Original Observatory build Split Ring Mount Steel Telescope pier 12Inch Cassegrain AE Mount Refurb New Observatory build Home


The Horseshoe mount project

This project is the result of an acquisition from Worthing Astrosoc in early 2010 of their 12" Newtonian and mount. This telescope they had replaced in their Society Observatory with a 12" Meade Cassegrain. The telescope consists of a Newtonian OTA bearing a 12" mirror and mounted in a fibreglass tube with a pair of focusers selected by rotating the secondary. The fibreglass tube sits on a wooden sledge which is attached to the bearer on the end of the DEC axis. The dec axis is driven using a hand-made gear hobbed from pulley wheel. The RA axis uses what appears to be a Beacon Hill 14" worm and wheel running on a 1 1/4" Stainless steel shaft. The underlying mount frame seems to be a piece of industrial machinery pressed into use with some additional pillow block bearings.

I basically bought this for the mirror set and worm wheels to use as a project scope, either by mounting the OTA in the observatory on the large AE systems mount or by re-building and integrating the optics into a completely new scope. In the end, I opted for the completely new scope idea. I had really wanted to take a good look at building a split-ring or horseshoe mount. The benefits of this type of mount are...

  • Its stable - all the weight is inside the axles
  • Its rigid - the axis distances are as small as can be
  • its easy to motorise
  • Its low profile in that the end of the rocker box is very close to the ground
  • It can be lightweight and portable, though this one isn't..
  • It is the nearest thing to a Dobsonian that is still equatorially driven (i.e. it's a dob with tracking platform built in)
  • I had the materials to hand

This is what it looked like before I started and while dis-assembling the parts.

12inch scope

The horse-shoe or split-ring

The intention was to size for a 16" scope and use 12" optics in the meantime. The horseshoe or split-ring I cut from a piece of 1.5 &inch; thick desktop material I had spare using a plunging router on a trammel. It's a 36 inch circle with a 16.5 inch wide u-shaped slot cut down the centre to within 5 &inch; of the opposite edge to leave some meat for strength. The U-slot cut is relieved at the corners to a 1 inch; radius to reduce the chance of crack failure in the wood. The bare edge of the horseshoe will be covered in plastic edging strip attached by impact adhesive to provide an even rolling surface for the horseshoe supporting rollers to support and drive the ring. The wood edge had already been sealed with a few coats of thinned out varnish, soaked into the chipboard.

12inch scope

The Polar axis

The polar axis of the horseshoe is a box structure, sized to the full width of the slot and with side stiffening supports out to the edge of the ring. The box needs to be both stiff to prevent splaying and to support the horseshoe now that a big chunk has been cut out the middle. On the back of this box (sized for a swing-through of 12 &inch; is bolted the polar (RA) axle. This a wheel hub recovered from a scrap Ford Sierra and bolted directly into place once the brake disk has been removed. More thought is needed about using the brake disk: I could true it up and use for the RA drive wheel if the horseshoe turns out not be circular enough. This RA box is made from 3/4 &inch; marine plywood glued and doweled together and then glued and dowelled to the horseshoe.

The rocker box

The box to carry the mirror on the dec axis is also made from plywood. I used 12mm sheet to make the octagonal end mirror-carrier and 6mm sheet for the side. Each of the 8 sides was glued together in pairs using overlapping formers and left to dry in jigs that held the pieces at the right angle before final assembly using the overlapping formers to hold the pieces together while gluing (only). The formers also provide the recessed lip for the top aperture and bottom mirror board. The gaps in the octagonal box were then filled with bodyfiller, sanded and coated in several layers of thinned varnish. The bottom of the rocker box is meant to be screwed on and carry the mirror, hence I recessed 4 sockets for 6mm wide-head socket screws to bolt the back board onto. The Worthing telescope already had its mirror on this type of board, so the issue was just to make sure that I calculated the swing clearance correctly based on the thick ness of board I thought required. The collimation of the mirror is achieved by adjusting 3 through-board screws which also have to swing through the horseshoe and RA box.

