Making a Diagonal

There is a competitive market in diagonals for primary mirrors up to 12 inches so if you make one you will work at a loss.
For mirrors larger than 20 inches you will be dealing with a specialty item and if you want a good one you will pay for it.

Diagonals from Scrap Glass

How to Make  Diagonals


Diagonals from Scrap Glass

We have had the window company make us some 1 1/2 inch by 2 inch squares of 3/16 window glass for our demonstration and the window company has sanded the edges.
The thickness is really too thin, it is hard to hold onto this thin glass without touching the first surface. At one time there was very flat plate glass on the surplus market.
That type of glass would be just right for square diagonals.

intrf01.jpg (139107 bytes) If you have ever looked at the spectrum of a cool white, CW, fluorescent lamp through a toy spectrometer you have seen that there is a very strong green spectral line. We will use this light to produce interference patterns to test our glass. We only need the single color so we will wrap our fluorescent fixture in green transparent film we bought at the party store.
intrf02.jpg (130073 bytes) We have numbered the edges of all six pieces with a marking pen so we can tell them apart and also tell one side from the other. There are many possible combinations of 12 sides so when we find a good combination we will write it down. We put two pieces of glass face to face. If we see straight interference lines it means that the space between the faces can be adjusted to be equal all over the entire surface. Before putting the glass together make sure there is no grit anywhere nearby.
intrf03.jpg (70202 bytes) At first we see nothing. After a bit the air escapes from between the glass and we see the interference lines or fringes.
If there is any fuzz between the glass they will never come together. These lines look fairly straight, lets try more surfaces.
intrf04.jpg (76864 bytes) OK, that center one is in the ball park, let's rotate the top glass 180 degrees and see what happens.
intrf05.jpg (78766 bytes) Oops, everything changed. You can begin to see how really small the wavelength of light is. If things were going well we would be seeing almost straight lines with any combination of surfaces. If you press on the glass in one spot the fringes will spread in that area. That tells us that when we mount our diagonal we don't want pressure in one spot.
intrf06.gif (171296 bytes) How well is this pair of surfaces matched? We take the picture into a graphics program and remove the blue so we can see the fringes better. Then we draw a straight line to see how straight the fringes are. The line crosses two fringes so there must be about one wave of variation since the light travels double distance from the top surface to the second surface and then back.
int-bmp.gif (138892 bytes) Actually if you want to be precise there is a shift of pi ( 180 degrees ) before the reflections from the two different surfaces recombine as shown in this illustration from 'Optical Shop Testing'.

The glass we have tried would work as telescope diagonal. By the time the light has traveled from the primary to the diagonal the image has partially formed so the diagonal is not quite so critical. On the other hand the coating charge will be 1/3 the price of a ready made diagonal and also if we test our mirror in the telescope this diagonal will be in the testing loop. If we want to tell errors in the diagonal from errors in the primary we will have to rotate the primary. Too frustrating, lets get a ready made diagonal. 

How to Make  Diagonals

I will tell you in advance that I am not an optician so what you read here is not from personal experience. Anyone who has made diagonals will tell you that it is easier to make a batch than to make just one. To make diagonals you will need a flat as a standard. To make a flat you will need to make three flats. If you test two candidate flats together and get straight fringes at any angle and offset they are both either flat or both spherical, one convex and the other concave. And so this is why we need the third candidate flat since the third flat cannot be both convex and concave spherical it will break a tie. If all combinations of three candidates tested together have straight fringes they are flat.

OK, lets presume we have a flat.

intrf07.jpg (42960 bytes) Here we stack our diagonals to be at a 45 degree angle with something like masking tape between each glass. We encase the stack in something like plaster of Paris. We mount a copper tube the size of our diagonal minor axis in a drill press in such a way that it does not wobble. Then using 80 grit and water we cut down through the stack. Now we have elliptical diagonals to make flat.

We have included a short excerpt, in a  .pdf  file, from the book:
"A Manual for Amateur Telescope Makers" Karine & Jean-Marc Leclaire"
Their book can be purchased from:



Return to calling page