Re: Flat fields are not the problem

[Michael Gutzwiller <deepsky@fuse.net>]

Tom,

Alignment could very well be the cause of the focus to be different
between the two edges.  The width of the light cone produced by a lens
that is slightly out of focus is equal to f (the focal ratio) times the
distance along the lens axis from the focal plane.  For the Mark III
lens f=3.5 so the light cone is 1 pixel wide (9 um) at a distance of
3.5*9um from the focal plane.  This comes out to 0.001".  From what you
wrote below an error of 0.001" across the CCD could be quite feasible. 
A misalignment would also explain why the PSF's are asymetric. That is
they flare to one side.

If the lens droops from gravity you would expect the deviation to be in
the worst direction for the Mark III, straight down in declination. 
Also since the lenses are bolted and glued into position, a droop would
seem inevitable to me.

Mike G.

droege@wwa.com wrote:
> 
> Mike,
> 
> Could this be an alignment problem?  The attempt has been to make
> everything "square" but there is no guarantee, and I do not measure
> how well things are alligned.
> 
> Early on, I did measure how well the glass face of the CCD chip was
> parallel to the base.  It was good to 0.001" which is as good as
> I could measure it.  To the same extent, the 3" idameter tube should
> be square (cut on a lathe it is hard to make it otherwise), and should
> be alligned to the base with less than 0.001" in 3" error.  The lens
> glued to the lens ring.  Again it should be good to 0.001" in 2" or
> so.
> 
> There have been no checks to assure that the lens does not droop, or
> that the axis of the lens is perpindicular to the CCD.
> 
> The lenses could just be drooping from gravity.  Or the optical axis
> could not be perpindicular to the CCD.  One test might be to rotate
> the lens in 90 degree steps as allowed by the 4 screw mount.  This
> would check the lens axis with respect to the glue joints, but would
> not check the droop from gravity.
> 
> Tom
>