ATM Diary

Refiguring a Meade 16" f/4.5 mirror!

Dave Nelson recently approached me to refigure a Meade 16" f/4.5 mirror for him. He is mounting this mirror in a telescope body that's he's in the process of fabrication. Dave graciously gave me permission to document the refiguring process here. This one should be fun!

05/26/03

Before starting to work on the surface of this mirror, I wanted to measure everything to provide a basis to judge the progress on this mirror. This mirror is currently aluminized, so I set up a Foucault tester to begin the process. First, the gross specifications on the mirror:

Diameter is 16-3/16", with a clear optical diameter of 16-1/8". I am a little surprised that this mirror is actually bigger than 16"!

Focal length is 72.25", so the radius is 144.5". This was measured to an accuracy of about 1/8". The focal ratio is thus f/4.48, but I'll refer to it as f/4.5 here.

Thickness is 1-5/8", and the mirror is Pyrex.

Overall Foucault (TEX) wave rating is 1/0.7 wave.

TEX Monte Carlo result is 100% certain the mirror is worse than 1 wave. Visual inspection: Mirror has a couple of small scratches that will not be removed during the refiguring. It also has quite a few cleaning sleeks that may or may not come out during smoothing. There is also surface damage from retaining clips that will not be removed during refiguring. The current surface has a 2" diameter center defect that will probably be eliminated during refiguring, but it should be entirely hidden under the secondary anyway. The surface currently has severe radial defects from machine polishing, and these should be improved greatly during refiguring. Most seriously, the mirror suffers from quite a bit of astigmatism. Note the quite different (and consistent) tests across the two different diameters. This mirror is thick enough that it is definitely surface difference, and not a measuring artifact.

05/27/03

Poured two dental plaster bases to use for fabrication pitch laps. I made one that is 11" in diamter, and one that is 8" in diameter. The plan is to use the big one for smoothing the mirror, and the smaller one for most of the figuring.

05/28/03

Poured pitch on the 11" lap, and channeled it with the tire grooving iron. I used #73 pitch, and it was really tough pushing it through the lap. Used this 11" tool with cerium oxide and worked the mirror for about three hours, using just a standard polishing stroke. The purpose of this was to remove the aluminum coating from the mirror. I had read that the aluminum would "come right off" but in reality it took three hard hours of polishing before the aluminum film was history. Next time I will consider removing the aluminum chemically!

Also, the pitch lap failed after three hours. I lost three facets during this effort. The pitch just didn't adhere well to the plaster tool. I believe I made a mistake by not using turpentine on the face of the tool, and by having the plaster be a little wet.

06/12/03

Back to work on Dave's mirror. Today I poured an 8" pitch lap on the plaster tool base. I tried to improve this one. I went to #64 pitch, as the #73 was too hard. Also, I cleaned the plaster base with turpentine before pouring, and I also roughed it up quite a bit by scraping a saw blade over the face of the plaster base. Poured pitch, and channeled it with the tire grooving tool. Very much easier pushing with the #64 pitch! Ended up with a beautiful lap.

06/13/03

Friday the 13th, but I decided to work on Dave's mirror anyway! I put in six 10-minute polishing sessions. The goal was only to try to smooth out this mirror. So I used a broad "W" stroke that overhung about 2" on the left and right, and went about 6" back and forth, total travel. I realize this is probably going to reduce the current (poor) paraboloid some, so I threw in a few bigger "W" strokes to keep some of the parabolization in place. Later note: ran a quick Foucault test. The surface is already quite a bit smoother, but I have flattened the parabola out, as expected. It's now a 1/0.5-wave mirror, and needs to be seriously deepened in the center.

06/14/03

Saturday, did four sets of 10 minutes each. S1 was half CoC and half CoC but slightly offset to the left. S2 was a big "W" stroke. S3 was CoC with a slight "W". S4 was a medium-large "W". All strokes should have deepened the parabola a bit. After testing, need to deepen the center more.

Session #2: Did 25 minutes, mostly CoC, a little with tool edge pressure over the center of the mirror. Need to deepen center even more yet.

Session #3: Did 10 minutes with pressure on the tool edge over mirror center, with a 6" stroke. Did 5 minutes same setup, but with an 8" stroke. Still need to deepen center more.

Session #4: Did 16 minutes, broken up among three shorter periods, all with edge tool pressure over mirror center, with about a 4" to 5" stroke. Center is now deep enough. Need to deepen Z3 and Z4 a bit, and deepen Z5 just a little. Mirror is now 1/3.1-wave according to TEX, after averaging two sets of measurements.

Session #5: Went to rouge. Did 3 minutes tangential over Z3, Z4 and Z5. Then 1 minute CoC, no pressure. Finished this session with 2 minutes in a broad "W", but avoiding the center. Let mirror cool down before testing, and averaged three tests. Mirror rated at 1/2.7-wave @Z2. Mirror is pretty smooth, but there is still a small defect at the center, left over from the original figuring by Meade. I measured center zone based on diagonal covering the very center of the mirror. Need to shorten up Z3 a fair bit, and a little in Z4 and Z5.

