Making a 12 inch Mirror

[Glasspusher]

Hi TeaDwarf,

thanks for the input. It is best not to use Cerium Oxide slurry with the greaseproof paper as the paper crinckles when wet and this causes small indentations in the lap, if localised these can result in contact problems between the mirror and pitch lap. Yes, good point about heating pitch slowly, heating too quickly can drive off volatiles and result in a harder lap.

Had a PM from Ron mentioning that when grinding with fine aluminum oxide sticking can occur, care needs to be taken at this stage to avoid this problem. Sticking is more likely to occur when grinding a mirror on a glass tool, another advantage of the tile tool is that the gaps between the tiles allow good circulation of abrasive which prevents sticking.

I am trying to be precise when making posts and may miss out some relevant points, so feel free to comment. I know there are a number of members who have experience in making mirrors and I would welcome their input at any point.

John

[MG1]

I hope this isn't a silly question...

How do you determine your optimum f-ratio for a given diameter of primary?

Obviously an issue with dobs is that the target will drift quite quickly if you're viewing at a high mag so I'm guessing that is one factor.

But would you get a better view through a longer scope with a wider ep or a shorter scope and tight ep?

Eg 12" f10 with 20mm EP gives approx x150 but so would an 12" f7 with a 14mm EP

Question is which is better and how does the 'speed' of the glass correlate to say coma or other intrinsic visual defects...?

Not withstanding practicalities of size, which is better?

Cheers,

Michael

[hunterknox]

Excellent thread, thanks John.

Can I ask about costs? How much would I be looking at to make a complete mirror, say an 8, 10 or 12", starting with no equipment?

[Glasspusher]

Michael,

This is the question that all amateur astronomers agonise over at some point and often as a result they end up with two telescopes, one a long focal length instrument for planetary work the other a short focal length for deep sky work. A compromise might be around F5 to F6. There are lots of ramifications to your question but I would like to comment from the point of view of a telescope maker. We are concerned with two aberrations here, spherical aberration and coma. Spherical aberration and coma increase with decreasing focal length but so does the field of view, and a wide field of view is often a requirement. The process of figuring or parabolising a mirror removes most of the spherical aberration but not the coma (which makes the stars near the edge of the field look like little fans or comets). There are commercially available ‘coma correctors’ which are quite effective in reducing the coma in fast Newtonian telescopes. Fast, short focus Newtonian telescopes have the very big advantage of being relatively compact compared to their long focus brothers. There are no ideal all purpose telescopes, just compromises, uncomfortable as it is it’s just a fact of life. Hope this goes some way to answering you question.

John

[Glasspusher]

Hunterknox,

thanks for the question, I am sure other have been considering this. Last time I seriously looked at this (a couple of months ago) I worked out that the cost of making a 14 inch mirror was less than 50% the cost of buying a new one. Much depends on where you buy materials and what you can beg and borrow. Generally speaking the bigger the mirror the more you are likely to save against buying a new one. Some 8 inch mirrors can be bought for little more than the cost of making one. Making a mirror is not solely about saving money, it is an engrossing, educational and rewarding process in it own right. Try it and see!!

John

[Beulah]

Hunterknox,

thanks for the question, I am sure other have been considering this. Last time I seriously looked at this (a couple of months ago) I worked out that the cost of making a 14 inch mirror was less than 50% the cost of buying a new one. Much depends on where you buy materials and what you can beg and borrow. Generally speaking the bigger the mirror the more you are likely to save against buying a new one. Some 8 inch mirrors can be bought for little more than the cost of making one. Making a mirror is not solely about saving money, it is an engrossing, educational and rewarding process in it own right. Try it and see!!

John

One can actually make one's focusser, so I was told recently.

Thank you John, for this outstanding thread. I really want to start on my 14" soon - and have bookmarked this thread for reference.

[barkis]

Making a mirror will test your patience and fortitude.

But think of the reward, an optical wonder fashioned by your own hands, and as perfect as your determination wants to make it.

All those who have nurtured the desire to do this, won't get a better opportunity. John has provided the perfect vehicle for it with this excellent tutorial. I believe the succes rate will be high.

Ron.

[Beulah]

Making a mirror will test your patience and fortitude.

But think of the reward, an optical wonder fashioned by your own hands, and as perfect as your determination wants to make it.

All those who have nurtured the desire to do this, won't get a better opportunity. John has provided the perfect vehicle for it with this excellent tutorial. I believe the succes rate will be high.

Ron.

That's a very trustworthy statement, coming from you.

