This blog is mainly about Telescope making, and some things about my politics. At last we finally have a President that can say "Fool me once; shame on you. Fool me twice; shame on me." instead of mixing up with an old Who song.

Monday, April 28, 2008

Mirror Testing

As part of my new Telescope venture, I had to make a device to test it. The amazing thing is that this device, known a Foucault knife edge tester can measure the figure of a mirror to well within 1/10th of a wavelength of an average green colored light. One wavelength of green light is around 500 nanometers - that's 500 billionths of an inch or 0.0000005 inches, and this device made of wood and metal can measure 1/10 of that or 0.00000005 inches. Pretty nifty, actually. And it was invented by a guy named Leon Foucault back in the 1830's, and it hasn't changed a lot since then.

Here is the idea... a concave mirror is gound to the shape of a part of a sphere. If you have a point of light at the center of that sphere aimed right at the mirror, then it should bounce right back to the center of the sphere. (even if there is no aluminum on it yet, its still shiny enough) If you move that point to the right a little bit then it should bounce back a little to the left. and if you put your eye there, the mirror will appear to be all lit up. That distance is known as the radius of curvature.

If you cut the returning cone with the edge of a knife, then it will appear to evenly darken from light to grey to black. In practice, its green, because we use a green LED for the light source.

Well this would be great, if all we wanted was to see the dot of light, but we really want to see the stars, which are infinitely far away by comparison. So when the light source is way out in space, the focus falls somewhere close to half the radius of curvature. Its commonly called the focal length of the mirror.

Well, its close, but it doesn't focus very well. Only a parabola will focus all the rays from infinity to one point. For very long focal lengths, it doesn't matter much, but for normal ones 8 times the mirror diameter or less, it matters a whole lot. And until old Hugo came along, telescopes were very very long. William Hershel, for example had telescopes with focal lengths 20 times the diameter and it was quite limiting for him. At the focal lengths I am using here, the difference between the shape of a sphere and the shape of a parabola is very slight. if you dig the center of the mirror a little deeper than a sphere, and you flatten out the edge a little, what you are left over with is a paraboloidal shape. Testing it at the average Radius of Curvature (ROC) with this tester, it becomes really appearent that the outside of the mirror is flattened out a little and that the center is dug a little deeper. You can actually see that the focus at the edge is a little longer than it is in the center and other places in between. If you measure this within a few ten-thousandths of a inch, you can get a really accurate picture of what the shape of the mirror is. And measurements like that are quite possible to do with tools you can make yourself.

With a design that I found in several places, most notably the Stellafane website, I constructed a design of my own. But you not only need the tester, you also need a way to safely hold the mirror while you are testing it. So there is a matching test stand for that.

One thing I want to avoid is screwing my back up again, so the last thing I wanted was to have to bend over with my head right near the top of my workbench staring at a mirror reflection. So I changed the design someone. I raised it on top of a box. And since its a hollow box, I put a lid on it with a hinge, so that the moving parts can be removed and placed inside for storage.

The photo to the left shows the box with the testing stage atop it. This is looking at it from where the mirror is. You can see the battery pack that lights the LED to its right, and further down and to the right is the switch to turn it on and off. right above the LED is the knife edge (a razor blade), which is mounted in a photographic slide I found at a vintage camera shop in Decatur. To the right is the knob that pivots the whole stage back and forth to cut the bundled cone of light rays bouncing back from the mirror. One thing here is that the hinge is not attached yet in this photo. Again, I took my time with these things, made sure they were sanded well and coated with at least three coats of polyurethane.

Here is a side view of it. In use, the mirror is to the left, and my head would be to the right. By turning the knob in, it raises the near side of the platform, and pivots it around the aluminum bar in the background. To the right, along the vertical back face, you can see a slot about two inches high, and wide enough to slip the slid with the knife edge through it into the rails behind the slot. The slide can be interchanged with one that contains a Ronchi grating (parallel lines) that are used for a slightly different kind of test.

Here is a close up of the pivot knob. Its really just a 1/4-20 carriage bolt. one turn is 1/20 of an inch, and at about 6 inches from the pivot, that translates to about 1/40 of an inch per turn at the knife edge, so the amount of control is tremendous. The whole thing goes left and right on that aluminum bar, and so the head of the bolt rides on top of a little teflon furnature slider I found at the hardware store.

Here is a close up of the knife edge in the plastic photographic slide resting in two little rails. Next to it is the LED socket, and below it is where the wires for the LED get routed through to the back. Also note the battery pack. You can't see it, but its stuck there with some velcro. I found a half a pack at Lowes that someone had pilfered, so the gal at the checkout line gave it to me for 90% off, since it was half missing. You can see my finger pointing out the slot for inserting the various slides for the knife edge and/or the Ronchi grating. You can make a Ronchi Grating with a good printer. It has to be at least 1200 dpi to work, and it has its wierdness. A better one can be made with a photographic reduction method. I'm thinking of going down to that vintage camera shop to see if they can do that for me. They do all kinds of actual black and white film in their actual photo lab. You don't see that much anymore, eh?

Here is the back side of the tester. I'm pointing to the slot again for the knife edge slide. To the right is the window with the knife edge and the LED mounted in its socket on a little stick of leftover pine scrap. Below my finger is the little switch to turn the LED on and off. I should probably also have a little dimmer there too, but I forgot. To the right is the homemade micrometer. Its divided up into 500 parts, and each turn is 1/20 of an inch, so each mark is .0001 inch. In practice, its a little shakey. But I bought a dial indicator at the disposable tool store to replace it. Haven't done that part yet.

This next picture shows the micrometer better. The head of the carriage bolt presses against the stage riding on the aluminum bar. The stage is held against it with a small spring visible underneath at an angle.

Again, from the other side. You can see here that it is almost all the way screwed in. On the left side of that plywood bracket is a "T" nut. The wheel of the micrometer is hardboard.

Here is the back side of the testing stage showing the LED in its socket and the resistor for it mounted on a little piece of scrap pine.

The stand is adjustable. I can test any standard mirror diameter from 12.5 down to 4.25 with it. It sits on a shelf atop a low base and its tipped back at 5 degrees so that the big mirror won't accidentally fall forward. The top of the box has to be aimed parallel to the axis of the mirror. I do that by proping the lid at a 5 degree angle too. So when testing the mirror, the tester is looking down at the mirror by 5 degrees, and the mirror is looking up at the tester at the same angle.

I'll show pictures of that later on when I edit this again.

No comments: