This report summarizes work done at Apache Point Observatory (APO) by Charlie Briegel, Steve Bastian, Bob Nagle, Jon Davis, and Bill Boroski during the week of November 10-13, 1998.
Bob Nagle worked with us this week at APO. Bob is a retired telescope engineer with over 20 years of experience with telescope control systems. We shared with him data on the servo measurements we've made in the past and spent quite a bit of time discussing the present state of the telescope control system. Bob feels that we're getting close but appreciates that there's still quite a bit of work to be done. Bob concurs with our decision to replace the rotator harmonic drive with a direct-drive system. He also shares our concerns with the azimuth drive bearing. More on this later. We'll continue to stay in touch with Bob and will share any tracking data we collect in the coming dark run.
Charlie made a number of measurements on the altitude and azimuth axes. In altitude, we are tracking with error rates of +/- 5 counts at velocities ranging from 10 cts/sec to 10,000 cts/sec. In azimuth, the picture is not so good. We continue to struggle with tracking at very low velocities. At velocites above 300 counts/sec, we can achieve tracking error rates of +/- 5 counts. However, at a velocity of 10 cts/sec, the position error rapidly increases until the integral term saturates and the servo falls out of closed loop. We suspect that this problem is related to friction in the azimuth drive bearing.
French recently looked at the azimuth drive bearing and noted that two-thirds
of the cylindrical rollers in the bearing had been removed. He also
noted that the remaining rollers were half-submerged in a bath of gear
lube. In researching this further, we have learned that gear lube
is not the most suitable lubricant for this application. Rather,
a high-performance, low-temperature bearing grease is preferred.
So before making additional azimuth drive measurements, we agreed to replace
the lubricant in the azimuth bearing.
Jon Davis, Bob, Steve, and I lifted the dust cover off the azimuth drive bearing. The gear lube was easily visible and had the appearance of clean motor oil. However, it was very sticky to the touch with an extremely low viscocity. When we removed the drain plugs, nothing came out. By working allen wrenches into the drain plug openings, we were able to extract a very thick, very dirty bearing grease. Jon said this appeared to be the grease that was originally in the bearing, before the gear lube was added. After several minutes of extracting the grease in this manner, the gear lube finally started to ooze from the openings. Its flow rate was comparable to molasses in winter, so we gathered some space heaters and placed them in the lower cone pit to warm the telescope body. Over the course of the day, the telescope warmed slightly and the flow rate increased. After approximately 12 hours, the level of the gear lube was below the bearing rollers. We replaced the drain plugs and the dust cover for the night to keep the bearing clean and moth-free, and then opened things up the following morning. We continued with the heaters until virtually no fluid was exiting the drain holes. I should note that there is still a rather thick, sticky film of this lubricant on all of the bearing surfaces that is of concern to all here. We plan to wipe off as much of this film as possible before installing the new grease.
Our original intent was to replace the gear lube with the bearing grease that Jon has been using on the 3.5-meter telescope bearings, since he has had good cold weather experience with it. However, before doing this we looked at the L&F drawing for the azimuth bearing assembly and saw that they had specified a synthetic bearing grease. Jon obtained a spec sheet from the vendor and noted that the grease has better properties than the grease being used on the 3.5-m. The grease is spec'd to -100 deg. F and is applicable for low RPM applications involving moderate to heavy loads. Jon has ordered a case of this grease and it is expected to arrive on Friday. It will be installed in the azimuth bearing on Friday afternoon and then Charlie will repeat the azimuth servo measurements.
Charlie continues to work on the interface between the TCC and MCP and has been in contact with Russell Owen to discuss some of the problems he is having. There appears to be a rounding problem between the two systems that results in an accumulated position error. Charlie is focusing his effort on fixing this problem.
Now for the good news. We managed to fix the latest wind baffle altitude oscillation problem. By turning on the altitude motor brake and pulling and pushing at the end of the wind baffle, we noticed that the main drive gear was moving back and forth by a few degrees. The drive chain at this point was very tight. Bob, Steve, and I removed the chain and disassembled the motor/planetary gear assembly and found that all of the bolts securing the planetary gear spline were loose. Steve Bastian cleaned the bolts and bolt holes with acetone and then applied Loctite to the bolt threads and reassembled the motor. After remounting the motor, we installed the drive chain following the procedure on the L&F windbaffle assembly drawing. The chain is taught, but not as tight as it has been. We moved the telescope in altitude inside of the building at a velocity of 10,000 counts/sec and noted no oscillation in the wind baffle. We also hung from the end of the wind baffle and could not make the wind baffle budge. The control loop held position quite well. We then opened the enclosure and ran the telescope up and down twice in altitude at velocities of 200,000 counts/sec. Again, we observed no oscillations in the wind baffle. Needless to say, we were very happy to find and fix this problem.
The bar code scanners for the alt/az fiducials have arrived, so Steve and I installed one of the scanners on the azimuth drive and tested it with one of the fiducial bar code labels. The scanner worked quite well. We also worked out the design of the mounting brackets for the azimuth and altitude tape encoder readheads. Steve will work on these brackets; our objective is to install the fiducial systems, the bar code labels, and the altitude bar code scanner during the week of Dec 8.
Glenn Federwitz had sent with me revised logic diagrams showing the recent changes made to the interlock system (stop buttons, etc). These diagrams have been placed in the interlocks binder at APO; the binder now reflects the current state of the system.
We would like to make special note of all of the help and support we
were given by Jon, Dan, Gretchen, Madonna, Mark, and the rest of the APO
staff. As always, everyone was great in helping us with our work.
We also want to thank Bob Nagle for all of us help with the azimuth bearing,
the windbaffle altitude drive, and for all of his input and advice.