ON Semiconductor KAC-12040 User manual

©Semiconductor Components Industries, LLC, 2016
March, 2016 − Rev. 5 1Publication Order Number:
KAC−12040/D
KAC-12040
4000 (H) x 3000 (V)
CMOS Image Sensor
Description
The KAC−12040 Image Sensor is a high-speed 12 megapixel
CMOS image sensor in a 4/3″optical format based on a 4.7 mm 5T
CMOS platform. The image sensor features very fast frame rate,
excellent NIR sensitivity, and flexible readout modes with multiple
regions of interest (ROI). The readout architecture enables use of 8, 4,
or 2 LVDS output banks for full resolution readout of 70 frames per
second.
Each LVDS output bank consists of up to 8 differential pairs
operating at 160 MHz DDR for a 320 Mbps data rate per pair.
The pixel architecture allows rolling shutter operation for motion
capture with optimized dynamic range or global shutter for precise
still image capture.
Table 1. GENERAL SPECIFICATIONS
Parameter Typical Value
Architecture 5T Global Shutter CMOS
Resolution 12 Megapixels
Aspect Ratio 4:3
Pixel Size 4.7 mm (H) ×4.7 mm (V)
Total Number of Pixels 4224 (H) ×3192 (V)
Number of Effective Pixels 4016 (H) ×3016 (V)
Number of Active Pixels 4000 (H) ×3000 (V)
Active Image Size 18.8 mm (H) ×14.1 mm (V)
23.5 mm (Diagonal), 4/3″Optical Format
Master Clock Input Speed 5 MHz to 50 MHZ
Maximum Pixel Clock Speed 160 MHz DDR LVDS, 320 Mbps
Number of LVDS Outputs 64 Differential Pairs
Number of Output Banks 8, 4, or 2
Frame Rate, 12 Mp 1−70 fps 10 bits
1−75 fps 8 bits
Charge Capacity 16,000 electrons
Quantum Efficiency
KAC−12040−CBA
KAC−12040−ABA 40%, 47%, 45% (470, 540, 620 nm)
53%, 15%, 10% (500, 850, 900 nm)
Read Noise
(at Maximum LVDS Clock) 3.7 e−rms, Rolling Shutter 25.5 e−rms, Global Shutter
Dynamic Range 73 dB, Rolling Shutter
56 dB, Global Shutter
Blooming Suppression > 10,000x
Image Lag 1.3 electron
Digital Core Supply 2.0 V
Analog Core Supply 1.8 V
Pixel Supply 2.8 V & 3.5 V
Power Consumption 1.5 W for 12 Mp @ 70 fps 10 bits
Package 267 Pin Ceramic Micro-PGA
Cover Glass AR Coated, 2-sides
NOTE: All Parameters are specified at T = 40°C unless otherwise noted.
www.onsemi.com
Figure 1. KAC−12040 CMOS Image Sensor
Features
•Global Shutter and Rolling Shutter
•Very Fast Frame Rate
•High NIR Sensitivity
•Multiple Regions of Interest
•Interspersed Video Streams
Applications
•Machine Vision
•Intelligent Transportation Systems
•Surveillance
See detailed ordering and shipping information on page 2 o
f
this data sheet.
ORDERING INFORMATION

KAC−12040
www.onsemi.com
2
The image sensor has a pre-configured QFHD (4 ×1080p,
16:9) video mode, fully programmable, multiple ROI for
windowing, programmable sub-sampling, and reverse
readout (flip and mirror). The two ADCs can be configured
for 8-bit, 10-bit, 12-bit or 14-bit conversion and output.
Additional features include interspersed video streams
(dual-video), on-chip responsivity calibration, black
clamping, overflow pixel for blooming reduction, black-sun
correction (anti-eclipse), column and row noise correction,
and integrated timing generation with SPI control, 4:1 and
9:1 averaging decimation modes.
ORDERING INFORMATION
Table 2. ORDERING INFORMATION − KAC−12040 IMAGE SENSOR
Part Number Description Marking Code
KAC−12040−ABA−JD−BA Monochrome, Micro-PGA Package, Sealed Clear Cover Glass with AR
Coating (Both Sides), Standard Grade. KAC−12040−ABA
Serial Number
KAC−12040−ABA−JD−AE Monochrome, Micro-PGA Package, Sealed Clear Cover Glass with AR
Coating (Both Sides), Engineering Grade.
KAC−12040−CBA−JD−BA Bayer (RGB) Color Filter Pattern, Micro-PGA Package, Sealed Clear Cover
Glass with AR Coating (Both Sides), Standard Grade. KAC−12040−CBA
Serial Number
KAC−12040−CBA−JD−AE Bayer (RGB) Color Filter Pattern, Micro-PGA Package, Sealed Clear Cover
Glass with AR Coating (Both Sides), Engineering Grade.
1. Engineering Grade samples might not meet final production testing limits, especially for cosmetic defects such as clusters, but also possibly
column and row artifacts. Overall performance is representative of final production parts.
Table 3. ORDERING INFORMATION − EVALUATION SUPPORT
Part Number Description
KAC−12040−CB−A−GEVK Evaluation Hardware for KAC−12040 Image Sensor (Color). Includes Image Sensor.
KAC−12040−AB−A−GEVK Evaluation Hardware for KAC−12040 Image Sensor (Monochrome). Includes Image Sensor.
LENS−MOUNT−KIT−C−GEVK Lens Mount Kit that Supports C, CS, and F Mount Lenses. Includes IR Cut-filter for Color Imaging.
See the ON Semiconductor Device Nomenclature document (TND310/D) for a full description of the naming convention
used for image sensors. For reference documentation, including information on evaluation kits, please visit our web site at
www.onsemi.com.

