[laulau]

I was concerned about the curvature of the dome however and how this might reflect / refract or even focus the sun light. Also, for the solar panel at least, would be very difficult to moiunt it directly under the dome with no air gap.

I'm almost sure that the curvature of the dome will cause problems. Another possible problem is making that enclosure resistant to weather elements (water, sun exposition, etc, etc.)

My preferred enclosure is like this one:






Unfortunately I never found one on ebay with a clear cover (without that white trace).

Those enclosures are very resistant, they are made to be installed in the ground. It's very easy to remove the switch at the back, and use the hole to pass the cable (just make small hole in the membrane, then after pass the cable seal it with silicone). There is plenty of space inside to install the solar / UV interface



Here is a picture of the unit I'm using since two years.



I bought mine at a local store. The cover is made of a 5mm glass, which has lowered the UV readings by around 30% (later compensated in the software, by comparing the readings with nearby weather stations, and values published on portuguese official meteo site). Not a very scientific method but enough to have readings inline with other sources.

I have mine in a very similar enclosure.
I kept the original PV cell and put a proto PCB into the plastic box that was inside. I reused one hole for the UV sensor. I tried to align the inside PV cell connections with the white trace of the cover.
   
   
Calibration hasn't been done correctly because i have to wait for a clear summer day!
Some questions (@ Werk_AG):
Is it possible to have the dimensions of the PCB and the PV cell from your solar kit to see if i can swap them at some point.
Where did you find the cover without the white trace ?
Is the UV sensor sensible to this white trace ? if no it's perhaps better to align the trace with the UV sensor!
Thanks
Laurent