Monday night, I decided to conduct a quick test to confirm whether or not the modifications (IR filter removal and cooling) of my Canon 350D had indeed introduced tilt of the CMOS assembly. Here is what I did:

1. I replaced my modified 350D with my non modified 450D;
2. Focused;
3. Took 15 unguided images of 10 seconds each, near zenith;
4. Analyzed the results with CCD Inspector.

Here are the results:

I am puzzled…it looks like my 450D is not capable of producing stars with similar FWHM across the image plane. I don’t even know how to translate these results?!?

So, after this test, I re-installed the 350D and did an automated run using CCD Commander and shot 38 guided subs of 4 minutes at ISO j800 of IC63. 32 of these proved to be good enough for stacking and processing. The results can be found in the images section of this site here.
The result is pretty good (i.e. 0.17 of flatness) for the final image. I also found something interesting: the firs sub-exposures had on average 0.25 flatness where the last sub-exposures had much more round stars. It might be due to the fact that there were shot near zenith where gravity is not pulling the camera sideways anymore.

Futhermore, tonight I inspected the focuser to field-flattener “connection”. I did found some play. Next steps: ensure that the field-flattener is squared and very well secured to the focuser and run some tests again.

Waiting for clear sky…

François