The 22" mapstar mirror

[barkis]

It's like staring into the faces of weird Aliens.

I never made a mirror large enough to make use of the Ronchi gratings, although I did 

use a home made one to get a general picture a finished 12" f6 parabola. 

The Foucault/ Couder, and Everest test methods were mostly my bag. 

A general overall view of the shape and smoothness, and edge diffraction were always  more apparent to me.

The more modern methods seem to have relegated the old methods to the second division now though.

Whichever one uses, matter little, as long as the goal is attained.

Ron.

[mapstar]

It's like staring into the faces of weird Aliens.

I never made a mirror large enough to make use of the Ronchi gratings, although I did

use a home made one to get a general picture a finished 12" f6 parabola.

The Foucault/ Couder, and Everest test methods were mostly my bag.

A general overall view of the shape and smoothness, and edge diffraction were always more apparent to me.

The more modern methods seem to have relegated the old methods to the second division now though.

Whichever one uses, matter little, as long as the goal is attained.

Ron.

Hi Ron

I use the Ronchi as I really struggle with the Foucault to take a photographic image.

Seems everything I do with it just doesn't seem to work?

I have managed another hour today which is a bit disappointing as I wanted to do more.

Back to work tomorrow but I will get some time in probably an hour or so?

Damian

[barkis]

HehHeh I understand Damien, Foucault shadows can be deceptive until you become completely adjusted to what they are indicating.

Sometimes an apparent hill, looks like a crater, and vice versa.

The shadows that define a very slow Newt. mirror, a six inch f 8 for example, can be very faint.

some may be tempted to leave such a mirror spherical to avoid the difficulty of figuring it, however,

that figure when created accurately, makes a huge difference to performance, everything else being equal of course.

Anyway, I wouldn't dream of dictating how you should do this job, my knowledge of Ronchi testing is far from good enough to guide you away from it.

I'm sure that like me, some of the other mirror making fanatics contributing to your thread, would like nothing better than to descend on your home with a need to help out. Whether or not we would be welcome is another matter of course HehHeh!

Too far away though, so you're probably pleased about that. Besides, with John on your side, you are in good hands.

Good Luck on the run in anyway matey, I think you have the will and determination to lick this big fella.

Ron.

[barkis]

Taking pictures of the various shadow patterns isn't usually difficult Damian.

Positioning your camera as close to the knife edge as you can when you are intercepting the focus of a particular zone on the mirror.

You saw Raymonds lovely parabaloid Image. That would be at 50% of the total longitudinal knife edge movement. which 

images the 71% zone of the mirror, if my memory serves me right.

I almost went to find the book, but If I'm wrong, someone will correct it for me.

Ron.

[Ajohn]

Trying to help again. The problems with Foucault testing are likely to be down to the tester used. There are lots of nvg info about however there is a sensible one shown on this page especially for some one with an x-y stage but that can be made up in all sorts of ways.

http://www.stathis-firstlight.de/atm/foucault_tester.htm?sa=X&ved=0CDwQ9QEwEzgoahUKEwiw0eCp_frGAhUOKtsKHa0IBFg

This one

The telescope is optional. Note the comment on the web page about sanding the led or using a diffuser. I would add a flat ended led, If it's round ended just sand it flat. Some sort of plastic diffuser as mentioned isn't a bad idea anyway.

What he doesn't mention is alignment. He's done many mirrors so will have some sort of bench set up so that the mirror can be placed on a stand and align well with the mirror. What I would do is fix the knife edge more securely and have a piece of cardboard with 2 holes in it. One to let the light through and the other smaller one to indicate where the knife edge is. If the centre of the led was say 10mm down from the top of the mounting this small hole should be 10mm above it. Best make that distance as small as possible but the return image or the led must clear the woodwork so the distance depends on the size of the led in some respects. If the cardboard is place in front of the tester it will catch the return image of the led from the mirror. At the ROC the image of the led and knife edge will be sharp and the set up can be manoeuvred around to place the image on the knife edge hole. Fiddle with the knobs and it should be fairly easy to null a near sphere now but the adjustments will be very sensitive. That can be helped by mounting say 50mm dia disks some how on the end of micrometer spindles.

When a mask is used for measurements the shadows have to move equi spaced around the centre of the mirror as the knife is moved back and forth. For initial rough measurement the shadows can be evened up side to side with the other axis but that needs to remain fixed for final figuring. Old time testing using a fixed source will already have taken care of that and people without an x-y stand can use this sort of tester in the same way to set it up - alter it's angle until the shadows move evenly. This can also be done roughly with the cardboard in front of the tester as the return image will move from side to side.

To use Ronchi just remove the knife edge and clip the screen on but set up in the same way.

