A 20 Inch Telescope           By Dennis Borgman

 

The beginning

Aperture fever had been getting stronger by the day.  My chance to pacify this desire came about quite suddenly in December 1992 when Ed Sczcepanski, a prolific buyer, builder and seller of telescopes in Houston, put his home built 20" DOB up for sale.  Upon informing me of it's availability, I immediately asked him to consider it sold and advertise no further.  I soon found out that my speedy reply was indeed fortunate, for several Houston amateurs had heard through the grapevine that it was for sale and would surely have snatched it up had I not been prompt with my decision to purchase.  My intentions for some time had been to purchase the optics and build the mount, but that of course, would have meant completing construction before I had a useable instrument.  Here, for little more than the cost of the optics, I had a useable scope and could build my own mount at a leisurely pace.  Ed's home-built scopes are seldom light-weights and this one was no exception.  The upper cage alone weighed in at over 40 pounds!  Nothing wimpy here!  The whole scope wasn't quite ready for the one-ton club, but it was close.  I sprained my back the first time I loaded it in the truck, and decided there must be a better way.   Wheel-barrow handles and loading ramps to ease the pain of transporting were only the beginning.

 

Personal background

Building and fixing things has been a passion of mine for many years.  In fact home shop machining was my primary hobby before I became interested in Astronomy, so I have a well equipped home shop and the basic skills to use the tools.  I am a computer technician by trade and a self taught hobby machinist and woodworker.

 

The design

I used Generic CADD software to do most of my design work.  CADD allows you to correct mistakes before they are discovered after hours of construction in the shop.  I can't claim a lot of original ideas in my finished scope, but I have diligently scoured the pages of telescope making publications and dutifully observed the craftsmanship and ideas of others.  Through this process I have gleaned the ideas that appealed to me most and attempted to apply them wherever possible in my own design.  So while there are many great ideas from others in my design, the overall finished product is my own.  My primary desire was to build a DOB that was reasonably easy to transport and assemble.  Telescopes that don't meet these criteria are seldom used, they just sit and collect dust!  Secondly, I wanted a tracking telescope.  Several years with a Celestron Ultima 8 had spoiled me with the ability to study star charts and return to the scope with the same star field still in the eyepiece. (I seemed to spend an undue amount of time back at the charts determining how to star hop with those mirrored images!)  After talking with Andy Saulietis about getting a copy of his 'Polar Roller' design, he informed me that there was something better and brand new, a true alt-az drive system.  This system was a 'CADD' (Computer Aided Dob Driver) system built by Tangent Systems and purchased through Andy Saulietis of DTG (Danciger Telescope Group).  He machined a pair of 17" worm wheels and matching steel worms to complete the basic drive package.  Finally, I wanted a scope that was easy to maintain and align.  Easy maintenance was accomplished by rugged construction and easy alignment would be augmented with electric collimation motors for the primary mirror and an "Easy Tilt" secondary holder.

 

The rebuild

First to be rebuilt was the upper cage assembly and serrurier truss system.  Both cage rings are identical octagonal segmented solid oak, with biscuit joinery to insure strength.     After glue-up of all but one of the eight joints, the inside radius of the ring was formed by fastening it to a one quarter inch thick backup panel to provide a center pivot and then band-sawing.  After removing the ring from the band-saw and the remaining backup panel, the final joint was glued and clamped.  The eight separating struts between the two cage rings are one(1) inch diameter .062" wall aluminum tubing with epoxied solid aluminum filler pieces where cross holes or tapped fasteners needed to be located.  Recesses, matching the tube diameter, are milled .062" deep into the oak cage rings to positively locate the tube struts.  To provide fastening points for the spider vanes that were rotated 90° to the eyepiece position, four cross bars were mounted horizontally between adjacent tube struts.  Short .25T x .5W x 1.5L bars fastened vertically at the center of each of the four cross bars provided a pair of fastening holes for the  threaded ends of the spider vanes. [Photo X]  A standard 20" Novak spider was modified to handle a beefier secondary mount rod.  The center piece was replaced with one containing a 5/8" x 20 TPI threaded hole.  I machined the secondary holder according to a Steve Watkins designed "Easy Tilt" with a steel 5/8" x 20 TPI threaded shaft to match the spider.

