IBIS4-6600 CYII4SM6600AB
6.6 MP CMOS Image Sensor
Cypress Semiconductor Corporation • 198 Champion Court • San Jose,CA 95134-1709 • 408-943-2600
Document Number: 001-02366 Rev. *G Revised April 1,2010
Features
■2210 (H) x 3002 (V) Active Pixels
■3.5 µm X 3.5 µm Square Pixels
■1 inch Optical Format
■Monochrome or Color Digital Output
■Frame Rate:
❐5 fps (2210 x 3002)
❐89 fps (640 x 480)
■High Dynamic Range Modes: Double Slope, Non Destructive
Read out (NDR)
■Electronic Rolling Shutter
■Master Clock: 40 MPS/40 MHz
■Limited Supplies: 2.5V and 3.3V
■-30°C to +65°C Operational Temperature Range
■68-Pin LCC Package
■Power Dissipation: 0.19W
Applications
■Machine vision
■Biometry
■Document Scanning
Description
The IBIS4-6600 is a solid-state CMOS image sensor that
integrates complete analog image acquisition, and a digitizer
and digital signal processing system on a single chip. This image
sensor has a resolution of 6.6 MPixel with 2210 x 3002 active
pixels. The image size is fully programmable for user-defined
windows. The pixels are on a 3.5 µm pitch. This sensor is
available in a Monochrome version or Bayer (RGB) patterned
color filter array.
The user programmable row and column start and stop positions
enable windowing down to 2x1 pixel window for digital zoom.
Sub sampling reduces resolution while maintaining the constant
field of view. The analog video output of the pixel array is
processed by an on-chip analog signal pipeline. Double
Sampling (DS) eliminates the fixed pattern noise.
The programmable gain and offset amplifier maps the signal
swing to the ADC input range. A 10-bit ADC converts the analog
data to a 10-bit digital word stream. The sensor uses a three-wire
Serial-Parallel (SPI) interface. It operates with a single 2.5V
power supply and requires only one master clock for operation
up to 40 MHz. It is housed in a 68-pin ceramic LCC package.
This data sheet enables you to develop a camera system, based
on the described timing and interfacing given in the following
sections.
Figure 1. IBIS4-6600 Image Sensor
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 2 of 34
Ordering Information
Specifications
Marketing Part Number Description Package
CYII4SM6600AB-QDC Mono with Glass
68-pin LCC
CYII4SM6600AB-QWC Mono without Glass
CYII4SE6600AB-QDC Color micro lens with Glass
CYII4SE6600AB-QFCH Color micro lens with IR Coating, High Grade
CYII4SM6600-EVAL Mono Demo Kit Demo Kit
CYII4SC6600-EVAL Color Demo Kit
General Specifications
Parameter Specification Remarks
Pixel Architecture 3T-Pixel
Pixel Size 3.5 µm x 3.5 µm The resolution and pixel size results in a 7.74 mm x 10.51 mm
optical active area.
Resolution 2210 x 3002
Pixel Rate 40 MHz Using a 40 MHz system clock and 1 or 2 parallel outputs
Shutter Type Electronic Rolling Shutter
Full Frame Rate 5 frames/second Increases with ROI read out and/or subsampling
Electro Optical Specifications
Parameter Specification Remarks
FPN (local) <0.20% RMS% of saturation signal
PRNU (local) <1.5% RMS of signal level
Conversion Gain Conversion Gain At output (measured)
Output Signal Amplitude 0.6V At nominal conditions
Saturation Charge 21.500 e-
Sensitivity (peak) 411 V.m2/W.s
4.83 V/lux.s
At 650 nm
(85 lux = 1 W/m2)
Sensitivity (visible) 328 V.m2/W.s
2.01 V/lux.s
400-700 nm
(163 lux = 1 W/m2)
Peak QE * FF
Peak Spectral Response
25%
0.13 A/W
Average QE*FF = 22% (visible range)
Average SR*FF = 0.1 A/W (visible range)
See the section Spectral Response Curve on page 3.
Fill Factor 35% Light sensitive part of pixel (measured)
Dark Current 3.37 mV/s
78 e-/s
Typical value of average dark current of the whole pixel array
(at 21°C)
Dark Signal Non Uniformity 8.28 mV/s
191 e-/s
Dark current RMS value (at 21°C)
Temporal Noise 24 RMS e- Measured at digital output (in the dark)
S/N Ratio 895:1 (59 dB) Measured at digital output (in the dark)
Spectral Sensitivity Range 400 - 1000 nm
Optical Cross Talk 15%
4%
To the first neighboring pixel
To the second neighboring pixel
Power Dissipation 190 mW Typical (including ADCs)
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 3 of 34
Spectral Response Curve
Figure 2. Spectral Response Curve
Figure 2 shows the characteristics of the spectral response. The curve is measured directly on the pixels. It includes the effects of
nonsensitive areas in the pixel, for example, interconnection lines. The sensor is light sensitive between 400 and 1000 nm. The peak
QE * FF is 25% approximately 650 nm. In view of a fill factor of 35%, the QE is close to 70% between 500 and 700 nm.
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
400 500 600 700 800 900 1000
Wavelenght [nm]
Spectral response [A/W]
QE 10%
QE 20%
QE 30%
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 4 of 34
Electro Voltaic Response Curve
Figure 3. Electro Voltaic Response Curve
Figure 3 shows the pixel response curve in linear response mode. This curve is the relation between the electrons detected in the
pixel and the output signal. The resulting voltage-electron curve is independent of any parameters, for example, integration time. The
voltage to electrons conversion gain is 43 µV/electron.
Table 1. Features and General Specifications
Feature Specification/Description
Electronic shutter type Rolling shutter
Integration time control 60 µs - 1/frame period
Windowing (ROI) Randomly programmable ROI read out
Sub Sampling Modes Several sub sample modes can be programmed (refer Table 7 on page 12)
Extended Dynamic Range Dual slope (up to 90 dB optical dynamic range) and nondestructive read out mode
Analog Output The output rate of 40 Mpixels/s can be achieved with two analog outputs, each working
at 20 Mpixel/s
Digital Output Two on-chip 10-bit ADCs at 20 Msamples/s are multiplexed to one digital 10-bit output
at 40 Msamples/s
Supply Voltage VDD Nominal 2.5V (some supplies require 3.3V for extended dynamic range)
Logic Levels 2.5V
Interface Serial-to Parallel Interface (SPI)
Package 68-pins LCC
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 5000 10000 15000 20000 25000
# electrons
Output swing [V]
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 5 of 34
Electrical Specifications
Absolute Maximum Ratings
Exceeding maximum ratings may shorten the useful life of the device. User guidelines are not tested. Stresses beyond those listed
under Ta ble 2 may cause permanent damage to the device. These are stress ratings only and the functional operation of the device
at these or any other conditions beyond those indicated in the operational sections are not implied.
All parameters are characterized for DC conditions after thermal equilibrium is established. Unused inputs must always be tied to an
appropriate logic level, for example, VDD or GND.
This device contains circuitry to protect the inputs against damage caused by high static voltages or electric fields. However, you must
take normal precautions to avoid applying voltages higher than the maximum rated voltages to this high impedance circuit.
Table 2. Absolute Maximum Ratings
Symbol Parameter Value Unit
VDD[1] DC Supply Voltage –0.5 to 3.3 V
VIN DC Input Voltage –0.5 to (VDD + 0.5) V
VOUT DC Output Voltage –0.5 to (VDD + 0.5) V
IIO DC Current Drain Per Pin; Any Single Input or Output ± 50 mA
TLLead Temperature (5 seconds soldering) 350 °C
TST Storage Temperature –30 to +85 °C
H Humidity (Relative) 85% at 85 °C
ESD ESD Susceptibility 2000 V
Recommended Operating Conditions
Symbol Parameter Min Typ Max Unit
VDD DC Supply Voltage 2.5 2.5 3.3 V
TACommercial Operating Temperature –30 24 +65 °C
Note
1. VDD = VDDD = VDDA (VDDD is supply to digital circuit, VDDA to analog circuit).
DC Electrical Conditions
Symbol Characteristic Condition Min Max Unit
VIH Input High Voltage VDD-0.5 V
VIL Input Low Voltage –0.6 0.6 V
IIN Input Leakage Current VIN = VDD or GND –10 +10 µA
VOH Output High Voltage VDD=min; IOH= –100 mA VDD-0.5 V
VOL Output Low Voltage VDD=min; IOH= 100 mA 0.5 V
IDD Operating Current System clock <= 40 MHz 70 80 mA
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 6 of 34
Sensor Architecture and Operation
Floor Plan
Figure 4. Floor Plan
Figure 4 shows the architecture of the designed image sensor. It
consists of the pixel array, shift registers for the readout in x and
y direction, parallel analog output amplifiers, and column
amplifiers that correct for the fixed pattern noise caused by
threshold voltage nonuniformities. Reading out the pixel array
starts by applying a y clock pulse to select a new row, followed
by a calibration sequence to calibrate the column amplifiers (row
blanking time). Depending on external bias resistors and timing,
typically this sequence takes about seven seconds every line
(baseline). This sequence is necessary to remove the Fixed
Pattern Noise of the pixel and of the column amplifiers
themselves (by a Double Sampling technique). Pixels can also
be read out in a nondestructive manner.