Setting up is a pain. A simple solution for most people if they haven't got a bench that is long enough and things can't be left in place, lack packing to raise the tester up etc might be a shelf on a wall. Maybe 2 one for the mirror and another for the tester so that mirror sizes and tester height can be at least part adjusted. I have seen a professional who made lots of mirrors daily work on a shelf. It seems to work well but he was mostly making 10in F6.2 mirrors so his tester was built for that size. Me I had a bench and a bookcase which made up the lengths The bookcase was taller than the bench so height adjustment wasn't a problem, just adjusted the height of the mirror. Once set up the mirror could then be just dropped in place, the tester checked for axial alignment and measurements taken.  I haven't got that any more so am looking for a wall for shelves that wont upset the wife.

One warning about bright sources. After a long session I found the pupil in my right eye was a tiny pin [removed word] and the other a lot larger. It took a while to settle down.

The telescope on the tester in the link - just used to make the mirror look bigger. A converse view is to use one backwards to improve the view of the shadows.  Pass never tried either.

Whoops - People need to be able to get their eye behind the tester. Might not be easy as shown. Remove one side / turn it around etc but turning it around would need some slots in the cardboard to clear the frame or what ever is used for that.  The knife will probably need moving forwards by 1/2 the thickness of the cardboard if the return image has been sharply focused.

John

-

Edited July 27, 2015 by Ajohn

[RAC]

The knife edge was moved off center in my image so I could see the near entire mirror just for that pic. I wanted a full image I could process to check for roughness.

[mapstar]

The knife edge was moved off center in my image so I could see the near entire mirror just for that pic. I wanted a full image I could process to check for roughness.

I was doing something similar last time I took some foucault pics.

I would be interested to see your testing set up Raymond if you could post a pic

And the way you managed to take your pics

Thanks

Edited July 27, 2015 by mapstar

[mapstar]

I have done a further half an hour today,

I've mainly been working on the centre out to about 60% and trying to blend the hill into the outer zone for a smoother curve.

Here are the images

Inside ROC

Outside ROC 

I will agree with AJohn that the foucault test is very sensitive to alignment along the optical axis. This I'm going to have to work on probably by adding some jacking screws on the test stand to tilt the mirror so that the plane of the mirror is vertical

Hopefully more work tomorrow 

Damian

Edited July 27, 2015 by mapstar

[barkis]

I do believe you are cooking with gas now Damian, that flippng glass is starting to listen to you at last  . 

 Keep your thinking cap on now.

Ron.

[pdols]

Really exciting to see the changes. Thanks for all the images of the testing as well - from a learner's point of view I've found it really helpful and interesting to see the test results along with the narrative. Keep going - home stretch

Sent from my iPhone using Tapatalk

[mapstar]

I do believe you are cooking with gas now Damian, that flippng glass is starting to listen to you at last .

Keep your thinking cap on now.

Ron.

I hope so Ron, slowly slowly does it. I have generated the images I need and hopefully I can match them to those.

Need to suss out the pesky Foucault test though

Damian

[mapstar]

Really exciting to see the changes. Thanks for all the images of the testing as well - from a learner's point of view I've found it really helpful and interesting to see the test results along with the narrative. Keep going - home stretch

Sent from my iPhone using Tapatalk

Thanks. I'll try my best to keep up with adding updates

[Ajohn]

The knife edge was moved off center in my image so I could see the near entire mirror just for that pic. I wanted a full image I could process to check for roughness.

That's odd really. The camera should work just the same way as the telescope does in the photo of the test rig I pinched of the web. It doesn't care  that the frame of the tester and the knife edge is blocking the view. The knife edge has to for the test to work. I'd guess that getting the camera on axis and square to it would be a bit tricky though and some zoom lenses can do odd things used on systems like this.

There was a mention of the Ross test. This can tempt people but on faster and larger mirrors the measurement accuracy needed form tester to lens and lens to mirror to make it worth while is very extreme.

John

-

Edited July 27, 2015 by Ajohn

[RAC]

It wasn't so the camera could see the entire mirror it was to get a nice shading across the mirror. I'll take some pics tonight.

[RAC]

This is how simple it can be. A focuser on an old engine flex plate and a dial gauge touching the focuser tube. The flex plate sits on three points. Two are bolts machined to points and the third(the brass one) is a thread with a hole down the middle so it can rest on a ball bearing, this keeps it very smooth as this adjustment effectively moves the knife edge across the mirror. I have the return beam come back about 10mm below the LED. When using a camera all you do is place it on a tripod behind with the center of the lens looking down the beam. The front lens of the camera is about 10mm behind knife edge. The camera does not need to move with the knife edge, it stays on the tripod.

You need to have the focuser very square to the mirror but this is easy, you know when it's not as when you move it out or in the mirror won't null out evenly with the couder mask.