 

Note:  The "Easy Tilt", for those unfamiliar with it, provides a simplified means of adjusting the secondary mirror.  While many secondary holders use a push-pull adjusting/locking design that can be a bear to adjust even in the daylight,  Steve's design is based on a single socket head cap screw at each of three evenly spaced radial points, with enough built in friction to eliminate secondary movement under telescope use or transport.  Belville spring washers hold the adjustment screws under constant tension to the mounting plate fastened to the secondary support rod.  Small threaded brass balls between compression plates allow for tilting of the portion of the holder glued to the secondary mirror with silicone adhesive.  This sounds complicated, but it is in fact quite simple.  The whole design depends on friction forces applied after the whole unit is assembled and works extremely well.  In practice, the screws are adjusted with a small hex wrench and are simply turned clockwise or counter-clockwise at each of the three adjusting points to align the secondary mirror WHILE you are looking through the focuser, alignment tube or autocollimator.  Some nice pictures of the holder may be seen in Sky and Telescope magazine, Vol 83, #1, January 1992, page 14.

 

I use a JMI 2" NGF focuser mounted to a 3/16" aluminum plate fastened to adjacent cage struts.  The TELRAD mounts to a 1/16" plate fastened between the lower oak cage ring and one of the horizontal spider mount braces.  The finder scope is an ORION Big Eye 11 x 80 with an illuminated reticule and a 45° amici prism.  A PULSE-GUIDE lights up the reticule in short, medium intensity bursts and along with the right-side-up views provided by the amici prism, is extremely pleasant to use.

 

 The serrurier truss system was rebuilt at the same time so I could use the new cage on the original 20" rocker box.  The poles use a combination of innovations, again designed by others.  I'm not sure who needs to get credit for the following ideas, so I'll leave this up to the editor.  The cage end of the poles are a ball and clamp design.  The ball ends are solid aluminum and were lathe turned with a ball turning attachment.  They are turned for a sliding fit to the truss pole tube ID and epoxied in place.  The clamp brackets each hold two poles and have matching socket recesses machined to positively locate and hold the ball ends.  A non-removable knob locks each clamp during telescope assembly.  The lower end of each pole has a cam-lock design machined from aluminum solid and epoxied in the truss pole tube.  The cam has a total offset of about .060" and locks securely with a quarter turn twist in closely machined steel sockets fastened in the mirror box.  The bottoms of the steel truss pole sockets are adjustable to compensate for slight differences in their placement and therefore allow truss tube interchangability.  The construction so far was completed in time for TSP 1993 and that is how the telescope arrived at the Prude Ranch with it's proud new owner!

 

After a successful TSP trip, I turned my attention to the design and replacement of the telescopes lower end, the mirror and rocker boxes.  Since my design of the telescope lower end was to include a drive system, I had to plan for room to mount 17" worm drives and clutches for both altitude and azimuth.  With this drive system the telescope is used like any traditional non-tracking DOB, just grab the telescope and point it.  Initial setup is very simple and requires only two steps.  After turning on the drive computer, point the telescope straight up by using a bubble level and press a "zenith button".  Second, using an accurately aligned finder, point the telescope at the North Celestial Pole and press the "NCP button".  From here on, the drive system uses encoders to feed back telescope position and tell the drive computer the precise sky position being pointed to, and therefore allow precise determination of altitude and azimuth drive rates.

 

