Testing Methods:
Total telescope light throughput can be measured in two different ways; either by measurement of the assembled optical system, or by measurement of the reflectance of each mirror (or reflective element), and the transmission of each refractive element in the optical path. In the case of a Schmidt Cassegrain telescope, there are two reflective elements (the primary and secondary mirrors), and one refractive element (the corrector plate, or Schmidt Corrector). See diagram below:
Assembled Telescope vs. Individual Optical Element Analysis:
To measure the throughput of the assembled telescope, a beam of light is passed through the telescope and compared to a beam of equal intensity light passing through air only. Total telescope throughput is then the ratio of light intensity measured through the telescope divided by the light intensity measured through air. This is easily said, but very challenging to execute correctly. Great care must be taken to ensure that the reference beam is of constant intensity, and that its light is collected in a manner which does not bias the results. Errors introduced by beam geometry (f ratio) at the entrance to the detector, less than perfect alignment of the optical elements, including placement and dimensions of internal light baffles, will tend to reduce the intensity of light measured through the telescope.
The second method of measuring total telescope throughput, by spectrophotometric analysis of each element in the optical path, is not susceptible to these sources of error. Furthermore, individual element analysis provides specific information about each optical element, while measuring the throughput of the assembled optical tube does not. Results obtained in this manner represent an upper limit to the actual throughput of the assembled telescope. Total Telescope Throughput (%TT) is less than or equal to Corrector Plate Transmission (%TC) times Primary Mirror Reflectance (%RP) times Secondary Mirror Reflectance (%RS).
Corrector Plate Transmission (%TC):
We use a Shimadzu UV1601 spectrophotometer for analysis of corrector plate transmission. This is a double beam instrument with a spectral range of 190 to 1100nm. Transmission data is typically collected in the visible region from 400 to 750 nm. Small samples of corrector material called witness plates are included in each corrector coating run. In order to minimize handling and the possibility of scratching a full size corrector plate, we use these witness plates to represent the transmission characteristics of our correctors.
Our instrument is capable, however, of measuring the transmission of correctors up to 8” diameter. If this is necessary, the corrector plate is measured at 4 points roughly 90° apart, and the results are averaged. Before and after each measurement, baseline (100%) measurements are made to ensure light source and/or detector drift is negligible.
Primary and Secondary Reflectance (%RP, %RS):
The preferred method of measuring reflectance of primary and secondary mirrors involves the use of witness plates as well. These are small (1” to 2” diameter) flat polished glass substrates, which are coated along with the primary and secondary mirrors. Since the coating process is the same, and the surfaces are equally well polished, the reflectance of the witness plate is the same as that for the primary and secondary mirror. The reasons for using flat witness plates are 1) the primary and secondary mirrors are not themselves subjected to a measurement process which can potentially cause scratches, and 2) very simple test methods and readily available reference standards can be used to measure the reflectance of flat surfaces.
Typically, the reflectance of a surface is measured against a standard reference
of known reflectance. Our standard reference is an enhanced aluminum coated quartz
flat, calibrated against a NIST (National Institute of Standards and Technology)
specular reflectance standard. To measure the reflectance of a flat sample, the
baseline measurement is made using this standard, and the reflectance of the
sample is compared to this baseline. The sample reflectance factor (%RS)
is
equal to its reflectance relative to the reference standard (%RSR) times the
reference
standard’s known reflectance (%RR):
However, if the sample to be measured has a curved surface like a secondary
or a primary mirror, and there is no witness plate available, then special
care must be taken to ensure that the method used to measure reflectance
is insensitive to this curvature. If we compared the reflectance of a curved
surface directly to that of a flat reflectance standard, our results would
not be accurate, since the converging or diverging beam generated by a curved
surface would direct either less light (in the case of a secondary mirror),
or more light (in the case of a primary mirror) onto the detector than was
directed by the flat reference standard.
The most widely used tool for measuring the reflectance of curved surfaces
is called an integrating sphere. This device collects and then measures
the intensity of light in a manner which is insensitive to beam geometry, hence,
insensitive to surface curvature of a reflective sample being measured. However,
integrating spheres can be quite expensive, and they are time-consuming to
set up and calibrate. We developed a method which is equally insensitive to
surface curvature, but much less costly and time consuming to perform. We made
our own reference standards from secondary and primary mirrors with the same
surface curvature as those we wished to test.
We obtained samples of the secondary and primary mirrors which we wished to
test, stripped the existing coating, and replaced it with one for which we
also obtained flat witness plates. These flat witness plates were calibrated
against a NIST specular reflectance standard. Since the flat witness plates
were coated along with the curved samples, and since we have adequate data
to show that our coatings are very uniform from part to part in any given coating
run, we can apply this reflectance data to our curved samples. Using these
curved surface reflectance standards we are able to measure other mirrors
of the same curvature just as we use our flat reflectance standard to measure
the reflectance of flat samples.
To perform these measurements, we use an Ocean Optics USB2000 Spectrometer
with an LS-1 Tungsten Halogen Light Source. This is a single-beam instrument
with a 0.3nm resolution, a scanning range from 340nm to 1024nm, and is equipped
with a fiber optic curved-surface reflectance measuring probe.
Reporting the Data:
Collecting the data and reducing it to yield total telescope throughput (%TT)
(system transmission) is simply a matter of multiplication. We find the average
of each data set (%TC, %RP, and %RS) for each
wavelength measured, and multiply them together.
|