KAC−12040
www.onsemi.com
3
DEVICE DESCRIPTION
Architecture
Figure 2. Block Diagram
G
B
GR
4000 (H) y3000 (V)
4.7 mm Pixel
G
B
GR
G
B
GR
G
B
GR
LVDS Bank 2 LVDS Bank 4 LVDS Bank 6
Even Row ADC, Analog Gain, Black-Sun Correction
Clk2 Clk4 Clk6
2D0− 2D6
4D0− 4D6
6D0− 6D6
8
88
8
88
Odd Row ADC, Analog Gain, Black-Sun Correction
LVDS Bank 3 LVDS Bank 5 LVDS Bank 7
(0, 0)
Clk7 7D0− 7D6 Clk5 5D0− 5D6 Clk3 3D0− 3D6
LVDS Bank 0 LVDS Bank 1
Digital Gain/Offset, Noise Correction
Clk1
1D0− 1D6
Clk0
0D0− 0D6
Timing Control, Sub-Sampling/Averaging
8
104
8
104
3.5 VA
3.3 VD
2.8 VA
2.0 VD
1.8 VA
Chip Clock (2 Pins)
TRIGGER
RESETN
CSN
SCLK
MOSI
MISO
Serial
Peripheral
Interface
(SPI)
ADC_Ref1
ADC_Ref2
VSS 0 V
4.02 kW±1%

KAC−12040
www.onsemi.com
4
Physical Orientation
Figure 3. Package Pin Orientation − Top X-Ray View
LVDS Bank 2 LVDS Bank 4 LVDS Bank 6
LVDS Bank 3 LVDS Bank 5 LVDS Bank 7
LVDS Bank 0 LVDS Bank 1
A
B
C
D
E
AA
AB
AC
AD
AE
272625242322212019181716151413121110987654321
272625242322212019181716151413121110987654321
1. The center of the pixel array is aligned to the physical package center.
2. The region under the sensor die is clear of pins enabling the use of a heat sink.
3. Non-symmetric mounting holes provide orientation and mounting precision.
4. Non-symmetric pins prevent incorrect placement in PCB.
5. Letter “F” indicator shows default readout direction relative to package pin 1.
Notes:

KAC−12040
www.onsemi.com
5
Table 4. PRIMARY PIN DESCRIPTION
Pin Name Type Description
AB09 RESETN DI Sensor Reset (0 V = Reset State)
E07 CLK_In1 DI Sensor Input Clk_In1 (45−50 MHz)
D08 CLK_In2 DI Sensor Input Clk_In2 (Connect to Clk1)
AB08 TRIGGER DI Trigger Input (Optional)
AA05 SCLK DI SPI Master Clock
AA08 MOSI DI SPI Master Output, Slave Input
AA07 MISO DO SPI Master Input, Slave Output
AA06 CSN DI SPI Chip Select (0 V = Selected)
AA14 ADC_Ref1 AO 4.02 kW±1% Resistor between Ref1 & Ref2
AA15 ADC_Ref2 AO 4.02 kW±1% Resistor between Ref1 & Ref2
AB07 MSO DO Mechanical Shutter Output Sync (Optional)
AB06 FLO DO Flash Output Sync (Optional)
E05 FEN DO Frame Enable Reference Output (Optional)
E06 LEN DO Line Enable Reference Output (Optional)
1. DI = Digital Input, DO = Digital Output, AO = Analog Output.
2. Tie unused DI pins to Ground, NC unused DO pins.
3. By default Clk_In2 should equal Clk_In1 and should be the same source clock.
4. The RESETN pin has a 62 kWinternal pull-up resistor, so if left floating the chip will not be in reset mode.
5. The TRIGGER pin has an internal 100 kWpull down resistor. If left floating (and at default polarity) then the sensor state will not be affected
by this pin (i.e. defaults to ‘not triggered’ mode if floated).
6. All of the DI and DO pins nominally operate at 0 V →2.0 V and are associated with the VDD_DIG power supply.
Table 5. POWER PIN DESCRIPTION
Name Voltage Pins Description
VDD_LVDS 3.3 V D C04, C05, C23, C24, D04, D24, E04, E24, AA04, AA24,
AB04, AB24, AC04, AC05 AC23, AC24 LVDS Output Supply
VDD_DIG 2.0 V D C18, C19, D18, D19, E18, AA18, AB18, AB19, AC18, AC19,
C20, C21, C22, D20, D21, D22, D23, E20, E21, E22, AA20,
AA21, AA22, AB20, AB21, AB22, AB23, AC20, AC21,
AC22, AB15, E08
Digital Core Supply
AVDD_HV 3.5 V A C11, D11, E11, AA11, AB11, AC11, C10, D10, E10, AA10,
AB10, AC10 Pixel Supply 1
Vref_P 2.8 V A C13, D13, E13, AA13, AB13, AC13 Pixel Supply 2
AVDD_LV 1.8 V A C17, D16, D17, E17, AA17, AB16, AB17, AC17 Analog Low Voltage Supply
Vpixel_low 0 V E09 Pixel Supply 3. Combine with VSS for
normal operation. Can be pulsed for
Extended Dynamic Range Operation.
VSS 0 V C12, C14, D12, D14, E12, AA12, AB12, AB14, AC12, AC14,
E15, D15, AA09, A02, A14, A26, B14, C03, C06, C25, D03,
D25, E03, E19, E23, E25, AA03, AA19, AA23, AA25, AB25,
AC03, AC06, AC25, AD14, AE02, AE14, AE26
Sensor Ground Reference
No Connect NA A01, AC09, E14, E16, C09, D09, D05, D06, D07, AA16,
AB05 Unused and test-only pins. These pins must be floated.

KAC−12040
www.onsemi.com
6
Pin Name Description
E01 1DCLK+
E02 1DCLK−
D01 1DATA0+
D02 1DATA0−
C01 1DATA1+
C02 1DATA1−
B01 1DATA2+
B02 1DATA2−
A03 1DATA3+
B03 1DATA3−
A04 1DATA4+
B04 1DATA4−
A05 1DATA5+
B05 1DATA5−
A06 1DATA6+
B06 1DATA6−
Bank 1
LVDS Clock
Bank 1
LVDS Data
Pin Name Description
C07 3DCLK+
C08 3DCLK−
A07 3DATA0+
B07 3DATA0−
A08 3DATA1+
B08 3DATA1−
A09 3DATA2+
B09 3DATA2−
A10 3DATA3+
B10 3DATA3−
A11 3DATA4+
B11 3DATA4−
A12 3DATA5+
B12 3DATA5−
A13 3DATA6+
B13 3DATA6−
Bank 3
LVDS Clock
Bank 3
LVDS Data
Pin Name Description
C15 5DCLK+
C16 5DCLK−
A15 5DATA0+
B15 5DATA0−
A16 5DATA1+
B16 5DATA1−
A17 5DATA2+
B17 5DATA2−
A18 5DATA3+
B18 5DATA3−
A19 5DATA4+
B19 5DATA4−
A20 5DATA5+
B20 5DATA5−
A21 5DATA6+
B21 5DATA6−
Bank 5
LVDS Clock
Bank 5
LVDS Data
Pin Name Description
A22 7DCLK+
B22 7DCLK−
A23 7DATA0+
B23 7DATA0−
A24 7DATA1+
B24 7DATA1−
A25 7DATA2+
B25 7DATA2−
B27 7DATA3+
B26 7DATA3−
C27 7DATA4+
C26 7DATA4−
D27 7DATA5+
D26 7DATA5−
E27 7DATA6+
E26 7DATA6−
Bank 7
LVDS Clock
Bank 7
LVDS Data
Pin Name Description
AA01 0DCLK+
AA02 0DCLK−
AB01 0DATA0+
AB02 0DATA0−
AC01 0DATA1+
AC02 0DATA1−
AD01 0DATA2+
AD02 0DATA2−
AE03 0DATA3+
AD03 0DATA3−
AE04 0DATA4+
AD04 0DATA4−
AE05 0DATA5+
AD05 0DATA5−
AE06 0DATA6+
AD06 0DATA6−
Bank 0
LVDS Clock
Bank 0
LVDS Data
Pin Name Description
AC07 2DCLK+
AC08 2DCLK−
AE07 2DATA0+
AD07 2DATA0−
AE08 2DATA1+
AD08 2DATA1−
AE09 2DATA2+
AD09 2DATA2−
AE10 2DATA3+
AD10 2DATA3−
AE11 2DATA4+
AD11 2DATA4−
AE12 2DATA5+
AD12 2DATA5−
AE13 2DATA6+
AD13 2DATA6−
Bank 2
LVDS Clock
Bank 2
LVDS Data
Pin Name Description
AC15 4DCLK+
AC16 4DCLK−
AE15 4DATA0+
AD15 4DATA0−
AE16 4DATA1+
AD16 4DATA1−
AE17 4DATA2+
AD17 4DATA2−
AE18 4DATA3+
AD18 4DATA3−
AE19 4DATA4+
AD19 4DATA4−
AE20 4DATA5+
AD20 4DATA5−
AE21 4DATA6+
AD21 4DATA6−
Bank 4
LVDS Clock
Bank 4
LVDS Data
Pin Name Description
AE22 6DCLK+
AD22 6DCLK−
AE23 6DATA0+
AD23 6DATA0−
AE24 6DATA1+
AD24 6DATA1−
AE25 6DATA2+
AD25 6DATA2−
AD26 6DATA3+
AD27 6DATA3−
AC26 6DATA4+
AC27 6DATA4−
AB26 6DATA5+
AB27 6DATA5−
AA26 6DATA6+
AA27 6DATA6−
Bank 6
LVDS Clock
Bank 6
LVDS Data
1. All LVDS Data and Clock lines must be routed with 100 Wdifferential transmission line traces.
2. All the traces for a single LVDS Bank should be the same physical length to minimize skew between the clock and data lines.
3. In 2 Bank mode, only LVDS banks 0 and 1 are active.
4. In 4 Bank mode, only LVDS bank 0, 1, 2, and 3 are active.
5. Float the pins of unused LVDS Banks to conserve power.
6. Unused pins in active banks (due to ADC bit depth < 14) are automatically tri-stated to save power, but these can also be floated.
Table 6. LVDS PIN DESCRIPTION

KAC−12040
www.onsemi.com
7
IMAGING PERFORMANCE
Table 7. TYPICAL OPERATIONAL CONDITIONS
(Unless otherwise noted, the Imaging Performance Specifications are measured using the following conditions.)
Description Condition Notes
Light Source Continuous Red, Green and Blue LED Illumination 1
Temperature Measured Die Temperature: 40°C and 27°C
Integration Time 16.6 ms (1400d LL, Register 0201h)
Readout Mode Dual-Scan, Global Shutter, 320 MHz, PLL2
Clamps Column/Row Noise Corrections Active, Frame Black Level Clamp Active
ADC Bit Depth 10 bit
Analog Gain Unity Gain or Referred Back to Unit Gain
1. For monochrome sensor, only green LED used.
Table 8. KAC−12040−ABA CONFIGURATION (MONOCHROME)
Description Symbol Wavelength
(nm) Min. Nom. Max. Unit Sampling Plan
Temperature
Tested at
(5C) Test
Peak Quantum Efficiency
Green
NIR1
NIR2
QEMAX 550
850
900
−
−
−
53
15
10
−
−
−
% Design 27
Responsivity − 84 − ke*
Lux @sDesign 27 20
Responsivity − 7.0 − V
Lux @sDesign 27 21
Table 9. KAC−12040−CBA CONFIGURATION (BAYER RGB)
Description Symbol Wavelength
(nm) Min. Nom. Max. Unit Sampling Plan
Temperature
Tested at
(5C) Test
Peak Quantum Efficiency
Green
NIR1
NIR2
QEMAX 470
540
620
850
900
−
−
−
−
−
40
47
45
15
10
−
−
−
−
−
% Design 27
Responsivity Blue
Green
Red
−
−
−
17
35
38
−
−
−
ke*
Lux @sDesign 27 20
Responsivity Blue
Green
Red
−
−
−
1.4
2.9
3.2
−
−
−
V
Lux @sDesign 27 21

KAC−12040
www.onsemi.com
8
Table 10. PERFORMANCE SPECIFICATIONS ALL CONFIGURATIONS
Description Symbol Min. Nom. Max. Unit Sampling
Plan
Temperature
Tested at
(5C) Test Notes
Photodiode Charge
Capacity PNe − 16 − ke−Die 27, 40 16
Read Noise ne−T −
−3.7 RS 25.5 GS − −e−rms Die 27 8 1
Total Pixelized Noise −
−4.5 RS 28.3 GS − −e−rms Die 27 19 1
Dynamic Range DR −
−73 RS 56 GS − −dB Die 27 1, 4
Column Noise CN−
−0.6 RS 3.0 GS − −e−rms Die 27 9 1, 6
Row Noise RN−
−1.0 RS 5.0 GS − −e−rms Die 27 10 1, 7
Dark Field Local
Non-Uniformity Floor DSNU_flr − −3.0 RS 21 GS − −e−rms Die 27, 40 11, 5
Bright Field Global
Photoresponse
Non-Uniformity
PRNU_1 − 1.5 − % rms Die 27, 40 2 2
Bright Field Global Peak to
Peak Photoresponse
Non-Uniformity
PRNU_2 − 6.5 − % pp Die 27, 40 3 2
Maximum Photoresponse
Non-Linearity NL − 6.3 − % Die 27, 40 11 3
Maximum Gain Difference
between Outputs DG− 0.3 − % Die 27, 40 12 8
Photodiode Dark Current IPD − 4.6 70 e/p/s Die 40 13 9
Storage Node Dark Current IVD − 1,200 5,000 e/p/s Die 40 14 5
Image Lag Lag − 1.3 10 −Design 27, 40 15
Black-Sun Anti-Blooming XAB −
−12 > 10,000 − −W/cm2
xllumSat Design 27 7 14
Parasitic Light Sensitivity PLS − 730 − − Design 27 6 10
Dual-Video WDR −
−140 RS 120 GS − −dB Design 27 1, 11, 12
Pulsed Pixel WDR
(GS Only) − 100 − dB Design 27 12, 13
1. RS = Rolling Shutter Operation Mode, GS = Global Shutter Operation Mode.
2. Measured per color, worst of all colors reported.
3. Value is over the range of 10% to 90% of photodiode saturation, Green response used.
4. Uses 20LOG (PNe / ne−T).
5. Photodiode dark current made negligible.
6. Column Noise Correction active.
7. Row Noise Correction active.
8. Measured at ~70% illumination.
9. Storage node dark current made negligible.
10.GSE (Global Shutter Efficiency) = 1 − 1 / PLS.
11.Min vs Max integration time at 30 fps.
12.WDR measures expanded exposure latitude from linear mode DR.
13.Min/Max responsivity in a 30 fps image.
14.Saturation Illumination referenced to a 3 line time integration.

KAC−12040
www.onsemi.com
9
TYPICAL PERFORMANCE CURVES
Figure 4. Monochrome QE (with Microlens)
Figure 5. Bayer QE (with Microlens)

KAC−12040
www.onsemi.com
10
Angular Quantum Efficiency
For the curves marked “Horizontal”, the incident light angle is varied along the wider array dimension.
For the curves marked “Vertical”, the incident light angle is varied along the shorter array dimension.
Figure 6. Monochrome Relative Angular QE (with Microlens)
Figure 7. Bayer Relative Angular QE (with Microlens)

KAC−12040
www.onsemi.com
11
Dark Current vs. Temperature
Figure 8. Dark Current vs. Temperature
NOTE: “Dbl” denotes an approximate doubling temperature for the dark current for the displayed temperature range.
Power vs. Frame Rate
The most effective method to use the maximum PLL2
speed (313 →320 MHz) and control frame rate with
minimum Power and maximum image quality is to adjust
Vertical Blanking. (register 01F1h). Unnecessary chip
operations are suspended during Vertical Blanking
conserving significant power consumption and also
minimizing the image storage time on the storage node when
in Global Shutter Operation.
Figure 9. Power vs. Frame Rate, 10 bit Mode
NOTE: The LVDS clock is ½the PLL2 clock speed.

KAC−12040
www.onsemi.com
12
Power and Frame Rate vs. ADC Bit Depth
Increasing the ADC bit depth impacts the frame rate by
changing the ADC conversion time. The following figure shows the power and Frame rate range for several typical cases.
Figure 10. ADC Bit Depth Impact on Frame Rate and Power

KAC−12040
www.onsemi.com
13
DEFECT DEFINITIONS
Table 11. OPERATION CONDITIONS FOR DEFECT TESTING
Description Condition Notes
Operational Mode 10 bit ADC, 8 LVDS Outputs, Global Shutter and Rolling Shutter Modes,
Dual-Scan, Black Level Clamp ON, Column/Row Noise Corrections ON,
1×Analog Gain, 1×Digital Gain
Pixels per Line 4,000
Lines per Frame 3,000
Line Time 8.7 ms
Frame Time 13.9 ms
Photodiode Integration Time 33 ms
Storage Readout Time 13.9 ms
Temperature 40°C and 29°C
Light Source Continuous Red, Green and Blue LED Illumination 1
Operation Nominal Operating Voltages and Timing, PLL1 = 320 MHz, Wafer Test
1. For monochrome sensor, only the green LED is used.
Table 12. DEFECT DEFINITIONS FOR TESTING
Description Definition Limit Test Notes
Dark Field Defective Pixel 30°C
RS: Defect ≥20 dn
GS: Defect ≥180 dn
40°C
RS: Defect ≥30 dn
GS: Defect ≥240 dn
120 4 1, 4, 5
Bright Field Defective Pixel Defect ≥±12% from Local Mean 120 5 2, 5
Cluster Defect A group of 2 to 10 contiguous defective pixels, but
no more than 3 adjacent defects horizontally. 22 3
Column/Row Major Defect A group of more than 10 contiguous defective pixels
along a single column or row. 0
Dark Field Faint Column/Row Defect RS: 3 dn Threshold
GS: 10 dn Threshold 0 17 1
Bright Field Faint Column/Row Defect RS: 12 dn Threshold
GS: 18 dn Threshold 0 18 1
1. RS = Rolling Shutter, GS = Global Shutter.
2. For the color devices, all bright defects are defined within a single color plane, each color plane is tested.
3. Cluster defects are separated by no less than two good pixels in any direction.
4. Rolling Shutter Dark Field points are dominated by photodiode integration time, Global Shutter Dark Field defects are dominated by the
readout time.
5. The net sum of all bright and dark field pixel defects in rolling and global shutter are combined and then compared to the test limit.
Defect Map
The defect map supplied with each sensor is based upon
testing at an ambient (29°C) temperature. All defective pixels are reference to pixel (0, 0) in the defect maps. See Figure 11 for the location of pixel (0, 0).

KAC−12040
www.onsemi.com
14
TEST DEFINITIONS
Test Regions of Interest
Image Area ROI: Pixel (0, 0) to Pixel (4015, 3015)
Active Area ROI: Pixel (8, 8) to Pixel (3999, 2999)
Center ROI: Pixel (1958, 1458) to Pixel (2057, 1557)
Only the Active Area ROI pixels are used for performance and defect tests.
Figure 11. Regions of Interest
G
B
GR
4000 (H) y3000 (V)
4.7 mm Pixel
G
B
GR
G
B
GR
8
88
8
88
8
104
8
104
0,0
8,8
Test Descriptions
1) Dark Field Local Non-Uniformity Floor (DSNU_flr)
This test is performed under dark field conditions.
A 4 frame average image is collected. This image is
partitioned into 300 sub-regions of interest, each of which is
200 by 200 pixels in size. For each sub-region the standard
deviation of all its pixels is calculated. The dark field local
non-uniformity is the largest standard deviation found from
all the sub regions of interest. Units: e−rms (electrons rms).
2) Bright Field Global Photoresponse Non-Uniformity
(PRNU_1)
The sensor illuminated to 70% of saturation (~700 dn). In
this condition a 4 frame average image is collected. From
this 4 frame average image a 4 frame average dark image is
subtracted. The Active Area Standard Deviation is the
standard deviation of the resultant image and the Active
Area Signal is the average of the resultant image.
PRNU_1 +100 @ǒActive Area Standard Deviation
Active Area Signal Ǔ
Units : % rms
3) Bright Field Global Peak to Peak Non-Uniformity
(PRNU_2)
This test is performed with the sensor uniformly
illuminated to 70% of saturation (~700 dn), a 4 frame
average image is collected and a 4 frame averaged dark
image is subtracted. The resultant image is partitioned into
300 sub regions of interest, each of which is 200 by
200 pixels in size. The average signal level of each sub
regions of interest (sub-ROI) is calculated.
The highest sub-ROI average (Maximum Signal) and the
lowest sub-ROI average (Minimum Signal) are then used in
the following formula to calculate PRNU_2.
PRNU_2 +100 @ǒMax. Signal *Min. Signal
Active Area Signal Ǔ
Units : % pp
4) Dark Field Defect Test
This test is performed under dark field conditions.
The sensor is partitioned into 300 sub regions of interest,
each of which is 128 by 128 pixels in size. In each region of
interest, the median value of all pixels is found. For each
region of interest, a pixel is marked defective if it is greater
than or equal to the median value of that region of interest
plus the defect threshold specified in the Defect Definition
Table section.
5) Bright Field Defect Test
This test is performed with the imager illuminated to
a level such that the output is at approximately 700 dn.
The average signal level of all active pixels is found.
The bright and dark thresholds are set as:
Dark Defect Threshold = Active Area Signal ⋅Threshold
Bright Defect Threshold = Active Area Signal ⋅Threshold
The sensor is then partitioned into 300 sub regions of
interest, each of which is 128 by 128 pixels in size. In each
region of interest, the average value of all pixels is found.
For each region of interest, a pixel is marked defective if it
is greater than or equal to the median value of that region of
interest plus the bright threshold specified or if it is less than