To provide a sturdy light weight foundation, my idea was to build the rocker box as an aluminum frame in place of the more traditional wood.  Telescope movement friction would be controlled by clutch plates on the drive gears.    With the telescope drive clutches completely released, telescope movement should be frictionless.  While the telescope would never be used with the clutches completely released, it made sense to start frictionless and use the clutches to provide full control of friction.  Before I could complete the design, I needed to know the balance point of the main telescope assembly.  The balance point would dictate the vertical dimensions of the rocker frame for clearances and the placement of altitude pivot bearings.  This need necessitated that the mirror box be completed first.  My mirror box design includes ample room for electric collimation drive motors and a cooling fan to speed equalization of primary mirror temperature to atmospheric temperature.  Oak veneer plywood panels with solid oak corners were used for the box construction.  Building close fitting joints for the 12 solid oak corners turned out to be the most challenging aspect of the mirror box construction.  I was forewarned by a close friend and cabinet maker, that this would be difficult.  Never the less I stumbled on and was for the most part successful.  I paid close attention to precise measurements and wound up with only a couple of joints that could have been improved upon.  The box is held together with glue only, and has shown no signs of failure at any of the joints.  The mirror box is finished with 3 coats of semi-gloss marine polyurethane and seems to protect the wood well against moisture.  The primary mirror cell is a slightly modified Novak unit.  Electrically driven primary mirror collimation necessitated the use of a different adjustment bolt scheme.  The lower casting normally used to fasten the cell to the bottom of the mirror box and carry the collimation screws was discarded.  The upper portion of the cell was modified to add ball and socket extension arms to the underside of the three main casting arms. [Photo X]  A .75" semi-spheroid socket is machined into the upper extended surface and a .625" through hole is centered in the depression.  The upper half of the socket is machined from a smaller piece with a similar semi-spheroid depression and through hole.  Three #10-32 machine screws fasten the upper socket half to the extension arm and are used to retain and control the captive tension on .75" brass pivot balls.  The brass balls have a .5" x 20 TPI internal thread.  The 4"L x .5"D x 20 TPI adjustment screws pass through the upper captivation plate, through the threaded brass ball and through the extension arm.  The collimation screws turn freely in ball bearings on each end supported by a bracket fastened to the mirror box.  The threaded collimation screws pass through the mounting bracket and bearing on one end to provide a stub to mount a small toothed drive pulley.  The motors are mounted to the side of the collimation screws and coupled with a 3:1 precision toothed belt/pulley system.  One turn of the 20 TPI screw will move the mirror cell assembly .050".  Coupled to a 100 step per revolution motor with a 3:1 drive reduction, provides .00017" movement per motor step.  Close tolerances were required in the collimation screw assemblies to minimize lateral mirror cell movement as the telescope is tilted.

 

The weight of the entire telescope is carried by altitude ball bearings in the rocker frame and 1 inch stainless steel shafts locked into aluminum plates fastened to the side panels of the mirror box.  The altitude balance point turned out to be about 1 inch above the top of the mirror box.  All the while I was building the mirror box, I was thinking about and refining my rocker frame design.  I pre-machined many parts that I new dimensionally would not change based on the rocker frame size.  Lots of small pieces like stainless steel pivot shafts fit to the altitude and azimuth bearings, clutch plate assemblies made from old 1/4" thick computer disk platters, bearing mounts, worm wheel hubs, stepper motor and worm gear mount assemblies, etc. filled the shorter evenings and left time on weekends for the big pieces.

 

My original intention to use aluminum rectangular tubing for the frame, posed another problem ... I had the MIG welder but I had no expertise to weld aluminum.  I could cut and fit all the parts and have it professionally welded or I could choose another construction material.  With TSP 1994 right around the corner, I decided to use steel tubing instead so I had a chance of getting the telescope back in useable condition in time for the trip.  This added significantly to the total weight of the telescope.  It weighs 340 pounds fully assembled. (So much for light weight!)  One delay led to another and finally with only a few days left, I scheduled some additional vacation so I could work continuously on the frame and remaining parts.  In three days the rocker frame parts were cut, fitted, and welded.  A number of other parts had to be fabricated as well, and I worked late each night to complete these.  The rocker frame pivots in azimuth on a heavy computer disk drive bearing and is further stabilized by three outrigger bearings fitted into the steel tube "ground board".  These 3 bearings ride on the underside of a three inch wide high density polyethylene ring.  This ring is fastened to the underside of the rocker frame. [Photo X]  HDP was chosen to provide a smooth tough surface  for the outrigger bearings.  This ring is about .75" thick and proved tough to cut.  I finally found the best cutting setup was a jig saw with a very coarse blade, running at it's slowest speed.  Friction from the blade even at the lowest cutting speed melted the HDP and welded the kerf back together behind the blade.  Boy! Talk about lots of effort and no progress!  I finally used a small trickle of water flowing from a garden hose to keep the material cool at the cut.  Cutting was still very slow and it took over two hours to cut the inner and outer circles of the 27" diameter ring!  Saturday noon rolled around (I've already missed my scheduled travel day to TSP) and I was ready to assemble the telescope to the rocker frame for the first time!  I can't remember the last time I was so elated by success!  But wait!  The mirror box is awful close to the azimuth pivot shaft.  Darn*#!@%&!  It doesn't clear!  I must have missed a calculation somewhere, because I had 1/2" clearance in the CADD drawings! ... PANIC! ... Can I cut some off the mirror box? ...  Wait!...  If I shorten the azimuth shaft and slightly redesign the clutch pressure adjustment screw it might clear!?  It works!  1/32" to spare!  Piece of cake!  The worm wheel/clutch assemblies were mounted and the worm wheels temporarily clamped to the rocker frame to provide traditional DOB positioning friction.  And so it went to TSP 1994.  The drive motors weren't ready yet, no gear covers to keep out the dirt, but at least I could use the telescope as an undriven DOB.  By 7:00pm Saturday evening I had packed all in the truck.  Now for a good nights sleep! ... Man!  I've got enough adrenaline pumping in me to last a lifetime!  Sleep? ... The heck with it!  If I leave now, I can be at TSP by Sunday morning!  I'll sleep when I get there.  And leave I did.  I arrived at the Prude Ranch about 4:30 am Sunday.  The only thing that kept me awake, was counting the deer along I-10.  If I counted 50, there must have been at least 150!  This kept the adrenaline going and me awake.  TSP 1994 was enjoyed immensely and was a much needed rest from telescope construction.

