Guide to Making Schmidt Correctors

Contributed by Bob Pfaff

A few notes on making a Schmidt corrector by the vacuum method. These were originally posted in 8 sections to the ATM listserver.

First - Why the vacuum method?

Part One - Literature and Formulae

The formulae and math needed to make a corrector plate are a bit much to describe here. Besides, they can all be found in an article - The Vacuum Method of Making Corrector Plates (Sky and Telescope, June 1972, pg 388-393).

Most of the information is by Dr. Edgar Everhart from his article in Applied Optics (May 1966, Making Corrector Plates by Schmidt's Vacuum method). The math is here to do the design. I used the partial vacuum method with 1/2 correction each side of the plate. No need calculating Young's modulus because the plate deflection will be measured with a dial indecator (see page 389 of the S&T article). Also do not add extra deflection for the weight of tool as the pan is filled with water; preventing extra deflection. Be sure to do the example on page 392 to check your math.

Other articles of Interest:
Related, but Not Using Vacuum Method:

Part Two - Tools

  1. The vacuum pan. I will describe this in the next section.
  2. One inch micrometer to measure the wedge of the corrector plate. We want to keep the wedge to less them then .001". Enco sells imported micrometers for as little as $10.00.
  3. Dial indicator that reads to .0001". The corrector plate will be deflected in the range of .002" to .007". Enco has one listed for $39.95.
  4. Convex grinding tool to grind the plate. I used a tile and plaster tool. The tool was generated by grinding the back of my primary mirror blank against it. To generate my larger 16" tool I glued a 1/2 inch plate glass disk to a plaster disk and ground it against the tool. Remember the tool will be convex in the range of .002" to .007" almost flat. Use the spherometer to check this accurately. You can stop fine grinding at about 500 grit.
  5. Special spherometer which has to be made to fit the plate diameter. The spherometer should be large enough to just set inside the edges of the plate. I built what I call a hybrid two footed spherometer. A third leg at 90 degrees from the center like a (T) shape is added so it will stand up leaving both hands free.

    You can replace the dial indecater with a micrometer head and have a great two footed spherometer to measure the sagitta of most any mirror. For this you will have to measure the distance accuratly between the balls. (Advanced Telescope Making Techniques, Vol 2, pg 100).
  6. Your lungs. No you do not need a vacuum pump to pull the little vacuum that will be needed.

Part Three - The Vacuum Pan

To make the vacuum pan a metal lathe is used. I machined my pan from 6061T6 aluminum plate stock (very common in the metal world). I purchased a square piece of plate (one inch thick for my 11" pan and a 1 1/2" thick for my 16" pan). The corners were cut off with a wood table saw. A cheap carbide tiped saw blade covered with lots of motor oil was used on the saw to do this. Wear a safety face shield for this operation! (A bandsaw with a metal cutting blade would be my choice - Matt)
Three blind 1/4" 20 holes are drilled and tapped in the back of the plate for bolting to lathe face plate. A spacer such as a washer is placed on each bolt between the face plate and the pan plate. I used 1/2" ball bearings for washers. Do not bolt directly to the face plate as this will distort the pan plate. Machine the pan 1/2" to 1" larger then the optical diameter of the Schmidt plate. On the front side of the pan the pan depth is turned to about .100" to .150" deep. The inside of the pan is the optical diameter of the Schmidt plate. In my case, I machined a 1/4" seal rim around the pan. This can be less but I think the extra width of the rim gives a better seal. Also, the 1/4" is covered by the Schmidt corrector cell which will cover any turned edge.

Part Four - More on the Vacuum Pan

Now that the vacuum pan machining in the lathe is finished the pan is removed. The next step is to drill an internal (L) shaped hole in the pan that enters the side of the pan and exits into the pan. (1/8" or less in diameter), see diagram.

A small valve such as a good gas valve is fitted to the entrance hole on the side of the pan. A valve that had 1/8" pipe thread was fitted to the pan by using a drill and pipe tap. I had problems with some valves leaking, causing loss of vacuum. To fix the problem I disassembled the valve, lapped the parts and reassembled it with laboratory vacuum jar grease. A short length of small diameter hose is attached to the valve which is used to suck the vacuum.

Last step in making the pan is to lap the seal rim. Lay a piece of plate glass on some carpet. With some fine grit grind the pan rim against the plate glass. Check the rim at many points around it for flatness. Get the rim flat to .0001" or less. Use the spherometer that was made to do this. This is the only critical part in making the pan and like mirror making it is ground in. Also important is to check this at about the temperature which you will be grinding and polishing. I painted the inside of my pan with black lacquer. This helps when checking for pits as the plate is polished out.

Part Five - Glass

I will start this part by stating that I have no knowledge of the glass or optical business. I can relate my experiences with finding glass for my two projects.

