Subject: minutes of September 1995 User's Committee Meeting

From: Michael Strauss

Submitted: Mon, 2 Oct 1995 13:54:59 -0400

Message number: 18 (previous: 17, next: 19 up: Index)

	Minutes to APO 3.5m User's Committee
	 Phone Conference, September 18, 1995

  Attending: Julie Lutz, Rene Walterbos, Bruce Gillespie, Ed Turner,
Michael Strauss, Don York, Chris Stubbs, Lew Hobbs

  Three basic items to discuss:
1. The current status report of the telescope from Bruce Gillespie,
posted as apo35-general message #15.
2. Ongoing engineering done by Chris Stubbs and colleagues to assess
the optical performance of the telescope.
3. Discussion of scheduling of fourth quarter.

Bruce Gillespie circulated a report (apo35-general message 15) of what
was accomplished during the recent engineering push, and where the
telescope stands at the moment. The high points are:
  - Azimuth drives now seem to be working reliably. 
  - Stalling Nysmith rotator motor now seems to be working.
  - Enclosure drive is rather erratic, and continues to be a source of
  - One of the three motors which drives the secondary (tilt and
focus) started to miss steps Sunday evening. It is not clear how long
this problem has actually been happening; it is possible that this is
related to some of the pointing problems that have come up recently.
This was worked on on Monday (editor's note: I observed Monday night,
and secondary motor was fine; telescope was in proper collimation.)
  - Recollimation of secondary was done.
  - Reflectivity of mirror surfaces measured at 60% (product of all
three mirrors, not each individually; this is at 6000 A). It was 73% when
everything was perfectly clean. This certainly affects throughput
  - Primarily with observations with Ed Kibblewhite's CHAOS
instrument, a variety of resonances have been discovered in the
support structure. The strongest of these is at 20 Hz, with a typical
amplitude peak-to-peak of 0.5" (which, when added in quadrature to the
intrinsic seeing, adds 0.1" to the PSF width). This resonance is
driven by the bit rate at which one of the stepper motor sends out
commands. However, analysis shows that if that bit rate is taken off
of resonance, the total power in vibrations does not diminish
significantly. A 10 Hz oscillation of several arc-second amplitude is
seen when the telescope points into the wind. 
  - The enclosure wheels remain a serious concern, although they have
not gotten obviously worse. It is possible that the manufacturers did
not harden the steel properly, making them liable for repairs. A new
wheel will arrive in mid-October.

We discussed the throughput of the DIS. Rene believes that his
measurements and those of Jim Gunn are in fact in agreement. Jim
measured the throughput of the spectrograph optics, while Rene
measured that of the whole system: telescope, optics, grating, and
slit. Ed reported that the gravitational lens monitoring program,
which has taken direct images with DIS every other night since the
telescope came on line, shows no evidence of a general decline of
throughput with time. 
  In imaging mode, the throughput is 15-20%, similar to what is found
with the DSC (see recent postings to apo35-dsc from Michael Richmond).
  Chris Stubbs has compared DIS images with images of the same field
from the Tololo 4m; he says we're getting 75% of what we expect. He
suspects possible scattered light problems. One clue is from DSC data
in hand, Michael Richmond (apo35-dsc #2) finds that 14\% of the light
from a bright star is between radii of 5.5" and 96".
  There was strong support for plans to measure the throughput of all
the instruments on a regular basis, and making the results available
to the APO community. 

  The guider continues to have problems; Spectrasource continues to
deliver unreliable electronics and software. The current plan is to
swap out the 1024x1024 chip currently there, and send it back to them
for electronics fixing, while we use a 512x512 chip with the resulting
smaller field of view. The pixels are 0.14"; it would make a lot of
sense to put in a focal reducer and thus increase the field of view.
People asked whether it would make sense to throw away the existing
guider and start over with another more reliable vendor; this was
estimated to cost $15-25K, money which is not clearly budgeted. But
Chris told us that the guider in fact does seem to work, if given a
bright star; it guided correctly to a few tenths of an arcsecond in
one hour, while an unguided exposure drifted by several arc-seconds in
the same amount of time.

Chris Stubbs and his colleagues at UW have put in a tremendous amount
of work to characterize the optics of the telescopes. They are writing
a series of reports, which will be distributed widely very soon. The
made a poor man's Hartman mask by closing the mirror covers and
opening the ports to the individual Nysmith foci (seven in all). They
found an enormous (~1') translation of the images as a function of
focus; could this be related to the problems the support motors for
the secondary were having? With these seven apertures, the individual
rays come only within 1" of a common focus, which is definite proof
of figure errors in the optics. Suspicion weighs most heavily on the
secondary. Chris is planning to build a real Hartman mask, and
continue similar tests. He will rule out problems with the figure of
the primary by doing the Hartman tests at prime focus. 

Chris has also looked at the time evolution of the pointing model; it
has drifted all over the place! This could conceivably be related to
the problems with the secondary drive motors. 

  We discussed the high fraction of engineering time, and how we will
know when it is time to slack off. No conclusion was reached. 

  We discussed the fourth quarter schedule, and the issue of how to
schedule in general. There was concern from UW about the amount of
time being spent in instrument changes; it would be good to limit it
to no more than one instrument change per night. The people at NMSU
felt strongly that time was being given in overly fragmented pieces,
and this should be avoided if possible. In particular, as
higher-ranked proposals get scheduled first, the lower-ranked
proposals are "fit into the cracks", and are often broken up into
little pieces. We discussed various ways to make the scheduling more
efficient. One plan that had some support was to have the APO director
schedule the 70% of the top proposals from each institution, and then
give the remaining 30% as blocks of time to each institution to
schedule themselves. This caused much discussion. Bruce Gillespie
pointed out that this was closely related to the issues of queue
scheduling, and service observing, which we might consider in the

 Michael Strauss re-emphasized the strong desire from Princeton to
have the throughputs of each of the instruments measured and

  The next meeting will be Monday, October 9, at noon EDT. 
				-Michael Strauss
APO APO APO APO APO  Apache Point Observatory 3.5m  APO APO APO
APO  This is message 18 in the apo35-general archive. You can find
APO  the archive in /u/strauss/apo/mailer/apo35-general on
APO  To join/leave the list, send mail to
APO  To post a message, mail it to