Filter Development in the CEPSR Clean Room

Many current CMB experiments use bolometers, which leads both to very high sensitivity and extreme susceptibility to out-of-bandwidth noise.  There is thus a high demand for narrow bandwidth millimeter and sub-millimeter filters.  We have been working for the past year on developing such filters using micro-etching processes in the CEPSR clean room at Columbia, and we have successfully produced prototype inductive cross metal-mesh filters at several frequencies on both mylar and HDPE substrates that are in good agreement with theoretical results that we have produced by computer simulation:

Filter I

Filter II

The experimental data was collected using a Fourier Transform Spectrometer at UMass, which, as you can see from the data, becomes innaccurate at lower frequencies.  We are working on building our own FTS with an intended range of 80-1000 GHZ, which will allow us to test filters in-house for a variety of IR/microwave frequencies that are needed in the field of CMB research. 

Here are some photos of our work:

150ghz filter with copper on mylar	280ghz filter with copper on HDPE
280ghz filter with copper on mylar	300ghz filter with copper on mylar

The small irregularities are due to imperfections in the mask that we use for exposures, but already the data above shows that we are getting good results.  Obviously any final work will be done with high-quality masks.

Here is a normal picture of a 280ghz filter: