Foucault / Ronchi tester

[woodbob123]

I am looking for advice on the "best" type to allow me to photograph the shadows, plenty of plans available with pin hole, slit and slitless.

Have a 24" x 1" thick blank ground and polished to f5 now comes the part that I have no experiance in figuring. Lucky enough to be Spain for 3 months so will do the job slowely (is there anyother way) asking advice on this wonderfull forum each step of the way, but will need to photograph my attemps so I can post and learn from your experiance. The stage has been built and I have leds of different colours and outputs also compact, dslr and video cameras. Any and all ideas welcome as if this works out will then build a bino scope of 8" or 10" apature I usually do things the wrong way round but had a bad case of apature fever

[Ajohn]

Why not make something flexible so that you can try all of them. Texereau describes how to make a slit using bits of brass which is very easy to work and do. This link includes diagrams, also the same set up using razor blades or what ever for the slit

drona.csa.iisc.ernet.in/~priti/HowToBuildTelescope.pdf

It's been modified a little from the original - it uses metal angle to support the knife rather than wood. I'd use mdf for all of the wooden parts but it's no problem for me to use aluminium. Blades can be attached any old how really as they are in many designs.

The main point is that it's a slit plate that is then attached to something. If one of the blades of the slit is extended and the other left off and a led is fitted as usual it's a slitless tester or the gap can be set with the other blade. It will still be slitless as far as aligning the blade with the slit is concerned. As it can be removed it's easy to set the slit parallel by viewing a bright light through it. If this is turned upside down, knife edge at the bottom a ronchi screen can be clipped to it. Or make a separate part for the ronchi screen without a knife edge. I would try various sized holes over the light source. Or even just clip the ronchi to the knife.

The Texereau diagram gives a distance between the knife edge and source. Having any distance introduces errors. For an 8" F6 mirror he will have chosen one that doesn't introduce significant errors. When led's are used the distance can be a lot shorter. It would also be easy to convert his arrangement to both knife and source moving with these  or come up with something that could do both arrangements. He describes a test for a sphere that needs a stationary source. No need for a ronchi screen really with that. Trying to null out a sphere needs very fine adjustments with the slit size he suggests.

Some have used an x-y stage for moving the tester about. They might use both axis initially to balance out adjacent shadows and later align the tester so well that it only needs moving back and forth. Texereau's rocking idea should do something similar but for final testing it needs to be aligned. That's easy with a slit. The image of the slit can be seen via a cheap eye cup type loupe so both the knife edge and it are bought into focus and the stage slid back and forth and things adjusted until there is no relative movement between the two. Should work out on slitless too but to get shadows the knife edge will need to block part of the image, as it has to with a slit. The easiest way to initially align is to catch the return image on a card and move stuff around until it falls on the knife edge. A mechanism to adjust the tilt of the mirror is a good idea. That way if the source is initially dimensionally central to the mirror axis it can easily be adjusted to be square on as well. I've packed the tester up with books. Some people support the mirror on a strap that runs half way round the edge making it easy to adjust it's height. Some stick the tester on a camera tripod. It could shake around when touched.

The web seems to have gone a bit astray when slitless testers are mentioned. They talk about the difficulty of aligning the knife with the slit. It's easy with Texereau's arrangement. The knife and slit image are viewed with the eyecup loupe and the knife angle adjusted to suit when it's half way through the image of the slit. The whole initial idea of using led's  was to obtain a larger source to avoid very dark shadows and diffraction effects. Texereau's tester in that guise might just have a 2.5mm hole rather than a slit. Pass I've no idea but faster mirrors produce darker shadows so some hole size should be more ideal than others. Too big and sensitivity will be lost. On the face of it there is no need for a slitless tester's knife to pass half way over the hole the led is in. It could just be above and central to it. In some ways that makes more sense. However it's done the knife edge needs to cut into the return image otherwise no shadows - so if it's half way across things need to be tilted.

LOL I'm 'issing into the wind bothering typing this out but one fact for sure people will spend some often a long time using their tester. It deserves a lot of attention especially to the stage the business end runs on. Once that is made various things can be stuck on it all pretty quick to make. The stage itself is likely to take longer. Texereau's is easy to make and works well, ok it need some bits of bar and brass plus something to weigh it down but it does work well. The spring isn't needed as pressure can be applied with a finger. A dial gauge or any of the other ideas for measurement can be used instead of a screw. If the rocking aspect is built in do use a very fine screw / put a large knob on it in case at some point you want to use a fine slit. It's also possible to mount another on top to get an x-y stage. I've done that to try the wire / caustic test. There is no ambiguity at all with that test. Ideally the x and y need to be perfectly square, the usual reason for people not using it. I decided to just set it as square as I could. The results tied in with other methods and very accurate readings can be taken easily.

Cameras - well every source says close to the knife where the eye would usually be. Not something I have done but I'd bet alignment issues will prevent that from working out. The other problem is the focal length of the lens - ideally it needs to fill the image on the sensor in the camera or at least give a decent sized image on it. It should be simple trig to work out what that needs to be based on the focal length of the mirror. I have used a red led in a test set up. A Dall null test. As they are pretty narrow band the chromatic effects were dramatic. I don't think this matter for knife/ronchi testing but it might make any diffraction effects worse if fine slits are used - pass.

Led's have optics in them. I'd file the end off going as close to the chip as I could and put a bit of slightly milky coloured kitchen food / sandwich bag over it as well. For all I know the optics in the led, reflectors and lens shape might mess up the test. Also initially play with slit size and also probably try a few sizes of hole. I wonder if figuring with Ronchi is a good idea especially for a first mirror. Using a faucault tester takes some practice. Maybe it's best to get stuck in first when it doesn't matter too much so that people will be ready for when it does.

Arggggggggg no preview - that's when I usually spot my weird typo's. Some are really weird at times.

John

-

 

[RAC]

Here's a shot of my setup. It's as simple as having half the led cut off with the knife edge and looking at the return beam below the led. There isn't a lot in it between pin holes, slits and slit less IMO.