
TLC8044
12-BIT ANALOG-TO-DIGITAL INTERFACE FOR
CHARGE-COUPLED DEVICE IMAGE SENSORS FOR SCANNERS
SLAS128 – JUNE 1997
20
POST OFFICE BOX 655303
DALLAS, TEXAS 75265
PRINCIPLES OF OPERATION
correlated double sampling mode (continued)
To perform CDS input sampling, the device should be set up in single-ended mode with the differential inputs
of each channel (xINP, xINN) connected together to the video input signal.
For positive going input signals, i.e., white signal level greater than the reset level, the offset DACs should be
set to the maximum negative value (DAC code 1FF hex). This configuration sets the zero signal level (video
input equal to the reset level) at the ADC zero-scale code transition. Increasing the DAC code towards zero
moves the zero signal level up from the ADC zero-scale code transition. For negative-going input signals, i.e.,
white signal level less than the reset level, the DACs should be set to the maximum positive value (DAC code
0FF hex). This configuration sets the zero signal level at the ADC full-scale code transition. The polarity of the
ADC output signal can be inverted under control of the serial interface data.
The multiplexing shown in Figure 12 refers to color operation, however the same overall timing scheme applies
to monochrome CDS operation, in that a single input sample is applied to the ADC per VSMP period. Thus the
maximum sampling rate in monochrome CDS mode is limited to one third of the maximum rate achievable in
normal monochrome sampling mode.
digital image processing
The digital image processing functions following the ADC as shown in the functional block diagram include the
following:
DC restore: This allows fine adjustment of the dc video level at the ADC output with adjustment values
being programmed through the serial interface.
Pixel-by-pixeloffsetcompensation:Thisusesoffsetcoefficientsthatareeitherexternallysuppliedatthe
multiplexed channel rate or supplied from internal default registers whose values are programmed
through the serial interface.
Compensation for pixel-by-pixel shading curve nonuniformity and photo response nonuniformity within
the sensor: Coefficients are externally supplied at the multiplexed channel rate. Default registers are
provided for use during calibration.
Global offset adjust: Offset adjust over the whole scan for each channel to give brightness control.
Values are programmed through the serial interface.
Global gain adjust: Independent gain adjust over the whole scan for each channel to give contrast and
color balance control. Gain values are programmed through the serial interface.
Programmableoutputwordlengthselection:Theoutputwordlengthcanbeprogrammedto8,10,12,or
16 bits through the serial interface.
Programmable threshold detector with independent thresholds for each channel.
Global adjustments are implemented after the pixel-by-pixel compensations allowing calibrations and
modifications in operational use without having to recalibrate the pixel-by-pixel factors.
DC restore
The dc restore block is used for fine adjustment of the dc signal level at the ADC output by adding a value stored
in an internal register. Separate level adjust registers are provided for each channel (color) with multiplexing
between channels controlled internally. The level adjust registers are programmed through the serial interface
as 12-bit 2s complement numbers with a range of
±0.5 of the ADC full scale, allowing 1-bit resolution in
adjustment of the ADC output. The dc adjustment registers are reset to zero.