- #1
seycyrus
Hello,
I have a Tungsten filament radiance source that has been calibrated in the UV region and would like to extrapolate the radiance to longer wavelengths. By calibrated, I mean, a NIST plot of spectral radiance v. wavelength at a certain set of operating conditions.
I'm new to this area, so i will start off with a big "I think" and apply it to everything I write from now on :)
I have an old lamp with a more complete (larger range of wavelengths) calibration. It is fairly obvious that I cannot treat it as a simple grey body as the old report contains words like "brightness temperature" at different wavelengths. In other words, one single T won't fit all the data.
That lamp HAD spectral radiance values in my region of interest, but that lamp is now broken :) Now I must make do with the source that has a much smaller range of calibration and would like to figure out what the spectral radiance would be in my range of interest.
I came upon a paper that indicated that the emissivity had a certain wavelength dependence and then used ORIGIN to do a multi-variable nlsf. I was pretty proud of myself when I could fit the broken bulbs data set "perfectly".
I though that I might be able to do the same sort of nlsf on the bulb with the smaller set of calibration data and then use the fit to get values for the wavelengths I wanted.
But when I tested this with a subset of the broken bulbs data, it didn't work. The peak position was way off.
Anyone have any idea on how to proceed? Is there a standard way to calculate the spectral radiance of a given metal at an arbitrary wavelength (besides paying NIST to do it).
Thank you for your time
I have a Tungsten filament radiance source that has been calibrated in the UV region and would like to extrapolate the radiance to longer wavelengths. By calibrated, I mean, a NIST plot of spectral radiance v. wavelength at a certain set of operating conditions.
I'm new to this area, so i will start off with a big "I think" and apply it to everything I write from now on :)
I have an old lamp with a more complete (larger range of wavelengths) calibration. It is fairly obvious that I cannot treat it as a simple grey body as the old report contains words like "brightness temperature" at different wavelengths. In other words, one single T won't fit all the data.
That lamp HAD spectral radiance values in my region of interest, but that lamp is now broken :) Now I must make do with the source that has a much smaller range of calibration and would like to figure out what the spectral radiance would be in my range of interest.
I came upon a paper that indicated that the emissivity had a certain wavelength dependence and then used ORIGIN to do a multi-variable nlsf. I was pretty proud of myself when I could fit the broken bulbs data set "perfectly".
I though that I might be able to do the same sort of nlsf on the bulb with the smaller set of calibration data and then use the fit to get values for the wavelengths I wanted.
But when I tested this with a subset of the broken bulbs data, it didn't work. The peak position was way off.
Anyone have any idea on how to proceed? Is there a standard way to calculate the spectral radiance of a given metal at an arbitrary wavelength (besides paying NIST to do it).
Thank you for your time