Things had been working quite well until last week when my guiding had gone west. When I looked at the guiding vectors in CCDSoft, they were suddenly rotated 20 degrees counter-clockwise and the X and Y values seemed to have been exhanged. Needless to say, the guidng sucked. I paused the program and manually re-entered the angles and all proceeded well.
I redid a calibration and initialization the other nite, and things proceeded well with the first series in H-a; when the next series in O3 started, guiding went west - I paused things again and found the vector angles off by about 17 degrees or more.
I did screen captures of the calibration results in CCDSoft AND then the result when CCCDAP3 rewrote them with the altered angles and scalars.

Dave

Please post logs of a reference good run of a few weeks ago and a recent one showing the problem.

OK, sent to your private email, including screenshots as WORD files so you can see calibration results.

Dave

Dave,

The log for 0831 shows numerous plate solve failures as well as run interruptions as you mentioned. Also, it doesn't appear you use precision slew to the target - always recommended. Also, CCDSoft stopped reporting guide errors at 00:07:55. This indicates a failure of the COM interface to CCDSoft.

The log for 0902 seemed to be going fine until it was aborted. I was unable to decipher what the various .doc files were indicating. The one named CCDAP Cal.doc showed two screen shots. The first one showed CCDAP generated vectors.

As usual, when something that had been working suddenly stops working, it is good to examine what has changed. A reboot is always a good idea. It may be simply that the star field around your current target is insufficient for accurate plate solves.

If you don't get accurate plate solves, you may not get accurate guide vectors. I suggest you enable Trace so that even successful plate solve files will be saved (they are normally deleted). Examine those files to insure a sufficient number of stars are used in the plate solve solution.

John: not sure what caused the plate solve failures that nite EXCEPT I think some clouds rolled in and out. With only a 7" refractor, things can get touchy. I expect the new routines in TheSkyX will help with this.
OK, I've enabled trace for all future runs. And it seemed to track fine last nite. Thanks for the advice. I'll want to talk with you in Pasadena to try to understand better what you are writing into the guider vectors.

Thanks,
Dave

Glad that resolved it for you, Dave. I won't be at PATS so see you at AIC.