REFLECTOMETRY: Measurement of light reflected from an object as a means of assessing its properties

Usually, reflectometry requires complex assemblies of lasers, lenses, mirrors, apertures, spectrometers, etc.. This works fine. Except that it often takes one or more Ph.D.s to set up and maintain the apparatus. We needed to design a system that could operate night and day in a factory environment, without such assistance. We "stole" technology that had been developed in other areas, and adapted it to serve our purposes. We replaced the laser and assorted optics with a light bulb directing light into an optical fiber and out a single lens into our crystal growth systems. That's the blue box, coil, and small cylinder at the lower left. To collect the light reflected from the sample, we used an identical lens and fiber coil to direct the light into a simple spectrometer. Spectrometers, alone, can be $10,000-20,000 dollar boxes, with lots of motorized gratings and mirrors. Instead, ours is the little white box at the right. It has absolutely no moving parts, not even an on-off switch. It takes the light directly from the fiber, bounces it off a tiny grating, separating it into its component colors. That fan of light is directed onto the open top of a single integrated circuit. This integrated circuit, a linear CCD array, converts the light, at each point, into electrons, then reads the bits out. Did we "invent" any of these components? No - that's the point! Every single one of these components came straight from one catalog or another. Further, not one of them required any adjustment or alignment: they simply snapped or twisted together. We had to work a bit to find the right components, but having done that, anyone could replace them, should the need arise. The cost of the total system (minus the overpriced laptop PC)? About $2000. What about the Ph.D.s? We incorporated our expertise into the software package, so we could move on to other challenges, and let the factory get on with its work.