When complete, this will be the largest telescope of its kind built by amateur astronomers.
Astronomical (use) telescopes are generally equatorially mounted. The LAT will use a much cheaper altitude azimuth (gun) mounting and uses computers to synthesize equatorial motion.
Telescope animation files; One for us with telephone MODEMS (254Kbytes (note file size)) small animation and for you technology 'haves', a larger version. (1.8 Mbyte (note file size)) big animation of LAT 'tracking' Arcturus.
Well, folks. This was the BEST weekend's worth of work on the mirror since we had our marathon sessions last year.
Prior to this weekend, our last optical test found the mirror in a strange condition. The last 4 to 6 hours of work on the edge that was SUPPOSED to make that zone more parabolic had, in fact, made it seemingly more SPHERICAL.
(Actually, on closer inspection, the outer zone is likely either hyperbolic or toroidal. I'm not sure which.)
Well, this would be bad enough but what made it worse was the fact that there was a sharp slope discontinuity 10" in from the edge between this strange outer zone and the fairly good parabola that we have in the inner 25" of the mirror. This discontinuity made in nearly impossible to accurately measure the shape of the glass.
(Perhaps I should say, it was nearly impossible for ME. Dick seemed to have no trouble with this at all and came to roughly the same conclusions quantitatively that the rest of us came to qualitatively.)
Anyway, the upshot of it all was that this edge at the 10" zone had to go. So, after some discussion, we decided to do about an hour on this zone. And, we decided to do it BY HAND.
The reason for doing only an hour by hand is that we were nervous about working this zone. You see, 10" in from the edge is almost EXACTLY over the 70% zone. And this is the zone that we want to have HIGH on the glass. So, overworking this zone could set us back by some months by taking OUT some of the correction that we have labored so hard to put IN.
Further, this was my first experience hand polishing. You see, when it comes to actually making mirrors with your hands, everyone on my polishing team has more experience than I do. And, I can tell you that polishing with a 9" tool that is well seated while leaning over a 70" piece of glass that is turning at 6 RPM is hard work.
We took turns every 5 minutes so no one would get too tired and went through the hour with no trouble at all.
After that, we cleaned up the glass and let it sit while we went to lunch. After lunch, we came back & stood it up for a test.
Except for the fact that the shop gets REAL hot during the daytime when you close it up to reduce air currents, the test went well. (Well, I had some trouble with the light source in my tester but we used Dave's instead.) It showed, as we expected, that we could work much longer in that zone without any danger.
So, we put the glass down again and went to Denny's to have our board meeting. During the meeting, someone suggested that we continue to polish today. After all, it was still early and we knew that we could polish in that zone safely. So, back to the shop and, with more confidence this time, we allowed the machine to polish that zone with the 9" tool for 45 minutes.
We cleaned up the glass for a second time and, since it STILL wasn't dark yet, we decided to wait for a couple of hours and do a second test.
The second test went better than the first (with the shop being cooler and my tester working better) and we got good wire readings all the way from the edge to the innermost zone 1.5" from the center. (Anthony notices some minor circular striations that the 9" tool and the machine put into this zone. I guess we'll have to change the stroke again next time.) We STILL have work to do on that zone but we'll wait for this last wire test to be reduced to numbers before we polish again.
(As it happens, several of us are unavailable for the weekend of the 21st so the next session will be on 6/28.)
I will be sending out the results later this week but for the moment I am excited by the possibility that we may be able to use this weekend as a model for polishing in the near future.
It appears that waiting just an hour or two is sufficient to eliminate any thermal effects that polishing may have on the optical test at the units of wavelengths level. When we get to within a fraction of a wavelength we will have to revisit this issue. But, for now, this observation allows us to polish, test, decide how to continue, polish again, and test again, ALL IN ONE DAY.
by Dan Zuras
Much of the Large Amateur Telescope progress to date has been in the grinding & polishing of the primary mirror.
After receiving the glass from Tasmania in late 1989, it took more than two years before we were able to get the money & expertise necessary to finish the design & fabrication of the grinding machine as well as the A-frame hoist & lifting bands needed to move the glass.
My notes show that during this time we : visited Lick Observatory to look for sites; designed & built the drip pans; built the A-frame out of scrap in Dick Vaubel's shop; worked out the computer controls strategy; estimated the electrical power needed; rewired the old shop for 220 3-phase; came up with a mathematical theory of the transformation of celestial coordinates to alt-az coordinates that includes fabrication errors; and experienced the Oakland fire that burned down the Medlocks' house. We were not exactly idle.
The actual grinding began on January 18th, 1992 at 3:30 with a mixture of 30 & 60 grit on the back.
We ground the back first for three reasons. First, we wanted to stress relieve the glass. Second, we wanted the back to be flat so it would sit well in the mirror cell. And, third, we wanted to get familiar with how to use the grinding machine in a way that wouldn't risk the optical surface of the glass.
In the two months we ground the back we went from the 30/60 mixture to 60 grit alone to 120 to 240 to 400 & finally to 30 micron grit. The 46.5 hours we put on the back resulted in a surface with a sagitta of one thousandth of an inch over 6 feet. (This works out to a radius of curvature of 10 MILES! Jack Zeiders characterized this as f-long.)
During this time we also : talked with Ethan Clifton about the possibility of using his patented Monoptic dome for our building (we eventually concluded that it would be far too expensive); picked up a horizontal mill from a man named Buckminster; worked out more computer control details; and Kevin Medlock & Rob Toby designed & built the edging machine.
March of '92 was spent getting the kinks out of the edging machine. The edger was attached to the grinding machine to turn it into a vertical lathe with a diamond wheel tool and an automatic feed of about one stroke in 8 hours.