KAC−12040
www.onsemi.com
15
or equal to the median value of that region of interest minus
the dark threshold specified.
Example for bright field defective pixels:
•Average value of all active pixels is found to be 700 dn
•Lower defect threshold: 700 dn ⋅12% = 84 dn
•A specific 128 ×128 ROI is selected:
♦Median of this region of interest is found to be
690 dn.
♦Any pixel in this region of interest that is
≤(690 − 84 dn) in intensity will be marked
defective.
♦Any pixel in this region of interest that is
≥(690 − 84 dn) in intensity will be marked
defective.
•All remaining 299 sub regions of interest are analyzed
for defective pixels in the same manner.
6) Parasitic Light Sensitivity (PLS)
Parasitic Light Sensitivity is the ratio of the light
sensitivity of the photodiode to the light sensitivity of the
storage node in Global Shutter. There is no equivalent
distortion in Rolling Shutter. A low PLS value can provide
distortion of the image on the storage node by the scene
during readout.
PLS +Photodiode Responsivity
Storage Node Responsivity (UnitlessRatio)
GSE (Global Shutter Efficiency) is a related unit.
GSE +ǒ1*1
PLSǓ%
Detailed Method: Photodiode Responsivity:
The sensor is set in global shutter serial mode (integration
time not overlapping readout) and the FLO signal is used to
control a 550 nm normal incident (or large f# focused)
illumination source so that the sensor is illuminated only
during photodiode integration time (not illuminated during
readout time). The integration time is not critical but should
be large enough to create a measurable mean during this
time. A 16 frame-average illuminated photodiode image is
recorded. A 16 frame-average dark frame using the same
sensor settings is captured and is subtracted from the
illuminated image.
Detailed Method: Storage Node Responsivity:
The sensor is set to a special characterization mode where
the PD signal is discarded and does not impact the storage
node. A long total frame time (storage node exposure time)
is used to increase the storage node signal. A 16
frame-average dark frame is captured. The sensor is
illuminated by the same 550 nm incident light source used
for the photodiode responsivity. A 16 frame-average
illuminated photodiode image is recorded; the dark frame
image is subtracted from this. The integration time is not
critical but should be set such that a significant response is
detected, typically several orders of magnitude greater than
the photodiode integration time.
7) Black-Sun Anti-Blooming
A typical CMOS image sensor has a light response profile
that goes from 0 dn to saturation (1023 dn for KAC−12040
in 10 bit ADC mode) and, with enough light, back to 0 dn.
The sensor reaching 0 dn at very bright illumination is often
called the “Black-sun” artifact and is undesirable. Black-sun
artifact is typically the dominant form of anti-blooming
image distortion. For the KAC−12040 the Black-sun artifact
threshold is measured at the onset of saturation distortion,
not at the point where the output goes to 0 dn. To first order
the onset of black-sun artifact for the KAC−12040 is not
proportional to the integration time or readout time.
The sensor is placed in the dark at unity gain and
illuminated with a 532 nm laser with the intensity of about
26 W/cm2at the center of the sensor. The laser is strong
enough to make the center of the laser spot below 1020 dn
without any ND filters. ND filters are added to adjust the
laser intensity until the signal in the region at the center of
the spot increases to > 1020 dn.
This illumination intensity at this ND filter is recorded
(W/cm2) as the Black-Sun Anti-blooming.
The ‘xIlumSat’ unit is calculated using and integration
time of 100 msec.
Exposing the sensor to very strong illumination for
extended periods of time will permanently alter the sensor
performance in that localized region.
8) Read Noise
This test is performed with no illumination and one line of
integration time. The read noise is defined as one standard
deviation of the frequency histogram containing the values
of all pixels after the excessively deviant pixels (±three
standard deviations) are removed.
9) Column Noise
After all rows are averaged together. Shading (low
frequency change wrt column address) is removed.
A frequency histogram is constructed of the resulting
column values. The column noise is the standard deviation
of the frequency histogram of the column values.
10) Row Noise
All columns are averaged together. Shading (low
frequency change wrt row address) is removed. A frequency
histogram is constructed of the resulting row values.
The row noise is the standard deviation of the frequency
histogram of the row values.
11) Maximum Photoresponse Non-Linearity
The photoresponse non-linearity is defined as the
deviation from the best fit of the sensor response using 70%
of saturation and zero signal as the reference points.
The different signal levels are determined by varying the
integration time. The sensor saturation level is (1023-dark
offset). The dark offset is subtracted from the image for the
following MAVG and LAVG.

