Solar Sensors and Cloud Conditions

[TokWW]

This discussion was Off Topic in the 1st August Weather Reports so I have decided to start it here as a new Technology Topic.

A solar sensor I purchased was a non oem addition to my weather station. It was a Hobby-Boards one (Maxim-IC chips) and matches individual components connected together in a CAT 5 network with or without a field hub. The device was a cheap add-on which Weather Display software can read and "add" to the otherwise independent Wx stn equipment. It can come as a combo with humidity and temperature.

LINK:
Humidity/Temp/Solar [HTS3-R1-A] - $55.75 : Hobby Boards : Complete 1wire Solutions

This solar sensor gives a mV signal amplified from this photodiode that is a broad optical spectrum sensor - totally uncalibrated and is connected via a network (of Maxim "1-wire" devices- other devices can be chained on) to the PC and Weather Display software (ex Windy) can read the devices (humidity sensors, temp sensors, wind anemometers, lightning strike detectors etc) and knowing Lat and Long his software can generate an expected 100% value of the Sun's radiaton for the time of day and year but you have to calibrate the device signal in mV as a max at zenith as 100% value in the software. And based on the time of year etc, Windy has calcuated a W/m2 value as well. However, sunshine hours can also be calculated based on >80% of expected max value for that time of day.

However, with these devices, they do need a translucent shield over them to give an even performance, but then it can give error values >100% due to cloud reflectivity etc and I guess you could get light reflected off buildings and roofs too. If they are left to themselfes in our strong NZ light, they can over saturate the signal and go down to zero! So they do need a reduction in strength of the light that these shields (eg a ping pong ball shroud) give, and thus a more linear response too at the left hand side of their curve. At some point sloping to the right, there is a knee on the curve heading to zero based on excessive light signal.


So while not perfect or truly calibrated, it is enough to give a variable rate of cloud cover and verbal parsing to give sky conditions as can be seen on the mesomap where they may say, for instance: overcast, partly cloudy, cloudy or cloudy2 - which we still need clarification on. I have had all of the first three conditions today already. HTH... Thanks for the question!!


Note: this is not a device for measuring UV which needs a narrow UV band sensor and calibrated accurately in a lab and are quite expensive usually.