Two DACs are added to make the offset level of the pixel values
adjustable and equal for the two output buses. A third DAC is
used to connect the buses to a stable voltage during the row
blanking period, or reset the buses continuously in case of a
nondestructive readout.
Two 10-bit ADCs running at 20 Msamples/s convert the analog
pixel values. The digital outputs are multiplexed to one digital
10-bit output at 40 Msamples/s. Note that these blocks are
electrically completely isolated from the sensor part, except for
the multiplexer, for which the settings are uploaded through the
shared address and data bus.
The x and y shift registers have a programmable starting point.
The possibilities of the starting point are limited because of
limitations imposed by subsampling requirements. The start
address is uploaded through the serial to parallel interface.
Most of the signals for the image core shown in Figure 4 are
generated on-chip by the sequencer. This sequencer also allows
running the sensor in basic modes, not fully autonomous.
res et
select
analog output
(2)
eos_yl
eos_yr
tri l
tri r
DAC in
ADC, 10 bit
ADC, 10 bit
sequencer
IMAG E COR E
SENSOR
SPI
pixel array
2210 x 3002
(excl. dark +
dummy pixels )
addressable x-shiftregister + sub-sampling
addres s able y-s hift register + sub-sampling
column amplifiers
res et and s elect drivers
res et and select drivers
clk_y
sync_yl
clk_x
sync_x
Pixel (0,0)
DAC
addres s able y-s hift register + sub-sampling
clk_y
sync_y
r
Dig. logic Dig. logic
addres s &
data bus
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 7 of 34
Pixel
Architecture
The pixel architecture is the classic three-transistor pixel, as
shown in Figure 5 The pixel is implemented using the high fill
factor technique patented by FillFactory (US patent No.
6,225,670 and others).
Figure 5. 3T Pixel Architecture
FPN and PRNU
Fixed Pattern Noise correction is done on-chip. Raw images
taken by the sensor typically feature a residual (local) FPN of
0.35% RMS of the saturation voltage.
The Photo Response Non Uniformity (PRNU), caused by the
mismatch of photodiode node capacitances, is not corrected on
chip. Measurements indicate that the typical PRNU is about 1.5%
RMS of the signal level.
Color Filter Array
The IBIS4-6600 can also be processed with a Bayer RGB color
pattern. Pixel (0,0) has a green filter and is situated on a
green-red row. Green1 and Green2 are separately processed
color filters and have a different spectral response. Green1 pixels
are located on a blue-green row, and green2 pixels are located
on a green-red row.
Figure 6. RGB Bayer Alignment
Figure 7 shows the response of the color filter array as function
of the wavelength.
Figure 7. Typical Response Curve of the RGB Filters
selec
Vdd
M1
M2
M3
reset
output
(column)
select
0.000
0.020
0.040
0.060
0.080
0.100
0.120
0.140
400 500 600 700 800 900 1000
wavelength [nm]
blue
green 1
green 2
red
mono
spectral response [A/W]
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 8 of 34
Figure 8. Relative Response Graph
Dark and Dummy Pixels
Figure 9 shows a plan of the pixel array. The sensor is designed in portrait orientation. A ring of dummy pixels surrounds the active
pixels. Black pixels are implemented as "optical" black pixels.
Figure 9. Floor Plan Pixel Array
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
400 500 600 700 800 900 1000
wavelength (nm)
blue
green 1
green 2
red
relative response
Dummy ring of pixels ,
surrounding complete pixel
array. not read
R ing of 2 dummy pixels,
illuminated, readable
array of active pixels, read
3002x 2210
R ing of dummy pixels,
covered with bla ck layer,
readable
2222
3014
2210
3002
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 9 of 34
Pixel Rate
The pixel rate for this sensor is high enough to support a frame
rate greater than 75 Hz for a window size of 640 x 480 pixels
(VGA format), and 23 pixels over scan in both directions. Taking
into account a row blanking time of 7.2 µs (as baseline, refer the
following calculations), this requires a minimum pixel rate of
approximately 40 MHz. The final bandwidth of the column
amplifiers, output stage, and more is determined by external bias
resistors. Taking into account a pixel rate of 40 MHz, a full frame
rate of a little more than 5 frames/s is obtained.
The frame period of the IBIS4-6600 sensor is calculated as:
=> Frame period = (Nr. Lines * (RBT + pixel period * Nr. Pixels))
In this equation:
Nr. Lines: Number of Lines read out each frame (Y)
Nr. Pixels: Number of pixels read out each line (X)
RBT: Row Blanking Time = 7.2 µs (typical)
Pixel period: 1/40 MHz = 25 ns
Example: Read out time of the full resolution at nominal speed
(40 MHz pixel rate):
=> Frame period = (3002 * (7.2 µs + 25 ns * 2210)) = 187.5 ms
=> 5.33 fps.
Region of Interest (ROI) Read Out
Windowing is easily achieved by uploading the starting point of
the x and y-shift registers in the sensor registers (refer Table 10
on page 17). This downloaded starting point initiates the shift
register in the x and y-direction, triggered by the Y_START
(initiates the Y-shift register) and the Y_CLK (initiates the X-shift
register) pulse. The minimum step size for the x-address is 24
(only even start addresses can be chosen) and 1 for the
Y-address (every line can be addressed). The frame rate
increases in an almost linear manner when fewer pixels are read
out. Table 3 lists the achievable frame rates with ROI read out.
Output Amplifier
The output amplifier subtracts the reset and signal voltages from
each other to cancel FPN as much as possible (shown in
Figure 10). The DAC that is used for offset adjustment consists
of two DACs. One DAC is used for the main offset (DAC_raw).
The other enables fine tuning to compensate the offset difference
between the signal paths arriving at the two amplifiers A1 and A2
(DAC_fine). With the analog multiplexer, the signals S1 and S2
from the two buses can be combined to one pixel output at full
pixel rate (40 MHz). However, the two analog signals S1 and S2
can also be available on two separate output pins to allow a
higher pixel rate.
The third DAC (DAC_dark) puts its value on the buses during the
calibration of the output amplifier. In case of nondestructive
readout (no double sampling), bus1_R and bus2_R are
continuously connected to the output of the DAC_fine to provide
a reference for the signals on bus1_S and bus2_S.
The complete output amplifier can be put in standby by setting
the corresponding bit in the AMPLIFIER register.
Figure 10. Output Amplifier Architecture
Table 3. Frame Rate vs. Resolution
Image Resolution (Y*X) Frame Rate [frames/s] Frame Readout Time [ms] Comment
3002 x 2210 5 187.5 Full resolution
1501 x 1104 14 67 ROI read out
640 x 480 89 11 11
+
+
bus1 S
bus1_R
bus2 S
bus2_R
DAC_raw /
DAC_fine
analog
multiplexer
programmable
gain amplifiers
Pixel output
Pixel output 2
A2
A1
S2
S1
1
1
output
drivers
DAC_dark
Stage 1
Stage 2
Stage 3
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 10 of 34
Stage 1: Offset, FPN Correction, and Multiplexing
In the first stage, the signals from the buses are subtracted and
the offset from the DACs is added. After a system reset, the
analog multiplexer is configured for two outputs (see the bit
settings in the AMPLIFIER Register on page 22). In case
ONE_OUT is set to 1, the two signals S1 and S2 are multiplexed
to one output (output 1). The amplifiers of Stage 2 and Stage 3
of the second output path are then put in standby. The speed and
power consumption of the first stage can be controlled through
the resistor connected to CMD_OUT_1.
