Courtesy of
Sky & Telescope
March 1994


Smooth Focusing with JMI's NGF

by Dennis di Cicco

     Focusing is second only to aiming on the list of fundamental telescope operations. The requirements of a focusing system, however, vary according to how the telescope is used. Visual observers, for example, can tolerate a focuser with some mechanical play, since the eye can adjust slightly if the eyepiece is not exactly at the point of best focus. Astrophotographers, on the other hand, require a high degree of mechanical precision because film is an unforgiving detector that must be rigidly held at a telescope's exact focal point.
     But even with photography there are focusing variables. Faster focal ratios require stricter tolerances than long-focus instruments. So do fine-grain black-and-white emulsions, which reveal out-of-focus images more readily than do color emulsions whose multilayer structure naturally diffuses and softens an image.
     But all these concerns pale when compared to the incredibly tight focusing tolerances necessary to get optimum performance from CCD cameras. In fact, no sooner had I begun experimenting with CCDs than I found myself reevaluating the way I focused a telescope. The slightest change in focus, even that arising from the mechanical contraction of a telescope's tube as the nighttime temperature dropped,* was readily apparent in a CCD image. The reasons for this exceptional sensitivity to focus stems from the fact that a CCD's image is highly magnified when displayed on a computer monitor (often as much as 60x the equivalent of enlarging a 35-mm negative until it is more than 7 feet wide) and because a faint star's image spreads dramatically in a CCD detector as its out-of-focus blur spills even slightly onto adjacent pixels.
     Maintaining a sharp image is not the only problem that arises from using a CCD with an inadequate focusing system. The image shift that often occurs when the focusing knob is turned in the opposite direction, especially with Schmidt-Cassegrain telescopes, added to my woes. This was particularly true while attempting eyepiece-projection astroimaging of the Moon and planets, since highly magnified shifts could move the image completely off a tiny CCD detector.
     It was obvious that working with CCDs would be a lot easier with a focusing device that could be precisely positioned and had very little mechanical slop. Enter the Next Generation Focusers (NGF) manufactured by Jim's Mobile, Inc. (JMI), in Colorado. To put it bluntly, if there's a better line of commercial focusers made for imaging, I don't know about it.
     Based on the Crayford focuser developed in the early 1970s by English amateur astronomer John Wall, the NGF units have ultrasmooth motion and virtually zero lateral shift when reversing their direction of travel. At present the NGF line includes four models intended for Newtonian and Cassegrain telescopes. They range from the $109 economy unit with a single focusing knob and sleeve-type bearings to the $300 top-of-the line focuser with electric drive and stainless-steel ball bearings throughout. All accept 2-inch eyepieces and accessories and have a maximum focusing travel of 1-7/8 inches.
     There is also model NGF-S, which I tested. Although it looks almost identical to the others, it is designed specifically for Schmidt-Cassegrain telescopes. It costs $289, features an electric drive and stainless-steel ball bearings, and accepts both 2- and 1-inch eyepieces. It also comes with an adapter for the threaded, rear-cell ring on Celestron and Meade Schmidt-Cassegrains, so that all accessories for these telescopes fit the NGF-S. The unit's drawtube has a relatively short, -inch travel, since rough focusing is accomplished with the telescope's conventional focusing system.
     Midway along the drawtube's travel, the NGF-S is essentially a 2-1/8-inch extension to the telescope's rear cell. This could pose problems for large accessories that need to swing between the arms of a fork-mounted telescope. It can also present a problem for people using focal reducers, since these devices are usually designed to be attached directly to the telescope's rear cell. One telescope I tried this on would not reach focus with the reducer placed on the outer end of the NGF-S and, even if it had, vignetting would have been unacceptable. The solution was to leave the focal reducer on the telescope and attach the NGF-S to it (all done with adapters supplied). This arrangement, however, may require custom fittings on the business end of the NGF-S, since accessories must be placed a specified distance behind some focal reducers for optimum performances.
     JMI states that the NGF-S will carry a 5-pound load. The unit tested bettered this mark by more than a half pound before the focusing drive began to slip. This is more than ample to handle even the heaviest camera equipment most amateurs are likely to use. More critical, however, is how far this equipment extends beyond the focuser. For example, one setup I tried had an SBIG ST-6 CCD camera on an eyepiece-projection unit attached to the NGF-S. Although the total weight was well below the 5-pound limit, the leverage created by having the heavy camera so far from the focuser made the NGF-S drawtube pivot off its bearings. Aiming an equipment-laden telescope near the horizon intensifies the problem and reduces the distance equipment can extend behind the focuser.
     Apart from these operational caveats, the NGF-S far exceeded my expectations. I found it impossible to detect any lateral shift when reversing the focusing direction. With a CCD camera the images of stars fell on exactly the same pixels as the focus was adjusted in and out!
     The electric focus has a continuously adjustable speed control that moves the drawtube from 0.020 inch per second to 0.031 inch. The motion stops almost instantly when you release the push-button control. With the focuser on a test stand, I measured how accurately I could stop the focuser at a predetermined position. Using the slowest drive rate and holding the motor control in one hand with my thumb operating the drive button got me within 0.0015 inch of the target. However, if I held the drive control in one hand and "tapped" the buttons with the finger of the other hand, I could consistently come within 0.0003 inch of the mark! This means that it should be easy to position the NGF-S within even the tight, 0.0022-inch focusing tolerance demanded by an f/5 optical system.
     The NGF-S works so well that it opens the possibility for a simple accessory that reads the position of the focuser's drawtube, perhaps like the dial indicator I rigged to the refractor as pictured above. With it I was able to make a series of CCD test exposures at various settings, determine which one was focused the best, and return exactly to that point. It cut the time I spent focusing the system from what with experience had become a 15- to 20- minute operation down to just 5 minutes or so. It would also be possible to develop a table of a telescope's focus shift versus temperature change and make simple adjustments according to the thermometer rather than having to refocus from scratch every so often.
     Taking this thinking a step further, I can imagine an NGF fitted with a computer-readable position encoder. With the computer also able to control the NGF's focus motor, it would be relatively simple to develop a program that would make a series of exposures moving the focuser a small amount between each one, calculate the exact focal point from the data, and position the focuser accordingly. Auto focus for a CCD camera! Far from frivolous, this possibility should definitely whet the appetite of the growing number of people working with CCD cameras. I, for one, would get in line to buy such a device should it ever become available.

* For a Cassegrain telescope the focus shift equals the length of the tube's contraction times the square of the secondary mirror's amplification factor. Thus, with today's popular f/10 Schmidt-Cassegrains, each 0.0001 inch of mechanical contraction translates into a focus shift of about 0.025 inch well beyond the 0.0088-inch "depth -of -field" tolerance considered acceptable for an f/10 optical system.