[spudstrawb]

I was liaising with someone just recently about UV sensors etc and they made the following comments:

"I initially made my UV sensor cover from high quality PTFE due to its excellent optical properties, but my sensor was blind as a bat. Further research uncovers the optical properties are wonderful for visible light right through NIR and up to a couple of thousand nm, however it's about 97% attenuating at wavelengths below about 390nm, which includes all of the UVA, UVB and UVC bands!"

"I ended up getting some 25mm dia JGS1 which is around 95% transmission down to about 205nm. The sensor needed only a tiny little tweek in software for correct the miniscule loss. Problem I then had was that without any diffuser, I had poor angular response. Mechanical frosting of the glass was considered, but sandblasting it leaves a surface that attracts and holds oil and really doesn't behave well outdoors, and is difficult to clean. Laser etching leaves microfissures that are as bad or worse. In the end, I resorted to hydrofluoric acid to chemically etch both sides of the glass. It seems far more weather-stable, and is giving me near-perfect results. I'd urge anyone considering this to be VERY VERY careful of HF, to read up about it very well beforehand, and to take every possible precaution with its handling."