KAC−12040
www.onsemi.com
16
•The integration time is varied until the integration time
required to reach the 70% saturation is determined.
MAVG = the active array mean at the 70% saturation
integration time.
•The integration is set to 1/14 (5% exposure point).
LAVG = meant at the 5% exposure point.
•PRNL (@ 5% saturation) = ((LAVG/MAVG) ⋅(14/1) −1)
⋅100
12) Maximum Gain Difference between Outputs
The sensor contains two ADC and four channels of analog
data in its highest frame rate configuration. The sensor is
factory calibrated to reduce the gain differences between the
channels. The gain variations are manifest as a row oriented
pattern where every other row uses a different ADC. Using
triple scan read out mode, an additional two analog channels
are introduced resulting in a four row pattern. With one
channel (‘Top Ping’) used as the reference, the residual gain
difference is defined as:
ǒBottom Ping Row Average
Top Ping Row Average *1Ǔ@100
ǒTop Pong Row Average
Top Ping Row Average *1Ǔ@100
ǒBottom Pong Row Average
Top Ping Row Average *1Ǔ@100
13) Photodiode Dark Current
The photodiode dark current is measured in rolling shutter
read out mode using 105 ms integration time and an analog
gain = 8. The value is converted to electrons/pix/sec using
the formula:
Photodiode Dark Current +Aver. Signal (DN) @el−per−DN (gain=8)
0.105 seconds
where ‘average signal (DN)’ is the average of all pixels in
the sensor array, and ‘el-per-DN (gain=8)’ is measured on
each sensor using the photon transfer method.
14) Storage Node Dark Current
The storage node dark current is measured in global
shutter read out mode using a special timing mode to prevent
the photodiode dark current from being transferred to the
storage node. In global shutter mode, the integration time of
the storage node is the time it takes to read out a frame. The
sensor analog gain is set to 2:
Storage Node Dark Current +Aver. Signal (DN) @el−per−DN (gain=2)
0.138 seconds
where ‘average signal (DN)’ is the average of all pixels in
the sensor array and ‘el-per-DN (gain=2)’ is measured on
each sensor using the photon transfer method.
15) Lag
Lag is measured as the number of electrons left in the
photodiode after readout when the sensor is illuminated at
70% of Photodiode Charge Capacity.
Analog gain is set to 8. With no illumination a 64 average
dark image is recorded (Dark_ref). The ‘el-per-DN’ is
measured using the photon transfer method.
Illumination is adjusted blink every other frame such that
the mean image output is 70% of the Photodiode Charge
Capacity for even frames, and with no illumination for odd
frames. A 64 frame average of Odd Dark Frames is recorded
as Dark_Lag.
Lag +(Dark_Lag *Dark_Ref)@el−per−DN
Units : Electrons rms
16) Photodiode Charge Capacity
The sensor analog gain is reduced to < 1 to prevent ADC
clipping at 1023 dn. The ‘el-per-DN’ is measured using the
photon transfer method. The sensor is illuminated at a light
level ~1.5x the illumination at which the pixel output no
longer linearly changes with illumination level.
The Photodiode Charge Capacity is equal to the average
signal (DN) ⋅el-per-DN. Units: electrons rms.
17) Dark Field Faint Column/Row Defect
A 4 frame average, no illumination image is acquired at
one line time of integration. Major defective pixels are
removed (> 5 Sigma). All columns or rows are averaged
together. The average of the local ROI of 128 columns or
rows about the column/row being tested is determined. Any
columns/rows greater than the local average by more than
the threshold are identified.
18) Bright Field Faint Column/Row Defect
A 4 frame average, 70% illumination image is acquired at
one line time of integration. Major defective pixels are
removed (> 5 Sigma). All columns or rows are averaged
together. The average of the local ROI of 128 columns or
rows about the column/row being tested is determined. Any
columns/rows greater than the local average by more than
the threshold are identified.
19) Total Pixelized Noise
This test is performed with no illumination and one line of
integration time. A single image is captured including both
Temporal and Fixed Pattern Noise (FPN). A spatial low pass
filter is applied to remove shading and deviant pixels
(±three standard deviations) are removed. The Total
Pixelized Noise is defined as one standard deviation of the
frequency histogram.
20) Responsivity ke−
/lux-sec
This number is calculated by integrating the
multiplication of the sensor QE by the human photopic
response assuming a 3200K light source with a QT100 IR
filter. This is a sharp 650 nm cutoff filter. If the IR filter is
removed a higher response value will result.
21) Responsivity V/lux-sec
Voltage levels are not output from the sensor. This metric
uses the pixel output in volts at the ADC input for 1x Analog
Gain.

KAC−12040
www.onsemi.com
17
OPERATION
This section is a brief discussion of the most common
features and functions assuming default conditions. See the
KAC−12040 User Guide for a full explanation of the sensor
operation modes, options, and registers.
Register Address
The last bit of any register address is a Read/Write bit.
Most references in this document refer to the Write address.
All SPI reads are to an even address, all SPI writes are to an
odd address.
Sensor States
Figure 12 shows the sensor states, see the KAC−12040
User Guide for detailed explanation of the States.
Figure 12. Sensor State Diagram
RESETN low or
reset Reg 4060h
RESET
STANDBY
CONFIG
IDLE
RUNNING
WAKE−UP
(50 ms)
<35µ s
<2µ s
<2µ s
150µ s
RUNNING mode OR
TRIGGER pin
<50µ s
End of acquisition AND
IDLE mode AND
No TRIGGER
Slave Integration Mode
<50µ s
TRIG_WAIT
<2µ s
End of
acquisition
READOUT
TRIGGER Active Edge
EXT_INT
TRIGGER Inactive Edge

