It’s finally clear out. Not really clear, not a good sky for imaging, and only a couple of days past the full moon. But there are enough stars visible to at least capture some data for periodic error testing and programming.
So, I just spent 1/2 hour outside opening the dome and getting things connected. (Startup is still delayed by an annoying problem with Windows Vista not remembering where certain drivers are, making me manually reinstall them at each boot.) Got the scope going, calibrated, and focused, and started a PEMPRO data aquisition run, then came inside to monitor by VNC. I plan to gather a good hour’s data, about 15 worm cycles, then do a mount program, then do a test of the improved tracking.
It’s nice sitting indoors in the warmth, watching the data gather. I like being outside at the telescope while imaging, to hear the filters changing and see that everything is pointing well. But watching a graph grow on the screen is less exciting, and is something that tolerates remote viewing just fine.
With 15 worm cycles of data, I programmed the scope PEC then did a test run. It got a *lot* worse – 10 arc-seconds or more off. Then I remembered reading somewhere that programming Gemini Level-4 from PEMPRO requires inverting the curve first. I found a button in PEMPRO that does that, and re-programmed the mount, then did another test run.
By the time this was done, the sky was filled with spotted clouds, so I closed up for the night, and analysed the data the following day.
Even better, the error is almost entirely periodic at 1 worm revolution, so there is good reason to hope PEC will correct it.Slightly troubling is the little bump at 76 seconds, which is not a multiple of worm frequency. We’ll be relying on guiding to get that error out.
After PEC correction was used for a new data run, the error is reduced to +/- 3 arc seconds. While this is pretty good, and easily treatable with guiding, I expected better from the PEC, so I’m planning to repeat this another night, paying more attention to the various tuning parameters.
Maybe this is as good as it will get – this frequency chart shows the 1-worm error is now only slightly larger than the other fundamentals; only twice as large as the non-correctable 76 second error.Anyway, another clear night, I’ll work at this a bit more and see if a little more improvement can be tweaked out.I know from experience last year, for example, that the 76-second error can be eliminated entirely by putting more slop in the RA backlash. However, this comes at the expense of looseness in the RA balance, making the mount more susceptible to flopping when imaging near the zenith (the point where the balance shifts from East to West, crossing the space across the gear backlash).
I’m hoping to live with the tightness I presently have, but may try loosening things up a touch more – probably waiting until spring for that, to avoid fiddling with adjustments and shims in the cold.
So, status: I have period error reduced to +/- 3 arc seconds without guiding, and expect guiding to handle that easily. I hope to improve it a little more – the next night that is good for calibrating but not good for imaging will be when that happens.