For the 11" plate I was able to obtain from an optical friend a BK7 plate that was a clear glass with a coating used as a filter. The wedge was only .0002" and the coating ground off as the plate was ground. No problems here. As for the 16" plate I was not that lucky. In this case I needed a 16" 1/2" plate. Plane parallel BK7 would do the job. In this size I was quoted 1K plus dollars which was out of my low budget range. I talked with another glass supplier that had a glass called water white. He said it was plane parallel and was for optical windows and filters. It sold for $100.00 per sq. ft. but only came in 1/4" thickness. Next I located a glass dealer that makes glass table tops. I ordered a 16" table top of a glass that the dealer called crystal white. Also ordered was a 16" disk of plate glass. Both were unedged or edge polished (big error as the edge had 1/8" chips in the disks). The crytal white was almost white and cost twice as much as plate glass but it had .015" wedge! As not being up to grinding .015" wedge from a 16" disk I ordered a another 16" plate glass edged and polished disk. This one had only .0004" wedge. Yes in a 1/2" thick plate it does have a green tint but in my Schmidt camera this is no problem.

I think any good glass can be used in a plate if the thickness is in the correct range and the wedge is .001" or less. (more if you don't mind grinding the wedge out). The index of refraction only needs to be in the ball park. How about those solar filters? The coating will grind off. For you low budget people remember the original 48" Schmidt had a plate glass corrector. A lot those SCT's out there have green glass correctors also (looks like window glass to me).

Part Six - Mounting the Plate to the Pan

The first thing to do is to zero the spherometer dial indicator. Adjust the dial indicator in the spherometer so that there is enough range to read the concave plate and the convex tool. Lay the plate on some carpet then place the spherometer on top and zero the indicator. Now lay the spherometer on a flat surface such as a machinist's granite flat. Take a reading and record it. Any stable flat could be used if you mark it where the spherometer will sit for a reading. This is needed to calibrate the spherometer in case you bump it or lose calibration. This reading may be different than was recorded on the glass plate because the glass may not be flat.

Next the plate is sealed to the pan. I put a thin coat of laboratory vacuum jar grease on the rim of the pan. Other thick greases could be used. A friend had good luck with non-hardening Permatex. The pan is filled with distilled water. The water should be preheated and then cooled to drive the air from the water. Hard water can etch glass. Lay the plate on the water filled pan. Try to keep out as many bubbles as possible. A few bubbles will be no problem. Place the spherometer on the plate. With safety goggles on, as plates sometime break, suck on the small hose connected to the pan valve while watching the dial indicator. When the plate is deflected the correct amount, close the valve. I put RTV around the edge of the plate to prevent vacuum leaks. After a few hours the plate will loose some of its deflection due to squeezing out the seal grease. Resuck the deflection. It should be stable for 24 hours before any grinding is done. A .0002" change from day to day is normal. Note the size of the bubbles under the plate. If they increase in size there is a vacuum leak.

Once the plate deflection is stable start grinding with the preformed tool. 500 grit is a good place to start. The pan-plate is on the bottom but if you overshoot you can put the tool on the bottom. Use very light pressure in your grinding. Check the tool with the spherometer frequently as it will have to be reground to shape now and then. Also check the plate deflection to be sure it is holding. Grind through the grits like making a mirror down to 5 - 3 micron. Polish out with a standard pitch lap. After and beginning each session recheck the deflection in case it changes due to leaks.

It has been suggested by some ATM's to rotate the plate a few times during grinding to minimize astigmatism. I prefer not to break a good vacuum seal and loose the calibration. Because the plate is to be ground on both sides I mark the plate and pan. When side 1 is done I release the vacuum, turn the plate over and rotate 90 degrees with respect to side 1. I think this should cancel most of the astigmatism.

Now reseal and deflect the plate. Grind and polish as before. Release the vacuum and the plate is finished!

Part Seven - Testing

A simple way to test a Schmidt corrector plate used in a classical camera is shown below:

A test rig is made of wood. The corrector is at the RC of the mirror. The flashlight is near the focus of the mirror, adjustable and pointing at the mirror. The flashlight is covered with a ronchi grating. Now look at the corrector from a long distance (150' or more), the more the better. You should see straight lines. If the lines flare at the top or bottom the collimation is off. Unless the bands are grossly bad make any corrections on the primary. My two plates needed no correction. I could only see some aberration with one ronchi line showing which was probably the primary mirror. Just for fun remove the corrector and take a look. The lines not only bowed but the outer lines went into circles. For correctors in SCT's or Wrights which are some percentage of a true Schmidt, the corrector will be tested in the complete system. A flat (water, mercury or glass) would be ideal. An artificial star as described in Star Testing Astronomial Telescopes by Suiter with a ronchi grating should work. As in the case of the true Schmidt the correction, if needed, should be done on the primary.

Part Eight - A Few Extra Notes

Some things that I should mention: Anyway this will end my ramblings on this subject. I hope this will inspire someone to try a plate. A little math, machining, working with clear glass and willing to try something different can be fun. :-)

Bob Pfaff
4505 Merrill Street
Torrance, CA 90503