KAC−12040
www.onsemi.com
18
Encoded Syncs
To facilitate system acquisition synchronization the
KAC−12040 places synchronization words (SW) at the
beginning and at the end of each output row as indicated in
the following Figure 13. This is performed for each of the
8 LVDS output banks providing frame, line, and output
synchronization. See the KAC−12040 User Guide for
additional detail on LVDS and Encoded Sync output.
Figure 13. Encoded Frame Syncs
V Blanking Period
V Blanking Period
DataSOL
SOF
EOL
EOF
H Blanking Period
Line Length (LL)
Line Time
This Datasheet presumes the recommended startup script
that is defined in the KAC−12040 User Guide has been
applied. The KAC−12040 defaults to Dual-Scan mode. In
this mode the LVDS data readout overlaps the pixel readout
and ADC conversion time. The Pixel and ADC conversion
time are fixed (for 10 bit operation) and total ~8.66 ms.
The LVDS time will be dependent on the PLL2 frequency
selected. If the PLL2 < 313 MHz, then the LVDS data
readout will dominate the row time. For PLL2 > 313 MHz,
the Pixel + ADC will set the minimum Line Time. The Line
Time is not impacted by the selection of Rolling Shutter or
Global Shutter mode.
The KAC−12040 architecture always outputs two rows at
once, one row from the top ADC, and one from the bottom
ADC. Each ADC then divides up the pixel into 1 →4
parallel pixel output LVDS Banks. The default is 4 output
banks per ADC for a total of 8 parallel pixel outputs to
minimize the LVDS data output time. Since the sensor
always outputs 2 rows at a time the timing and registers are
based on a Line Time (LT) or Line Length (LL) where one
LT = the time to readout 2 rows in parallel (one even row and
one odd row).
Figure 14. Default Line Time (Dual-Scan) with PLL2 = 320 MHz
Pixel n+1 10 bit ADC n+1 Line
Wait
8 Bank LVDS Output n
LVDS Clock = 160 MHz
Line n Time = Line Length Register (0201h) = 1400
4.30 ms 4.36 ms
8.44 ms
8.75 ms

KAC−12040
www.onsemi.com
19
Frame Time
The frame time is defined in units of Line Time. 1 Line
Time unit = 2 output rows. To first-order the frame rate is not
directly impacted by selection of Global Shutter, Rolling
Shutter, Dual-Scan, or Tri-Scan.
The Frame Time is made up of three phases:
1. Integration Phase
2. Readout Phase
3. Frame Wait Phase (Vertical Blanking, VBLANK)
By default the Integration Phase overlaps the Readout and
Frame Wait Phases. If the Integration Phase is larger than the
Readout + Frame Wait time, then the Integration Phase will
determine the video frame rate. Otherwise the frame rate
will be set by the Readout + Frame Wait time. In other words,
if the programmed integration time is larger than the
minimum readout time (and vertical blanking) then extra
vertical blanking will be added and the frame rate will slow
to accommodate the requested integration time.
Figure 15. Default Frame Time Configuration (Frame A)
Single Frame of Video Overhead
Integration Phase Frame mIntegration Phase Frame m+1 Integration Phase Frame m+2
Readout Phase Frame mReadout Phase Frame m+1
Frame
Wait Frame Wait
13.80 ms by Default
Video Frame Time = Readout + Wait
8.34 ms by Default (952 Line Times)
13.79 ms by Default
(1576 Line Times) ~0.01 ms
(1 Line Time)
If the Integration Phase is less than the Readout Phase then
the start of integration is automatically delayed to minimize
the storage time and dark current.
Figure 16. Frame Time with Extended Integration Time
Integration Phase Frame m+2
Readout Phase Frame m+1 Frame
Wait
Video Frame Time = Integration Time
Integration Phase Frame m+1
Readout Phase Frame mFrame
Wait
Integration Phase Frame m
If the Readout Phase (+ VBLANKING) is less than the
Integration Phase, then the readout occurs as soon the
integration is complete to minimize the storage time and
dark current.
See the KAC−12040 User Guide for detailed calculation
of the Integration Phase, Readout Phase, and Frame Wait.
To first-order the Readout Phase is equal to the number of
rows ⋅row_time.

KAC−12040
www.onsemi.com
20
Global Shutter Readout
Global Shutter readout provides the maximum precision
for freezing scene motion. Any motion artifacts will be
100% defined by an ideal integration time edge. Every pixel
in the array starts and stops integration at the same time.
Figure 17 illustrates a Global Shutter Frame readout
assuming the recommended Start-up Script defined in the
KAC−12040 User Guide (8 LVDS banks, Dual-Scan,
8.75 ms line time). The Frame Wait Phase is not shown due
to its small default size (1 LL) and for clarity.
Global Shutter readout mode is selected using Bits [1:0]
of Register 01D1h.
Images can be initiated by setting and holding the
TRIGGER input pin or by placing the sensor into
RUNNING mode by writing 03d to register 4019h. If the
TRIGGER input pin is true when at the start of the
integration time for the next frame then the sensor will
complete an additional frame integration and readout. In the
case shown in Figure 17 two frames will be output.
Figure 17. Illustration of Frame Time for Global Shutter Readout
Integration Time = 8.33 ms
Integration of Next Frame Overlaps
Readout of Previous Frame
Frame Readout = 13.80 ms
Effective Frame Time (Video) = Readout Time
Trigger Pin: True = 2.0 V
Time/Col Address Axis
Row Address Axis