|
LINK: APPLICATION NOTES Signal Processing Techniques:
At one time, Radiation Thermometry was fairly simple. You selected the appropriate type of RT to use, determined what the best emissivity probably was, and if you designed things right, you’d get some reasonable readings on most products. Then some simple techniques were developed such as peak-picking. When trying to look through occasional smoke or steam that obscured the sight path, a peak picker continually selects the strongest signal received and holds it for a while, which vastly improves the ability to look through the obscuration. It’s a fairly simple circuit electronically. Valley pickers (much less used) work oppositely, selecting for the lowest signal. This could be useful if some source of periodic radiation added to the RT signal. With the advent of cheap A/D cards and computers, it became obvious that a computer could be used to log the information from the RT. For instance, if a rotating or rocking mirror was place in front of the RT (or the RT itself was rotated) then product could be scanned and the temperature profile could be displayed and stored or peak temperatures could be computed. (A signal of mirror or RT position is usually required as an additional computer input). Such devices are easily built and cheap if you have a spare old computer lying around. I’ve built quite a few. Then things got a bit more complicated in the search for greater accuracy. If the product being viewed had a variable emissivity that could somehow be measured (for instance by a laser-based surface roughness analyzer) then a small (often embedded) computer could be used to compensate the temperature readout for the varying emissivity. Or, correlated material properties could be analyzed (again, through a computer) at several different wavelengths to develop material-specific compensation and obtain better temperature readings. Such techniques include several different patents I hold. Williamson is one of the few companies offering such devices commercially. Another type of application uses laser-TV systems to determine product shape or angle and combines this information with the RT to adjust emissivity for better readings. The number of applications where improved signal processing can be applied are limited only by our imaginations. I could not even begin to describe applications without writing a book, and by the time I wrote the book it would be outdated! Two specialized methods for measuring the temperature of low emissivity materials near room temperature deserve mention as without these methods, the measurement would be almost impossible: One is what is called a driven source method. An appropriate long wavelength RT, with its emissivity set to 1.0, sights the surface through a hole in a plate spaced close to the material and which has been machined and blackened to have an emissivity close to 1.0. The plate, of course, has an embedded thermocouple. The temperature indicated by the RT is used as a setpoint for a plate heater which drives the plate to the same temperature as that indicated by the RT. Naturally as the plate gets hotter, the RT (through reflected radiation) reads a higher temperature and the whole system finally stabilizes (via computer control) at the point where the plate temperature tracks the temperature of the viewed surface. The other method uses radiation polarization with a driven source, which is beyond the scope of this discussion. You could probably learn about this by using scientific search engine referencing “Polaradiometer” and my name or you can buy a book called “The Theory and Application of Radiation Thermometry”, the Bible of RT.
|