Re: The promised format

[Herb Johnson <hjohnson@pluto.njcc.com>]

I agree with Andrew, the columns 1, 3, and 5 have too little variance
to be PIXEL measures. In other words, they only vary by one or two ADU
counts; all the other columns, even in the dark images, vary much more
as one would expect. There is some value in measuring amplifier voltages,
or the reset value of the amplifier, as well as a few of the dark
covered pixels. Measuring the dark covered pixels in every image provide
a check as to the constantcy of the camera. For instance if the CCD
is warmer the covered pixels will have a higher value; if the timing
or exposure changes their value will change, etc. I disagree with Andrew
about delaying work on these columns.

Tom, you might try some simple checks to see what the source of the
measured values is, such as disconnecting some circuit paths to see
what changes. Keep in mind you already have a pretty good A/D and 
controller in-circuit, think of ways you can use IT to make the 
measurements. Maybe you can make a simple probe and connect it
directly to the A/D, and probe around to see what values are what.
I find that looking at old circuits in new ways like this may 
provide work-arounds to your instrumentation limitations.
Another possibility: disconnect pins of the ROM that sequence
the A/D's measurements, again see the results and thinkk through WHY
these occurred. New data may yield new insights. Hope these remarks
are helpful, they are not new ideas but few ideas are new.

Herb Johnson

Herbert R. Johnson              http://pluto.njcc.com/~hjohnson
hjohnson@pluto.njcc.com         voice 609-771-1503, New Jersey USA

             amateur astronomer and astro-tour guide
     classic S-100 computers restoration & parts as "Dr. S-100"
   rebuilder/reseller of classic Macs for your computing pleasure
          and senior engineer at Astro Imaging Systems

On Thu, 7 Dec 2000, Andrew Bennett wrote:

> On Thu, 7 Dec 2000 08:24:03 -0500, "Michael Gutzwiller" 
> <deepsky@fuse.net> wrote:
> 
> >I have looked at the images and can give some clues about where the problem
> >may lie and where the imaging pixels are.  I analyzed the first 7 columns
> >and last 6 columns of image H3R1836.529.  Here are the results:
> >
> >Column	Median	Noise
> >   0		 7860		 95.8
> >   1		 6375		  1.8
> >   2		 7271		  7.1
> >   3		 6374		  2.1
> >   4		 7424		 32.2
> >   5		 6375		  2.1
> For what it is worth, my theory is that 1, 3 and 5 are
> full-scale negative for the amplifier ahead of the
> digitizer. My guess is that the timing is screwed up
> and these are three extra samples being taken at times
> when the digitizer is not actually connected to anything.
> 
> Tom: don't be in too much of a hurry to fix it. I have
> all the numbers for the present format firmly built in
> to my programs!
> 
> Andrew Bennett, Avondale Vineyard, Nova Scotia, Canada.
> 
>