Stage 2: Programmable Gain Amplifier
The second stage provides the gain, which is adjustable
between 1.36 and 17.38 in steps of approximately 20.25 (~1.2).
An overview of the gain settings is given in Tab le 4 . The speed
and power consumption of the second stage can be controlled
through the resistor connected to CMD_OUT_2.
Stage 3: Output Drivers
The speed and power consumption of the third stage can be
controlled through the resistor connected to CMD_OUT_3. The
output drivers are designed to drive a 20 pF output load at
40 Msamples/s with a bias resistor of 100 k.
Offset DACs
Figure 11 shows how the DAC registers influence the black
reference voltages of the two different channels. The offset is
mainly given through DAC_raw. DAC_fine can be used to shift
the reference voltage of bus 2 up or down to compensate for
different offsets in the two channels.
Figure 11. Offset for the Two Channels through DAC_RAW and DAC_FINE
Table 4. PGA Gain Settings
Bits DC Gain Bits DC Gain
0000 1.36 1000 5.40
0001 1.64 1001 6.35
0010 1.95 1010 7.44
0011 2.35 1011 8.79
0100 2.82 1100 10.31
0101 3.32 1101 12.36
0110 3.93 1110 14.67
0111 4.63 1111 17.38
DAC_raw
DAC_RAW_REG<0:7
out
DAC_fine
VDDA
out
10K
200K
50K
50K
10K
200K
blackref
bus1
blackref
bus2
GNDA
DAC_FINE_REG<0:7
rcal
rcal
pad RCAL_DAC_OUT
floating
RCAL
V
CAL
+
Note that in this figure, “K” represents K
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 11 of 34
Assume that Voutfull is the voltage that depends on the bit values that are applied to the DAC and ranges from:
Externally, the output range of DAC_raw can be changed by connecting a resistor Rcal to RCAL_DAC_OUT and applying a voltage
Vcal. The output voltage Vout of DAC_raw follows the relation (R = 10 k).
Special case:
Rcal = then Vout = Voutfull (for example, for DAC_fine)
Rcal = 0, Vcal = GND ............................. then Vout = Voutfull/2
A similar relation holds for the output range of DAC_DARK (RCAL_DAC_DARK can be used to tune the output range of this DAC).
Analog to Digital Converter
The IBIS4-6600 has a two 10-bit flash analog digital converters. The ADCs are electrically separated from the image sensor. The
inputs of the ADC must be tied externally to the outputs of the output amplifiers. One ADC samples the even columns and the other
samples the odd columns. Alternatively, one ADC can also sample all the pixels.
Setting the ADC Reference Voltages
Figure 12. ADC Resistor Ladder The internal resistance has a value of approximately 577 . Only
277 of this internal resistance is actually used as reference for
the internal ADC. This causes the actual ADC voltage range to
become half of the voltage difference between VHIGH_ADC and
VLOW_ADC. This results in the values listed Ta bl e 6 for the
external resistors.
Table 5. ADC Specifications
Parameter Specification
Input Range Set by External Resistors
(Refer the section Setting the ADC Reference Voltages)
Quantization 10 Bits
Nominal Data Rate 20 Msamples/s
DNL(Linear Conversion Mode) Typ. < 0.4 LSB RMS
INL (Linear Conversion Mode) Typ. < 3.5 LSB
Input Capacitance < 2 pF
Conversion Law Linear/Gamma corrected
)11111111()
2
1
1()00000000(0:
8
valuesbitVDDAvaluesbitV
outfull
o
cal
cal
outfull
cal
cal
out V
RR
R
V
RR
RR
V
22
150 Ohm (ESD)
RADC = 577 Ohm
Internal
GND
VLOW_ADC ~ 0.42V
VHIGH_ADC ~ 1.5V
277 Ohm
VDDA_ADC
High reference voltage
used by ADC
150 Ohm (ESD)
Low reference voltage
used by ADC
Table 6. ADC Resistor Values
Resistor Value ()
RVHIGH_ADC 560
RInternal 577
RVLOWADC 220
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 12 of 34
Sub Sample Modes
To increase the frame rate for lower resolution and regions of
interest, several sub sampling modes are implemented. The
possible sub sample modes are listed in Ta bl e 7. The bits can be
programmed in the IMAGE_CORE register (refer Table 10 on
page 17). To preserve the color information, two adjacent pixels
are read in any mode. The number of pixels that is not read
varies from mode to mode. This is designed as a repeated block
24 pixels wide, which is the lowest common multiple of the
modes described. Including the dummy pixels and the two
additional rows/columns, the number of starting coordinates for
the x and y shift register is 99 in the X direction and 138 in the Y
direction. The total number of pixels, excluding dummy pixels, is
a multiple of 24, and two additional pixels to have the same
window edges independently of the sub sampling mode.
In the X direction, two columns are always addressed at the
same moment, because the signals from the odd and even
columns must be put simultaneously on the corresponding bus.
In the Y direction, the rows are addressed one by one. This
results in slightly different implementations of the sub-sampling
modes for the two directions (Refer Figure 13 and Figure 14 on
page 13).
Figure 13. X-Sub Sampling
Table 7. Subsample Patterns
Mode Bits Read Step Description
A 000 2 2 Default mode
B 001 2 4 (Skip 2)
C 010 2 6 (Skip 4)
D 011 2 8 (Skip 6)
E 1xx 2 12 (Skip 10)
A
B
C
D
E
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register
Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
Shift register Logic selecting 2 collumns
24 column amplifiers
bus1_S
bus1_R
bus2_S
bus2_R
scan direction
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 13 of 34
Figure 14. Y-Sub Sampling
Figure 15 on page 14 shows the pixels read out in each color sub sampling mode.
shift re
g
isters on pixel pitch
Lo
g
ic selectin
g
1row
AB
C
DE
scan direction
Table 8. Frame Rate vs. Sub Sample Mode
Mode Ratio Resolution (Y*X) Frame time [mS] Frame time [mS]
A 1:1 3002 x 2210 187.4 5.3
B 1:4 1502 x 1106 52.3 19.1
C 1:9 1002 x 738 25.7 38.9
D 1:16 752 x 554 15.8 63.2
63.2 1:36 502 x 370 8.2 121.2
VGA (p) 640 x 480 12.3 81.5
VGA (p) + 23 663 x 503 13.1 76.4
VGA (l) 480 x 640 11.1 89.9
VGA(l) + 23 503 x 663 11.9 83.7
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 14 of 34
Figure 15. Pixel Readout in Various Subsample Modes
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
01234567891011121314151617181920212223
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Mode A Mode B
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Mode C Mode D
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
01234567891011121314151617181920212223
Mode E
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 15 of 34
Electronic Shutter
An electronic shutter similar to a rolling curtain is implemented on-chip. As shown in Figure 17, there are two Y shift registers. One
shift register points to the row that is currently being read out. The other shift register points to the row that is currently being reset.
Both pointers are shifted by the same Y-clock and move over the focal plane. The integration time is set by the delay between both
pointers.
Figure 16. Electronic Shutter
In case of a mechanical shutter, the two shift registers can be combined to simultaneously apply the pulses from both sides of the
pixel array. This is to halve the influence of the parasitic RC times of the reset and select lines in the pixel array. This can result in a
reduction of the row blanking time. This is the case when FAST_RESET in the SEQUENCER register is set to 1, or in the
nondestructive readout modes 1 and 2.
Figure 17. Electronic Rolling Shutter Operation
Integration
time
Readout
pointer
Reset
pointer
Time axis
Line number Reset sequence
Frame time Integration time
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 16 of 34
High Dynamic Range Modes
Double Slope Integration
The IBIS4-6600 has a feature called double slope integration to
increase the optical dynamic range of the sensor. The pixel
response can be extended over a larger range of light intensities
by using a "dual slope integration" (patents pending). This is
obtained by adding charge packets from a long and a short
integration time in the pixel during the same exposure time.
Figure 18 shows the response curve of a pixel in dual slope
integration mode. The curve also shows the response of the
same pixel in linear integration mode at the same light levels,
with a long and short integration time.
