Per our recent discussions, I had a chance to image last night under fair to good seeing (3 out of 5 to 4 out of 5 on Clear Sky Chart). I'm doing 20 minute exposures, in this case M13, to test guiding with a model active. The settings are as follows:
Mount - Mach2 Encoder mount
Scope - 10 inch F6.3 Mak-Cass astrograph, 1600mm focal length
Model points - 8 points each on 2 Dec lines on either side of the object.
Guide settings - 5 second guide exposures, 0.02 sec (0.3 arc sec) Min move
Guider scale - 0.95 arc sec per pixel
The mount is not precisely polar aligned, so there is drift in both axes as the object moves toward the western horizon. The measured drift at the midpoint of my imaging session was 24 arc sec per hour RA and 60 arc sec per hour Dec. The model created a curve that compensated for this drift, and without guiding the 20 minute exposures had approximately 0.75 arc second drift error. Without the model the Dec would have drifted 20 arc seconds in a 20 minute period.
Adding the guider into the loop increases the accuracy somewhat, depending on the seeing. Last night varied a bit but was generally good until the object was down in the western sky at midnight. Below are some guide results along with guider settings that I was using. With .02 sec Min Move the system will only send a correction pulse when the guide star has exceeded 0.3 arc seconds of error. On perfect nights (which are extremely rare here) I can set the Min Move to .01 sec (0.15 arc sec) and can have guide graphs with well under 0.1 rms guide errors. On average to poor nights I set the Min Move to 0.03 seconds.
I hope this gives a bit of insight into the different ways you can set up your guiding during an imaging session. If polar alignment is good, and you are not far from the zenith, and your focal length is short, you might be able to get decent results without any kind of guiding. If you want better precision, then guiding by itself can produce good results. In my case I'm testing the resolution capabilities of a 1600 mm focal length instrument and want to have as tight a guiding as possible. In that case I'm willing to spend a bit of time creating a quick model so that the guider only has to tweak the guide star every once in a while.
First and second images show the settings and results of a typical 20 minute exposure. The 3rd image shows a 20 minute unguided exposure with just the model running.
20 minute unguided exposure: