Mirror Reflectivity
The reflectivity of the 6 first surface mirrors at various angles:
30 degrees:
45 degrees:
60 degrees:
75 degrees:
Note that mirrors 1 and 4 correlated with eachother as separate from the other mirrors. All 6 had the same coating, and nothing was visibly different between them. A geometric issue would have been a constant offset everywhere, and this does not seem to be the case. As such, both mirrors 1 and 2 are compared to ECI and thor. I am not sure where the high wavelength wiggles came from, but they are present in every run. My suspcision is that they have to do with the difraction grating and overtones, as they have consistent peaks and valleys per angle.
Below compares the 4 different coatings:
30 degrees:
45 degrees:
60 degrees:
75 degrees:
The ECI and the thor mirror perform better at lower angles and wavelengths. The thor mirror does not perform as descibed on the website I am not sure if this is due to my handing of it or the geometric affect that may have caused the wiggles. At 60 degrees both the thor and the ECI had above unity reflectance at low wavelengths (though not by much. I am not sure what caused this the reference runs were taken per angle, but this is far above the expected performance of both.
Despite the systematic concerns related to the wiggles and the 60 degree data, I integrated the reflectances against our expected wavelength distribution to get an average reflectance for each mirror:
The ECI mirror only outperforms the glass mirrors at low angles. This is expected - the coating is specifically made to enhance the UV reflectance at low angles, and the distribution of wavelengths is peaked at ~ 350nm. That said, until the systematics are better understood, there is some uncertainty on these results.
A reminder of some expected reflectivities:
Thor:
ECI:
Glass mirror (see enhanced):
Angles for the glass mirror and ECI are not known, but presumed to be small.
Reflectivity measurements made at bates (using LEDs and a ccd camera, with the caveate that the purple spectrum isn't known, nor are we certain that the low lambda LEDs have a pure spectrum):