The radius of Curvature is very easy to find with the foucault test as all you do is null out the center of the mirror and measure from the knife edge to the center of the mirror. Enter that into Figure XP and away you go.

[Soupy]

Damian it's looking great so far! Bet you feel you are now getting somewhere with it!

Raymond, you make it sound reasonably simple, but I'm still bewildered you can get it so accurate!

[mapstar]

This is how simple it can be. A focuser on an old engine flex plate and a dial gauge touching the focuser tube. The flex plate sits on three points. Two are bolts machined to points and the third(the brass one) is a thread with a hole down the middle so it can rest on a ball bearing, this keeps it very smooth as this adjustment effectively moves the knife edge across the mirror. I have the return beam come back about 10mm below the LED. When using a camera all you do is place it on a tripod behind with the center of the lens looking down the beam. The front lens of the camera is about 10mm behind knife edge. The camera does not need to move with the knife edge, it stays on the tripod.

You need to have the focuser very square to the mirror but this is easy, you know when it's not as when you move it out or in the mirror won't null out evenly with the couder mask.

IMG_1990.JPG

IMG_1992.JPG

The radius of Curvature is very easy to find with the foucault test as all you do is null out the center of the mirror and measure from the knife edge to the center of the mirror. Enter that into Figure XP and away you go.

Thanks for posting that Raymond. I will have a good look through it.

[Stub Mandrel]

Is there a difference in that most cameras have a much bigger aperture than the eye's pupil?

[RAC]

The best canon lens I have used for imaging with foucault is an old cheap canon 75-300 zoom. I have a canon 15-85 and also a 200mm f2.8L but those are no good for some reason.

[barkis]

Looks a decent setup Raymond, I too used a DTI gauge, although there are simpler methods, I preferred the Gauge.

I used to use a very tiny, and hopefully star like pin hole, but I surrendered to a slit source for better illumination.

I found that fresnel fringes could be a nuisance with too small a light source.

Ron.

[Mark62]

What a change!! It's looking really good Damian. The blending of the hill seems to working really well. Keep going like you are and it will be done for Galloway

Mark

[Ajohn]

Thanks for mentioning off setting the knife.   I've been webbed. That tester crops up all over the place and is what I intend to make but it can't work on axis, the mirror will return the image of the source right back above it and the knife wont cut it at all so it needs offsetting to cause the return beam to hit the knife. Leaves me wondering if it would be better to leave the led unobstructed and just have the knife central too it and above it. Minimising the offset does improve the accuracy.

This is what I have used. It was in a loft room when we had the roof done. Covered in dust and got damp. It a fixed source version but the idea for the sliding table can be used with any type.

It's easy to make. MDF etc would do but the sliding table needs some weight added even in aluminium. The tilt screws just run on a bit of metal plate under them. I added my own thoughts. As I used a slit I tilted the source rather than having a separate piece that can be tilted mounted on the knife. I also added a zero facility for the mic spindle - set mic to zero, slide table about to pick up the centre and then slide the bar up against the mic spindle and lock it in position. I didn't add a spring to pull back the table onto the mic, just used finger pressure on the table. Most people don't mount the bar the table runs on in blocks and just screw it down onto a board of some sort. In that case the mic spindle can be arranged to press on something on top of the table. It is best to make the V's that rest on the bar out of brass. It slides very smoothly. The slide going across has the V's made of aluminium and seemed ok.

The other gubins that moves the knife across is for a version of the caustic test. For that a wire is mounted on the knife and  I just held a 10x eye cup type loupe in my eye focused on the wire by moving my head. This test does need a slit. I've lost that but it was made as per Texereau. His way does work. No shadow reading at all with this test as it uses diffraction. I used it as a final check. The all thread - for playing with a Dall null test.  I might play with that again and get a better lens for doing it. Abuse of the Ross tests looks easier though. Just move the lens and tester around until the mirror nulls flat and look for ripple and small zonal errors etc.

John

-

Edited July 28, 2015 by Ajohn

[Stub Mandrel]

It might be worth using a cheap digital calliper instead of a micrometer head to measure movement.

[barkis]

There are many variations of the Foucault tester, but you  rarely find one as basic as the one pictured here

from N E Howards Telescope Making Book.

Of course it is crude, but the Barr Scale was used to measure the longitudinal 

readings quite accurately. How to make it also appears in the Book.

I think a Micrometer, or DTI  Gauge are preferable though  .

The gap between the light source and knife edge is a bit too wide,

but of course the Image is purely to show the simplicity of the device.

Ron.                                                                                                                                

       

[mapstar]

There is a multitude of Foucault testers all over the net and as Ron says they don't need to be complicated. 

Thanks for the suggestions guys I'm going to work on what I have at present for now but will build something to make further tests.I have not made any progress today as I have been concentrating on getting the other bits that I need together.

I'm really hoping to get back on with it tomorrow though

Damian