Date: Wed, 19 May 1999 07:52:44 -0600 From: long@apo.nmsu.edu (Dan Long) Subject: 2.5m obslog 990518 ( MJD 51317 ) TUESDAY MAY 18, 1999 APACHE POINT OBSERVATORY SDSS 2.5m OBSERVING LOG This afternoon into the evening the secondary was recollimated. This included moving the vanes enough to allow a collimated range of elevation for the secondary from 30 to 90 degrees. The engineering camera went on shortly before 21:00. It has the flexure test apparatus on it as well. Like last night we used the guider fibers to connect to the engineering camera. There were problems using the engineering camera fibers which may or may not be related to the fibers themselves. In any case, it seems a good idea to just use the guider fibers as a general rule. Since the same Mac runs both the guider and engineering camera, it makes more sense to make the swap in one obvious place ( at the instruments on the rotator ), rather than at a less obvious place ( either the patch panel downstairs or in the computer room ). If the guide camera doesn't have fibers going to it, it's a good bet that is why you can't talk to it. The imager had another episode of nitrogen diarrhea. Connie worked intensely for nearly two hours to keep the imager cold and try to restore the fill system to working order. Bob Nichol was also trying to get the spectrographs ready for flexure testing. The both cameras were cooled, but given the time constraints, it was decided not to mount the red camera on the spectrograph ( the blue camera was already in place ). I got the telescope balanced for the engineering camera plus flexure test equipment. There are now clamshells on the front truss. With this setup, the counterweights were around 195 for best balance. I then switched to computer control from the hand paddle, and set the fiducials for all three axes. I then slewed to a FK5 star just east of the meridian and at about the equator. At this point the secondary decided to stop communicating. The TCC could no longer talk to it, and a "host" session got no response either, not even an new line upon carriage return. However, a telnet session to the port on the terminal server could talk to it fine. A telstop/telrun cured the problem. After the telstop/telrun, I released the last stop button. As sometimes happens, the axes became active immediately... no axis init or reset was issued. I then set the instrument to engcam. As I was typing, I heard the sound the alt lvdt makes when the windscreen changes position with respect to the altitude. I looked up with a flashlight to see oscillation in altitude. By the time I got to a stop button, the oscillations were so violent that the turnbuckle rods in the windscreen were slapping the sides of the windscreen making a loud clang. At least that is what I hope the noise was and not the telescope hitting the windscreen. The oscillations would occur outside of the range of elevation between 35 and 50 deg. Between 35 and 50 there would be oscillation upon startup (e.g. releasing stop button or brake), but it would settle down. Another note about the windscreen. With only two motors in azimuth ( number 3 got put on alt yesterday ), the windscreen glided smooth as silk in azimuth. Could it be that something about the third motor is different than the other two and causes the "lumpy" motion in azimuth? In preparation for checking primary collimation I changed the number of actuators in the MIR block from 4 to 6 for the primary. This is only temprorary as I did it via SET BLOCK MIR. As it turned out, this was not sufficient. Changing from 6 to 4 might work, but the min and max limits are set to zero for the E and F actuators, as well as the offsets. And when you go from 4 back to 6, it doesn't reset them to the values in the MIR.DAT file. This resulted in the setmir command all zeros giving the galil position setmir 0 0 0 0 0 galil 4600 -2000 0 -12900 0 0 instead of 4600 -2000 0 -12900 -7200 -7200 This also explains why the response to the setmir command for all zeros was a complaint that it couldn't move the first time I issued it, and no longer complained after I homed the primary ( E and F were then within tolerance, e.g. both zero ). When writing the log and realizing there was something wrong, I went into the file MIR.DAT and edited it so that 6 actuators were available Even with the odd settings, the images did not look bad. No astigmatism, and little coma ( see sdsshost /data/spectro/guider/990519_all_zero_500.fits ). Unfortunately there was little time left of the night, and I didn't notice this discrepancy in the all zeros position until after we closed ( the seeing was so poor, we didn't pursue coma correction very long ). With what little time remained, I obtained an instrument block for the engineering camera. In case either I or inst_data_fit goofed, I've listed the data below. scale ( pixel values from FINDSTARS are binned 2x2 ) inst offsets x y 0 0 202.3 143.1 .01 0 319.7 147.0 -.01 0 84.3 139.3 0 .005 200.2 202.9 0 -.005 206.6 83.0 .01 -.005 324.3 85.4 -.01 -.005 85.3 79.7 -.01 .005 82.3 198.2 .01 .005 320.5 203.6 rotator ( pixel values from FINDSTARS are binned 2x2 ) obj 0 202.6 140.5 30 182.2 142.7 60 169.5 131.3 90 173.0 139.2 120 176.3 131.6 150 183.3 117.0 180 200.3 105.7 210 224.5 104.6 240 239.1 114.8 270 226.1 94.8 300 224.2 106.5 330 218.5 123.8 The results looked reasonable, so I implemented them. The rot_inst_ang went from 88.3 to 90, and the rot_inst_xy moved about 4 arcseconds in x and 2 arcseconds in y ( they used to both be zero ). The telescope was closed up at 05:30. Note that only hand paddle operation of azimuth is recommended until the oscillation problem is solved. Dan