This is a short history of my quest to become, in the terms of Sky and Telescope magazine some thirty years ago, a Telescope Nut, and therefore to append the letters TN to my name.

Before I became an old-fashioned single-handed GP¹ I worked for some five years in the hospital service, and rapidly became friends with the manager of the hospital workshop which was well equipped with lathes, drill presses and all the bits and pieces of a general purpose machine shop. It was also well equipped with a wondrous 'junk heap' round the back, where all manner of cast-offs from the workings of a large hospital ended up.

I'd been interested in astronomy since childhood, and grew up making all sorts of gadgetry, and I decided to try telescope making after meeting a professional astronomer at a wedding: we spent much of the evening chatting about his portable 'Rich Field' telescope with a 6in mirror ground in the garden shed. I was hooked, and got hold of the standard book of the time, as well as pestering the local library to lend me the three volumes of the Amateur Telescope Making reprints of Sky and Telescope articles - a veritable encyclopedia of the art.

Preparations and Equipment

I took over a perfect workshop outside the doctors' mess in the hospital, a brick-built storeroom some 15 foot by 8 foot, and borrowed, from the hospital tip, a couple of empty oil drums, to make the standard work-bench of the well-equipped amateur. Filled with water, one at each end of the workshop, these made a perfect optical testing bench. I fitted one with a wooden clamp to hold the glass blank to be ground, while the other, somewhat later, would hold the vertical clamp for holding the mirror during the various test procedures.

I bought a pair of 1 inch thick and 8 inch diameter glass plates from an astronomical supply house, together with an assortment of carborundum powders, and set to work.

Mirror Grinding

Making a telescope mirror can be done in three distinct phases: grinding of a spherical surface, polishing of this to a perfect sphere, and finally turning this spherical surface into a paraboloidal surface.

Grinding a Spherical Surface

First up, clamp one of the pieces of glass on top of the oil drum, sprinkled with coarse carborundum powder and water, and place the other piece of glass (the one destined to become the final mirror) on top. Now starts the long period of rubbing the two together, at random, while walking round the oil drum. By a miracle of geometry, and the use of successively finer grades of grinding grit, the glass clamped to the drum becomes convex, whilst the piece on top becomes concave.

What's more, because two surfaces cannot slide over each other without being part of a sphere (a flat surface can be counted as a sphere of infinite radius...) the mirror blank on top gradually takes on a perfectly spherical surface. The radius of this sphere can be judged initially by measuring the depth of the centre compared to the edge. This initial measurement, the 'sagitta', can be roughly measured by placing a steel rule across the surface and measuring the depth to the curved surface directly - in my case, this came to just under a tenth of an inch. From this, you can make a rough estimate of the radius of the sphere of which the mirror is a part. I aimed for a radius of about 100in to make an f6 mirror: the focal length is half this figure, or 50in. Use the following equation:

Sagitta = r²/2R

as an approximation for a large radius of curvature, where r is the radius of the mirror blank and R the radius of curvature of its face.

Polishing the Spherical Surface

When the centre of the mirror is at roughly the right depth, further grinding with successively finer grades of grit takes the surface to a smooth matt finish. Now comes the start of a series of optical tests actually using the half-finished surface as a mirror. Wet the glass and place it on edge on the oil drum at the far end of the workshop, to enable you to make a more accurate measurement of the radius of curvature by shining light at the 'mirror' and looking at the reflection of the light source back at the centre of the sphere - at the first oil drum.

To carry out this operation, use the first of the tests named after Monsieur Foucault. The theory (and the practice) of this is simple: a point source of light which illuminates the mirror will be returned to a point only if the light starts at the centre of the sphere and is viewed, after reflection, back at the centre of the sphere. The really effective part of this test can not only measure the radius of curvature, but can also give an indication of any deviation of the surface of the mirror from being truly spherical.

At this stage however, all we are interested in is measuring the radius of curvature: finishing of the mirror now awaits the process of polishing. Here another hunt round hospital departments revealed a source of beeswax, pitch and pure turpentine oil where these were used in preparing specimens for microscopy. From the original astronomical supply house came the powders used for polishing: cerium oxide and optical rouge.

The convex glass plate, hitherto fixed to the top of the oil drum, now becomes a polishing tool, with the mirror-to-be now clamped in its place. The preparation of the polishing tool is initially quite complicated: the pitch is heated with turpentine (in an old saucepan discovered on the junk heap) and poured over the convex surface of the glass polishing plate. A turn of sticky brown paper tape round the edge keeps the pitch in place while it cools. Now comes the tricky bit... before it has cooled too much, and while it is still nicely malleable, you cover its surface with rouge and liquid soap and place the mirror on top, weighted down with several bricks.

