This article discusses how to build a monochromatic light source to determine the flatness of a surface optically. It uses a high pressure sodium lamp housed within a box to shroud the light source, designed around a diagram from the book Amateur Telescope Making by Albert G. Ingalls. High pressure sodium lamps are the bulbs used in most street lights, they emit visible light within a narrow band of wavelengths.
In my previous post I described observing Newton’s rings using my computer monitor as a light source to show the flatness of a TEC. A flatter TEC has a better thermal contact with its heat sink, allowing faster transfer of heat away from the hot side and hence faster cooling of the cold side. The light I used previously was not monochromatic and so all of the wavelengths of light interfere constructively and destructively at different points, creating rings of different colours blurring into each other. While this allows you to see if it is flat or not, it doesn’t let you do so quantitatively. By replacing the non-monochromatic light of my PC monitor with a monochromatic source, I should in the future be able to determine how flat the TEC is.
The following descriptions and photos discuss the why and how of building a source to measure flatness.
Steve pointed me to a section in a book called Amateur Telescope Making by Albert G. Ingalls which detailed methods of determining the flatness of a surface. The book describes four methods of varying difficulty. The diagram below is the one I chose, which compromises clarity of image for simplicity.
The method from the book expects you to view the rings in the glass plate at 45 degrees as the image is reflected off of it. I built a box from 18mm ply to contain a monochromatic light source, the electronics to run it and a glass plate I salvaged from an A4 scanner. After building it I found that it didn’t work as indicated in the above diagram and that the image I saw of the rings was much stronger when viewed directly at a shallow angle than when I looked at the reflection in the glass plate. In reality I found that the glass plate was very useful to hold some sheets of plastic dispersers/polarisers that were salvaged from a broken PC monitor. Bear this in mind while following the instructions below; the wiring and box details may be useful but you may prefer to take different steps to secure a method of light dispersal which I found necessary to prevent the bulb overpowering the image seen in the optical flat.
For the light source I used a 70W high pressure sodium lamp (SOX) with a ballast and ignitor.These were wired as per the diagram printed on the side of the ignitor. The diagram didn’t show the earth connections though these were stamped on all of the components. The picture below shows the circuit diagram printed on the side of the ignitor which is mounted in the aluminium bracket I made for it (click to enlarge):
Below are the electronics mounted in the beginnings of the box I constructed, with a small bracket to support the bulb halfway along its length. An aluminium bracket was also constructed to hold the ignitor which is an awkward shape to mount without it. The mounting hole doubles as its earth point. (click to enlarge)
A stainless steel plate made to hold the IEC socket. Not necessary but a satisfying touch when I fancied making something in the workshop one day and it helps with storage if there’s not always a plug and lead dangling from the box. Fixed to the box with self tapping screws:
The almost finished box without its front panels and lid (below). This picture shows the arrangement of electronics well and the spacers used to keep the bulb central(ish) above the glass plate. The box could have been made narrower to avoid this but then the electronics would have had to have gone on the outside at the back.
Below is the finished light box without its 3mm ply wood lid. After this picture was taken thin films of plastic from a broken computer monitor were cut to the size of the glass and laid between it and the SOX bulb to diffuse the light, dullthe reflection of the bulb on the optical flat and hence make the rings clearer.
I had intended to paint the interior of the box white to reflect the light down and the back of the box behind the glass black to reduce the reflection of the bulb on the glass. Upon finding that the image of the rings were clearer when viewed directly I didn’t think this was necessary.
With this setup I can see Newton’s rings between the flat and my TEC much more clearly than is apparent in the title photograph. The brightness of the SOX caused a lot of problems in getting a clear shot with my point and shoot camera. Even with a lot of manipulation the image shows the rings much less clearly than with the naked eye.
In a subsequent post I hope to discuss the use of this light box to improve the flatness of the TEC and its water cooled block, including methods of lapping with a lapping plate and grits as well as approximating the flatness by knowing the wavelenth of the light emitted by the light source.