Making a 110mm f7 mirror.

1: Preparations

The challenge for this mirror is its small size combined with low speed (f/7) and the fact that it has to go in an existing Vixen tube. I believe this tube was originally meant for an 4.24" f/8 mirror. The work started on August 29, 2007.

cutting setup As usual, first the disk has to be cut out of a piece of 19mm glass I had still in stock. A biscuit cutter made from a plywood disk and some copper strips, mounted in a drillpress, surrounded with plenty plastic, do the trick.

cutting setup

Above, the blank has just released. You can see the putty dam, and the mess it has made.

Cutout blank

Circular scratches on the back of the blank indicate the wear when it came loose. Next step is to grind flat the back. Float glass is usually very flat, and especially for a blank of this size back grinding is not really neccessary. However, last two mirrors I forgot to do this, and found it is especially annoying during the testing. So grab a tile, left over form the kitchen floor, and grind the back of the mirror with some SiC180 and water. A single session does the trick, and when prolongued, the grit has broken down sufficiently to give the back a #600 look.

Grinding the back  Ground back


2: Rough grinding

This was started on the same day the blank was cut. The pudding bowl left over from the 150f4 was used as tool.

Hogging setup  Initial cavity

As you can see, the hogging affects the center most. Special care must be given to the edges, since the sagitta is reached quickly, it's only 1mm. In the right picture the sag is almost there, but I will do a bit more roughing to get a better approximation of a sphere: this may save some time during fine grinding.


3: Fine grinding

The calculated sagitta of 1mm has been reached. Well, in fact I've overshot it as later focal length measurement revealed, because my calculation was faulty: filled in F instead of RoC...

Sagitta reached  Start fine grinding

I made a tile tool, still ignorant of any overshoot: in the right image the setup can be seen with the plaster still curing. The #80 carbo is waiting in the background to do the first mating wets.
After contact is reached the tile, tool and blank are rinsed, and the fine grind commences with #180 SiC. Tool on top first, to work the edge regions a bit more. The left image above shows the usual flat edge left over after hogging. Full contact is tested with sharpie test, and this is also the time to find out about the overshoot, by wetting the mirror and imaging a ceiling halogen light back on the ceiling.
Float glass cuts fast, however, so the focal length was quickly extended while goind down through the grits: #180, #360, #800, #1200. It all takes just a couple of hours with this glass and this size.

Scratches  No scratches

Toward the end of fine grinding the mirror surface is carefully studied, and yes: there are the scratches! (left image) Luckily quite shallow, and one more gentle wet later the scratches appeared to be made during the #1200: they are all gone now...

Red-out test The final test for the quality of the fine grind is the red-out test. Any more-or-less smooth surface reflects an image up to some limiting angle. With rougher surfaces the angle will be more shearing, or toward 0°. Fine ground surfaces, such as a mirror after #1200 grit, will reflect up to 30°-40°. In the imgage on the left, the foucault tester slit is reflected up to about 30° with restect to mirror surface, and at this point it just starts to "red-out". Therefore, I call the fine grinding phase done!


4: Polishing

Polishing always starts with making a lap, no difference for this mirror. Like for the grinding tool, a plaster base is cast on top of the mirror blank. To prevent scratches and enable easier release, the mirror is first covered in plastic clingfoil. When cured, a bonding layer is made by soaking the top of the lap base in Richards' Secret Sealer sauce, a mix of pitch chunks dissolved in turpentine.

Pitch foundry  Fresh pitch lap

On the left the pitch foundry can be seen, while on the right the fresh lap surface is shown. Pitch viscosity is more critical for small laps, as I found out: if only a little too fluid, a lot will spill over the edge. Excess pitch has been cut off with a knife under running water.

Pitch foundry  Channeling setup

I channel with an old 100W soldering iron, which has a V-shaped piece of thin sheet metal is attached. The grooved are plowed out of the pitch surface. Daily maintenance is further done with a sharp knife under a running tap. First warm press the lap into contact, and then the poishing may commence!

Polish 15 minutes  Polish 45 minutes

Polishing goes fast on a small mirror like this, and with the softish floatglass. Apparently the final #1200 has done a good job; after 15minutes of polishing the edge is already clearing up. Half an hour more clears most of the surface, except for a ring around the center region. An early foucault test reveals a fairly good sphere, and a radius of curvature of 1550mm.


5: Figuring

Tool and mirror during polishing Tool reflects like glazed

The tool and mirror covered in Cerium Oxide, ready for a polishing session. On the right the tol surface is shown to be very glazy. This condition is bad for the even polishing action, and must be treated with brushing or scoring the pitch surface.

Not yet spherical As can be seen on this Foucault image, the surface is quite rough and not entirely spherical; it has a central hole. In fact this is probably caused by the mirror not being completely polished out yet. This condition is addressed by wide TOT strokes, about 1/2 by 1/2 D.
Getting better now Getting better now The surface is getting better, smoother and more spherical, after a few rounds of heavy TOT treatment. The Foucault and Ronchi images show only slight deviation from spherical. Note that the contrast has been increased with respect to the former image, by narrowing the light slit. However this results also in less light and more noise.
Now some wide and long MOT will attck the remaining center roughness.
Good enough Good enough The sphere is there, after only an hour of work: parabolization may commence. The hooks on outer bands of the Ronchi image are a bit misleading: these are caused by overlaying diffraction image due to a fairly narrow slit light source.
At first a trial MOT parabolization round is done to see how the mirror reacts. It appears to react quite slowly, which is good because the required correction for such a small F7 mirror is only minimal.
First parabolization attempt The first attempt to parabolize almost hits the target, except for a high 70% zone. After getting the feel for the tool in the former trial round, this was done with three rounds of dual sideswings, MOT parabolization.
Second attempt: done The correction of the high zone was done with TOT, accented pressure with tool edge over the 60-90% zones. Single round, singe swing, and test more accurately again. The result is good enough for me: off to the aluminisation chamber!