It has been again six months since I wrote one of these. It seems like only yesterday. A lot has happened. The Mark IV prototype has taken a lot of data. It is not yet good data as the lenses have been found to have a coma problem. We hope that we now understand it and that we have a fix. So this note will contain only a data sample.
Fig 1 shows the Mark IV wiring. The neat and tidy wiring of the Mark IV provides ... er ... the scramble wiring of the Mark IV provides low cross talk while assuring that we can maintain the precision of the 16 bit data signals. OK, the wiring is a mess. We have designed covers that will cover it all up so it will look sort of neat when it is all done. This is a complicated system with lots of interconnections. I don't like connections and connectors as they are a source of failure. The Mark IV was designed with a lot of connectors and a lot of interconnections as a way to build it with one person and to have a way to test and check it out. So far, I have had little trouble with all the connections, but this does not mean that I like them.
There is a production line going. Fig 2 shows six dual telescope carriages lined up on the work room floor. The floor tiles are one foot square to give a size comparison. All the telescope parts are really crowding me out of my work space.
I have mounted the Mark IV prototype in the south west tower. Fig 3 shows me standing by it. You can't see the lenses which are hidden by the large paper lens shields.
As you can see in Fig 4, the telescope must be mounted in the tower at an angle. Mike is posed by it. This is because the tower faces south west. With the doors fully open and the telescope on the floor, one can just see over the trees to the south. Point too high in declination and one is looking at the roof of the tower. Sigh! But it is fine for testing, and I can get all around the telescope to work on it. later I will put it in "Dead Barney", part of which can be see to the left of this picture. The floor level would be the floor level for a third floor if I had one.
We have taken a lot of exposures with the camera in this position. It is quite convenient as one just has to go up the spiral stair and open the doors and one is ready to map the sky. So far, we have no really good pictures to show. There was an error in the specification of the lens tube, and most of the lenses had an improper spacing between the first three and the last two lens elements. All the pictures taken so far display coma in the corners, and probably less than ideal focus.
A collection of geese is a gaggle, a group of quail is a covey, it looks like a collection of broken lenses is a quilt. The lenses were all stored in my garage during a normal Chicago winter. When I brought one out to work on the spacing problem, I discovered that the image end elements were broken. On examination many lenses were found to be broken. Fig 5 is a line up of the lens cells that were found to have one or more broken lenses. The quilt squares are about 6".
The optics were designed by Elliot Burke. When he became busy, we had Steve Wright design the mechanics of the 5 element assembly. This is, of course, asking for trouble. Some little bit of information failed to be transmitted. The spacing between the first three and the last two lens elements is off by about 0.2". The lenses and lens tubes were then fabricated in China at, I am told, a laser factory. The design called for gluing the lenses into the cells seen in Fig 5. Some were glued with 3 dots of glue. Some were glued all around. Those glued in three places survived. Most of those glued all around were broken when the winter temperature in my Chicago garage caused the lens cell to contract more than the glass.
It is interesting to see how this happens. Glass is quite strong in compression. The lens cell acts like a metal band around the lens which squeezes it from all directions. When this happens the glass has no place to go other than to bulge out from the median plane. The glass actually fails in tension. It just splits down the middle. This is shown in Fig 6. The glass tends to split in a ying-yang pattern. The lenses were glued in place with some marvelous glue. Nothing seems to attack it, and it stands high temperature. I picture the laser factory as being in one wing of a plant where they manufacture supersonic aircraft. When they needed glue, they just went over into the next bay and got a pot of the same stuff used to glue the wings on the supersonic aircraft.
We had to try to rescue the lenses as we expected colder temperatures on a mountain than in my garage. The way (after testing various solvents) we finally removed the lenses was to just put them in the oven and raise them up to warm oven temperature (350 F) as shown in Fig 7. This reverses the problem. Instead of crunching the glass, the cell wants to pull away from it. Lower temperatures than 350 F did nothing, and even here the glue was still hard. At 350 F the differential expansion was large enough to break the glue seal. So it was not epoxy. I was told it was a UV cure glue. Once I had them up to 350F, I was able to push the lenses out of the cell with an oven mit. Most that had not already broken survived. A few were simply pulled apart by the glue.
Since the factory failed to blacken the lens edges, I tried this out on a couple of "paperweights". Using the recommended magic marker, I coated the edges with black ink. The ink wicked into the cracks and make them quite visible, even though most are hard to see without some sort of aid. You can see the shear plane on the lens at the left in Fig 8, and also several starting ying- yang cracks.
Just so you all recognize that this is an astronomy project, we include a star picture (above). This is a 200 second exposure taken through the I filter. Alpha Serpens Caput at RA 15h 44m and Decl. +6d 25m is the bright star that trails almost entirely across the page. North is to the top, and west is to the right. The raw image was 2k x 2k at 16 bit resolution. It has been binned 2 to 1 in x and y, and then put into .gif format to conserve space. Many of the stars are saturated in this image. That is OK for us. We are interested in the stars that are not saturated, which is most of them. This is just to give all an idea of what we will be working with in tass. These are large images. They contain a lot of stars.
I have included two small portions of a Mark IV frame to demonstrate the coma problem. Each are 1/100 of a frame and are 200 x 200 pixels. Corner (the left-hand picture) is taken near the upper left corner of a frame and shows the stars a little "moths". Center (the right-hand picture) is taken at the center of the frame and shows the stars as round dots. We hope that we have this fixed, but do not yet have a picture to prove it.