The rocker box is suspended on the mid-line of the horseshoe using simple pillow blocks. At the moment I am expecting to just use one each side and have a very small clearance between rocker box and pillow block to prevent the self-alignment of the bearings taking up mis-alignments of the axis of rotation in a non-consistent way. To force the issue I may have to adopt two pillow block either side. The axles in each of these pillow blocks are taken inside the rocker box to a bearing hub which uses a compression taper to firmly and concentrically hold the shaft. The bearing hub is bolted to the inside of the rocker box using 6 bolts on a 110mm PCD pitch, the standard for 1610 series shaft hubs. Right now it is only attached using two bolts. Once I am happy that the sockets align and rotate the rocker box normally, I will drill and complete the set. The rocker box upper face will carry the supports for the truss tube frame connecting the secondary box, one in the centre of each octagon face.

One other option that is open to me is to attach bearings to the slot sides for a telescope tube to simply drop in. The tube would have to have attached to it a pair of large roller disks, one each side that would sit on these bearings. To tell the truth I never thought about this until after I had done most of the work of design and assembly. In this case the drive wheel for dec would be one of these bearings replaced with a motor drive wheel and maybe the opposte replaced twith an encoder.

Mirror Cell

Since coming up against the balancing issues I re-engineered this part. The mirror cell is carried on the end plate fro the octagonal tube, Previously, the end-plate carried the mirror cell. The mirror cell had three bolts with push springs for mounting and collimation. By the time you added the thickness of the cell and the springs the mirror front surface was a way up the rocker box, contributing to the balance problems. I redesigned this to use a thicker end plate, the same cell and turned the collimatipon around to use pull-springs. THe collimation is now achieved using push-screws against the back of the mirror cell which is pulled hard down on the screws using spring pressure. This saves about 1/2" of CofG. The centre of the back plate is routed for a circle, the circle allows a 4" cooling fan to be mounted on the back of the mirror cell. The speed of the fan is controlled by a small PWM circuit from a HP printer. At some point that will be rigged to operate from the temperature difference between the air in the tube and the air outside. The mirror cell itself used to be a simple container with carpet backing for mirror support. I updated this to an 18-point mirror cell using 1/8" thickness aluminium triangles suspended on 3 10mm square box beams. These box beams are centrally drilled through and rock on a rubber grommet. The mirror support triangles have the same arrangment at the end of each of the beams. I'm left wondering if there will be any effect due to changing compression on the grommets as the scope moves in declination. I guess I'll find out. The mirror is retained horizontally by a wall of narrow ply, glued around the round mirror cell and lined with baize. Fingers cut from Al angle are used to retain the mirror vertically. The contacting surface is lined with inner-tube rubber and the attaching holes are slotted for vertical adjustment.

The truss tube frame

The truss tube frame on this mount is intended to keep weight down to a minimum with optimal stiffness and strength. There is no mechanical design to speak off - it uses 1 inch diameter 18 gauge aluminium tube with push-in plastic tube inserts at both ends of each. The tube inserts hold a captive M8 nut, into which is threaded a 14mm rod-end ball which also tightens the insert. The ball at each end of the tube is then held captive at each end in spherical recesses in strong angle mounts and held there by a metal bar, tightened by a wingnut on a spring. Assembly is to insert and tension. Disassembly is undo and pull apart. The recesses are created using a 14mm ball-end milling cutter on a 50 mm pitch between centres. The spring just helps hold captive the balls before final tightening. The tubes themselves are purchased from the aluminium Warehouse in 5 m lengths, so a mitre cutting guide is used to hacksaw the tubes to the correct length. Calculating the corect length is slightly optimistic but consists of mirror focal length - (box depth of mirror + truss mount height (x2) + camera height + focuser height + half tube diameter

The secondary box

The secondary box is made from a pair of disks of 9mm plywood separated by dowels. The dowels are stiffened and hold captive 4 panels of 6mm plywood to provide shadow to the user at eyepiece and a surface to mount the focuser on. The disks have an inner diameter of 13 &inch;. For re-use on a 16 &inch; mirror set I'll have to make a new secondary. The upper surface will support the secondary mirror spider, the lower surface will support the space frame brackets.

