P&T Webcam Scripting Magnetometry Lightbox Artificial stars Home


Mirror Making

Mirror making is something I have moved into after owning 3 Newtonian and 2 Cassgrain telescopes. There's definitely something mystical about bringing a curve into nanometre perfection on glass in your garage. That said in 2008, a whole 12" [Skywatcher/Meade Dobsonian]telescope could be had for less than the price of a retail 12" mirror in the UK. SO what's the point ?

Well mostly I like to make things regardless of the price drivers ... this is what I am doing

I have a Mirror-o-maticmirror making machine that I swapped with MikeMS for a Newtonian telescope that was no use to me. With this I have so far attacked a 6" mirror blank and taken it to the polishing stage. This is where I am up to...

Mirror-o-Matic mirror grinding machine



The mirror machine came to me as a result of astronomical horse-trading, swapping a large GEM mount and a 8" Newtonian OTA for the machine as it was and a start kit consisting of two 6" blanks and the grit, rouge and pitch to manufacture a mirror.

The grinding machine had been built to the design of the mirror-o-matic design of Dennis Rech ( link ). While the bulk of the components were in place it wasn't finished: some linkages were not in place, the tool drive arm was unfinished and a few other bits weren't there. However the big stuff was in place and well made. I finished off the machine, building parts and providing linkages from my bits box. For example, the eccentric driving linkage used a bicycle quick-release clamp to allow rapid changing of the eccentric motion without tools.

powerhead details power train with belts in place. The grinding tool holder arm The underside of the mirror platter showing pads which fit holes in the pulley wheel Assembled machine, waiting for the mirror platter Mirror platter waiting for mirror mount 'top hat' Mirror platter with spherometer, showing adjusting clamp slots 12.5 inch mirror blank in place - no clamps

The Testing Equipment

While grinding the first mirror I discovered that there is a range of test equipment that I needed to aquire or build to analyse the steps along the way. The spherometer is used to measure the mirror curvature as it roughed out, the mirror stand to hold it for testing, the foucault tester to do the focal plane testing, a light source for the foucault tester, tile grinding tools to grind the mirror and more. These are looked at individually.



The spherometer is a crucial tool to allow the measurement of the depth of ground mirror to be measured. This figure, called the sagitta, allows the measurement of the radius of curvature and thus the mirrors current focal length by the r&pow2;/2R relationship.

This one was made from a piece of flat Aluminium bar about 6 inches long by 1" by 3/4". I drilled a clear hole in the centre for the dial guage and several threaded holes on 1" pitches along the centre line out towards the long edges of the bar. These holes were tapped M6 and two pieces of steel bar turned with centre drills to allow ball bearings to be glued in the hollow. One of these is screwed in place and the other is adjusted on a flat plate until the dial reading is zeroed. Alternatively, adjust he height of the dial gauge while on a flat plate, to read zero.

I measured the pitches of pairs of leg-holes using micrometer verniers and taking the average of the inside and outside measurements. Divide by two and these are the R values. The r value is the depth measured by the dial guage. The point of the adjustable pitch is to allow the mean radius ovwer the whole mirror to be measured for different size mirrors and also to allow zonal radius comparison measurements to be made across a mirror.

Measurements are taken by zeroing before use, and then tilting from side to side on the mirror until the largest reading is obtained. Three legged types are also available but less easily made.


Mirror mount

For testing you need something to hold the miror upright and flat - at right angles to the testing rig you want to use. So I made this - its a right angle board with triangle buttresses at the back, a hole in the middle for when (IF) I eventually get to auto-collimation testing and the ability via the two-point strap hanging to accomodate different sizes of mirrors. The final points are the top mirror clamp to prevent the mirror falling forward and the rubber blobs on the board to push the mirror forward and flat against its weight in the strap.


Foucault tester

This foucault tester is a blade on a pivot that moves back and forth in a steel slideway, at the end of a 10:1 lever reduction by means of a pivot. The slide is nice and smooth and the blade is on a spring loaded arm. I didn't make this though - I found it on Ebay. However I did make the light source which is a 12v 24W coiled halogen bulb mounted in a Aluminium tube with a removable pinhole. The idea about the tube was to provide a a right angle mirror and a ronchi screen. I haven't completed that bit yet. The light source mounts on to the foucault platfrom near to the blade slideway. More recently I have bought Francis Milsoms Foucault tester from him for a modest sum while purchasing another mirror blank. This foucault tester is more aligned to the screw-based x-y mount as demonstrated by Stellafane ATM pages here


Ronchi Grating


First Grinding log

To setup for a grinding session I needed to provide the mirror mount to hold the mirror on the grinding base and the tool to hang off the eccentric drive arm. The mirror mount I provided for by putting three inserts in the grinding table at 120 degrees on a 6" PCD. This allows any size of mount disk to be bolted into place. The mount disk is mirror-dependent, each different size mirror is expected to use its own. Each mount disk has routed cuts at 120 degrees in which the mirror clamp bolts sit. These are cut at 6.3mm and 4mm to provide a ledge for the M4 bolt to clamp to, the 6mm cut stops the nut from spinning. The clamps hold the mirror centrally on the grinding platter, Freeing one should allow the mirror to be removed, cleaned and reset with out disturbing the centering.

The tool for grinding was made from cutting a disk to the same size as the mirror from 3/4" plywood and using duck tape to create a dam around it. The insides were filled with 1" square tile segments set on edge. Small tile segments were used to fill the odd spaces. These were always on the inside to prevent small bits falling off afterwards. The tiles were set using a runny cement and water mixture.

Once set, the top of the cement was filed flat and body-filler user to attach it to the plywood disk. The plywood disk had a bronze bushing socket pressed into the center of the top surface to locate the eccentric drive pin. The final touch prior to grinding is to cut a piece of carpet to sit the mirror on, on top of the mirror mount. The single reason for this was that the aluminum angle mirror clamps were taller than the mirror. The carpet provided some grip and some padding to lift the mirror above the clamps.

The first 6" mirror was rough-ground using this arrangement and tracked using the spherometer ( construction as detailed below ). I found that tool-on-top and full range of movement on the eccentric was too much and opted for 1" overhang at start and end of sweep. My powerhead drives the platter at 12rpm for the large one and 23 rpm for the eccentric.

During grinding I put two changes in place - I changed the eccentric for a more-or-less fixed position just off centre which I found easier to feed grit and water to and changed the tool mount to a swivel foot embedded in a recess in the tool. The first was due to the grinding reaching the right curve so I moved the eccentric to maintain it while it smoothed. The second was because the rod in the copper bush was too resistant to flexure while following the mirror curves so I wanted something more flexible. The hole was drilled with a normal forstner bit and the body filler was used to backfill using the foot as a mould so the foot would fit tightly when inserted. Since the foot has a swivel joint I want the foot to be snug and let the joint do the work.

The initial grinding was performed using #80 grit and went very quicky using the eweighted eccentric ( upto 7 Kg applies using dumbbell weights on the eccentric arm). This rapidly progressed to 120, 400C and the #303E. On the way I used the sharpie test and visual examination for pits and sparkles under bright light. With the machine I used wets of 5 minutes or so before grit refresh and complete clean off after every 10. So an hour is about 4 wets.

  1. Have ground up to here with large tile tools and small tile tool (75%) diameter of blank but not recorded very well.
  2. 25 Jan 2010: 1 wets at #303E with 2Kg weight to verify ready for polishing. Discovered introduced scratches (2) due to lack of proper cleaning - the machine has been standing in the garage for several months.
  3. 27 Jan 2010: 2 wets of 10 mins each with refills of 400C at 5 mins with 3Kg weights. These wets were with sharpie tracing scratches remains and across other areas to check for contact. Sagitta remains at 0.40mm. Ready for #303 again. PIcture below shows state of mirror blank, still wet from cleaning.

Further Links