I bought a GUVAs12sd light sensor from Aliexpress and wanted to calibrate(it seemed far off) and put a unit on its readings. I know there are multiple UVs, UVA,UVB, and UVC. Is there a unit that covers them all? If so how do I calibrate and make my readings become like that? Thanks.
Watts per square metre.
Is it a raw sensor or does it have an oboard op amp like this one?: Analog UV Light Sensor Breakout - GUVA-S12SD : ID 1918 : Adafruit Industries, Unique & fun DIY electronics and kits
Do you have a datasheet?
Note: measuring radiation UV IR or otherwise is strongly dependent on the angle.
So if you want meaningful data you must compensate for that angle.
Google Lamberts Cosine Law
See - GitHub - RobTillaart/BH1750FVI_RT: Arduino library for BH1750FVI (GY-30) Lux sensor how it could be done.
Do you really want to spend maybe more that $500 for test equipment just to calibrate a sensor that cost $2
Never seen one for hobby use.
Common sensors detect UVA/B. You need separate sensor for UVC. For example GUVC-S10GD can sense far-UVC.
Compare your readings with the numbers from a nearby airport or weather station.
Do that for at least a dozen days with different "weather" conditions.
Should give you several points to compare.
Use a spreadsheet to find the correlation
(or use - GitHub - RobTillaart/Correlation: Arduino library to determine the correlation between X and Y datasets.)
I was looking at this website and found:
I don't know if this is accurate but it looks like it can measure 50% of UVC, all of UVB, and 75% of UVA.
What I know is that it outputs analog voltage but don't know exactly how to convert that to meaningful readings. I found from the same website the following table:
I don't know if it is accurate, but I could go off of it. one thing I noticed was that the sensor has variable readings between runs that are +- 20 when taking analog values, is that normal? I am taking values from inside my house where there isn't too much UV.
I tried to sense germicidal UVC light (254nm peak) with that sensor. It was total failure for me.
OK, thanks for letting me know. I don't need it to monitor all the UV types, would it be the safest if I just say it measures UVB only?
Depends on what you are looking for. I think your sensor has quite good sensitivity to irradiation used to calculate UV index for example. Afaik UVB is dominant there.
No, the photodiode responds to a fairly wide range of wavelengths, overlapping UVA, B and C. So, it does not "measure" UVB, for example in sunlight.
To use that photodiode to measure UVB, a bandpass filter is required.
GUVA-S12SD responsivity curve:
UVA (315-400 nm)
UVB (280-315 nm)
UVC (100-280 nm).
If you compare that to solar uv spectrum, it's quite linear. In practice UVC is insignificant in both. If you do what @robtillaart suggested, you can "calibrate" your sensor against local UV index. If that is what you are looking for...?