Dual slope integration is obtained by feeding a lower supply
voltage to VDD_RESET_DS (for example, apply 2.0V to 2.5V).
Note that for normal (single slope) operation, VDD_RESET_DS
must have the same value as VDD_RESET. The difference
between VDD_RESET_DS and VDD_RESET determines the
range of the high sensitivity, and as a result the output signal
level at which the transition between high and low sensitivity
occurs.
Put the amplifier gain to the lowest value where the analog output
swing covers digital input swing of the ADC. Increasing the
amplification too much may boost the high sensitivity part over
the whole ADC range.
The electronic shutter determines the ratio of integration times of
the two slopes. The high sensitivity ramp corresponds to "no
electronic shutter", thus maximal integration time (frame read out
time). The low sensitivity ramp corresponds to the electronic
shutter value that is obtained in normal operation.
Figure 18. Double Slope Response
NonDestructive Read Out (NDR)
The default mode of operation of the sensor is with FPN correction (double sampling). However, the sensor can also be read out in
a nondestructive method. After a pixel is initially reset, it can be read multiple times, without being reset. The initial reset level and all
intermediate signals can be recorded. High light levels saturate the pixels quickly, but a useful signal is obtained from the early
samples. For low light levels, use the later or latest samples. Essentially an active pixel array is read multiple times, and reset only
once. The external system intelligence interprets the data. Table 9 on page 17 summarizes the advantages and disadvantages of
nondestructive readout.
Figure 19. Principle of NonDestructive Readout
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0% 20% 40% 60% 80% 100%
Relative exposure (arbitrary scale)
Output signal [V]
Dual slope operation
Long integration time
Short integration time
time
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 17 of 34
Sequencer and Registers
Figure 4 on page 6 showed several control signals that are
needed to operate the sensor in a particular sub sampling mode,
with a certain integration time, output amplifier gain, and more.
Most of these signals are generated on-chip by the sequencer
that uses only a few control signals. These control signals must
be generated by the external system:
■SYS_CLOCK, which defines the pixel rate (nominal 40 MHz),
■Y_START pulse, which indicates the start of a new frame,
■Y_CLOCK, which selects a new row and starts the row blanking
sequence, including the synchronization and loading of the
X-register.
The relative position of the pulses is determined by a number of
data bits that are uploaded in internal registers through a Serial
to Parallel interface (SPI).
Internal Registers
Tab le 1 0 lists the internal registers with a short description. The
registers are discussed in more detail in the following sections.
Table 9. NDR: Advantages and Disadvantages
Advantages Disadvantages
Low Noise, because it is true CDS. In the order of 10 e- or below. System memory required to record the reset level and the
intermediate samples.
High Sensitivity, because the conversion capacitance is kept rather
low.
Requires multiples readings of each pixel, thus higher data
throughput.
High Dynamic Range, because the results include signals for short
and long integrations times.
Requires system level digital calculations.
Table 10. List of Internal Registers
Register Bit Name Description
0 (0000) 11:0 SEQUENCER
register
Selection of mode, granularity of the X sequencer clock, calibration,
Default value <11:0>:"000100000000"
0 NDR Mode of readout:
NDR = 0: normal readout (double sampling)
NDR = 1: non-destructive readout
1:2 NDR_mode 4 different modes of nondestructive readout (no influence if NDR = 0)
3 RESET_BLACK 0 = normal operation
1 = reset of pixels before readout
4 FAST_RESET 0 = electronic shutter operation
1 = addressing from both sides
5 FRAME_CAL_MODE 0 = fast
1 = slow
6 LINE_CAL_MODE 0 = fast
1 = slow
7 CONT_CHARGE 0 = normal mode
1 = continuous precharge
8 GRAN_X_SEQ_LSB Granularity of the X sequencer clock
9 GRAN_X_SEQ_MSB
10 BLACK 0 = normal mode
1 = disconnects column amplifiers from buses, output of amplifier equals dark
reference level
11 RESET_ALL 0 = normal mode
1 = continuous reset of all pixels
1 (0001) 10:0 NROF_PIXELS Number of pixels to count (X direction). Max. 2222/2 (2210 real + 12 dummy pixels).
Default value <10:0>:"01000000000"
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 18 of 34
2 (0010) 11:0 NROF_LINES Number of lines to count (Y direction)
Max. 3014 (3002 real + 12 dummy pixels)
Default value <11:0>:"101111000110"
3 (0011) 11:0 INT_TIME Integration time
Default value <11:0>:"000000000001"
4 (0100) 7:0 DELAY Delay of sequencer pulses
Default value <7:0>:"00000011"
0:3 DELAY_PIX_VALID Delay of PIX_VALID pulse
4:7 DELAY_EOL/EOF Delay of EOL/EOF pulses
5 (0101) 6:0 X_REG X start position (0 to 98)
Default value <6:0>:"0000000"
6 (0110) 7:0 Y_REG Y start position (0 to 137)
Default value <7:0>:"00000000"
7 (0111) 7:0 IMAGE CORE
register
Default value <7:0>:"00000000"
1:0 TEST_mode LSB: odd, MSB: even
0 = normal operation
4:2 X_SUBSAMPLE sub sampling mode in X-direction
7:5 Y_SUBSAMPLE sub sampling mode in X-direction
8 (1000) 9:0 AMPLIFIER register Default value <9:0>:"0000010000"
3:0 GAIN<3:0> Output amplifier gain setting
4 UNITY 0 = gain setting by GAIN<3:0>
1 = unity gain setting
5 ONE_OUT 0 = two analog outputs
1 = multiplexing to one output (out_1)
6 STANDBY 0 = normal operation
1 = amplifier in standby mode
7:9 DELAY_CLK_AMP Delay of pixel clock to output amplifier
9 (1001) 7:0 DAC_RAW_REG Amplifier DAC raw offset
Default value <7:0>:"10000000"
10 (1010) 7:0 DAC_FINE_REG Amplifier DAC fine offset
Default value <7:0>:"10000000"
11 (1011) 7:0 DAC_DARK_REG DAC dark reference on output bus
Default value <7:0>:"10000000"
Table 10. List of Internal Registers (continued)
Register Bit Name Description
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 19 of 34
Register Descriptions
SEQUENCER Register
a. NDR (Bit 0)
In normal operation (NDR = 0), the sensor operates in double
sampling mode. At the start of each row readout, the signals from
the pixels are sampled, the row is reset, and the signals from the
pixels are sampled again. The values are subtracted in the
output amplifier.
When NDR is set to 1, the sensor operates in nondestructive
readout (NDR) mode (refer Ta b l e 11 ).
b. NDR_mode (Bit 1 and 2)
These bits only influence the operation of the sensor in case
NDR (bit 0) is set to 1. There are two modes for nondestructive
readout (mode 1 and 2). Each mode needs two different frame
readouts (setting 1 and 2 for mode 1, setting 3 and 4 for mode
2). a reset/readout sequence (reset_seq) and then one or
several pure readout sequences (called read_seq hereafter).
Tab le 11 gives an overview of the different NDR modes.
Mode 1
In this mode, the sensor is readout in the same method as for the
nondestructive readout. However, electronic shutter control is
not possible in this case, that is, the minimal (integration) time
between two readings is equal to the number of lines that has to
be read out (frame read time). The row lines are clocked
simultaneously (left and right clock pulses are equal).
Mode 2
In this mode, it is possible to have a shorter integration time than
the frame read time. Rows are alternatingly read out with the left
and right pointer. These two pointers can point to two different
rows (see INT_TIME register). The integration time between two
readings of the same row is equal to the number of lines that is
set in the INT_TIME register multiplied by 2 plus 1, and is the
minimal one line read time.
In setting 3, the row that is read out by the left pointer is reset and
read out (first Y_CLOCK), and the row that is read out by the right
pointer is read out without being reset (second Y_CLOCK).
In setting 4, both rows are read out without being reset (on the
first Y_CLOCK the row is read out by the left pointer; on the
second Y_CLOCK the row is read out by the right pointer).
For both modes, the signals are read out through the same path
as with destructive readout (double sampling), but the buses that
are carrying the reset signals in destructive readout, are set to
the voltage given by DAC_DARK in nondestructive readout.
