R. Millan-Gabet

Dates: June 9 - 23, 2000
Observers: Rafael Millan-Gabet, John Monnier, Wes Traub

This notes do not include comments/report on the 10 days dedicated to testing the new VxWorks control system.

(1) The NICMOS dewar needed to be opened because the previous observer (D. Segransan, May 2000) used it in the classical table (w.o. lens), and we needed to put the lens back to observe w. FLUOR. I took the opportunity to do a COMPLETE check of the NICMOS camera electronics. See notebook notes. Result: I didn't find any problem that could explain the noise pattern that appeared in Nov. 1999 and has been with us since.

(2) In May 2000 G. Perrin made some replacements in the NICMOS filter wheel. Some confusion was then reported by D. Segransan about the filter positions. When we arrived, we noted that when aligning the filter knob with the filter positions in the label, the wheel was in fact being placed *in between* filter positions. This was fixed. There is an additional label on the dewar indicating the replacements made by G. Perrin, and the alignment of the label and filter positions is now correct. To summarize, G. Perrin eliminated the aluminum block (which is now a clear hole) and replaced the J filter with one of [2.5,2.6 um] bandpass.

(3) The short delay line (SD) has been a big problem lately. After our problems in April 2000 were presumably fixed (under G3 control, problems persisted under Quadra control), G. Perrin had a flawless run (under G3 control). However D. Segransan could not observe at all due to SD problems (friction and inability to send to correct position, under Quadra control). Then M. Lacasse and C. Papaliolios fixed it again (under Quadra control), by replacing one pad and restoring the heat sinks in the Eltrol card, which appeared to be over-heating. During our run we then had problems again (G3 and Quadra control). There was an initial friction problem which we fixed by replacing another pad (East-South) and re-balancing the table (Marc). However, we then observed that both the G3 and Quadra were unable to send the SD to the commanded position (several cm errors when sending to about 1m and re-homing). This happened even under smooth motion of the table (as monitored via the overflow error signal), so it pointed to an electronics problem. We noticed false pulses being triggered at the Q outputs of the U2A and U10A chips of the Harvard velocity servo lock circuit, due to noisy output at pin 3 of U33A. Since replacing the chips didn't work, we eliminated the noise and false pulses by adding an RC at U33A(pin3) (R=33 ohm, C=330 pF). After that we were able to command (w. G3) the SD to about 2 m and re-home with an error < 1mm, typical of G3 performance. Note that when we do this, the SD moves at "high" speed (1 cm/s?). However, during our next observations the fringes were showing very large OPD fluctuations (just about under 100 um from fringe to fringe). We hooked the scope back on the Harvard card and noticed that we still had false pulses, but only when the SD moves at the slow speeds typical of fringe tracking while we take data. This problem is intermitent. In several instances, we clearly correlated the appearance of the false pulses, overflow errors and the appearance of an approx. 1 MHz interference signal on the U33A output. This interference came and went, and its amplitude grew from 0 to 2 volt monotonically in a couple of secs, and stayed for a few minutes before dissapearing. We have not been able to identify its origin. Conclusion: we think there is something fundamentally flaky about the SD control, which is intermitent and appears both when driving it with both the G3 and the Quadra, altough the G3 seems to have more luck with it.

(4) We borrowed the MMT reflectometer (contact: Bill Kindred) so that we could measure the reflectivity of our siderostats and other optics, and begin to asses what our optical losses are and where they originate.

NB: reflectivity measured for 45 degrees incidence; reflectometer beam about 1 cm diameter
results accurate to 0.5-1%; for each siderostat we measured at several locations
 

(5) As part of our setup for this run, we used Wes' interferometer device for alignment of the telescopes. We thought we did a good job based on the laser interferograms we observed. However, during actual observations, the star images seen through the alignment telescope appeared *very* comatic. The B telescope is worse than A, but both are pretty bad.