSTRAY is less than 10 pF, which results in an attenuation of about 0.8% when CIN is 1200 pF. The gain error will increase accordingly as the value of CIN is decreased. 2. Linearity The input capacitance of the AD9816 is shown in Figure 8 as CSTRAY. A small portion of this capacitance is junction capacitance, which will vary nonlinearly as the input voltage to the AD9816 changes. When the input voltage is attenuated by the combination of CIN and CSTRAY, there will be a small nonlinear component caused by the input junction capacitance. The magnitude of the junction capacitance will cause a 1 LSB (0.024%) nonlinearity over the 3 V input range when a 1200 pF CIN is used. This nonlinearity will increase if a smaller CIN is used. 3. Droop The input bias current of the AD9816 is typically 10 nA and is constant regardless of the AD9816’s input voltage. The droop of the voltage across CIN can be calculated with the following equation: 1.0kV 16 of the AD9816 can also handle an input signal down to AVSS – 0.3 V without any saturation recovery issues. Although an input level below zero volts will be clipped to the ADC’s fullscale output code, the input stage can respond quickly enough to accurately process the next pixel that falls into the linear input range. Any signals below AVSS – 0.3 V will turn on the input protection diodes, and recovery from the saturated condition may take up to several milliseconds. 3V OFFSET 17 18 CDSCLK1 CDSCLK2 dV = Figure 13. CDS Mode Input Circuit (All Channels Identical) Input Signal Range for CDS Mode An input dc bias level of 3 V allows a maximum 3 V p-p signal swing from the CCD. Figure 14 shows a typical full-scale input waveform to the AD9816, illustrating the allowable input range. With a reference level of 3 V, the AD9816 can tolerate up to 2 V of reset feedthrough above the reference level. The inputs i BIAS ×(t) CIN where t is the time between clamp intervals. Between the adjacent pixels of a scanned line, this droop will be insignificant. Between scanned lines, a 1 ms delay will produce a droop of about 10 mV, which can be easily clamped on the first pixel of the next line. If the value of CIN is reduced, the droop will increase accordingly. 5V MAX RESET FEEDTHROUGH 3V REFERENCE LEVEL (SET BY INPUT CLAMP) MAX PEAK-PEAK SIGNAL 0V MAX DATA LEVEL –0.3V MAX SATURATED DATA LEVEL Figure 14. CCD Input Signal Clamped to 3 V REV. A –11– AD9816 Line Clamp Programmable Gain Amplifiers If a line clamp technique is implemented (see Figure 5 for timing), the value of CIN should be increased to more than 1200 pF. The main requirement for line clamp is to keep the signal droop below 1 LSB across a scanned line. For example, if a CCD with 5400 effective pixels is clocked at 2 MHz, then t = 2.7 ms. One LSB at 12 bits with a 3 V full scale is 732 µV. Rearranging the above droop equation: The AD9816 has three programmable amplifiers, one for each channel. The gain is variable from 1 V/V (0 dB) to 5.98 V/V (15.5 dB) in 256 increments. Figure 16 shows the PGA gain transfer function. The gain of the PGA can be calculated according to the equation: CMIN = Gain Code PGA Gain = 1+ 51.2 i BIAS ×t dV 6 In this case, CMIN = 37 nF, and a convenient standard value of 0.047 µF will be adequate. 