In a couple of hours, with a bit of luck, the mirror can be removed, leaving a nice convex surface of pitch of the required curvature. To convert this to a random polishing tool, you divide the surface of the pitch up into squares with a tenon saw, in a pattern which makes sure that all polishing operations remain random. A thin coating of beeswax completes the polishing lap.

Now, at last, the first few minutes of polishing can commence. And wonder of wonders, the mirror surface, until now looking like any other miscellaneous piece of ground glass, quite suddenly becomes an optical surface: polished smooth and reflective.

The Foucault tester now comes into its own. The light source shines through a narrow slit between the edges of two razor blades. Just beside this is another razor blade mounted so that it can be slid sideways to block the light reflected from the mirror, as well as to and fro, inside and outside the actual centre of the imaginary sphere of which the mirror is a small part. The basic principle of the tester is that with the second edge at the exact centre of the sphere, the illuminated surface of the mirror will become uniformly dark only if the edge of the blade is moved exactly over the centre of curvature, only if the mirror has a perfectly spherical surface.

Of course, nothing is quite so easy: in general the surface does not darken evenly. Hills and valleys show up on the surface, magnified some 100,000 times by the tester. A warm thumb on the mirror for a few seconds shows up as a mountain gradually subsiding into the background as the heat dissipates and the local area of the glass cools. The whole process is quite phenomenally sensitive.

Depending on the pattern you observe with the testing apparatus, polishing can be concentrated on the centre or the edge of the mirror: with the polishing focused on top, the spherical surface tends to be flattened, and with it underneath the centre tends to become deeper. Eventually though, a perfect test result can be obtained: the surface of the mirror is spherical to a tiny fraction of the wavelength of light. A millionth of an inch - one twentieth of the wavelength of green light - was my goal.

Grinding a Paraboloidal Surface

Only now can you begin the long, long trek of polishing the spherical surface to convert it into a paraboloidal surface. The difference between a sphere and a parabola of revolution with an f8 mirror is remarkably small, yet is crucial to the performance of the final mirror. If you left the mirror spherical, it would bring incoming light not to a point focus, but to the familiar cycloid pattern seen on the surface of a cup of tea in sunlight.

The difference in the surface configuration is only a few millionths of an inch, with the centre a little deeper and the edge a little flatter, yet the Foucault tester can show all this. The basis of testing for the paraboloid involves measuring the radius of curvature for several concentric zones of the mirror. The zone near the centre will have a greater curvature, therefore the razor edge of the tester must be moved closer to the mirror for the zone to become uniformly dark; conversely, zones nearer the edge mean moving the razor further away. By relatively simple maths, the respective distances required can be calculated, and polishing directed accordingly.

During all this the optical rouge gets everywhere, staining hands and clothing, while the process of polishing becomes more and more fraught with tension, the closer you get to the ideal figure of the mirror. Finally the measurements will appear satisfactory, and other tests can be made to confirm them too, including inspection of diffraction patterns of a star, assuming the mirror can be set up in an appropriate mounting.

So far, the surface you're testing is that of the plain glass blank, albeit carefully brought to the correct shape. Now comes the process of aluminising, to bring up the reflectivity of the mirror from about 5% to 95%. This is done by specialists in the field, by electrically depositing in a vacuum a layer a few atoms thick of pure aluminium on the surface. Then follows a variety of other coatings to protect the aluminium - I am informed that beryl-quartz coats my mirror. Whatever the coating, it is thin enough not to change the shape of the polished surface; and in fact, after a few months with the surface continually, though slowly, evaporating, it starts to appear semi-transparent.

Constructing the Telescope Body

While you're waiting for the mirror to return from the specialists, you can start on the construction of the rest of the final telescope. I chose a fork mounting, using surplus ball-races and water pipe fittings, set up in a heavy aluminium U-section framework. I had an aluminium tube rolled and welded by a local firm, and purchased (remarkably cheaply) an elliptical flat secondary mirror and a couple of eyepieces, which I mounted in a cannibalised microscope tube to complete the Newtonian configuration.

The first time I assembled and tested the whole telescope happened to coincide with the first quarter moon. I shall never forget watching the slow progress of sunrise in the tiny craterlets on the floor of Plato, and the changing shadows of the lunar Appenines.

All in all, definitely one of my most satisfying projects.

This entry was written by beeline's father.