A PRECISION DOUBLE ARM DRIVE
Designed and constructed 2002 by Evan Williams
Original invention by Dave Trott
This is a modified type 4 double arm drive scaled to half of usual size. All critical dimensions are held to an accuracy of 1 part in ~10000. Full allowance is made for differential expansion and contraction over a temperature range of +20c to -20c. The design temp is zero celsius but the arm length can be adjusted for different conditions.
Materials are anodized aluminum, brass, steel and stainless steel with ball and roller bearings. All fasteners are high strength steel. All but two fasteners fix in about 85 drilled and tapped holes.
Power for the 1/2 rpm synchronous drive motor is supplied by a crystal controlled inverter of my design, with an error of less than 10 parts per million, checked with my frequency counter which was calibrated by me using a triple atomic clock reference output. The inverter uses CMOS integrated circuits and VMOS power FETS in the output stage. It draws only 45ma at 12 volts and will run reliably down to 8.5 volts. Eight AA nicads will run it for about 10 hours per charge.
The alignment scope is half a roof prism binocular. It has been stopped to a three degree field with an illuminated reticle added.
The counterweight allows consistent downforce to be applied to the drive arms. It is adjustable both by sliding the weight and by rotating the tube holding the weight shaft.
The wedge is adjustable without tools from ~30 deg to ~70 deg.
The drive screw is supported by stacked thrust bearings so the motor carries no load and the shaft is threaded 3/8 by 20 t.p.i. The drive screw and the follower nut were cut on the lathe to a tight fit and then lapped together with fine grinding compound. The follower nut pinions and the motor mount pinions are perfectly coaxial when closed and the length of the primary and secondary arms is accurate to +/- .001 inch at zero degrees celsius. The arm pivots, motor pivots, follower nut pivots and the first to second arm rolling contact point lie in the same plane when the arm is closed.
The drive is designed to close in operation with the motor automatically disengaging the drive screw when the arm bottoms out. The driven (second) arm is carried on the drive arm by a ball bearing that rolls over a piece of highly polished tool steel. Drive screw length is enough for two hours of tracking and reset is facilitated by the crank handle at the top of the screw shaft. A small 1/8" thick washer may be placed on the bottom of the shaft to disengage it from the motor while still supported by the thrust bearing so the drive can be hand operated. Rigidity is high making hand operation very practical.
The tripod is a surveyors tripod with ground spikes, stable and simple.
My original intent when constructing this drive was to reach the theoretical accuracy of the concept, which is less than 1 arc second error over two hours.
I estimate that the device should be capable of carrying about three to four kilos without trouble. The primary use will be film and webcam astrophotography. It is possible to mount a small telescope on it. The cost to build is difficult to estimate as I have been saving materials for years as I come across them. The main tools used to build it are a vintage 1934 9" South Bend Workshop Lathe, a drill press and assorted hand tools. The project took about 250 hours of construction time over three months. There are no plans or significant sketches as I keep everything in my head.
The drive may be removed from the wedge in about 30 seconds. It is held in place by two thumb screws. The wedge may then be used for other small equipment such as polar mounting my ETX 60.
Click on the pictures for larger versions.
A minor point about some of these photos: The indoor pictures were taken with a light source rich in infrared (quartz-halogen). Because of the dye used in the anodizing process it does not look black under infrared light. When seen outside under natural light all the anodized parts are jet black.
1: As seen from south side (see note 7). The mounting plate on top has a selection of holes, some threaded, for mounting equipment and adapters.
2: South view. The thumb screws on the wedge allow securely locking the wedge in position by hand.
3: "A" and "B" in the third picture refer respectively to the bearing that is attached to the second arm and to the thrust bearing. The second arm receives drive from the first via the ball bearing. The thrust bearing supports the drive screw and prevents any vertical loads from being transmitted to the motor.
4: This shows the drive screw in the disengaged position. When the arm is fully closed the motor continues to drive the screw up until it disengages from the motor shaft. This prevents problems when the drive runs out of travel.
5: The drive screw in the engaged position. The bottom of the drive screw has a hole in the end with a slot cut across to fit over a small roll pin on the end of the motor shaft. It is extremely important that the motor shaft and the thrust bearing supporting the drive screw be coaxial to avoid periodic error.
6: The follower nut and fork assembly. The pinons for the follower nut are made from high strength steel and thread into the fork. The pinions are held in place by two setscrews. The setscrews do not bear directly on the pinions as this would damage the threads. A small slug of lead is placed in the setscrew hole. This conforms to the pinion threads and allows the setscrew to be tightened to hold the pinion without damage. The fork is mounted on the end of the first drive arm with a fine thread steel bolt that may be adjusted for length. It is secured in the same manner as the pinions. A simple secret to high precision design is to allow for adjustment of parts and then make sure that the adjustments stay set.
7: The wedge is designed to be adjustable without tools. The long threaded shaft is 1/2" stainless steel and is threaded through the lower brass support. The brass jamb nut and thumb screws may be loosened for adjustment and polar alignment. The base of the wedge has a three inch diameter disc of nylatron (not visible) that fits the top plate of the surveyors tripod. Alignment is quick and easy. Note that this drive has the polar axis parallel with the wedge declination surface so the wedge is set up backwards compared to use with a polar mounted scope such as my ETX 60.
8, 9, 10, 11: Set up and ready to use at Mt. Kobau.
Does it work?
The first test I did was a drift test. After aligning the drive I attached a 200mm camera lens to it with an adapter and a 4mm eyepiece. This gives 50x. I put Arcturus on the edge of the field and it was still in the same place 20 minutes later.
The definitive test is photography. The first field test was at the 19th Annual Mt. Kobau Star Party. This was not a severe test as all my shots were with a 50mm lens (because of wind). However, It performed well as while I was at Mt. Kobau the wind blew both nights at 20 to 40 kmh. The mount is so rigid that the wind has no perceptible effect. Most people were not able to do any photography and in one instance a telescope blew over, smashing an expensive eyepiece.
The picture below is a 50mm view of M31 followed by enlargements from the same photo. It was only a six minute exposure but M31 is about 60 degrees from polaris, a good drift test subject. Even at high magnification the stars are round. That I was able to image below magnitude 13 shows that not only were conditions very good (except wind) but that all the light (of an object) was falling on the same place on the film, meaning that the drive works very well.
M31 with 50mm, F2.8, Agfa 800, six minutes.
Also imaged is M110 with a surface brightness of only 14.46.
Further analysis shows objects as deep as magnitude 14!
The position of the catalog labels was not interpreted from a star chart.
Instead, the chart was copied at the same scale and overlaid on the image.
The magnitude labels were then added manually.
A 50mm lens is usually considered to have a limiting magnitude of about 11.5 under excellent conditions.
It is obviously possible to do better.
This photo is a composite of my photo and a ESO Skycat DSS 1 photo so that I could determine what I really imaged. The blobby shapes are from my photo and the sharp star images are from the Skycat photo. Where they line up I can then determine what is a real image. The deepest that I can confirm is the 13.9 E_USNOA U1275_00432544 object in the centre. There does appear to be some light captured from a 15.2 object.
The irregular star shapes are due to seeing effects and artifacts produced during computer processing, not to mention using only a 50mm lens.
If you have questions or comments please e-mail Evan Williams at firstname.lastname@example.org