According to the spec sheet, the VEML6075 sensor from manufacturer Vishay delivers UV-A and UV-B intensity. And since the sensor, a photodiode, has cross sensitivity to visible and infrared light, the sensor offers additional signals, which supposedly can be used for correction.
But this correction is not working properly, and this is where I need help.
Unfortunately, this sensor is out of production, almost nobody can deliver, but I got lucky at AliExpress - at a rather high price :-( .
The sensor is based on I2C, and I use it on a Raspi with smbus in Python3. It is easy to program.
It is also lacking the right documentation. I have available the Application Note "Designing the VEML6075 into an Application" from 2015, and the data-sheet "UVA and UVB Light Sensor with I2C Interface" from 2016. But apparently things have changed from 2015 to 2016, and while the Application Note provides the formulas for correction, they are absent in the data-sheet, and the old formulas give negative values, which is physically impossible. For example, I get:
UVA: 103
UVB: 108
UVD: 0
UVCOMP1: 26
UVCOMP2: 13
UVAcomp: -16.08
UVBcomp: -22.91
UVI : -0.023
The terms are from the documents. UVA and UVB make sense at least when I watch their course over the day. UVD (dark current), UVCOMP1 (visible light), and UVCOMP2 (infrared light) are delivered by the sensor and are meant to be used for correction. The UVD is always zero, without an exception, and in the newer document is marked as "Reserved". I think this means it is no longer valid anyway. But as it is zero it does no harm within the formulas.
Applying this correction according to the documents and what I get is shown as UVAcomp, UVBcomp, and UVI (UV Index) values. All three are negative, which is impossible for real-world values.
Does anyone have guidance, ideas, documents, comments, which could help me further?