[Glasspusher]

Before going on to start polishing we need to thing about how we are going to test the mirror as we progress. I am going to suggest that we use the Ronchi test as I think this is easier to implement for a novice mirror maker. It also represents a powerful way of assessing a finished telescope. The star test is often referred to as the ultimate test for a telescope, however it is a test that requires some experience if the best is to be had from it. It is possible to use the ronchi test as an alternative to the star test, it is easier to perform and easier to interpret the results. A mirror maker who has a completed telescope to slip the mirror into for testing can use the ronchi star test to help produce a very high quality mirror. With the ronchi star test a perfect mirror shows straight bands so the interpretation of the mirror’s figure is very easy. More about this later, let’s begin by considering the ronchi test in the bench test format.

The Ronchi test is named after is inventor, Vasco Ronchi. The test is performed at the Radius of Curvature (twice the focal length), and is qualitative, that is it does not give us any numbers to help us define the surface profile of the mirror under test. The key requirement of the test is a grating or ruling as it is called, which consists of a series of parallel lines on a transparent surface such as glass or plastic. The cheaper versions are usually black ink lines printed on a clear plastic sheet, such rulings are perfectly useable. The number of lines on the ruling can vary, the minimum being 80 lines per inch (lpi) up to several hundred lpi. An 80 lpi grating is fine, but a ruling with 100 or 120 lpi will be more advantageous in the test to be described.

The test apparatus (which I will describe later) is setup at the ROC and the reflection of the light source is viewed through the Ronchi grating. The appearance of the lines, or bands as they are often called, gives us valuable information about the surface of the mirror under test. The patterns we view through the Ronchi grating are like the contour lines on a map, they can tell us about high and low areas however, we must take care when interpreting a pattern. A smooth spherical mirror shows straight Ronchi bands either side of the radius of curvature. As seen with the ronchi ruling the surface appears flat, when in reality it is concave, this is due to the fact that like the Foucault test, the Ronchi test is looking at slopes on the mirror surface. As a result of this the Ronchi patterns seen give us an apparent surface profile, not a real one.

If we start with the ruling inside of the ROC, we observe a number of bands; they may appear straight or curved. As the grating is drawn back away from the mirror the number of bands appears to reduce and those remaining broaden. As we pass through the ROC the number of bands appears to increase again and they appear thinner.

What do the bands tell us about the surface of the mirror? If the bands appear straight inside the ROC and remain the same through the ROC and beyond it, the mirror is spherical. Often this is not the case and some interpretation of what is seen is needed.

Here is a video showing the ronchi test for a spherical mirror;

Here is a video showing the ronchi test for a parabolic mirror;

UK source for Ronchi gratings: AWR Technology (Astronomy - Electronics) Miscellaneous

Link to a page about Ronchi testing:

The Ronchi Test

John

[Glasspusher]

The well respected American telescope maker Mel Bartels has written a computer programme to simulate the appearance of the Ronchi bands for any specified mirror. The software may be downloaded from: http://www.bbastrodesigns.com/ronchi.html, this software will be invaluable when it comes to assessing a mirror. Upon running the software the user is presented with a window as shown in the attached images.The user is invited to enter the following data; the mirror diameter in inches, the grating lines per inch, the radius of curvature of the mirror, the grating offset from the radius of curvature and the rays to plot. Please note that the software does not offer the user a choice of units, it accepts only imperial units so conversions may have to be made. Note that the software will cycle through various ronchi grating offsets, showing the appearance of the bands unless the user interrupts the process by entering data in any of the boxes. It is important to realise that the grating offset may be a negative or positive value. It is a negative value when the grating is located inside the radius of curvature, in other words the grating is closer to the mirror than the radius of curvature. The ‘Rays to plot’ box has a default value of 10000; this simply defines the number of points displayed in the image of the bands. Decreasing this value will allow the image to be drawn more quickly but will reduce the resolution of the resulting image. The default value works well.

The Matching Ronchi test.

It will be realised that the images produced by this software can be used to make an accurate assessment of the figure of a telescope mirror. The assumption is made that the figure is a smooth one.

In order to perform the test we will need to know the radius of curvature of the test mirror quite accurately, I will describe how to do this in a later post. We also need to generate, using Bartels ronchi software, a series of images at various offsets for a mirror of the diameter to be tested. We will need to know the number of lines per inch for the grating to be used. It may be useful to print these generated images out to have at hand during the test process. Suitable offsets might be from -0.3 inches to 0.4 inches, but there are no hard and fast rules. The attached images show some of the generated reference images.