06/15/03

Sunday, did an eight-minute session with rouge. Did 3 minutes edge pressure over Z2, then 5 minutes edge pressure and tangential over Z2, Z3, Z4 and Z5, slightly over Z6. Made a new Couder mask tonight, using a central spot to cover the diagonal shadow. Inside radius of Z1 is now 1.25", and outer radii of zones are: 3.483, 4.764, 5.767,6.620,7.375, and 8.060. Perfect zone separations are now 42,36,39,35,35 or so, as found by experiment, not calculation, for a 1/20-wave surface. New test shows the center is too deep, so I need to bring Z2 through Z6 down.

06/16/03

Did a six-minute session, working on Z6, and in general over the 70% zone. Used a tangential stroke on Z5 and Z6, then moved it over Z3 and Z4. Result was good, moving from 1/2 to 1/3 wave. Still need to deepen the edge a bit, and shallow up the center a bit, and it should start to look good. Basically repeat this session, although from the measurement, it looked like the center isn't going in the right direction. We'll see! BTW, I whacked the pitch tool on the edge of the sink after this session, and knocked off some pitch from a couple of facets on one edge. I'll give it a good hot press and see if it can go on. Looks like a 7-year-old with a couple of teeth missing! BTW2, finally sat down and calculated a set of perfect reading separations for the zones. They are listed in the log below.

Second session of the night. Did the hot press, and then did 4 minutes over Z5 and Z6 (especially Z6) then 3 minutes on Z2, and 1 minute with a "W" over Z2 through Z6 to smooth it a bit. Got a very good test, 1/4.7 wave. Need to take a little more glass off all the outside zones, and a fair bit off Z2, as the center of the mirror is the only problem.

06/17/03

I put in a quick session today before lunch. I did 3.5 minutes with edge pressure over Z5 and Z6, then 2 minutes with edge pressure over zones 4, 5 and 6. Then 2 minutes with even pressure with the center of the tool doing a "W" over Z4/5/6 to smooth a bit. Just had time for a preliminary test, but that looked pretty good. Will let the mirror sit for a few hours, then do a multiple-pass Foucault test to see what I've actually got right now. Preliminary say 1/5.2 wave @Z1, but that was immediately after figuring.

Later... I let the mirror sit on the test stand for a few hours, and tested it in much darker conditions. The average of three sets of readings gives my a 1/3.1 parabola, Z3. Generally I need to take off glass in Z2 through Z6.

Session #2: I did 7 minutes on Z6/Z5, 4 min on Z5/Z4, and 2 min on Z3/Z2. Waited 50 minutes to test. Test showed the center much improved, nearly perfect actually, but the edge zones are all too long. Need to put in some serious edge effort, before it gets out of control.

Session #3: Did 10 minutes on Z6/Z5, then 3 minutes on Z5/Z4. Trying to shorten the focal length on the outer zones. Not working. Outer zones getting longer, overall correction growing. In other words, things are going the opposite direction expected.

06/18/03

Re-cut the channels on the 8" tool with the tire grooving tool. Then used the brass brush to scour the surface of the lap. Gave it a hot press with screen to micro-facet the face of the lap. Then cold pressed until it was ready to go. Did four sessions, 4 min, 5 min, 5 min, 6 min (20 minutes total) doing a "W" stroke with the edge of the lap 1" to the right of the mirror's right edge. Then "W" to the left until it's 1" inside the edge, for a 2" total left-right travel. Stroke height was about 2 or 3 inches. Got a fairly good test. Mirror is now TEX-rated at 1/4.4 wave @Z5. The total correction on the mirror shortened from 207 to 198. Need more work on Z6 to shorten it.

Second session tonight. Did two stints on the outside edge, one for 6 minutes and one for 7 minutes. Stroke was a very narrow "W", starting with tool edge 1/4" outside the edge of the mirror, and stroking further out until the tool was overhanging 1" or a bit more. Put pressure on that edge of the tool, and lifted up slightly on the part of the lap nearer the center of the mirror. After those two sessions, I did a "W" stroke with the edge of the tool over Z2, with pressure on that edge.

06/19/03

Did a four-minute session with a radial stroke. Started with the top edge of the lap 1" below the top edge of the mirror, then pushed it up until it overhang by about 2" or a little more. Then did 1 minute with edge pressure over Z1 and Z2. Good test, improving, but need to shorten the radius on Z5 and Z6, and then deepen Z1 and Z2, as Z3 and Z4 are too shallow. Overall correction on the mirror is about perfect, I just have to redistribute a little glass.

