XRD98L61 Datasheet, PDF (22/37 Pages) Exar Corporation – CCD Image Digitizers with CDS, PGA and 12-Bit A/D

XRD98L61
Preliminary
To get the maximum color resolution and dynamic
range, the XRD9861 uses a digitally controlled cali-
bration circuit to correct for offset in the CCD signal as
well as offset in the CDS, PGA & ADC signal path. This
calibration is done while the CCD outputs Optical
Black (OB) pixels. In the “line” timing mode, the OB
pixels at the start or end of each scan line are used for
calibration.
Hot Pixel Clipper
CCD’s occasionally have hot pixels. These are defec-
tive pixels, which always output a bright level. To
ensure the Black Level is not affected by hot pixels in
the OB area, the Hot Pixel Clipper limits pixel data from
the ADC to a maximum value of 511 (1FFh). This
clipping only affects the data used by the internal
calibration logic. Data on the ADC output bus,
DB[11:0], is not clipped.
4096
AVG[2] AVG[1] AVG[0]
0
0
0
0
0
1
0
1
0
0
1
1
1
0
0
1
0
1
1
1
0
1
1
1
# of Pixels
to Average
4
8
16
32
64
128
(default)
256
512
Table 3. Programming the Pixel Averager
Offset Difference
Next, the Offset register value, OB[7:0], is subtracted
from the OB pixel average. If the difference is positive,
the offset DACs are adjusted to reduce the effective
ADC output code. If the difference is negative, the
offset DACs are adjusted to increase the effective
ADC output code. The FAST_CAL and DNS options
will affect how the DAC adjustments are made.
511
0
0 511
2048
ADC Data
4096
Coarse & Fine Accumulators
The Coarse and Fine Accumulators are the registers
which hold the digital codes for the Coarse and Fine
Offset DACs. The Offset DAC adjustments are made
by adding or subtracting to the value in the Fine
accumulator. If there is an overflow or underflow in the
Fine Accumulator, the Fine Accumulator is reset to it’s
mid-scale value, and the Coarse Accumulator is
incremented or decremented accordingly.
Figure 9. Hot Pixel Clipper
Pixel Averager
After the clipper, the logic takes an average of Optical
Black pixels. The number of pixels to be averaged can
be selected as one of the following 4, 8, 16, 32, 64, 128,
256, or 512. The AVG[2:0] bits in the Calibration
register are used to program the number of pixels to
average. This averaging function filters out noise and
prevents image artifacts. The calibration logic will
average OB pixels over as many lines as required to
get the programmed number of pixels to average.
Rev. P4.00
22