The spider I made from metal strapping off a delivery and a turned up piece of aluminium round section. I agonised for a long time over how to make this. I wanmted it to be re-usable and easy to adjust but it was getting heavier and more complex in my head and I also wanted to get it done quickly. After a monbthor so I settled on milling the round section with four flats for the vane seats and turning a lip on the round for a retained screw adjuster for vertical motion. The angular adjustment of the mirror is done traditionally using three screws with finger grips. The mount for the mirror was cut from Nylon stock I bored out and tapped the end for an aluminium circular plate. That way, I can use differnt size mirrors for different scopes just by swapping the mirror and backing mount. Again I looked for a good way of cutting an easy and accurate 45 degree angle in bar stock and in the end stuck with the simple - using a mitre block. I also experimented with making various diameter mirror holders from aluminium tube - either using rolled sheet and beating the end round a former or by cutting large diameter tube and providing a retaining lip. Neither worked well. The secondary mirror is glued to the nylon bar stock using butyl pond sealant. Three blobs of the black, viscous sticky stuff and then leave to go off overnight. Seems to have a good grip.

Base frame

The base frame supports the polar axle at the back and the roller plate that the horseshoe rides on at the front. The frame was built to be capable of being levelled and cope with a small degree of polar axis alteration for polar alignment. The frame has adjustable feet for levelling and the bearing support at the horseshoe end runs on a slotted mount to allow up andf down movement by about 1.5 inches - enough to give about 5 degrees latitude variation. (1.5 inches in 24 radius). At this distance, movement along the tangent to the circle changes the radius by a small amount only. The frame uses 3x2 inch mild steel tube to make an A-frame with an intention to operate like a wheel barrow with a wheel on the end that once put down is no longer in contact with the ground so I can push it to its place of use and set it up fairly quickly with levelling screws and bubble level. For polar alignment both the azimuth of the polar axis and the height of the roller plate (altitude) need to be adjusted. I expect this to be achieved by..

  1. levelling the platform,
  2. lining up on polaris using a polar reticle mounted off the horseshoe for azimuth
  3. final small adjustment in altitude at the northmost levelling foot.
  4. The roller board height will only need changing for significant changes in latitude.

Balancing the scope

When I did the first trial assembly it was clear that this was no dobsonian with a fist full of friction at the bearings to hold the scope when its released. Instead the pillow blocks are so smooth that any imblance will cause the scope to dip.

I put together the table below and came to the realisation that I needed to make changes - remove weight from teh secondary cage, from the tubes themselves, and add counterweights due to the short distance in this configuration of the mirror from the fulcrum. All in all, to be in balance, the mass adjustment to the rocker box needs to be around 12Kg!. To this end, I purchased 6 3KG diving weights from Ebay. 2 were attached to the back of the mirror board and 4 added around the inside of the rocker box. I also swapped the tubes from the original 14 gauge to the lighter 18 gauge, and re-made the secondary ring to be a single circle of wood with an underslung focuser mount. The issue I face with that last change is that the focuser weight could cause the ring to flex in use and that will ruin collimation, so some form of stiffening is likely to be required. But stiffening adds mass again....
ItemMassdistance from C of GMoment (Kg.mm)
TRuss tubes
Rocker box
Secondary Cage
Secondary Mirror
Primary Mirror assembly
Finder scope

Motors and encoders

The split ring mount is well suited for motorising to a driven telescope mount. For the polar/RA axis, the drive can be a roller on the big horseshoe arc itself or a disk attached to the polar bearing on the back. For the Dec axis drive, a disk or worm and wheel can be attached to one of the axle stubs and driven from a motor mounted on the horseshoe surface, on top or preferably underneath for protection. The other easy thing to do is to attach encoders to these disks to give pointing information even when not driven, so you always know where you are pointing. I have a pair of Magellan encoders standing by for this purpose, purchased from Scopes and Skies Astroboot.

All-up assembly

Putting the scope together consists of
  • Putting the base frame where it is needed
  • dropping the horseshoe box onto the frame
  • tightening up the polar bearing clamps
The tube will come separate from the horseshoe if the dec axis pillow blocks are unscrewed. the entire scope can be wheeled around like a wheelbarrow by lifting up the legs of the 'A' and using the front wheel.

This is what it looks like together:

thm_DSCN3372.JPG thm_DSCN3373.JPG - 6kb thm_DSCN3374.JPG - 6kb


Place to record build issues.


Place to discuss performance - optical, stability, drive, mobility, weight etc.


What do I need to change ?

  1. The swing-through distance on the polar axis box is too deep - this is a major system size driver - everything else is scaled to support it.