12 (1100) 10:0 ADC register Default value <10:0>:"00000000000"
0 STANDBY_1 0 = normal operation
1 = ADC in standby
1 STANDBY_2
2 ONE 0 = multiplexing of two ADC outputs
1 = disable multiplexing
3 SWITCH if ONE = 0: delay of output with one (EXT_CLK = 0) or half (EXT_CLK = 1) clock cycle
if ONE = 1: switch between two ADCs
4 EXT_CLK 0 = internal clock (same as clock to X shift register and output amplifier)
1 = external clock
5 TRISTATE 0 = normal operation
1 = outputs in tristate mode
6:8 DELAY_CLK_ADC Delay of clock to ADCs and digital multiplexer
9 GAMMA 0 = linear conversion
1 = 'gamma' law conversion
10 BITINVERT 0 = no inversion of bits
1 = inversion of bits
13 (1101) Reserved
14 (1110) Reserved
15 (1111) Reserved
Table 10. List of Internal Registers (continued)
Register Bit Name Description
Table 11. Overview of NDR Modes.
Setting Bits NDR mode Sequence
1 00 1 reset
2011read
3 10 2 reset
4112read
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 20 of 34
c. Reset_black (Bit 3)
If RESET_BLACK is set to 1, each line is reset before it is read
out (except for the row that is read out by the right pointer in NDR
Mode 2). This may be useful to obtain black pixels.
d. Fast_reset (Bit 4)
The fast reset option (FAST_RESET = 1) might be useful in case
a mechanical camera shutter is used. The fast reset is done on
a row-by-row basis, not by a global reset. A global reset means
charging all the pixels at the same time, which may result in a
huge peak current. Therefore, the rows can be scanned rapidly
while the left and right shift registers are both controlled
identically, so that the reset lines over the pixel array are driven
from both sides. This reduces the reset (row blanking) time
(when FAST_RESET = 1 the smallest X-granularity can be
used). After the row blanking time, the row is reset and
Y_CLOCK can be asserted to reset the next row.
After a certain integration time, the read out can be done in a
similar method. The Y shift registers are again synchronized to
the first row. Both shift registers are driven identically, and all
rows and columns are scanned for (destructive) readout.
FAST_RESET = 1 puts the sequencer in such mode that the left
and right shift registers are both controlled identically.
e. Output Amplifier Calibration (Bit 5 and 6)
Bits FRAME_CAL_MODE and LINE_CAL_MODE define the
calibration mode of the output amplifier.
During every row-blanking period, a calibration is done of the
output amplifier. There are two calibration modes. The FAST
mode (= 0) can force a calibration in one cycle. However, it is not
accurate and suffers from kTC noise, while the SLOW mode (=
1) can only make incremental adjustments and is noise free.
Approximately 200 or more "slow" calibrations have the same
effect as one "fast" calibration.
Different calibration modes can be set at the beginning of the
frame (FRAME_CAL_MODE bit) and for every subsequent row
that is read (LINE_CAL_MODE bit).
f. Continuous Charge (Bit 7)
For some applications, it might be necessary to use continuous
charging of the pixel columns instead of a precharge on every
row sample operation.
Setting bit CONT_CHARGE to 1 activates this function. The
resistor connected to pin CMD_COL is used to control the
current level on every pixel column.
g. Internal Clock Granularities
The system clock is divided several times on-chip.
The X-shift-register that controls the column/pixel readout, is
clocked by half the system clock rate. Odd and even pixel
columns are switched to two separate buses. In the output
amplifier, the pixel signals on the two buses can be combined to
one pixel stream at 40 MHz.
The clock that drives the X-sequencer can be a multiple of 2, 4,
8, or 16 times the system clock. Table 12 lists the settings for the
granularity of the X-sequencer clock and the corresponding row
blanking time (for NDR = 0). A row blanking time of 7.18 µs is the
baseline for almost all applications.
h. Black (Bit 10)
If BLACK is set to 1, the internal black signal is held high contin-
uously. As a result, the column amplifiers are disconnected from
the buses, and the buses are set to the voltage given by
DAC_DARK. The output of the amplifier equals the voltages from
the offset DACs.
i. Reset_all (Bit 11)
If RESET_ALL is set to 1, all the pixels are simultaneously put in
a 'reset' state. In this state, the pixels behave logarithmically with
light intensity. If this state is combined with one of the NDR
modes, the sensor can be used in a nonintegrating, logarithmic
mode with high dynamic range.
j. Nrof_pixels Register
After the internal X_SYNC is generated (start of the pixel readout
of a particular row), the PIXEL_VALID signal goes high. The
PIXEL_VALID signal goes low when the pixel counter reaches
the value loaded in the NROF_PIXEL register and an EOL pulse
is generated. Due to the fact that two pixels are addressed at
each internal clock cycle, the amount of pixels read out in one
row is 2*(NROF_PIXEL + 1).
k. Nrof_lines Register
After the internal YL_SYNC is generated (start of the frame
readout with Y_START), the line counter increases with each
Y_CLOCK pulse until it reaches the value loaded in the
NROF_LINES register and an EOF pulse is generated. In NDR
Mode 2, the line counter increments only every two Y_CLOCK
pulses and the EOF pulse shows up only after the readout of the
row indicated by the right shift register
INT_TIME Register
When the Y_START pulse is applied (start of the frame readout),
the sequencer generates the YL_SYNC pulse for the left Y-shift
register. This loads the left Y-shift register with the pointer loaded
in Y_REG register. At each Y_CLOCK pulse, the pointer shifts to
the next row and the integration time counter increases
(increment only every two Y_CLOCK pulses in NDR mode 2)
until it reaches the value loaded in the INT_TIME register. At that
moment, the YR_SYNC pulse for the right Y-shift register is
generated, which loads the right Y-shift register with the pointer
loaded in Y_REG register (shown in Figure 20 on page 21).
Table 12. Granularity of X-Sequencer Clock and Corresponding Row Blanking Time (for NDR = 0).
Gran_x_seq_msb/lsb X-Sequencer Clock Row Blanking Time Row Blanking Time [µs]
00 2 x sys_clock 142 x TSYS_CLOCK 3.55
01 4 x sys_clock 282 x TSYS_CLOCK 7.05
10 8 x sys_clock 562 x TSYS_CLOCK 14.05
11 16 x sys_clock 1122 x TSYS_CLOCK 28.05
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 21 of 34
Figure 20. Syncing of Y-shift Registers
Treg_int: Difference between left and right pointer = integration
counter until value "n" of INT_TIME register is reached =
INT_TIME register
In case of NDR = 0, the actual integration time Tint is given by
TintL: Integration time [# lines] = NROF_LINES register - INT_TIME
register + 1
In case of NDR = 1, NDR mode 1, the time Tint between two
readings of the same row is given by:
Tint:Integration time [# lines] = NROF_LINES register + 1
In case of NDR = 1, NDR mode 2, the times Tint1 and Tint2
between two readings of the same row (alternatingly) are given
by:
Tint1: Integration time [# lines] = 2 * INT_TIME register + 1
Tint2: Integration time [# lines] = 2 * (NROF_LINES register + 1) - (2
* INT_TIME register + 1)
DELAY Register
The DELAY register can be used to delay the PIXEL_VALID
pulse (bits 0:3) and the EOL/EOF pulses (bits 4:7) to synchronize
them to the real pixel values at the analog output or the ADC
output (which give additional delays depending on their settings).
The bit settings and corresponding delay are indicated in
Tab le 1 3.
X_REG Register
The X_REG register determines the start position of the window
in the X-direction. In this direction, there are 2208 + 2 + 12
readable pixels. In the active pixel array, sub sampling blocks are
24 pixels wide and the columns are read two by two. Therefore,
the number of start positions equals 2208/24 +2/2 +12/2 = 92 +
1 + 6 = 99.
Y_REG Register
The Y_REG register determines the start position of the window
in the Y-direction. In this direction, there are 3000 + 2 + 12
readable pixels. In the active pixel array, sub sampling blocks are
24 pixels wide and the rows are read one by one. Therefore, the
number of start positions equals 3000/24 + 2/2 +12 = 125 + 1 +
12 = 138.
Image_core Register
Bits 0:1 of the IMAGE_CORE register defines the several test
modes of the image core. Setting 00 is the default and normal
operation mode. If the bit is set to 1, the odd (bit 0) or even (bit
1) columns are tight to VDD. These test modes can be used to
tune the sampling point of the ADCs to an optimal position.
