S90 - A CRAZY GOOD CONDITION for ASTRONOMY // PHOTOS ATTACHED
This could be a first ... an S90 running "perfectly" under CHDKPTP without all of these (Fig 1):
a) No Flash (left). You get a warning on PUP but camera continues run.
b) No button panel (foreground). There are Hall-effect sensors for dial, push-button switches and activity LED.
c) No LCD (standing up background).
d) No ring function (front camera cover, background, lying down).
e) No top cover selector switch (between button panel & LCD). PUP defaults to AUTO, but you can naturally select the mode you want with the dropdown in CHDKPTP. Here I have it running in M mode.
CHDKPTP liveview is in the bottom left of the screen.
CONCLUSIONS
As it stands at present, I could safely say the S90 CCD is useable in primary focal plane architecture with excellent access to the chip. For most astronomy applications I feel there is enough room to swing out the CCD and make it face either perpendicular to the camera body or parallel in front of the large black chip to the right of its connector. Hence either way there is much space to engineer a temperature control cavity, and in my experience, you won't need anything lower than +15C to handle hot pixels reasonably well right out to the extended Tv of 64s offered by CHDK. Even in the parallel position, there is 1 cm of clearance between the main board and the back of the CCD.
For my project specifically, I attempted to disconnect the lens signal connector (Fig 2) in a desperate hope that the camera would continue running with the rationale that a full disconnect would be detected by firmware as a maintenance/test condition i.e., not be "as bad" as a jammed lens (entirely possible). Alas, not it was not to be ... camera booted up but then failed with REC attempt and would not boot again unless I reconnected lens control. AFAIK nothing was damaged in this test. So for me personally, it is still an uphill battle. These new results (a to e above) are nonetheless encouraging and are in fact a preface to Phase III, which of course I am convinced now will have to be undertaken. Despite that, these results here give me good access to tracing the signal lines with the scope. The lens even when not extended takes up more than 1/2 the volume of the electronics and no matter what position I put the CCD in is not useable as-is. Getting rid of the lens assembly effectively reduces the entire electronics with the CCD to a very tiny ~1 cu in allowing me excellent flexibility in optical positioning and thermal design in the tight space demanded by the instrument.
For its small 1/1.7" format, this is a superlative low-light CCD and signal processor combination for 12-bit RAWs. I look forward to someone getting spectacular images of Mars ... search for a part of the planet under overcast cloudy skies or better polar ice if you can see it, and set your white balance to pure white. With that I expect we should see publication-quality true planet surface colors never before delivered by NASA. Please let me know. The white balance settings might work well for Saturn too.
~SticK~