5 PGA GAIN – V/V SHA Mode Operation When the AD9816 is configured for SHA mode operation, the OFFSET pin functions as an offset adjustment input. Figure 15 shows a simplified diagram of the AD9816’s inputs when SHA mode is selected. A positive dc voltage may be applied to OFFSET which will be subtracted from all three input channels in the input stage of the AD9816. The maximum input voltage to the analog input pins or the OFFSET pin in SHA mode is 3 V. The OFFSET feature is provided to allow coarse offset adjustment of the input signal. If the signal is sampled with respect to ground, any positive offset on the input signal will subtract from the dynamic range of the ADC. For example, an input signal that spans from 1.5 V to 2.5 V cannot utilize all of the available dynamic range, using either the 1.5 V or 3 V span. However, by applying a dc value of 1.5 V to the OFFSET pin, the input signal will be level-shifted down to 0 V to 1 V. This would allow the use of the 3 V span and a PGA gain of three to use the entire ADC dynamic range. If no dc offset adjustment is desired, the OFFSET pin should be grounded. The input signal will be sampled with respect to ground. AD9816 4 3 2 1 0 51 102 153 204 GAIN REGISTER CODE – Decimal 255 Figure 16. PGA Gain Transfer Function The analog outputs of the three PGAs are multiplexed to the input of the 12-bit ADC. The differential output of the MUX is also buffered and externally available at Pins 43 and 44 (PGAOUT_C and PGAOUT_T, respectively). The timing diagrams, Figures 1 through 4, show the timing relationships between the analog inputs, CDSCLK2, ADCCLK, and PGAOUT_T and PGAOUT_C. The CDSCLK2 pulse resets the outputs of all three PGAs to an internal bias level. The first rising edge of ADCCLK after the rising edge of CDSCLK2 will switch the MUX to the red PGA output. The second ADCCLK rising edge switches the MUX to the green PGA output, and the third rising edge switches the MUX to the blue PGA output. PGA Outputs VINR SHA BUFFER VING SHA BUFFER VINB SHA BUFFER OFFSET 12kV CDSCLK1 CDSCLK2 Figure 15. SHA Mode Input Circuit The PGAOUT_T and PGAOUT_C signals represent the differential input to the ADC, and are complementary. Both signals will reset to 3.5 V while CDSCLK2 is high. The voltage swing of each output is equal to one-half of the ADC’s full-scale voltage, centered at 3.5 V. Table V shows the relationship between the analog input voltage, the PGA output voltage and the ADC input voltage. Figure 18 shows the PGA output voltages for three different color pixel amplitudes. In this example, the red pixel has the largest amplitude, and the blue pixel has the smallest amplitude. Because the PGAOUT_T and PGAOUT_C outputs are internally buffered by source followers, they are not an exact representation of the differential ADC input signal. PGAOUT_T and PGAOUT_C should only be used during evaluation; performance of the AD9816 is only guaranteed with these two pins unconnected. –12– REV. A AD9816 PGAOUT_T Analog-to-Digital Converter PGAOUT_C The AD9816 uses a high speed 12-bit ADC core. This CMOS converter is designed to run at 6 MSPS with good linearity and noise performance. Figure 19 shows the INL and DNL performance of a typical AD9816 device, running at 6 MHz in 3-channel CDS mode using the timing shown in Figure 1. The following timing parameters were used: tCRA = 500 ns, tADCLK = 83 ns, tC1 = 20 ns, tC1C2 = 170 ns, tC1 = 80 ns, tADC2 = 3 ns, tC2AD = 83 ns, and tC2C1 = 230 ns. RED PGA 3:1 DIFF MUX GREEN 12-BIT ADC PGA SELECT 2 BLUE The digital outputs of the AD9816 follow a straight binary coding scheme. Table VI shows the digital output coding for the 3 V input span. ADCCLK MUX CONTROL CDSCLK2 PGA 0.2 Figure 17. PGA/MUX Circuit Configuration 0.0 –0.2 MAX INL 0.18 MIN INL –1.46 –0.4 INL PIXEL n BLUE –0.6 –0.8 –1.2 ANALOG INPUTS GREEN –1.4 –1.6 0 RED 400 800 1200 1600 2000 2400 2800 3200 3600 4095 1.5 MAX DNL 0.31 MIN DNL –0.33 1.0 DNL CDSCLK2 ADCCLK 0.5 0.0 –0.5 RESET RED(n) PGAOUT_T RESET 4.25V 3.5V GREEN(n) PGAOUT_C PGAOUT_T PGAOUT_C Differential ADC Input 2.75 3.50 4.25 3.125 3.50 3.875 4.25 3.50 2.75 3.875 3.50 3.125 1.5 0.0 +1.5 0.75 0.0 +0.75 1600 2000 2400 2800 3200 3600 4095 Input Voltage1 Digital Outputs 3.0 – 1 LSB 3.0 – 2 LSB 0.0 + 1 LSB 0.0 1111 1111 1111 1111 1111 1110 0000 0000 0001 0000 0000 0000 NOTE 1 Analog input voltage in CDS mode is the difference between the CCD’s reference and data levels. NOTES 1 Analog input voltage in CDS mode is the difference between the CCD’s reference and data levels. 