Bits 2:7 of the IMAGE_CORE register define the sub sampling
mode in the X-direction (bits 2:4) and in the Y-direction (bits 5:7).
The sub sampling modes and corresponding bit setting are given
in the section Analog to Digital Converter on page 11.
Sync of left
shift-re
g
ister
Sync of right
shift-re
g
ister
Line n
T
reg_int
Last line, followed by
s
y
nc of left shift-re
g
ister
T
int
Syn
c
Table 13. Added Delay by Changing the DELAY Register Settings
Bits Delay [# SYS_CLOCK periods] Bits Delay [# SYS_CLOCK periods]
0000 0 1000 6
0001 0 1001 7
0010 0 1010 8
0011 1 1011 9
0100 2 1100 10
0101 3 1101 11
0110 4 1110 12
0111 5 1111 13
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 22 of 34
AMPLIFIER Register
a. Gain (Bits 0:3)
The gain bits determine the gain setting of the output amplifier.
They are effective only if UNITY = 0. The gains and
corresponding bit setting are given in Table 4 on page 10.
b. Unity (Bit 4)
If UNITY = 1, the gain setting of GAIN is bypassed and the gain
amplifier is put in unity feedback.
c. One_out
If ONE_OUT = 0, the two output amplifiers are active. If
ONE_OUT = 1, the signals from the two buses are multiplexed
to output OUT1. The gain amplifier and output driver of the
second path are put in standby.
d. Standby
If STANDBY = 1, the complete output amplifier is put in standby.
This reduces the power consumption significantly.
e. Delay_clk_amp
The clock that acts on the output amplifier can be delayed to
compensate for any delay that is introduced in the path from shift
register, column selection logic, column amplifier, and buses to
the output amplifier. Setting '000' is used as a baseline.
Dac_raw_reg and Dac_fine_reg Register
These registers determine the black reference level at the output
of the output amplifier. Bit setting 11111111 for DAC_RAW_REG
register gives the highest offset voltage; bit setting 00000000 for
DAC_RAW_REG register gives the lowest offset voltage. Ideally,
if the two output paths have no offset mismatch, the
DAC_FINE_REG register must be set to 10000000. Deviation
from this value can be used to compensate the internal mismatch
(see the section Offset DACs on page 10).
Dac_raw_dark Register
This register determines the voltage level that is put on the
internal buses during calibration of the output stage. This voltage
level is also continuously put on the reset buses in case of
nondestructive readout (as a reset level for the double sampling
FPN correction).
ADC Register
a. Standby_1 and standby_2
If only one or none of the ADCs is used, the other or both ADCs
can be put in standby by setting the bit to 1. This significantly
reduces the power consumption.
b. One
If OUT1 and OUT2 are both used and connected to ADC_IN1
and ADC_IN2 respectively, ONE must be 0 to use both ADCs
and to multiplex their output to ADC_D<9:0>. If ONE = 1, the
multiplexing is disabled.
c. Switch
If the two ADCs are used (ONE = 0) and internal pixel clock
(EXT_CLK = 0), the ADC output is delayed with one system clock
cycle if SWITCH = 1. If the two ADCs are used (ONE = 0) and
an external ADC clock (EXT_CLK = 1) is applied, the ADC output
is delayed with half ADC clock cycle if SWITCH = 1.
If only one ADC is used, the digital multiplexing is disabled by
ONE = 1, but SWITCH selects which ADC output is on
ADC_D<9:0> (SWITCH = 0: ADC_1, SWITCH = 1: ADC_2).
d. Ext_clk
If EXT_CLK = 0, the internal pixel clock (that drives the X-shift
registers and output amplifier, that is, half the system clock) is
used as input for the ADC clock. If EXT_CLK = 1, an external
clock must be applied to pin ADC_CLK_EXT (pin 46).
e. Tristate
If TRISTATE = 1, the ADC_D<9:0> outputs are in tri-state mode.
f. Delay_clk_adc
The clock that finally acts on the ADCs can be delayed to
compensate for any delay introduced in the path from the analog
outputs to the input stage of the ADCs. The same settings apply
for the delay that can be given to the clock acting on the output
amplifier (see Ta bl e 14 ). The best setting also depends on the
delay of the output amplifier clock and the load of the output
amplifier. It must be used to optimize the sampling moment of the
ADCs with respect to the analog pixel input signals. Setting '000'
is used as a baseline.
g. Gamma
If GAMMA is set to 0, the ADC input to output conversion is
linear, otherwise the conversion follows a 'gamma' law (more
contrast in dark parts of the window, lower contrast in the bright
parts).
h. Bitinvert
If BITINVERT = 0, 0000000000 is the conversion of the lowest
possible input voltage, otherwise the bits are inverted.
Table 14. Added Delay by Changing the DELAY_CLK_AMP
Bit Settings
Bits Delay [ns] Bits Delay [ns]
000 1.7 100 Inversion + 8.3
001 2.9 2.9 Inversion + 9.7
010 4.3 110 Inversion + 11.1
011 6.1 111 Inversion + 12.3
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 23 of 34
Serial to Parallel Interface
To upload the sequencer registers, a dedicated serial to parallel interface (SPI) is implemented. 16 bits (4 address bits + 12 data bits)
must be uploaded serially. The address must be uploaded first (MSB first), then the data (also MSB first).
The elementary unit cell is shown in Figure 21. Sixteen of these cells are connected in series, having a common SPI_CLK form the
entire uploadable parameter block. Dout of one cell is connected to SPI_DATA of the next cell (maximum speed is 20 MHz). The
uploaded settings on the address/data bus are loaded into the correct register of the sensor on the rising edge of signal REG_CLOCK
and become effective immediately.
Figure 21. SPI Interface
Timing Diagrams
Sequencer Control Signals
There are 3 control signals that operate the image sensor:
■SYS_CLOCK
■Y_CLOCK
■Y_START
These control signals must be generated by the external system
with the following time constraints to SYS_CLOCK
(rising edge = active edge):
■TSETUP >7.5 ns
■THOLD > 7.5 ns
It is important that these signals are free of any glitches.
Figure 22. Relative Timing of the Three Control Signals
DQ
C
DQ
C
SPI_DATA
To address/data bus
Dout
SPI_CLK
REG_CLOCK
16 outputs to address/data bus
SPI_CLK
SPI_DATA
Unity Cell
E ntire uploa da ble addre s s block
REG_CLOCK
A
3
A
2
A
1D0
REG_CLOCK
SPI_CLK
SPI_DATA
Internal register
upload
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 24 of 34
Basic Frame and Line Timing
The basic frame and line timing of the IBIS4-6600 sensor is shown in Figure 23.
The pulse width of Y_CLOCK must be a minimum of one clock cycle and three clock cycles for Y_START. As long as Y_CLOCK is
applied, the sequencer stays in a suspended state.
Both EOF and EOL can be tied to Y_START (EOF) and Y_CLOCK (EOL) if both signals are delayed with at least 2 SYS_CLOCK
periods to let the sensor run automatically.
Figure 23. Basic Frame and Line Timing
T1 Row blanking time: During this period, the X-sequencer generates the control signals to sample the pixel signal and pixel
reset levels, and start the readout of one line. It depends on the granularity of the X-sequencer clock (see Table 12 on page 20).
T2 Pixels counted by pixel counter until the value of Nrof_pixels register is reached. Pixel_valid goes high when the internal
X_sync signal is generated. In other words, when the readout of the pixels is started. Pixel_valid goes low when the pixel
counter reaches the value loaded in the Nrof_pixels register. Eol goes high Sys_clock cycle after the falling edge of Pixel_valid.
T3 EOF goes high when the line counter reaches the value loaded in the NROF_LINES register and the line is read (PIXEL_VALID
goes low).
T4 The time delay between successive Y_CLOCK pulses needs to be equal to avoid any horizontal illumination (integration)
discrepancies in the image.
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 25 of 34
Pixel Output Timing
Using Two Analog Outputs
Figure 24. Pixel Output Timing using Two Analog Outputs
The pixel signal at the OUT1 (OUT2) output becomes valid after
four SYS_CLOCK cycles when the internal X_SYNC (equal to
start of PIXEL_VALID output) appears (see Figure 24). The
PIXEL_VALID and EOL/EOF pulses can be delayed by the user
through the DELAY register.
