CCDs are revolutionizing amateur astronomy because, with homebuilt telescopes (and homebuilt CCD cameras, too!) they allow us to reach sky objects long considered beyond the realm of the amateur. However, CCDs place unprecedented optical and mechanical demands on telescopes. Reflectors must be light tight for CCD imaging, refractors must offer superior color correction, and all types of telescopes must form clean, tight images to produce satisfactory images with the new generation of CCD chips with very small pixels.
But long before new CCD users run into optical difficulties, they discover that they must deal with the tracking demands of CCDs. Both commercial and homebuilt mounts and clock drives have, for many years, been designed primarily to satisfy the needs of visual observers. If a drive could hold a celestial object in 2 or 3 minutes of arc of the center of a high-power field of view for half an hour, the drive was considered entirely adequate. Astrophotography was possible with drives of this quality with careful, attentive, and continuous guiding, and it is a tribute to their skill and patience that an entire generation of astrophotographers did as well as they did.
With CCDs, the tracking times are shorter -- typically just 2 to 4 minutes -- but the standards of tracking accuracy are also higher. The first experience many new CCD users have is getting trailed and smeared images whenever exposure run longer than 15 to 20 seconds. Getting round, sharp star images with 10-micron CCD chip on a telescope having a focal of 40 inches requires tracking to somewhat better than 2 seconds of arc during the whole integration time. The drive errors in homebuilt telescopes -- telescopes well suited for visual observing -- are often 10 times, and sometimes as much as 100 times, larger than required for sharp, round images. The performance of commercial drive units, especially those on Halley-era telescopes, are in the same ballpark. Yet better tracking -- and great CCD images -- is often possible with a clock-drive tune-up.
The first step in improving any drive system is to characterize and quantify the drive errors. Fortunately, CCDs make this process quite easy. What you must do is to set up the telescope, polar align as accurately as possible, and then take a series of short exposures extending ever several turns of the slowest element in the drive train. With worm-gear drives, if the worm turns once every 4 minutes, then you should record images for 15 to 20 minutes. Rotate the camera so that RA lies along the left-right axis of the image, and declination is up-down. Select a field near the equator and on the meridian and adjust the drive rate so the telescope follows a star without drifting ahead or behind significantly.
With the Cookbook camera, you can do this automatically in multiple exposure mode. Program the camera to take 10-second integrations at 15-second intervals, which will give you around 60 images. Measure the x,y position of a star in each image and then plot the position against time in each coordinate. You should get a graph showing the motion of the star image.
Next, quantify the plot. If you are taking images with a TC-211 CCD (such as the Cookbook 211, Lynx-PC, ST-4, and Electrim ED-1000 have), each pixel is 16 micrometers wide. For example, given that the telescope has a focal length of 1,200 millimeters, then one pixel corresponds to 0.016/1200 = 13.3 microradians or 2.7 seconds of arc. If the RA graph shows a clear sinusoidal variation, of 11 pixels peak-to-peak, then the drive has a periodic error of 29 arcseconds, which is fairly typical of a good-quality drive made for visual observing. (Of course, if the error curve shows nothing but very small errors, don't change anything. Count yourself lucky and get on with your imaging!)
In CCD imaging, however, if a star drifts by more than half a pixel during the integration time, star images may appear elongated. In this example, the drive error is roughly twenty times larger than desirable. What to do next?
The first step is to examine the whole drive system looking for loose, misaligned, or worn components. To function well, drives must be tight and square – after all, on an 8-inch diameter worm wheel, a one-pixel drive error corresponds to 0.00004 inches of loose or misaligned metal. Check that the bearings are tight. Many of the older commercial German equatorials used Teflon or Nylon bushings on the polar shaft and these materials wear down. They can be replaced with new bushings. Is the worm gear square on the worm wheel or tilted? (Tilts cause periodic error.) Disengage the drive motor and turn the shaft worm shaft by hand. If you detect a significant variation in friction as you turn the shaft, you are feeling the root cause of the periodic error. A bit of dry molybdenum disulfide or graphite brushed on the worm and wheel can make a big difference.
Once you have eliminated any obvious mechanical faults in the drive system, set up the telescope and repeat the test. You will probably see some improvement, and perhaps even see a lot of improvement. Quite often the component parts of the drive are more precise than their assembly; after they have been carefully reassembled, the same parts give considerably better performance. In commercial units, there is a reasonable chance that you are a more skilled mechanic than the original assembler. A missing or misadjusted tension screw, for example, might pass unnoticed during years of visual observing only to show up during CCD imaging, so check out every detail of the drive and set it right.
After the second test, reassess. If you can live with the residual drive errors, forget about further testing and get on with some real imaging. If the drive errors have persisted, you need to isolate the cause of the problem and cure it. At the level of accuracy required for good CCD tracking, every drive is unique, so you're on your own. However, I'll tell you how I tweaked up my Byers 812 so that I can run my standard 4-minute integration with a Cookbook 211 or Cookbook 245 and get nice round star images almost every time.
I had used my Byers 812 (manufactured to carry 8 to 12-inch telescopes) with a 6-inch f/5 Newtonian telescope for a lot of conventional guided film-type astrophotography. (By the way, it always helps to use what might normally be considered an "oversize" mounting. The drive hardly feels the weight of the telescope.) The drive would glitch occasionally, but I just guided it out. When I started shooting CCD images, the mounting had sat in an observatory little used for several years and the mechanical parts had a few isolated patches of corrosion, but nothing big. After getting the drive rate zeroed in, I found I could shoot perfect 5-minute images about two-thirds of the time, but the remaining third of the time the images were badly smeared. I ran a 20-minute sequence of images that showed very little periodic error with odd "spikes" in the tracking graph.
The spikes meant that the drive appeared to slow almost to a halt for about ten seconds and then to bounce back. The excursions did not seem to show a regular pattern. After studying the drive in action, I finally realized what was going on. The Byers 812 has a 30-degree sector gear that is fairly thin perpendicular to its plane of rotation. As the worm turns, therefore, the gear sector bends slightly but, so long as the coefficient of friction between the worm and sector remain constant, the polar axis turns at a uniform rate.
However, when a patch of dirt or corrosion comes between the metal surfaces, the sector would flex and rotation of the polar axis would stop until the patch moved past of the point of contact. Stars would trail on the CCD. Then the sector would spring back and catch up to the moving sky. I put a small amount of engine assembly grease with molybdenum disulfide on the gear and it eliminated the glitch. Since then the drive has performed very well.
The bottom line is: don't be surprised when you find that your telescope
doesn't track as well as you hoped it would. There is a lot you can do
to tune up the performance of a drive.
Back to the Cookbook Home Page