 

After returning from the Texas Star Party, I busied myself with completing the stepper motor mounts.  The steel worm gear as supplied by Andy Saulietis, was mounted in a block of PVC material machined to very close worm shaft tolerances.  The block of PVC was split such that all end play of the worm could be eliminated by squeezing the two halves together against the ends of the worm.  I wanted a spring loaded design that would maintain full worm engagement in the PVC worm wheel during the pressures of telescope pointing induced by the friction drive clutches.  The PVC worm wheels are rather soft, and I was afraid that it would strip the teeth if the worm wasn't kept fully engaged.  The cautious design has apparently paid off in that the worm wheels show no signs of failure after two years of service.   The worm is kept centered by Teflon lined guide plates that ride on the faces of the worm wheel.  The PVC worm bearing block slides between these same guide plates.  Adjustable guides along the edges of the block control radial movement.  A heavy spring maintains pressure on the worm bearing block to maintain constant worm engagement.  The whole worm/stepper motor assembly pivots laterally to follow any slight irregularities in the flatness or alignment of the worm wheel/clutch assembly. This whole system has worked well, but there is an annoying springiness in the altitude movement.  During periods of gusty breezes, it is difficult to hold the telescope stable, especially at higher powers.  Aside from this the drive system has been exceptionally pleasant to use.  How sweet it is to walk away from your telescope for a late night snack, and return 30 minutes later to find that faint fuzzy object still centered in the field at 480 power!

 

Drive enclosures came next.  A band roller was designed and built to facilitate construction of the .125" thick x 1.5" high curved walls of the drive enclosures.  The back covers are .062" aluminum.  Plexiglas front covers were used to complete the enclosures and provide visible access to the drive assemblies. An aluminum enclosure was also constructed for the altitude encoder with a thin Plexiglas front cover for visual access. The azimuth encoder is mounted inside the azimuth drive enclosure, so no separate protection was needed.  The azimuth encoder is driven with a thin flat belt from a pulley locked to the stationary azimuth shaft. This allowed the encoder to be mounted off to the edge of and in line with the azimuth worm wheel.  This maintained the much needed clearance between the mirror box and the azimuth shaft.  With an identical pulley on the encoder, one complete revolution of the rocker frame (with encoder attached) rotates the encoder shaft one revolution.

 

The only piece of my design that I have yet to complete is the microprocessor drive for the primary collimation stepper motors.  Adjustment is currently done by simply reaching into the mirror box from the top and turning the appropriate collimation screw.  My idea for the drive is a joy stick control that would be hand held while looking through a collimation Cheshire.  I have used a Motorola 6805 CMOS microprocessor coupled to MOSFET drivers in several such projects before and it shouldn't be too difficult to do.  Between Club projects and work, there never seems to be enough time to get everything done.  But I guess that's why they call this a hobby!

 

Clear Skies and Happy Observing

Dennis Borgman

Houston, Texas