chling@xxxxxxxxxxxx writes:
<< If my calculation is right a 4 times improvement is equal to a D-range
of around 0.6, it is a huge improvement.
C.H.Ling >>
Warning engineering content follows:
CH,
This is just my speculation but it is likely given the high light
levels from the lamp that system performance is limited by amplifier noise
and A/D convertor resolution not CCD performance.
If we assume no amplifier noise and consider CCD noise (relative to A/D
resolution) then since Noise is related to signal power it adds as Root Mean
Square (RMS). The cooling effect is thus not as large as you might imagine
for a small temp range. If we assume the system is well designed so the CCD
noise is at the same RMS level as half the least significant bit (LSB) of the
A/D at room temperature, then if the dark current increases by a factor of
four as the device heats up, this actually only produces sqrt(4)*0.5LSB or 1
LSB of additional noise fluctuation.
Noise in the CCD may be limited by other factors than just dark current in
the detector layer. For example clocking out the data adds significant noise.
I seem to remember reading in Photonics and Spectra that the noise in CCD
chips does not decrease much as temperature decreases unless the chips are
specialy optimized for use cooled down. In other words for example if clock
noise say, increases with decreasing temperature (chip gets faster ) it may
partially or completely cancel detector layer dark current noise decreases.
Regards,
Tim Hughes
>>Hi100@xxxxxxx<<
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|