Type 4 Double-arm Drive
Construction Notes and Photos

| Jefferson Lab Home Page | Science Education Home Page | Measurements and Overview Page |

Construction Notes

It isn't pretty, but it works.

This drive was built specifically for comet Hale-Bopp. Since the photographic windows of opportunity weren't very long (an hour and a half or so of dark skies a night, at best) I built this drive large enough to carry two 35mm SLR cameras. This allowed me to take two exposures at a time, using different films, lenses, etc... or, if a friend tagged along, their camera could hitch a ride.

Although it performed well, my tracker suffers from several construction flaws that I hope to work out in a new drive.

The base board, drive arm and camera platform are all made from a sheet of aluminum that, in a former life, was a traffic sign of some sort (bought legally from a scrap yard, not 'found' on a pole somewhere). I thought the aluminum would provide a durable, rigid, light weight base but it proved to be remarkably flexible. Strips of extruded aluminum were pop-riveted to the aluminum sheets to stiffen everything up, but with limited success.

The tripod base board was made by gluing together two quarter inch thick sheets of plywood. The grains of the two sheets were set perpendicular to each other to help stiffen the board. The board then mounts to a tripod that usually holds an 8 inch Celestron telescope. After a night or two of struggling to align the mounting holes, I threaded the mounting bolts up through the bottom of the tripod. It was then a simple matter to plunk the tracker onto the three exposed bolts and secure them from above with threaded rod connectors. I used threaded rod connectors in place of regular nuts wherever I could since I found them much easier to handle in the dark, especially with gloves on.

Three threaded rods attach the drive base board to the tripod base board and allow for polar alignment. The base boards are hinged at the opposite end with two large barn door hinges.

The drive arm and camera platform are attached to the base board with sections of piano hinge. Piano hinge is also used on the hinged section of the drive arm. These hinges are far too wimpy and introduce a lot of flexure. More robust hinges will be used in my next drive.

The drive is driven with a small, 1.8 degree per step stepping motor. The control circuitry was bought as a kit from a science surplus house and is housed in a large black box mounted on the base board. I added a 555 timer circuit on a 'daughter card' with the correct components to drive the motor at 200 steps per minute (one rpm). I also added a small hand-held unit to adjust the driving rate. A variable resistor controls the steady state driving frequency and two buttons allow for 'correction modes' of roughly twice the driving frequency and one-half the driving frequency. The circuit box also has an on/off switch, a forward/reverse switch and a 'run flat out' button that was intended to rapidly reset the drive but is way, way too slow. There is also a red LED 'heartbeat' indicator that flashes each time the motor is pulsed. This would allow me to confirm that there was a drive signal if the motor itself ever failed to step, provided me with a little light to read my watch, marked my location so I could find my way back when I wandered off during long exposures and gave the hunters something nifty to shoot at. Power was supplied from my car's cigarette lighter.

The drive rod is a length of 1/4-20 rod. The rod is attached to the stepping motor with two threaded rod connectors. I drilled and tapped six holes in one connector to accommodate set screws. The holes were placed in two sets of three. One set of three (faces 1, 3 and 5) secure the motor shaft to the rod connector and the other set of three (faces 2, 4 and 6) secure the threaded rod. If the motor axle and drive screw are not co-axial, a wobble will result that causes the drive to alternate between running fast and then slow. The average speed is still correct, but the wobble can be sever enough to cause tracking errors. The six set screws can be used to eliminate any wobble. The second threaded rod connector is tightened against the first one and helps lock the first one in place.

Captain Tom Krajci has worked out an Excel spreadsheet for determining the accuracy of a barn door tracker by projecting a laser pointer at a distant wall. He has generously allowed me to make it available here as a uuencoded file. For further information, please see his letter posted in the Astrophotography Mailing List Archives.

The drive rod is attached to the base board with two hideously over-sized pillow-box bearings. A threaded rod connector was fitted into a hole drilled through a wooden dowel that runs between the two bearings. The drive rod runs through this connector and the bearings allow for the required pivoting action.

I discovered that it isn't always possible to look through the camera without bashing my head on something. Also, when the comet was in the northeast, one camera would see the back of the other camera. I solved this by making an extension arm out of a section of 2x4. It's isn't very elegant looking, but it allowed me to aim and focus the cameras while actually looking through them.

I monitored the tracker's progress with a 500mm spotting scope mounted on the southeast corner of the camera platform. This restricted my choice of guide stars to those in the southwestern to western sky.

Polar alignment was achieved with an Orion polar scope. Aligning the polar scope's axis with the camera platform's was a bit tricky as I had to remove a bar that helped stiffen the camera platform. The bar would collide with the drive arm and not allow the camera platform to swing through a useful enough arc.

I was able to fit all of my gear into a Ford Festiva for the drive to a site in North Carolina. After arriving on site (and after several sessions) I could be ready to shoot in 15 to 20 minutes. The longest lens I shot with was 200mm, although one of the people who tagged along one night used a 400mm lens with no apparent tracking error. Exposure times were as long as 20 minutes.

Some photos of the completed tracker follow. Click on the images to get a larger version of that image.

Any suggestions for improvements to this page are welcomed as are any construction tips/hints you care to share.

Thanks go out to Lamond North for cutting the aluminum plate and attaching the extruded aluminum as well as for constructing two perfectly good bearing mount assemblies that were never used. Sorry about that.

(click to enlarge)