Time wrote:
> I would guess you would have trouble to get the spacing right
> since the ccd is likely to be soldered down like a chip to a
> printed circuit
> board. The peltier coolers are about 3mm thick and to make them work
> effectively you need to heatsink the back side. This is because the
> efficiency is quite low so you want the back to stay as close to room
> temperature as possible. Otherwise the delta temperature is split
> between the
> self heating ,heat pumping etc. rising above ambient. Adding a
> heatink makes
> it even thicker and more difficult to fit. Typically the best
> delta T you can
> achuieve pumping almost a zero heat load is of the order of
> 20-25C delta T.
Yes, I got some of this info from the link somebody posted.
> Another problem is that condensation may cause leakage problems
> on the ccd
> device as well as blurring the front window. In the typical military
I guess the condensation depends on the delta T plus prevailing humidity -
most likely would be a problem.
> applications where peltier cooling is used on detectors, they cascade two
> coolers to get a bigger delta T and place the whole assembly in
> an evacuated
> container to prevent condensation on the front window as well as reduce
Takes me back to my days using a semiconductor radiation detector !
> conductive losses on the front. Unless the CCD device is running very hot
> (seems unlikely since the dissipation is not huge usually) you don't even
> gain much with a 20C increment. Typically the dark current halves
> approximately every 10C reduction so 20C is a 1/4 of the noise
> current. Not a
> huge improvement especially as the circuit may be limited by the A/D
> convertor resolution or other factors not the detector.
Of course - the CCD may not be the culprit.
Thanks for the detailed analysis.
...Wayne
< 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 >
|