more pointing data, please!

i finally went through the pointing data that petri took for me. as you can see from the plots below, there is definitely a significant cyclical error. this further confirms what i saw in the salticam data i analyzed as well as what amanda and i saw with the startracker data this past winter. this data is a bit cleaner, though. the top plot shows the altitude and azimuth residuals modeled with sines and cosines. the bottom plot shows the residuals from the tpoint analysis. here are the best-fit tpoint parameters (units are arcseconds):

1 IA +0.000 +438.59 21.765
2 IE -0.000 +81.29 13.257
3 AN -0.000 -136.51 14.616
4 AW -0.000 +392.59 14.501

Sky RMS = 82.03
Popn SD = 86.35

the terms IA and IE are basically encoder offsets while AN and AW are misalignments of the azimuth axis in the north-south and east-west directions, respectively. these numbers agree qualitatively with what i found with the salticam data in the sense that there's a significant axial misalignment of the order of 5-6 arcminutes. the measured amplitudes of AN and AW don't quite agree within the errors. also, the tpoint residual plots show some systematic errors in the residuals that are not being modeled (probably to do with the phase offset i found in my sin/cos fits).

the upshot is that i need more data. so if the seeing is not good enough to do IQ tasks, please take more pointing data for me. the more, the better. the procedure is pretty easy. pick catalog stars with good coordinates that have tracks as centered as possible. then tweak the telescope's pointing to place the star as close as possible to the position of the center of rotation. i believe petri determined this and it should be in the logs somewhere (near X=223, Y=273). once the star is there, take image and save the data. also remark in the log which images were used for pointing. seeing and focus can be bad. i only care that the star gets centered onto the center of rotation.

thanks,
tim