When performing the test the observed image, as seen through the test apparatus, is made to match the generated image at zero offset, in other words the image as seen at the radius of curvature. The grating is then moved either towards or away from the mirror by exactly 0.1 of an inch. The bands observed with the test apparatus are now compared to the computer generated image at that offset. The process is repeated at each offset either side of the radius of curvature to see if the images match. If the observed bands are less curved than the computer generated ones then the mirror is under corrected which means that the curve needs to be made deeper. If the observed bands appear to be more curved than the computer generated ones then the curve is over corrected, or too deep. If the curvature and number of bands observed matches the computer generated images at all positions then the mirror is good. It should be said that performing this test, like all methods of testing optical surfaces, does require patience and practise but it can be done. With experience this test is capable of producing a mirror that will meet the Rayleigh criterion. One important point to be remembered is that Bartels’ software has been developed for use with test apparatus that uses a moving light source, this means that as the ronchi grating is moved towards and away from the test mirror the light source moves with the grating. I will be describing the test apparatus in a future post. Please note that I have not attached the images in the correct order, the offset figure will give you the right order!!

John

[Glasspusher]

Thought I would add a some descriptions of terms used in mirror making:

Figure. This is a description of the overall shape of the surface of a mirror, a telescope mirror for a Newtonian telescope has a parabolic figure.

Zone. This is an area on a mirror’s surface that is higher or lower than the ideal figure. We can talk about an edge zone or a central zone and so on.

Oblate. This describes a figure that is shallow relative to a sphere, in order to make it into a sphere it has to be deepened.

Under corrected. Describes a figure that is shallower than a parabola for a Newtonian telescope mirror.

Over corrected. Describes a figure that is deeper than a parabola for a Newtonian telescope mirror.

Turned down edge. This condition in which the edge of the mirror is lower than it should be.

Rolled Edge. Describes a condition similar to a turned down edge only it is more extensive.

I will be using some of these terms in future posts.

John

[Glasspusher]

To make a simple ronchi tester you will need the components shown in the first attached image, they are, a ronchi grating, a plastic or wooden base to mount the parts in, battery pack for two double A batteries, a connector and wire and a bright red led light source. In addition some clear tape and black electricians tape will help with assembly. Cut a piece of plastic or wood thick enough to house the led, approximately 10mm thick should be OK. Next drill two holes, one the same diameter as the led, the other slightly larger about 6-8mm as shown in the attached image. The led sits in the smaller lower hole, that larger hole above it being used to view the image of the mirror. Once the led is in place cover the back of the hole with black electricians tape to avoid stray light interfering with the view of the test mirror. On the front of the assembly place a small piece of tracing paper over the led to help diffuse the light. Next the ronchi grating is secured with clear tape so that it covers both holes. Connect up the led to the battery and you are ready to go.

Some means of holding this assembly will be needed so that it can be moved up and down, as you can see I use and old laboratory stand with a boss and clamp. Some means of making accurate lateral movements towards and away from the mirror is also needed. This could be a micrometer arrangement, but a simple slide and an accurate ruler will do. This will enable measurements to be made when performing the matching ronchi test.

Please feel free to ask any questions!

John

[Glasspusher]

It is very important to ensure that the contact between the lap and mirror is good before starting to polish. Using a full sized lap it is important to polish 50% of the time with the lap on top (LOP) and 50% with the mirror on top (MOT). It is a good idea to polish for a couple of hours then assess the state of the mirror and determine its radius of curvature (ROC). One hour MOT and one hour LOP would be a good start then check the mirror. Dry off the mirror with kitchen towel and hold it up to eye level and catch the reflection of a light with the surface. The mirror should appear fairly evenly illuminated across its surface. Any areas greyer than the rest of the surface will indicate poor contact between the lap and the mirror. Next thing is to perform the ronchi test on the mirror. I like my mirrors to hang in a sling during testing, standing a mirror on its edge can result in the mirror bending, a sling gives a nice even edge support. Stand close to the mirror and pick up the reflection of a torch with the eye, now move away from the mirror until the reflection of the torch light fills your eye. This is the place to set up the ronchi tester. Adjust the tester until the reflection of the red led light source returns through the aperture above it. Move the tester back and forth until the ronch bands are seen. They may appear curved or straight and more bands will be seen the further from the ROC you are. Move the tester until one line of the ronchi ruling is seen on the mirror. The distance from the tester in this position to the mirror is a pretty accurate measurement of the ROC. Remember, the focal length is half the ROC. I have attached a ronchigram of the first test on a 12 inch mirror with an ROC of 129.5 inches. Two important things are seen; firstly the ronchi bands are roughly straight indicating a spherical mirror which is ideal at the stage. Secondly, the centre of the mirror appears a little dark indicating under polishing due to poor lap contact. The reflection of a light caught in the mirror revealed a grey centre. The attached picture of the lap clearly shows a lack of contact in the middle and also the channels beginning to close up. This lap needs the channels cutting with a hand saw and then rinsing with water to remove the debris, it then needs to be warmed in hot water and the mirror pressed against it to improve contact. After this polishing can be resumed.