Second session. Goal here is to shorten the edge considerably, which will take quite a bit of overall correction off the mirror. This will look like a step backward, but it prepares the surface to be able to do the final few figuring steps. Did 8 minutes on Z6, with a narrow "W" going from OH 1/4" to OH 1". Then did 2 minutes edge pressure over Z1 and Z2. After testing: correction went from 186.5 down to 173.5. Edge needs just a slight bit more reduction, then bring up Z5 and deepen the center quite a bit, and we'll be home!

Ran a fairly complex session to get the mirror moving back in the right direction. Did 4 minutes "W" past Z6 OH 1/4"-1". Then 4 minutes radially, OH -1: to +2". Then 4 minutes with edge pressure on Z2. And 3 minutes with edge pressure on Z1. I let the mirror rest overnight, and then tested a set of readings to see about where it stands. This quickie test shows the overall correction is182, with roughly 187 being my final goal. The mirror needs to have the edge shortened up a touch, and then a slight bit of radial work to lengthen out Z5. Then a quick touch to deepen Z3 a bit, and I'll be pretty close, I think!

06/20/03

This session I put in 4 minutes on Z6 (1/4" to 1" OH "W") and 3 minutes with a radial stroke, -1" OH to +1.5" OH. Then I did 1 minute with edge pressure on Z3, and 1 minute edge pressure over Z1 and a little on Z2. Test shows I still need to shorten the edge, deepen the middle, and redistribute the middle zones. Average is 1/5.2 wave at zone 3.

06/21/03

Did a seven-minute session deepening Z1, Z2, Z3 and Z4 a bit. Used a "W" stroke with edge pressure over those four zones, then a little CoC, slightly offset, to smooth things out. The edge lengthened out more than I would have liked, and the center isn't quite deep enough yet, but the overall test stands at 1/6.3 wave at Z5. BTW, I had a bit of a problem tonight. After spending an hour doing a "final" measurement on the mirror (3 diameters, 3 sets of measurements on each diameter) I discovered, to my horror, that the probe on my dial indicator on my Foucault tester was riding at about a 10-degree angle from the horizontal. I corrected it and tightened it up and remeasured one diameter. Sure enough, the mirror went from 1/8.3 wave to 1/6.3 wave. This may also be why Dave's 8" mirror tested a little better on my Foucault tester than on his tester at home. At least the problem is fixed now!

06/22/03

Getting close to the end, now. Did a session where I worked with tool edge pressure directly over the mirror edge, Z6. Did three short sessions there, that totaled 10 minutes. Plus I did a 3-minute effort with edge pressure directly over the center of the mirror. Got a great test, averaging 1/9.6 wave at Z1. Hopefully one last session will finish this mirror. I need to deepen the center reading by .003 and this thing goes to bed better than 1/10th wave. Will have to do a multiple-diameter set of tests to finalize the figure, but it's looking very good. Just don't chip it now!!!!

Finished Dave's mirror with a final 2 minute session, working with edge pressure right over Z1, the mirror's center. After cleaning, I let it sit on the test stand for 90 minutes, and then gave it four tests on each of two diameters. I saw a very good result, with the surface averaging 1/16.4 wave, at Z1.

Final Test Summary

This mirror is now quite good, averaging 1/16.4 wave according to Michael Lindner's TEX program. Originally the mirror was very rough, and had a strong artifact in the center. After refiguring with rouge, the surface is now nicely smooth, and the central defect has been very nearly eliminated. The original mirror had a fair amount of astigmatism, and that has been completely removed now. Comparing the average readings for the two diameters (four sets of measurements each, in six zones) the first diameter measures 1/13.8 wave, while the second diameter measures 1/14.9 wave. Looking at the overall correction, the difference between the Z6 and Z1 readings is 185.75, while on the second diameter the difference is 185.88. I also entered the Foucault test data into Andreas Reifke's FTA program, which determined the Strehl ratio to be 0.9901. It also reported at 80% confidence that the strehl ratio was 0.9739 or better.

This mirror refiguring went pretty smoothly, considering it was the largest mirror I've done to date. The project was finished in 22 refiguring steps. Actually, I might argue that there were only 20 "refiguring" steps, as the first step was removing the aluminum coating, and the second step was just an hour of regular polishing to smooth the surface. Total contact time between the pitch laps and the mirror was under 7.5 hours. Of course there was about three hours of lap making. I ran, I believe, 58 Foucault tests. That probably represents about 10 hours of time between the tester and the computer. And each of the 22 sessions had about an hour of elapsed time for hot and cold pressing, lap maintenance, rouge mixing, cleanup and the like. So the total effort was probably something like 40 hours of total time, spread over 11 evenings. That gives a ratio of 7.5 hours of actual figuring versus 40 hours of work. Interestingly, it took 16 sessions to get to an "acceptable" 1/5th wave surface, and only six more to achieve the nice 1/16th wave mirror.

Gordon Waite

Refiguring Log, Dave Nelson's 16" f/4.5 Mirror

*PTT=Powered Turntable

Click here to go home!