Re: Never Build a Flat Roof

[Tom Droege <droege@wwa.com>]

Well, it was worse than I thought.  The roof really is leaking, but
Jennifer went up on the roof and pulled off some of the snow, and this
should help. 

The real problem was that in the midst of all this I started a run to test
the scan program and in the process I pulled the return line out of the
garbage can.  Then I went off to eat lunch.  When I returned, I found the
floor covered with water.  I naturally thought it was from the leak in the
roof and spend some time trying to figure out how the water was getting on
to the floor. 

Well, it was not water but the water anti-freze mixture.  Very gunky stuff,
but I am using the pet safe kind, so the cats are OK.  I  did not notice
this until I noticed that the camera temperatures were high.  Sigh!
Fortunately the cameras are not destroyed without cooling water as there is
a lot of metal there to soak up the heat.

By this time I was pretty discouraged.  The cameras were not working, the
floor was covered with water, the roof was leaking, and I could see
synchronous noise on the exposures.

But I kept at it.  I removed the x6 gain in the pre-amps and did another
stab at a noise and gain measurement.  The number now made sense.  But now
I can really see the pattern noise.  It is now about 33 e-.  This should be
OK because I think I can get a x4 gain in the preamp, so this would almost
divide the pattern noise by 4.  In any case even the 33 e- is such that we
would not see it in normal data.  But I will try to make it better. 

I have just completed a focusing run with the new gain, and the data makes
sense.  It saturates about where it should - of order 90,000 e-, the data
sheet says 80,000.  Even the corners of the frame do not look too bad.
Smeared out with coma? bit usable I think.  At least much better than the
camera lens I was using before.  

As I type I am running a sequence where I just keep taking pictures at one
declination while tracking. 

Alls well that ends well, I hope.

Tom Droege