Improved Tracking with Gemini Level 3

My G11/Gemini L3 tracks pretty well, with the periodic error measured at 6 arcseconds peak-to-peak. For  long focal length imaging I was looking for improving tracking without investing in things like AO units and more expensive mounts. I believe I stumbled on a reasonably simple and effective solution.

The basic idea is to eliminate all the residual drift due to polar misalignment. Unless the polar alignment is exactly spot-on, there will be a slight drift in both, RA and DEC. With a polar alignment within 1 arcminute of the pole, I still find that a 15 minute CCD exposure at long focal length drifts sufficiently to cause star elongation.

An autoguider will usually guide out any drift due to polar misalignment. But, an  autoguider is a reactive system, and will only correct for errors that have already occurred, which is a bit too late, in my opinion <g>. The method described here compensates for exactly the right amount of drift.

The short summary is to use the Comet Tracking mode of Gemini to compensate for all the drift, in both RA and DEC. This is easily accomplished with an autoguider, and can even be done while an exposure is being taken. Here is how:

1. Set up and start the autoguided exposure, as usual.
2. While the autoguider is running, set Gemini to the Comet Tracking mode and begin training.
3. Wait 10-20 minutes (the autoguider will keep the star centered, thus letting Gemini know the amount of drift)
4. Press the Menu button to stop training and to activate the Comet Tracking mode.
5. That's it! Gemini will now apply a rate change to RA and automatically adjust DEC to compensate for all the drift, in real time, many times per second.

In my testing, I had the ST-4 guider in an OAG configuration on a 10" F/6.3 SCT. The focal length for the autoguider was 1600mm. In Sidereal mode, the guider had to issue corrections every 4 seconds or so to keep the star centered. After training and activating the Comet Tracking mode, the tracking errors were reduced by about 2x in magnitude, and the guider now corrected only every 20-30 seconds!

As a result, a 15 minute exposure produced almost exactly the same star FWHM as a single 1 second exposure (the raw FWHM in a 15 minute exposure was 2.9" at 0.88"/pixel scale, just like it was in a 1 sec exposure). Now that's good tracking!

There are still some questions to be answered with this method, but so far it appears to give significant improvements:

Q1. Does the Comet training work only in the region it was trained in, or can the same training be used in different parts of the sky? In some preliminary testing, I was able to get  significant improvement in many  parts of the sky after tracking around the east meridian.

Q2. Because Gemini compensates completely for DEC drift, my usual method of disabling the DEC corrections in only one direction will no longer work, because the fluctuations will occur equally in both DEC directions. I found that I can now enable both directions, and still get outstanding guiding.

If you find this method useful, or have suggestions or refinements, please e-mail me here

Copyright 2003 Paul Kanevsky