Subject: 950814 User's Cmte Minutes

From: Ed Turner

Submitted: Mon, 14 Aug 95 16:09:46 EDT

Message number: 10 (previous: 9, next: 11 up: Index)

Minutes of 3.5 meter users committee, 8/14/95, noon EDT. Ed Turner, presiding.
Present, by phone: B. Gillespie (APO); L. Hobbs (UC); J. Lutz (WSU);
M. Strauss (PU); C. Stubbs (UW); E. Turner (ARC); R. Walterbos (NMSU);
D. York (ARC).

Minutes compiled by York with very light editing by Turner.

Bruce Gillespie gave a status report, contained in the appendix to these
minutes. He reported that the telescope seems to be in fine shape, with
regard to recent problems. Pointing model is 3 arcsec, tracking is
0.1 arcsec for over 20 minutes, and enclosure drives and telescope
drives are all very quiet and motor currents low.

He noted several problems being addressed:
1) Various mechanical oscillations have been picked up, some of which
could be contributing to image size problems. They include  20 Hz,
30 Hz, and fraction of Hz signals. Possibilities of feedback from
commanding cycles of TCC, and windshake are being pursued. Bruce noted
that the secondary can be moved a few arcsec by a very light human touch,
so there may be several ways to get enough energy to power resonances. 
Ed Kibblewhite found the same 20 Hz signal with adaptive optics as the
one at that frequency found on the telescope.

2) A routine ultrasound inspection of the enclosure wheels was done
8/13/95. Compared to the same diagnostic done 6 months ago, the cracks
in the wheels have grown considerably, and penetrated softmetal through
the hardened edges of the wheels. Bruce summarized his plans:
a) Monitor enclosure sounds continuously, via intercom.
b) Get a metallurgist to evaluate condition of wheels.
c) There will be a telecon of relevant parties on Thursday to collect history
and discuss proper action.
d) Explore any quick fixes that might last for at least a short time (welding?)
e) Do another ultrasound in two weeks, to determine growth rate.
f) Evaluate and explain why these wheels seem to be failing so quickly.

3) We have been runnning a small telescope (R-naught sensor) to monitor
seeing. Contrary to earlier indications, there seems to be little
correlation between the sensor and our telescope seeing. Stubbs has examined
the data and reports several disturbing trends that cast suspicion on the
results. Stubbs and Gillespie plan to review the matter at APO this week,
and possibly move the R-naught sensor, arrange a cross calibration with
another unit, or take other steps as necessary.

R. Walterbos asked Bruce if, as a result of the recent repair work,
he felt we had a better understanding of the main telescope weak points,
especially as regards pointing and tracking, for the long haul.

Bruce responded that we have not finished evaluating the thrust bearing
replacement. (This bearing was the main culprit in the fall of 1994, but
regular maintenance may be adequate to prevent future problems.) He said
we had certainly fixed all known symptoms, but that some felt there
could be hidden causes that had not yet been found. The sorting out of
various pointing failures let to installation of moth scrappers on exposed
drive disks, to a better understood maintenance requirement set (different
from the expectations at installation per manufacturer's recommendations),
and to a bettter understanding of the set of spares needed.

Bruce has offered the instrument job to a qualified individual. Verbal 
agreements are in place, and the candidate will arrive 30 days after the
final paper work is completed at NMSU.

Responding to the last User's Committee meeting, Bruce is assigning his
staff as parents of different instruments, with responsibility to post a
status report bi-weekly to an e-mail exploder. M. Strauss, responding to
the same meeting, has set up exploders at Princeton. Users of a given 
instrument can tell Michael they wish to subscribe to the mail on
that instrument. Michael encouraged the members of the User's group to
advertise this service to their faculties.

GRIM: nothing new. The last known bias problem (subtle) is being fixed by
Bernie next week, having been lab tested.
ECHELLE: Lew reported that the vacuum problem is not fixed and that preventing
frosting at the ccd may require rebuilding the Dewar. In this case, steel may
be used instead of aluminum.
DIS: Continues to operate nominally. Red camera bias problems not consciously
fixed, but there are fewer reports of trouble. Still infrequent command
Rene Walterbos has posted a note on NMSU determination of instrument thruput.
It seems the throughput is 6% near 6000A, but virtually zero below 3900A.
Michael Strauss agreed to get with Gunn and compare these measures with
Gunn's measurements when he commissioned the instrument. There is some
suspicion that the UV problem, and the lower than hope optical transmission,
may have a time dependence. Rene was asked to use the direct images with DIS
taken when the spectroscopic calibrations were made to check through put
without the gratings. Ed Turner said he would make a plot of counts per second
in his lensing qso's over time, to get an independent calibration at the 
peaks (presumed photometric nights) and to look for time dependence.
Bruce said it might be appropriate to have the new instrument engineer 
responsibility for fixing the problem in the UV, in consultation with Gunn and 
Carr. Bruce said he planned to get Richard Lucinio to the site again to work
on DIS electronics, and event would be a time for the new engineer to get
versed in the electronics on DIS.
GUIDER: Bruce is putting the 512x512 ccd in the guider and sending the 1024x
1024 back to Spectrosource. (This was delayed by the recent mechanical
problems on the telescope, and the need to make a mounting bracket.)
We will have the same capability as now for pointing models, and higher
success with getting guide stars (by perhaps one magnitude-- the ccd is
smaller by a factor of 4, but the noise is down by a factor of 5).
A mode suggested some time ago has now matured, to help guiding without
a guide star. If slow drift is expected, a simple routine can be used to
offset to a preselected and observed FK5 star. If the star is not on
the same pixels as it was at the last check, an offset will occur that
will recenter the FK5 star, and the target star (in the field of the 
scientific instrument.) Lew Hobbs asked, and Bruce confirmed, that blind
offsets over 1 degree are accurate to 0.1 arcsec. Stubbs reported that
the software for finding bright guide stars that can be accessed by
small rotations of the science instrument (the guider is fixed wrt the
instrument in use) is nearly done.

ChAOS (adaptive optics): Nothing new. Ed Turner asked if there could have
been residual balance problems when ChAOS was mounted that triggerd some
of the tracking and guiding problems (noting that installation of ChAOS
and onset of degraded performance of the telescope are nearly coincident in
time). Bruce said the causes so far identified could not be tracked to balance

CALIBRATION: Nothing new, delayed by mechanical problems that had to be
responded to.

SOFTWARE/NETWORKING: Outages on a router in Washington, D.C. (MCI) have
caused more problems with internet. All institutions should set up
backup phone lines.
Fowler has distributed instructions to all.

There was a discussion of whether we should do anything. Some felt that
we had no control over internet and should do nothing. York wondered aloud
about going to NSF, who he thinks monitors the commercial contract for
internet, and complaining. Since we are a unique, real time user, we
might be told to go fly a kite. Matter pending.

AMBER: Walterbos described a new instrument that belongs to WSMR and
is being used by them for daytime missile tracking. He is looking into
using it for science on the telescope. It has an IR sensitive InSb array,
1-5.4 micron response, 512 squared detector array, less than 250 electron
readout noise, frame rates of 1 KHz, 0.15 arcsec pixels (f/20 only), a
dynamic range of 10,000. Rene is examining what filters could be added for
our use. One issue is whether the L filter used by WSMR can ever be removed.
Rene will continue investigating and invites interest from other potential
users be sent to him or Bruce G.
Strauss agreed to add Amber to the mail exploder menu.

SCHEDULING: Gretchen has a reservation system automated and all with
access to internet are asked to use it.

Ed Turner ended the meeting by asking a question of the user's, about
observing policy for the next 6 months. Should we deal with problems as
they arise, recognizing that we have increasing staff resources and increasing
scientist participation? Or, should we declare some special focused effort
to fix everything in some finite period of time? (Special shut down,
special additional appeal to scientists for help, etc.). After inviting
users to discuss the matter locally and report on e-mail, he asked for 
a first reading by polling the committee. The concensus that arose was
that all institutions would agree to having observers preempted by engineering
demands, providing that the specific plan for an interruption and the likelihoodof success because of the interruption were well advertised. There was 
a statement that we should respond fast to problems, as it would benefit
the observatory in the long run.

The next meeting is at noon, EDT, Monday, Sept. 11.


             APO 3.5-m Telescope, Instruments, and related systems

                           Status: 13 August 1995


After a long, hard haul, the telescope appears to be pointing and tracking
as well as it ever did since going operational last fall.  A week of
unscheduled engineering time plus a portion of last week's scheduled
shutdown were required to repair the azimuth and altitude drives.  Since
the az drive failure (reported last June), both az drive boxes have been
removed, had bearings replaced, been lubricated and were reinstalled with
new spare motors.  The drive disks that had been found to be out-of-round
were reground at the machine shop at NSO.  New drive disks and shafts were
ordered in June and are promised by L&F next week--they will be installed
at the first convenient engineering time.  The main thrust bearing at the
bottom of the "cone" has been relubricated, and a spare for this bearing
was ordered and received but not yet installed.  Coincidentally, the drives
for the enclosure began to exhibit an intermittent servo problem, which has
since been repaired after extensive troubleshooting.  Many APO and ARC
staff have given much to help us return to an operational state, and
especial credit belongs to Mark Klaene and Jon Davis for dedicating many
long days, nights, and weekends to solving these engineering problems.

Tests on image quality improvement centered on measurements of telescope
pointing jitter.  Early tests indicate some image motion in a few low-Hz
frequencies, and further tests planned to isolate and remedy.

A second ultrasound measurement of existing cracks in the four wheels that
the enclosure rotates on has indicated growth in these cracks.  Discussion
are starting on how to respond to this problem.

An experiment to test laser-cleaning of mirrors is being supported at the
3.5-m, involving small "witness" test mirrors that are cleaned with in situ
with LCO2 and others cleaned routinely with a prototype laser system after
shipment to Seattle.

Routine measurements of r-naught and cloud scanner images of the sky have
begun at APO, with initial results under analysis.


A verbal offer has been tendered and accepted by a highly ranked candidate
for the Scientific Instruments Engineer position at APO.  A formal offer
will be made this week after clearing NMSU administrative procedures, and
the candidate is expected to start at the site in early September.

We also expect to begin posting regular instrument status reports in
Michael Strauss' e-mail exploder (see APO www server for details)


There have been only minor changes in the DIS since last report.  Questions
of DIS throughput have been raised and are presently unresolved.  The
general operation of DIS has been fairly reliable since repairs were
effected by Princeton staff in May/June.  Some outstanding problems remain
(e.g., UV response, remote-control interface hangups, red camera bias
instability) and we are attempting to engage Richard Lucinio to come to the
site on an extended basis to trouble-shoot the control and analog problems,
as well as help familiarize the new instrument engineer with the DIS
electronics.  Also, a third slit wheel has been built by NMSU, mainly for
use of slit masks but is also generally available.


GRIM2 has been relatively trouble-free for the past two months, and the
preamp replacements done by Bernie Rauscher and Jim Fowler seem to have
mitigated the bias drift problem reported earlier.  Bernie is planning a
site visit this coming week to attempt to further improve the bias


Pretty much unchanged since June, with the exception of some additional
tests and baffling which improves stray-light rejection of camera when
mounted at Nasmyth.


Delivery to APO uncertain.  Problems continue with camera dewar temperatures.


Return date to APO is TBD.


1024^2 camera is in use for pointing models and offset corrections to FK5
stars.  512^2 echelle guide camera has been reworked by vendor and noise
characteristics improved considerably.  Plan to install 512^2 camera in
place of 1024^2, use new vendor s/w, and have 1024^2 electronics reworked
by vendor and then returned to the site.  Some recent tests of actual
guiding on telescope performed, indicating good performance with brighter
guide stars.


Was mounted on back left corner of primary mirror cell April/May, initial
tests and alignments completed.  More testing scheduled this month, and
plans being formed for possible use of laser guide star beacon later this
year.  Also planning to use ChAOS to provide measurement of optical
wavefront of telescope via its Hartmann "screen."


In process of building system that allows remote operation of primary
mirror cover.  This will allow calibration lamps to shine down from 2ndary
truss onto mirror cover for flat fields and wavelength calibrations,
without having to manually cycle the mirror covers and the enclosure
shutters.  Initial testing of this method shows adequate throughput and
flat field accuracy of ~1%.  Some detail work will need to be done related
to baffling and to the accurate representation of the entrance pupil by the
mirror cover, and add'l work needed to integrate these functions into the
remote user interface.  Schedule for this project is TBD, but the initial
capabilities were expected to be ready by now, but have been delayed by
repair work on telescope and enclosure drives.

S/W and network-

The remote operations have been fairly reliable since June, with one
notable exception a couple of weeks ago--the internet hub in Wash D.C. was
apparently overloaded, and caused several remote observing outages at APO.
We are in the process of finding out who to complain to when the internet
goes down (it's now a commercial system and it's difficult to get
names/phone#s of system administrators).  A backup modem capability has
been implemented at the site, but to be fully useful requires certain s/w
and phone lines need to be installed at the remote sites.  Jim Fowler has
sent instructions to users, but no one appears to have yet implemented
modem capability at the remote observing rooms.


        -Nasmyth rotator stalling:  this problem has gotten better recently
         (temperature?).  New, stronger motor has been received, will install
        -tertiary automation:  project still being scoped, schedule TBD
        -R-0 telescope: in routine operation.
        -Dark Skies:  Gave dark-skies presentation to Holloman facilities
         commanding officer, will follow up with invite to visit site.
        -www documentation: user information services under continuous
         maintenance and upgrading.  E-mail exploder service now described
         there, plus a STARVIEW service link, plus web form for requesting dorm
         accommodations at APO.

APO APO APO APO APO  Apache Point Observatory 3.5m  APO APO APO
APO  This is message 10 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