While this meant a very slow cutting rate, the machine could operate unattended safely and, thus, in two months it was possible to put 96 hours into edging the glass. We finished edging on May 14th 1992.
May 16th 1992 was a big day. We flipped the glass over onto the front that day. It took an hour & a half and a half a dozen supervisors but we did it. After that, it took another month before we began grinding as everyone went to Riverside to attend the RTMC that year.
Grinding on the front began on June 12th, 1992 with a mixture of 16 & 60 grit. We quickly abandoned that in favor of 46 grit. (The 16 grit was like grinding with rocks.) The first 53.5 hours of hogging was done with the 46 grit in which we flattened out the original f-2 curve to about f-2.8 by removing 5/8" more glass on the edge than the center.
In September, we decided we needed to replace the steel disc we were using as a grinding tool. At 30" it was too small & was thought to be riding over some of the larger irregularities.
So we bought 360 pounds of Quickcrete and cast a 42" diameter tool using a Sonotube as a mold. (Sonotubes are those thick cardboard tubes that are used to cast freeway overpass supports.) After waiting a couple of weeks for it to cure, we sealed it with Bondo and attached a number of Mexican terra cotta floor tiles to it as the grinding surface. We also attached a mounting plate with a trailer hitch ball in it to connect to the trailer hitch we use to move the tool.
It was in the process of seating this tool that I learned a hard lesson : when you are seating a new tool, always use a short stroke. I used a stroke that went over the edge of the glass & knocked a few chips off before we could stop the machine. (Until the tool seats, there is danger of the corners of the tiles nicking the edge of the glass.)
Grinding continued into 1993 with this tool for another 46 hours and, then..., the turntable bearing seized up. We had gone from 46 grit to 300 grit by this time & we believe that grit had worked its way into the bearing race and ruined the bearings.
This put us out of commission for a couple of months until Kevin could design & build a new turntable bearing out of the thrust bearings from a couple of trucks.
The new bearing was finally installed on May 22nd 1993. It worked better than ever right from the start & it is still in use today with no evidence of wear at all.
Grinding continued right down to 10 micron grit until we finished on August 14th 1993 at 6:30 PM. We had put 269 hours on the glass altogether in the year & a half of grinding. The back took 46.5 hours, the edge took 96, & the front took 126.5.
During this time : Rob Toby built a spherometer to measure the curvature of the glass; we checked out Morgan Manufacturing in Aptos to see if they could make the race for an observatory dome; worked on a grant proposal; rebuilt the turntable support; added a fan to cool the turntable motor; and rewired the grinding machine for slow start.
In September 1993, I went to ABE Dental in Belmont and bought 200 pounds of dental stone. (The man that sold it to me said, "Boy that must be SOME cavity you've got there!") We used this and the rest of the Sonotube to cast our first pitch lap.
During this time, it was decided to move our operation from Hayward to San Leandro to save money on rent. The Hayward shop was costing us $636/month. We now share a shop with Epoch Instruments for $200/month. While the extra distance added to my commute, at a savings of over $400/month, it is well worth it.
The move took us through the holidays and then we started polishing in the new shop on April 16th, 1994 at 2:00.
Within a month, we had a good enough polish to do our first optical test. Dave set up the test instrument and, sure enough, we had a sphere! This is exactly what we hoped for at this point and meant that we were OK to start figuring for a parabola.
During this time : Denni Medlock & I went to a chop shop & bought 3 motorcycle chains to replace the worn out drive belt; Dave Barosso wired up the new shop for 220 3-phase; we mounted the attachment plate on the pitch lap (TWICE!); we built an enclosure as a clean room for the polishing machine; Rob made a mold for the pitch squares; we cast the pitch onto the tool; and John Bower came up with a design for the mirror mount.
For a small mirror, the difference between a sphere and a parabola of the same radius is measures in tens of millionths of an inch. For our 70" f-3, the difference is 3 THOUSANDTHS of an inch!
While that may not sound like much, Anthony Stillman once calculated that one wavelength of glass removed from the entire surface of a 6' diameter disc amounts to 1.5 cubic centimeters of glass. And, that 1.5cc must be removed by POLISHING it away! Experience has shown that we can move anywhere from 1.5 to 6 wavelengths in a 12 hour polishing session.
Of the 136 wavelengths we must remove to go from a sphere to a parabola, as of the time of this writing, we have removed all but 24.5.
The last two years have consisted mainly of long hours polishing & testing. We try to polish in 12 hour marathon sessions lasting all day Saturday and test as often as we believe we need to.
I am often amazed at the dedication of the polishing team. Those marathon sessions mean getting up early on a Saturday morning, working all day, and getting home well past midnight.
In these last two years we have : changed out the eccentric drive pulley 2 or 3 times; gone through 2 pitch laps (thats not counting 3 pitch recasts & we're nearly ready for our third lap); changed the polishing compound; improved our cleanroom procedures; developed a new theory of mirror testing; seen Mark Thein design & build a roll down door for our cleanroom; refabricated & replaced the lap overarm; replaced the hitch plate after the failure of the old one; thought we had some astigmatism that turned out to be band strain; replaced the overarm pivot; performed 10 optical tests and rasped & channeled the lap & changed the polishing stroke countless times.
To date, we have put 228.5 hours into polishing and we believe that we have another 120 hours or so before we get to within one wavelength of our final figure. At the rate we tend to go, that works out to another 10 months or so before we can see the end of this task.
So, if all goes well, by this time next year, Group70 should be able to say that we have finished the primary mirror. All in all, a significant milestone in the creation of a new telescope.
Joe Kral (ME-10)
& Kevin Tibbs (ALT-AZ consultant)
(use the BACK button on your browser).
Last modified: Friday, 18-Mar-14