Polaroid SprintScan SS35, Руководство по ремонту

Страница: 154 / 187

Slide Scanner Repair Manual Diagnostics and Troubleshooting
154
Part one of the Uniformity test checks for dust or dirt on the CCD. Dust is identified as
deviations from a polynomial curve fit to the illumination profile that exceed the
Fine_uniformity_tolerance set in the constants file (initially at 9%). Failure of this test will
occur for a single violation of the threshold, and will generally require removal and cleaning of
the CCD sensor. The location of the dust spots can be seen by running the stand-alone
version of the Uniformity II test. An array of dust locations for each of the R,G,B channels is
displayed on the control panel. The same information can be seen on the fixed pattern
illumination plots as shown in Figure 5-19 or, by looking at the analog video signal on an
oscilloscope while the sensor is illumination but, not saturated.
Part two of the test lists a 5 point profile of the uncorrected uniformity profile for each channel
of the CCD. The 5 points come off of the curve fit of the slide scanner’s uniformity calibration
coefficients, and represent the basic illumination profile of the lamp and optics. Peak
uniformity is normalized to 100% for the profiles. The test fails if the left or right edge
illumination is less than the Gross_uniformity_tolerance, or 65%, of the center uniformity.
Typical failure is due to an aged lamp with dark edges, or some obstruction to the optical path.
The Mean Square Fit Error is a measure of high frequency deviation of the actual CCD
uniformity response from the curve fit. It is not tested against limits, but provides a metric of
the uniformity of the CCD’s response.
Slide scanner’s which pass the scanner’s built-in calibration, but fail either the Gross or Fine
uniformity test may perform near flawlessly in the factory, but will likely develop streaks in the
field as the lamp dims or darkens at the edges, forcing the raw response out the slide
scanner’s correction range. This is the reason for the replacement of the old target-based
uniformity test.
Illumination
The Illumination portion of this test fails if it detects the scanner is greater than 10%
over-saturated or, if the slide scanner is operating 10% below the normal operating range and
speed. A factory new slide scanner is typically at 100% illumination which covers a raw peak
signal in the range of ~0.36V to 2V off each channel of the CCD. A slide scanner will attempt
to operate anywhere between ~50% and 130% illumination unless some portion of the
calibration procedure fails. However, slide scanners which are significantly saturated will be
more likely to produce unstable output and/or streaks, and will take longer to calibrate. Slide
scanners which are operating with dim lamps will run slower than normal and be slightly
noisier than a nominal slide scanner. Either failure condition is a function of:
• Lamp and lamp driver
• ND/IR transmission
• CCD sensitivity and color balance
This illumination test result is identical to the Illumination result which was recorded, but not
tested, in the previous versions of the Scanner Tester for the SprintScan 35ES product. It
should not be confused with the uncorrected light uniformity profiles which are themselves
normalized to their peak response.