Like most websites, SGL uses cookies in order to deliver a secure, personalised service, to provide social media functions and to analyse our traffic. Continued use of SGL indicates your acceptance of our cookie policy.

Making a 12 inch Mirror

Recommended Posts

The final part of making a mirror for a Newtonian telescope is the figuring process; this involves changing an approximately spherical figure on the mirror to a parabolic figure. This is the most absorbing part of making a mirror. Most textbooks on the subject are somewhat dated and often describe methods more applicable to smaller mirrors with longer focal lengths; something like a 12 inch F5 mirror needs a different approach. The method I favour involves working the mirror face up with a sub diameter lap. A 50% diameter lap is ideal and should be prepared in the same way as described earlier for the full sized lap. A suitable base to pour the lap onto can be a plywood disk at least 20mm in thickness and coated with a good varnish prior to making the lap. Other suitable materials for the lap base include mild steel, aluminium and glass.

The aim of figuring is to deepen the centre relative to the edge in order to form a parabolic curve, if the curve is too shallow, somewhere between the spherical and parabolic curve it is an ellipse. If the curve is deeper than a parabolic curve it is hyperbolic, a condition we would hope to avoid.

The stroke to be employed in order to parabolise the mirror is not dissimilar to the stroke used with the 6 inch diameter barbell weight during hogging out of the curve. The 6 inch diameter lap is moved backwards and forwards across the mirror with a slight overhang at the end of each stroke of about ½ an inch. The mirror is rotated in the hands as it is pushed back and forth and the worker walks round the work bench (or as far as possible and back again). It is also important to rotate the mirror relative to the worktop that it is sitting on by about 20 to 30 degrees every 3-4 minutes; this should be a random process. The surface that the mirror is sitting on will not be flat and the mirror will bend to conform to its shape which will result in astigmatism in the glass. This random rotation will help alleviate the problem. Sitting the mirror on carpet will make the process of periodically rotating the mirror easier.

A very important factor is how far the centre of that lap is away from the centre of the mirror during each stroke. Pushing the lap centre through centre will, as you would expect, deepen the middle of the mirror quickly. We must take care to ensure that the whole of the surface is uniformly moved towards the parabolic shape. This means that we need to polish the whole surface with increasing effectiveness towards the centre. If we offset the centre of the lap from the centre of the mirror whilst working the polishing effect in the centre will be less aggressive in comparison to the edge. There will be a ‘sweet spot’ at which the offset is just right to reduce the spherical surface to a parabolic one. Patience is important here, we must avoid the temptation to deepen the curve too quickly at the expense of the edge and end up with a hyperbolic curve which is difficult to correct. I would suggest beginning with an offset of the lap of about 1.5 to 2 inches, work for 15 minutes and test. If the stroke is effective you should see ronchi bands gently bowing outwards inside of the ROC and the reverse outside of the ROC. The bands should curve smoothly along there length. When viewed from inside of the ROC if the bands are bowed more strongly in the middle than the edge then a hole is being dug in the middle of the mirror and the lap offset needs increasing.

I have attached some pictures to illustrate the stroke by showing the position of the lap at the start, middle and end of the stroke. Notice that the mirror is sitting on a circle cut from a yoga mat, its purpose is to aid in the periodic rotation of the mirror relative to the base it is sitting on.

John

Share on other sites

Just a correction to my last post, it is the tool that is rotated in the hands, not the mirror which is sitting face up during this process. Many thanks to Tea Dwarf for bringing this to my attention!

John

Share on other sites

hi gp any chance of some video of you actually showing some strokes , would be good to put a face to your name , thanks again for a brilliant post,regards tony

Share on other sites

Hi Tony, You are right videos can be very helpful to demonstrate the strokes and I had considered this, but I just have not had the time to produce them. I did add some links on a previous post to some youtube clips showing basic strokes, you could try searching youtube and see what you come up with.

John

Share on other sites

A spherical mirror suffers from an inability to form a sharp image due to spherical aberration; this is particularly noticeable in mirrors with shorter focal lengths. The shorter the focal length the more severe the spherical aberration. A six inch mirror working at F9 can be left spherical and still have diffraction limited performance. Unfortunately, this is not the case with a 12 inch F5 mirror; the purpose of parabolising the mirror is to reduce the spherical aberration to acceptable limits. As we deepen the mirror at the centre relative to the edge we are adding ‘correction’ in order to reduce the spherical aberration. The matching Ronchi test offers a useable method of tracking how much correction has been added. After about 30 minutes work with the sub-diameter lap the mirror should be accurately assessed with the matching ronchi test (MRT). If the curvature of the bands is less than predicted by the software, the mirror is under corrected and more work needs to be done. If the bands are more strongly curved than predicted by the software, the mirror is over corrected and the centre is too deep. This is a condition that MUST be avoided and reinforces the need for regular testing.

The figuring process is continued until the observed ronchi bands at each position match those predicted by Mel Bartels ronchi software.

I have attached some pictures showing the predicted and actual ronchi image at various offsets. Photographing the images proved quite difficult as the predicted images are shown with one ronchi band coincident with the middle of the mirror. In reality it is not easy to set up the grating at the correct offset with one line coincident with the middle of the mirror in order to photograph it. This is not an issue when making a visual assessment of the actual and predicted patterns. The ronchi grating can be nudged laterally to move the bands on the mirrors surface to help assess the match.