John

[weareborg]

I have read many mirror making articles, both online and in books, and I must say this is by far the best I have come across. What really makes it easy to follow are the images that go with each post. I keep checking this thread every day to see what progress you are making. It looks fantasic.

Thanks again for posting John.

Andy

[Glasspusher]

Thanks Andy, I really appreciate your comments, when do you hope to start your 14 inch mirror?

John

[weareborg]

Well as usual whenever I plan on getting the mirror, something always pops up. We've got our lads birthday due at the end of the month, so it won't be until after then. My misses and I are also trying to pack in smoking so that we can save up for my scope project and our wedding next year. She's good like that is Michelle. When I first mention to her about building my own scope (including making my own mirror) she was very enthusiastic about the whole concept. Far more so than my ex wife, who used to moan like hell whenever I went into the back yard till the early hours observing, lol.

Andy

[Glasspusher]

The next stage of the process is to fully polish out the mirror whilst maintaining a basically spherical figure. This should be relatively easy to accomplish using a 1/3rd centre through centre stroke and spending 50% of the time LOT (lap on top) and 50% of the time MOT (mirror on top). The stroke should not be too quick, push the lap or mirror back and forth at a comfortable rate, you should feel the drag indicating good contact and a good polishing action. Do not allow the channels to close up and do not allow the lap to dry out. The progress of the polish can be followed by examining the surface of the mirror with a loupe or inverted low power eyepiece, a bright led torch helps if held close to the mirror. At the start of polishing the mirrors surface will appear grey when examined with the eyepiece, as the polishing progresses the grayness will gradually disappear indicating that the mirror is fully polished. The centre and edge should be checked carefully at regular intervals throughout polishing. It is difficult to give an accurate estimate of the time needed to fully polish out a 12 inch mirror but I would estimate a minimum of 12 hours.

The first attached image shows a ronchigram after working with the MOT. The ronchigram, taken just inside the ROC, shows a turned up edge and a centre that is higher than that of a sphere. Remember a spherical mirror shows straight ronchi bands, bands that bow inwards towards the centre indicate a ‘high’ surface relative to a sphere. The indication here is that the stroke was a little short; with the MOT a longer stroke tends to deepen the centre of the mirror. Working by hand tends to produce a slightly random action which is beneficial in blending in any zones on the mirror. Machine work is highly regular and as a result tends to produce zones.

The second attached image shows a ronchigram (taken just inside ROC) after working LOT, again with a 1/3rd stroke as was the case above. The result is a nice smooth mirror which is basically spherical. The turned up edge and high centre seen previously have been removed simply by working LOT. This mirror is now fully polished and ready for the figuring process in which the spherical mirror is changed to a parabolic surface appropriate to the mirrors diameter and ROC.

If good technique is observed producing an approximately spherical mirror is not too difficult a proposition, remember:

Maintain good contact between mirror and lap at all times.

Work 50% of the time MOT and 50% LOT.

Use a 1/3rd centre through centre stroke (about 2 inches overhang at the end of each stroke for a 12 inch mirror).

Do not polish in very cold or very warm conditions.

John

[ponytale]

well, GP i have received the grinding powders, and made my first grinding tool, i have made a table top that sits in my webber bbq base and it seems very sturdy , so i am ready to get on with some grinding, i have made my ronchi/focault tester , maybe tomorrow as i have a split shift with 3.5 hrs in the day where i might get a chance, i am taking pics as i go , wish me luck!!

[Glasspusher]

Hi PT, glad you have taken up the challenge, what diameter and focal ratio are you going for? Obviously, I have not covered all aspects of mirror making in these posts so if you have any problems please don't hesitate to ask. Keeping a record of your progress will prove very helpful, especially if you go on to make another mirror. I am looking forward to seeing the pictures!

Good Luck.

John

[ponytale]

hi gp as i have 2 mirrors partly done i am going for an F5 / ish and an F 6.5 -F7.5 i am going to regrind as you suggested in an earlier post , the f5 will go in my existing dob and the other will be in a different tube to fit the same dob base if all goes to plan? progress so far in pics,

[Glasspusher]

PT that is looking really good! I would strip the tiles off the backing web before sticking them into the cement tool this way you will find it easier to clean between grits.

John

[ponytale]

hi John , i have removed mesh backing and stuck them on with resin , do i fill in the grooves with resin/cement to almost full or leave them empty,? regards Tony

[Beulah]

Ponytale,

Silly question, I know, but can you please tell me where you bought your tiles?

[ponytale]

my local kitchen/tile shop in stroud, £4.77 for square foot sheet, resin from carparts shop/halfords

[Glasspusher]

Tony, you do not need to fill the spaces between the tiles, they need to be open to allow the abrasive slurry to circulate during grinding.

John