T1: Row blanking time (see Table 12 on page 20)
T2: 4 SYS_CLOCK cycles.
Multiplexing to One Analog Output
The pixel signal at the OUT1 output becomes valid after five
SYS_CLOCK cycles when the internal X_SYNC (equal to start
of PIXEL_VALID output) appears (see Figure 25). The
PIXEL_VALID and EOL/EOF pulses can be delayed by the user
through the DELAY register.
T1: Row blanking time
T2: 5 SYS_CLOCK cycles.
Figure 25. Pixel Output Timing Multiplexing to One Analog Output
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 26 of 34
ADC Timing
Two Analog Outputs
Figure 26 shows the timing of the ADC using two analog outputs. Internally, the ADCs sample on the falling edge of the ADC_CLOCK
(in case of internal clock, the clock is half the SYS_CLOCK).
T1: Each ADC has a pipeline delay of 2 ADC_CLOCK cycles. This results in a total pipeline delay of four pixels.
Figure 26. ADC Timing using Two Analog Outputs
One Analog Output
Figure 27 shows the timing of the ADC using one analog output. Internally, the ADC samples on the falling edge of the ADC_CLOCK.
T1: The ADC has a pipeline delay of 2 ADC_CLOCK cycles.
Figure 27. ADC Timing using One Analog Output
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 27 of 34
Pin Information
The following table lists all the pins and their functions. There are a total of 68 pins. All pins with the same name can be connected
together.
Table 15. Pin List
Pin Pin Name Pin Type Expected Voltage [V] Pin Description
1 CMD_COL_CTU Input 0 Biasing of columns (ctu). Decouple with 100 nF to GNDA.
2 CMD_COL Input 1.08 Biasing of columns. Connect to VDDA with R = 10 k and
decouple to GNDA with C = 100 nF.
3 CMD_COLAMP Input 0.66 Biasing of column amplifiers. Connect to VDDA with
R = 100 k and decouple to GNDA with C = 100 nF.
4 CMD_COLAMP_CTU Input 0.37 Biasing of column amplifiers. Connect to VDDA with
R = 10 M and decouple to GNDA with C = 100 nF.
5 RCAL_DAC_DARK Input 1.27 at code 128
DAC_DARK reg
Biasing of DAC for dark reference. Can be used to set output
range of DAC.
Default: Decouple to GNDA with C = 100 nF
6 RCAL_DAC_OUT Input 0 Biasing of DAC for output dark level. Can be used to set output
range of DAC. Default: Connect to GNDA
7V
DDA Power 2.5 VDD of analog part [2.5V]
8 GNDA Power 0 GND (&substrate) of analog part
9V
DDD Power 2.5 VDD of digital part [2.5V]
10 GNDD Power 0 GND (&substrate) of digital part
11 CMD_OUT_1 Input 0.78 Biasing of first stage output amplifiers. Connect to VDDAMP
with R = 50 k and decouple to GNDAMP with C = 100 nF.
12 CMD_OUT_2 Input 0.97 Biasing of second stage output amplifiers. Connect to
VDDAMP with R = 25 k and decouple to GNDAMP with
C = 100 nF.
13 CMD_OUT_3 Input 0.67 Biasing of third stage output amplifiers. Connect to VDDAMP
with R = 100 k and decouple to GNDAMP with C = 100 nF.
14 SPI_CLK Input - Clock of digital parameter upload. Shifts on rising edge.
15 SPI_DATA Input - Serial address and data input. 16-bit word. Address first. MSB
first.
16 VDDAMP Power 2.5 VDD of analog output [2.5V] (Can be connected to VDDA)
17 CMD_FS_ADC Input 0.73 Biasing of first stage ADC. Connect to VDDA_ADC with
R = 50 k and decouple to GNDA_ADC with C = 100 nF.
18 CMD_SS_ADC Input 0.73 Biasing of second stage ADC. Connect to VDDA_ADC with
R = 50 k and decouple to GNDA_ADC.
19 CMD_AMP_ADC input 0.59 Biasing of input stage ADC. Connect to VDDA_ADC with
R = 180 k and decouple to GNDA_ADC with C = 100 nF.
20 GNDAMP Ground 0 GND (&substrate) of analog output
21 OUT1 Output Black level: 1 at code 190
DAC_RAW register
Analog output 1
22 ADC_IN1 Input See OUT1. Analog input ADC 1
23 VDDAMP Power 2.5 VDD of analog output [2.5V] (Can be connected to VDDA)
24 OUT2 Output Black level: 1 at code 190
DAC_RAW register
Analog output 2
25 ADC_IN2 Input See OUT2. Analog input ADC 2
26 VDDD Power 2.5 VDD of digital part [2.5V]
27 GNDD Power 0 GND (&substrate) of digital part
28 GNDA Power 0 GND (&substrate) of analog part
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 28 of 34
29 VDDA Power 2.5 VDD of analog part [2.5V]
30 REG_CLOCK Input - Register clock. Data on internal bus is copied to
corresponding registers on rising edge.
31 SYS_CLOCK Input - System clock defining the pixel rate (nominal 40 MHz, 50% ±
5% duty cycle)
32 SYS_RESET Input - Global system reset (active high)
33 Y_CLK Input - Line clock
34 Y_START Input - Start frame readout
35 GNDD_ADC Power 0 GND (&substrate) of digital part ADC
36 VDDD_ADC Power 2.5 VDD of digital part [2.5V] ADC
37 GNDA_ADC Power 0 GND (&substrate) of analog part
38 VDDA_ADC Power 2.5 VDD of analog part [2.5V]
39 VHIGH_ADC Input 1.5 ADC high reference voltage (for example, connect to
VDDA_ADC with R = 560 and decouple to GNDA_ADC with
C = 100 nF)
40 VLOW_ADC Input 0.42 ADC low reference voltage (for example, connect to
GNDA_ADC with R = 220 and decouple to GNDA_ADC
with C = 100 nF)
41 GNDA_ADC Power 0 GND (&substrate) of analog part
42 VDDA_ADC Power 2.5 VDD of analog part [2.5V]
43 GNDD_ADC Power 0 GND (&substrate) of digital part ADC
44 VDDD_ADC Power 2.5 VDD of digital part [2.5V] ADC
45 VDD_RESET_DS Power 2.5 (for no dual slope) Variable reset voltage (dual slope)
46 ADC_CLK_EXT Input - External ADC clock
47 EOL Output - Diagnostic end of line signal (produced by sequencer), can
be used as Y_CLK
48 EOF Output - Diagnostic end of frame signal (produced by sequencer), can
be used as Y_START
49 PIX_VALID Output - Diagnostic signal. High during pixel readout
50 TEMP Output - Temperature measurement. Output voltage varies linearly
with temperature.
51 ADC_D<9> Output - ADC data output (MSB)
52 VDD_PIX Power 2.5 VDD of pixel core [2.5V]
53 GND_AB Power 0 Anti-blooming ground. Set to 1V for improved anti-blooming
behavior
54 ADC_D<8> Output - ADC data output
55 ADC_D<7> Output - ADC data output
56 ADC_D<6> Output - ADC data output
57 ADC_D<5> Output - ADC data output
58 ADC_D<4> Output - ADC data output
59 ADC_D<3> Output - ADC data output
60 VDD_RESET Power 2.5 Reset voltage [2.5V]. Highest voltage to the chip. 3.3V for
extended dynamic range or 'hard reset'.
61 ADC_D<2> Output - ADC data output
62 ADC_D<1> Output - ADC data output
63 ADC_D<0> Output - ADC data output (LSB)
Table 15. Pin List (continued)
Pin Pin Name Pin Type Expected Voltage [V] Pin Description
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 29 of 34
Note on Power On Behavior
At power on, the chip is in an undefined state. It is advised that the power on is accompanied by the assertion of the SYS_CLOCK
and a SYS_RESET pulse that puts all internal registers in their default state (all bits are set to 0). The X-shift registers are in a defined
state after the first X_SYNC, which occurs a few microseconds after the first Y_START and Y_CLOCK pulse. Before this X_SYNC,
the chip may draw more current from the analog power supply VDDA. It is therefore favorable to have separate analog and digital
supplies. The current spike (if there are any) may also be avoided by a slower ramp up of the analog power supply or by disconnecting
the resistor on pin 3 (CMD_COLAMP) at startup.
64 BS_RESET Input - Boundary scan (allows debugging of internal nodes): Reset.
Tie to GND if not used.
65 BS_CLOCK Input - Boundary scan (allows debugging of internal nodes): Clock.
Tie to GND if not used.
66 BS_DIN Input - Boundary scan (allows debugging of internal nodes): In. Tie
to GND if not used.
67 BS_BUS Output - Boundary scan (allows debugging of internal nodes): Bus.
Leave floating if not used.
68 CMD_DEC Input 0.74 Biasing of X and Y decoder. Connect to VDDD with R = 50 k
and decouple to GNDD with C = 100 nF.
Table 15. Pin List (continued)
Pin Pin Name Pin Type Expected Voltage [V] Pin Description
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 32 of 34
Glass Lid Specifications
Monochrome Sensor
A D263 glass is used as protection glass lid on top of the IBIS4-6600 monochrome sensors. The refraction index of the D263 glass
lid is 1.52. Figure 30 shows the transmission characteristics of the D263 glass.
Figure 30. Transmittance Curve of the D263 Cover Glass Lid
Storage and Handling
Storage Conditions
Handling and Soldering Conditions
Special care must be taken when soldering image sensors with
color filter arrays (RGB color filters), onto a circuit board,
because color filters are sensitive to high temperatures.
Prolonged heating at elevated temperatures may result in
deterioration of the performance of the sensor. The following
recommendations ensure that sensor performance is not
compromised during end-users assembly processes.
Board Assembly
Device placement onto boards must be done in accordance with
strict ESD controls for Class 0, JESD22 Human Body Model, and
Class A, JESD22 Machine Model devices. Assembly operators
must always wear all designated and approved grounding
equipment. Grounded wrist straps at ESD protected
workstations are recommended, including the use of ionized
blowers. All tools must be ESD protected.
Manual Soldering
When a soldering iron is used, the following conditions must be
observed:
■Use a soldering iron with temperature control at the tip. The
soldering iron tip temperature must not exceed 350°C.
■The soldering period for each pin must be less than 5 seconds.
Reflow Soldering
Figure 31 on page 33 shows the maximum recommended
thermal profile for a reflow soldering system. If the
temperature/time profile exceeds these recommendations,
damage to the image sensor may occur. See Figure 31 on page
33 for more details.
Precautions and Cleaning
Avoid spilling solder flux on the cover glass, because bare glass
and particularly glass with antireflection filters may be adversely
affected by the flux. Avoid mechanical or particulate damage to
the cover glass.
It is recommended that isopropyl alcohol (IPA) be used as a
solvent for cleaning the image sensor glass lid. When using other
solvents, it must be confirmed beforehand whether the solvent
can dissolve the package and/or the glass lid.
0
10
20
30
40
50
60
70
80
90
100
400 500 600 700 800 900
Wavelength [nm]
Transmission [%]
Table 16. Storage Conditions
Description Minimum Maximum Maximum
Temperature –30 +85 °C
[+] Feedback
IBIS4-6600 CYII4SM6600AB
Document Number: 001-02366 Rev. *G Page 33 of 34
Figure 31. Reflow Soldering Temperature Profile
RoHS (Pb-Free) Compliance
This section reports the use of Hazardous chemical substances as required by the RoHS Directive (excluding packing material).
Pb-Free Soldering
IBIS4-A-6600-M2 (serial numbers beyond 3694): This product is
successfully tested for Pb-free soldering processes, using a
reflow temperature profile with maximum 260°C, minimum 40s
at 255°C, and minimum 90s at 217°C.
IBIS4-A-6600-C2: This product does not withstand a lead-free
soldering process. Maximum allowed reflow or wave soldering
temperature is 220°C. Hand soldering is recommended for this
part type.
Note that "Intentional content" is defined as any material
demanding special attention is contained into the inquired
product by following cases:
1. A case that the previously mentioned material is added as a
chemical composition into the inquired product intentionally,
to produce and maintain the required performance and
function of the intended product.
2. A case that the previously mentioned material, which is used
intentionally in the manufacturing process, is contained in or
adhered to the inquired product.
The following case is not treated as "intentional content":
A case that the previously mentioned material is contained as an
impurity into raw materials or parts of the intended product. The
impurity is defined as a substance that cannot be removed
industrially, or is produced at a process, such as chemical
composing or reaction and cannot be removed technically.
Table 17. The Chemical Substances and Information about Any Intentional Content
Chemical Substance Intentional Content Portion Containing Intentional Content
Lead NO -
Cadmium NO -
Mercury NO -
Hexavalent chromium NO -
PBB (Polybrominated biphenyls) NO -
PBDE (Polybrominated diphenyl ethers) NO -
[+] Feedback
Document Number: 001-02366 Rev. *G Revised April 1,2010 Page 34 of 34
All products and company names mentioned in this document may be the trademarks of their respective holders.
IBIS4-6600 CYII4SM6600AB
© Cypress Semiconductor Corporation, 2005-2010. The information contained herein is subject to change without notice. Cypress Semiconductor Corporation assumes no responsibility for the use of
any circuitry other than circuitry embodied in a Cypress product. Nor does it convey or imply any license under patent or other rights. Cypress products are not warranted nor intended to be used for
medical, life support, life saving, critical control or safety applications, unless pursuant to an express written agreement with Cypress. Furthermore, Cypress does not authorize its products for use as
critical components in life-support systems where a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress products in life-support systems
application implies that the manufacturer assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Any Source Code (software and/or firmware) is owned by Cypress Semiconductor Corporation (Cypress) and is protected by and subject to worldwide patent protection (United States and foreign),
United States copyright laws and international treaty provisions. Cypress hereby grants to licensee a personal, non-exclusive, non-transferable license to copy, use, modify, create derivative works of,
and compile the Cypress Source Code and derivative works for the sole purpose of creating custom software and or firmware in support of licensee product to be used only in conjunction with a Cypress
integrated circuit as specified in the applicable agreement. Any reproduction, modification, translation, compilation, or representation of this Source Code except as specified above is prohibited without
the express written permission of Cypress.
Disclaimer: CYPRESS MAKES NO WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, WITH REGARD TO THIS MATERIAL, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. Cypress reserves the right to make changes without further notice to the materials described herein. Cypress does not
assume any liability arising out of the application or use of any product or circuit described herein. Cypress does not authorize its products for use as critical components in life-support systems where
a malfunction or failure may reasonably be expected to result in significant injury to the user. The inclusion of Cypress’ product in a life-support systems application implies that the manufacturer
assumes all risk of such use and in doing so indemnifies Cypress against all charges.
Use may be limited by and subject to the applicable Cypress software license agreement.
Document History Page
Sales, Solutions, and Legal Information
Worldwide Sales and Design Support
Cypress maintains a worldwide network of offices, solution centers, manufacturer’s representatives, and distributors. For more infor-
mation on Image sensor products, please contact imagesensors@cypress.com.
Cypress offers standard and customized CMOS image sensors for consumer as well as industrial and professional applications.
Consumer applications include the fast growing high volume cell phone, digital still cameras as well as automotive applications.
Cypress' CMOS image sensors are characterized by very high pixel counts, large area, very high frame rates, large dynamic range,
and high sensitivity.
Document Title: IBIS4-6600 CYII4SM6600AB 6.6 MP CMOS Image Sensor
Document Number: 001-02366
Rev. ECN Submission
Date
Orig. of
Change Description of Change
** 384900 See ECN FWU Origination.
*A 402976 See ECN FWU Preliminary notice removed.
Electro optical spec updated to characterization data.
*B 418669 See ECN FVK Table 15. ADC resistor values changed. ADC section added.
Figure 33 p41 corrected
*C 502551 See ECN QGS Converted to Frame file
*D 642596 See ECN FPW Ordering information update
*E 2649816 03/17/2009 PCI/AESA Final data sheet. Changed title from “IBIS4-A-6600 CMOS Image Sensor” to
“IBIS4-6600 CYII4SM6600AB 6.6 MP CMOS Image Sensor”. Updated Typical
Response Curve of the RGB Filters on page 7 and added Relative Response
Graph on page 8. Updated Package Diagrams and data sheet template.
*F 2766198 09/19/09 NVEA Updated Ordering Information table
*G 2902961 04/01/10 NVEA Removed inactive part from the ordering information table.
[+] Feedback
