MDM Trouble Report for 2011 Aug 26

Telescope: 1.3m
Observer(s): Jules Halpern (Cccolumbia U)
Instrument: R4K

Problem(s) Encountered:
The 1.3m produces double images.  I have made two example images
available at http://www.astro.columbia.edu/~jules/double_image.tar.gz
They were obtained consecutively using the same setup.
One is a good image, and one shows the double image.
It would be interesting for you to blink them as you read on . . .

I'm asking you to put on your thinking caps, as I am really
puzzled by this.  In the 1 hour that I was open before clouds 
and rain came in, I verified that the 1.3m is producing a very 
repeatable double image.  This is NOT the occasional bad image,
but is now the default.  Out of ~30 images, I was only able to
get two good ones.  Consider these observations, which I have
made over the past couple of nights:

1. The images are always separated by 2.0-2.5 arcseconds in the EW
   direction (along the rows of the R4k CCD, separated by 6-8 of the 
   0.315 arcsecond pixels).

2. The double images are of equal intensity and individually seem
   to match the PSF of the good single image.  They are uniform
   over the 5.5' field that I am using.

3. The bias pattern of the affected and unaffected images are identical.

4. The extension of the image can also be seen on the guider camera,
   but it is less obvious because the pixels are larger and I have
   not yet played with its intensity scale.

5. The shape is independent of exposure time, which has ranged
   from 1-300 seconds.

6. They appear in all filters.

7. The double images appear with or without guiding.  There is
   no temporal oscillation visible on the guider, with its 1 s
   exposure time in my setup.

Together, these facts as well as other characteristics of the
observing conditions here seem to rule out the following hypothesis:

a) poor focus
b) bad tracking
c) bad guiding
d) wind shake
e) optical reflections
f) residual afterimage in CCD
g) corrupted CCD readout

What is left?  Idle speculation leads to . . .

1. If an oscillation, the frequency must be higher than 1 Hz.
Could the secondary mirror be rocking like this because
of vibration in the focus motor, which in turn was caused
by lightning damage to its controller circuit?  If so
this might be seen by shining a laser pointer on it.
Alternatively, can the power be turned off to the focus
motor (while on the sky) to see if the problem goes away?

I couldn't feel a vibration by touching the telescope, although I
thought of giving it a good whack with a hammer.  When the sky
clears I will use the guide camera with shorter exposures than 1 s
to look for an oscillation.  However, the readout of the camera 
is not much less than 1 s, and its pixels are large, so I don't
know how useful this will be.  Can John T. tell me how to save
images from the guide camera?

2. If there is some opto-mechanical instability, it would seem to
have to be something in the primary mirror support that would change
the shape of the primary.  The double images don't look like the
usual coma from poor collimation of the secondary or the primary.
The primary mirror seems a less likely cause because the image changed
from bad to good and back without moving the telescope.  We don't
have an active mirror support.  Although the lateral mirror supports
have swinging counterweights, I don't hear them moving.

------------------------------

Submitted on 2011 Aug 26 [22:16:07]