Determining TEC Flatness with Newton’s Rings – part 1

With a 340W, 50x50mm TEC (thermoelectric cooler) arriving at work this week for a project we’re working on, Steve asked me to look at using Newton’s Rings to determine if this new TEC was flat enough to use as it was, or if we’d need to lap it with diamond lapping paste to improve the contact between it and the water cooled copper block it will be clamped to.

A large part of the problems I had when writing my dissertation for my degree came down the the flatness of the thermoelectric cooler (TEC) I was using and the clamping force I could achieve on it. Without both being as flat as possible, the thermal contact between the TEC and the heat sink was poor, and so the TEC didn’t heat or cool as well as when both were lapped flat.

Newton’s Rings provide a way to determine the flatness of two surfaces optically. With a piece of glass you know is flat, you can tell whether another surface is flat by looking at the interference patterns between the two surfaces. With a monochromatic light source and optical flat on order, Steve and I armed ourselves with our impatience, a telescope mirror blank (maybe almost flat), this excellent PDF and the TEC and copper block to see what we could see.

Newton's Rings

So with me in anticipation of seeing “flatness” optically for myself for the first time, and Steve having not seen Newton’s Rings since his college days, there we were, me perched on my desk and him on his, leaning towards the strip light in his office, tilting the glass and copper pair this way and that to see if we could see anything. Thoroughly dejected at our lack of success, I put the pair on my desk as he slumped in his chair and I started to clamber my way off my desk. His shout of triumph almost helped me the rest of the way down, since where he’d sat, the light from my computer monitor let him see Newton’s Rings on the pair that I’d placed in front of my monitor.

And so with MS Paint open to give us a convenient red or white screen, we could see the bands of constructively and destructively interfering light. Each transition from light to dark represents one half wavelength of difference in distance between the two surfaces.

With my (as yet) novice take on these pictures, one of both of the surfaces aren’t flat (well duh!!). If both were flat and the glass was spaced off the TEC or copper block by a piece of tissue paper, the lines of interference should run in straight lines parallel to the edge lifted off the lower surface by the tissue paper.

Looking at the pictures I got, I’d guess that the glass plate we had was slightly concave, since the bands curve around into circles and that when the two pieces of glass we had were stacked the lines that appeared curved towards the lifted point (where the tissue sticks out in the photo below), indicating concave.

With this introduction to flatness testing using Newton’s Ring’s I’d really like to take a further look at the difference between the flatness of TECs that are sold as pre-lapped and those that come unlapped once the flat and monochromatic light source arrive, since there’s a marked difference in price between lapped and unlapped. I’d also like to compare effectiveness of these lapped and unlapped TECs.

Leave a Comment