If the mirror is judged to have the correct figure as predicted by the MRT there is one further test that can be undertaken if the optical tube assembly and mount are available. The test is a variation on the star test which is generally recognised as the ultimate test for any telescope. The test can be performed on an uncoated mirror, the secondary should be coated. The telescope is pointed at a bright star; Polaris is convenient as a drive system is not required. The star is focussed using an eyepiece. The eyepiece is removed, looking into the draw tube will reveal the mirror fully illuminated with starlight. Introduce the ronchi ruling and move it and/or the head to reveal a series of bands silhouetted on the mirrors surface. Moving the grating back and forth will result in the bands increasing or decreasing in number, just as they did when testing at the radius of curvature. The mirror is now being tested at the FOCUS and NOT the ROC. The light entering the telescope is sensibly parallel, under these conditions a parabolic mirror will test as though it was a spherical mirror, in other words the ronchi bands will be straight. Ideally, with two or three bands visible on the mirror, no curvature should be seen if the mirror is to give diffraction limited performance. If the bands are seen to bow inwards the mirror is under corrected and needs more parabolising strokes. If the bands bow outwards the mirror is over corrected, avoid this at all costs.

John

Share on other sites

A couple more images from the MRT.

Share on other sites

If you have been following this thread and wondering if you are up to the task of making your own mirror and need a little inspiration, you need go no further than Bernhard Schmidt. The Schmidt Cassegrain Telescope (SCT) so popular today is a tribute to his genius. What is surprising is that this master crafsman lost his right arm as a boy yet was still able to make many mirrors and lenses. For more see...

John

Share on other sites

though that also shows why you shouldn't try to make your own fireworks at home...

Share on other sites

this has been a fascinating process. am I living in cloud cuckoo land thinking I could buy a 16" blank as a first project? I am not scared of hard work, reasonably skilled at DIY but have never made a mirror. my issue with buying anything smaller is that I already have a 6" and 12" dob and 'aperture fever is (but finances are not particularly) strong in this one' (well, not for a £1600-2000 completed mirror set anyhow).

Share on other sites

Moonshane, making a 16 inch mirror as a first project is doable. I guess that you would be using 25mm thick plate glass to keep costs down. I would advise that the mirror is worked face up throughout the grinding and polishing, this is to try and control flexure, and hence astigmatism in the glass. Working face up the mirror can be properly supported from underneath and rotated on a regular basis relative to the surface it is sitting on. After hogging out with an 8 inch barbell weight, I would grind with a 14 inch tile and cement tool with an off centre stroke to compensate for the smaller tool.

It is unlikely that you will produce a mirror suitable for high power planetary work, but it is not too difficult to produce a mirror good enough to give spectacular views of deep sky objects at medium powers from a dark site.

If you do decide to go ahead with the project do some research on the internet first I am sure it will help. If I can help at any point I would be happy to do so.

John

Share on other sites

In a previous post I discussed the use of a ronchi grating in a variation of the star test, there exists a bench test version of this known as the autocollimation test. This test, also known as the double pass null test, involves the paraboloidal mirror being tested in conjunction with an optical flat. Here is a video demonstrating the test:

I performed this test on the 12 inch F5.3 mirror that I had figured using the MRT (matching ronchi test) and described in a previous post. I have attached two images showing the mirror under the autocollimation test. Straight fringes indicate a good mirror, this mirror shows at worst 1/8 wave surface error which corresponds to ¼ wave wavefront error.

John

PM,D GP

Responded to PM

John

Share on other sites

video of my mirror grind and set up, let me know if i'm doing it wrong http://www.youtube.com/watch?v=9umyTZGAA-Q

Edited by ponytale

Share on other sites

Tony, nice video your technique looks OK but keep walking round the 'barrel'.

John

Share on other sites

thanks for your comments john , i will post more videos as i go,

Share on other sites

this is great!

I reckon I am gonna take the plunge, be brave and go for a 16" plate glass version just to 'have a go'. I reckon that if I make a good enough cell, the 25mm thickness won't matter too much, especially if I make it about f5.5. I am six foot three and reckon that this will make the EP about my eye level on this basis.

Share on other sites

Correct me if I'm wrong, but has Ponytale not got the grinding tool on top in his video, or is he making the curve in the mirror shallower?

Ron.

Share on other sites

Ron, Yes PT has got the tool on top. I have been suggesting that people work 50% of the time mirror on top and 50% of the time tool on top, after the correct curve depth has been reached. This ensures effecient grinding and maintains a constant radius of curvature. Hope that explains what is happening.

John

Share on other sites

Probably one of the best threads on this forum. The barbecue pit is a great idea!

Great video Ponytail...I have got to get going with my 14" blank!

It's be great to have it set up permanently so that I can have a go at it when I have a spare half hour in the day...

Edited by Beulah

Share on other sites
Ron, Yes PT has got the tool on top. I have been suggesting that people work 50% of the time mirror on top and 50% of the time tool on top, after the correct curve depth has been reached. This ensures effecient grinding and maintains a constant radius of curvature. Hope that explains what is happening.

John

Cheers John. I didn't know what progress PT had made, so apologies. I should have made sure before interrupting.

I was aafraid he may have finished up with a convex mirror, and wondered why.

It's a great thread, and it's continuity is an important aspect of it.

I should have sent a PM rather than posted.

Ron.

Share on other sites

Fascinating stuff... ive loved every post..

Share on other sites
video of my mirror grind and set up, let me know if i'm doing it wrong

Ponytale - I think you should work a little more methodically. ie. a specific number of strokes (not too many though - say between 5 & 10 perhaps) , walk round a bit, same number of strokes, walk round a bit etc

(I refigured my 13 inch and used a turntable on the floor.) It's also important to start & finish at the same point around the mirror. Good luck - it's great to make your own mirrors,

Alan

Share on other sites

In case anyone was wondering, I'm still following the thread. I just haven't made way to start doing any work on my blank yet.

Hopefully I'll get round to it before the arrival of the people from planet X.........

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

×   Pasted as rich text.   Paste as plain text instead

Only 75 emoji are allowed.

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×