Cumulus Support

I have a WS80 and on bright solar days I get the following effect on the charting of the solar radiation:

This is no coincidence, the measured curve is biased in the morning hours. From 10h to 13h30 (DST) it is biased upward around up to 30%. It seems directional: it starts at East and once the sun has passed South the values become reasonable again.

Anybody?

Does there exist a filter to avoid this effect?

Does you WS80 have a small white filter over the central area of the top where the solar sensor sits? Mine does and it has improved it but I still need to calibrate the output by 0.75 to get the levels down to reasonable. I still find CMX gives more sunshine hours than Weather Display though. Could there be something to the East which causes reflections?

Stuart

No, you look straight into the sensor, no filter to be seen. Your calibration is roughly the same as mine. Sunshine hours are OK here (have been checking for some time, seems close).

And I have been thinking about possible causes for this effect but can't think of any. No reasonable reflection.
Maybe polarisation through the cover glass?

Yes my old WS80 (replaced under warranty) did not have the white filter but the new one does. I'm still trying to work out why my Weather Display has fewer sunshine hours than CMX using the same data as input from the sensor. The expected max values are different between the two which may well account for that but I'm still investigating.

One other thought I suppose your WS80 is vertical and not leaning over at all? Should not make that much difference but thought I'd mention it. Also I assume it is clean?

Stuart

Yes, it's clean. Have to check the leveling I guess. Good point.

Wrt sunhours you probably have to compare the methods. I assume in both applications it is an algorithm (in CMX it is) and it is configurable.

Yes I know, I helped write the expected max code in WD some years ago. I'm talking to Mark about the CMX interpretation.

Stuart