2 3.0 V Input Range. 3 1.5 V Input Range. REV. A 1200 Table VI. Digital Output Format Table V. Voltage Swing of PGA Outputs 0.00 1.502 3.002 0.003 0.753 1.503 800 Figure 19. Typical Linearity Performance Figure 18. PGA Output Voltages (ADC Input Range = 3 V) 2 400 2.75V BLUE(n) BLUE(n–1) Analog Input Voltage1 –1.0 0 GREEN(n–1) –13– AD9816 APPLICATIONS INFORMATION CDS Mode Circuit 10mF + 0.1mF AVDD 2 AVSS 3 CAPT 4 CAPT 0.1mF AD9816 6 CAPB 7 VREF 21 DVDD 1200pF 20 DVSS RED_IN GREEN_IN 19 ADCCLK 1200pF 15 AVDD 11 VING 18 CDSCLK2 10 AVSS 16 OFFSET 1200pF 17 CDSCLK1 9 VINR 1.0mF 14 AVSS 0.1mF 13 VINB + 8 CML 12 AVSS 10mF 34 DB7 DB6 36 35 DB9 DB8 37 38 39 OEB DB11(MSB) DB10 42 5 CAPB 0.01mF DRVSS 33 DB5 32 DB4 DB3 DB2 DB1 DB0 (LSB) DVSS SLOAD SDATA SCLK 22 NC 0.1mF 1 43 0.01mF VDD PGAOUT_T PGAOUT_C NC OEB DB11 (MSB) DB10 DB9 DB8 DB7 DB6 DRVDD 0.1mF 44 VDD 41 The recommended circuit configuration for CDS mode operation is shown in Figure 20. The input coupling capacitor value of 1200 pF is recommended, but this value may be adjusted to suit a particular application (see Circuit Descriptions). A single ground plane is recommended for the AD9816. A separate power supply may be used for DRVDD, the digital driver supply, but this 40 supply pin should still be decoupled to the same ground plane as the rest of the AD9816. The loading of the digital outputs should be minimized, either by using short traces to the digital ASIC, or by using external digital buffers. All 0.01 µF and 0.1 µF decoupling capacitors should be located as close as possible to the AD9816 pins. Also, the 1200 pF input capacitors should be located close the AD9816’s analog input pins. 0.1mF DB5 DB4 DB3 DB2 DB1 DB0 (LSB) 31 30 29 28 27 26 25 24 SLOAD SDATA SCLK 23 NC = NO CONNECT BLUE_IN VDD 0.01mF VDD 0.1mF 0.1mF 0.01mF ADCCLK CDSCLK2 CDSCLK1 1.0mF 0.1mF 1kV VDD 1.5kV Figure 20. Recommended Circuit for CDS Mode –14– REV. A AD9816 10mF + 0.1mF AVDD 2 AVSS 3 CAPT 4 CAPT AD9816 6 CAPB 7 VREF 21 DVDD 15 AVDD 14 AVSS RED_IN GREEN_IN 13 VINB 12 AVSS 11 VING 16 OFFSET 10 AVSS 20 DVSS 9 VINR 1.0mF 19 ADCCLK 0.1mF 18 CDSCLK2 + 8 CML 17 CDSCLK1 10mF 34 DB7 DB6 36 35 DB9 DB8 37 38 39 OEB 42 DB11(MSB) DB10 DB4 DB3 DB2 DB1 DB0 (LSB) DVSS SLOAD SDATA SCLK 5 CAPB 0.1mF 0.01mF DRVSS 33 DB5 32 22 NC 0.1mF 1 43 0.01mF 0.1mF VDD PGAOUT_T PGAOUT_C NC OEB DB11 (MSB) DB10 DB9 DB8 DB7 DB6 DRVDD 44 VDD 41 The circuit configuration for SHA mode is identical to CDS mode except for two differences: the analog inputs should be dc-coupled, and the OFFSET pin is tied to ground or a desired dc voltage (see Circuit Descriptions). In CIS applications, the 40 reference black level of the CIS can be connected to the OFFSET pin, to remove the dc offset. Removing the coarse offset of the CIS signal will allow the dynamic range of the AD9816 to be maximized. SHA Mode Circuit 0.1mF 28 DB2 DB1 27 DB0 (LSB) 29 26 25 SLOAD SDATA SCLK 24 23 NC = NO CONNECT ADCCLK CDSCLK2 CDSCLK1 R1 0.1mF 0.1mF 0.01mF VDD OPTIONAL DC OFFSET R2 GROUND-REFERENCED SAMPLING Figure 21. Recommended Circuit for SHA Mode REV. A DB4 DB3 30 VDD VDD 1.0mF DB5 31 BLUE_IN 0.01mF –15– 0.1mF AD9816 OUTLINE DIMENSIONS Dimensions shown in inches and (mm). C3324a–0–10/98 44-Lead MQFP (S-44) 0.529 (13.45) 0.510 (12.95) 0.096 (2.45) MAX 0.398 (10.1) 0.390(9.90) 0.041 (1.03) 0.029 (0.73) 44 34 1 33 SEATING PLANE 0.333 (8.45) 0.327 (8.3) TOP VIEW (PINS DOWN) 11 0.01 (0.25) MIN 23 12 0.083 (2.1) 0.077 (1.95) 0.031 (0.80) BSC 0.018 (0.45) 0.012 (0.30) PRINTED IN U.S.A. 0.009 (0.23) 0.005 (0.13) 22 –16– REV. A