[hornychz]

Congratulations! Good job.

[41south]

Hi Philip

I'll be interested to hear how you go with the TSL2591. I have one here too but it tops out at about 88000 lux and normal sunlight can be up to 110-120k Lux. I have the analogue GA1A12S202 Log-scale Analog Light Sensor sitting outside on test at the moment, but it too tops out well before full sunlight. I now have it covered with a very opaque cover to lower the light range. Do you intend to just report lux or try and convert it to w/m2? I have a feeling that because they are so sensitive at low light light levels the light curve is going to be "interesting"


Cheers
Colin

[philipgm]

Re theTSL2591.
I am using a conversion to w/m2. I haven't monitored the lux reading so I will watch that. I have the sensor under clear plastic. I will let you know the results that I am getting fot the lux reading.
My uv readings are a little high. I am using a little bit of 'uv clear'  plastic over that sensor.
Cheers 
Philip 

[JT118]

Might be of interest re using a lux meter to measure solar irradiance.

Measuring solar energy is complicated stuff. A lux meter measures the spectrum visible to the human eye and is biased to the green spectrum, PAR meters bias to the red and blue spectrum(400-700nm) with a built in reduction in response to the green spectrum as plants use blue and red: pyranometers, try to get a fairly flat response between 300 and 2800nm, nett radiometers can be thermopiles measuring low frequency energy reflected back from the earths surface etc.

The lux is one lumen per square metre (lm/m2), and the corresponding radiometric unit, which measures irradiance, is the watt per square metre (W/m2). There is no single conversion factor between lx and W/m2; there is a different conversion factor for every wavelength, and it is not possible to make a conversion unless one knows the spectral composition of the light. ( https://en.m.wikipedia.org/wiki/Lux )

Good book on this is Alex Ryer, Light Management Handbook.

http://www.google.co.uk/url?sa=t&source=...pOwhkP99cV

[JT118]

Just to add this,

http://www.clearskycalculator.com

Is a very useful tool to check the solar radiation output from any solar sensor. On a clear, pollution free day (which I am sure you get plenty of in Picton) your sensor should read no more than the figure the calculator calculates on any given day of the year, latitude etc. and an offset applied to the readout if it does.

[41south]

Just to add this,

http://www.clearskycalculator.com

Is a very useful tool to check the solar radiation output from any solar sensor. On a clear, pollution free day (which I am sure you get plenty of in Picton) your sensor should read no more than the figure the calculator calculates on any given day of the year, latitude etc. and an offset applied to the readout if it does.

Yes that's my "go-to" calculator for all my testing with these different devices. I suspect that both the ones mentioned saturate quickly with full sun. The analogue one from Adafruit I'm testing also outputs a logarithmic curve, so the math is even more interesting but well supported (for Lux output) in the code examples available. I found 2 sources that use a factor of Lux * 0.0079 to convert to w/m2, but I agree with you that it is a little more complicated than that. However it may be ok for hobby use

All good fun experimenting though.

Colin

[JT118]

It is all interesting and useful. 

It was quite interesting at a meteorology meet one of the discussions was how data collection is changing from a relatively few manual stations around the world to using the data from 100's of 1000's of pws and other stations and how to build algorithms that could sort out the huge variation in the accuracy of the data and deal with the volume.

Looking at a wunderground map it is remarkable how consistent the readings are across numerous stations, but invariably there are outliers which are feeding the results from failed sensors to wunderground. I have one reading which records 13000mm of rain in a day and have been unable to remove it.

I think we all add to that data in a positive way and all contribute to this mass of data whilst also adding a little to the confusion for good measure.

6 weeks to go before I head your way, and maybe a weatherduino station will appear in Beach Haven.

[philipgm]

It is all interesting and useful. 

It was quite interesting at a meteorology meet one of the discussions was how data collection is changing from a relatively few manual stations around the world to using the data from 100's of 1000's of pws and other stations and how to build algorithms that could sort out the huge variation in the accuracy of the data and deal with the volume.

Looking at a wunderground map it is remarkable how consistent the readings are across numerous stations, but invariably there are outliers which are feeding the results from failed sensors to wunderground. I have one reading which records 13000mm of rain in a day and have been unable to remove it.

I think we all add to that data in a positive way and all contribute to this mass of data whilst also adding a little to the confusion for good measure.  

6 weeks to go before I head your way, and maybe a weatherduino station will appear in Beach Haven.

I have converted a Davis Vantage pro weather station as the old electronics was failing.
The solar sensor seems to be fine so I am about to start testing it with the Weatherduino.
This should give me far better solar data. Watch this space.  
Are you shifting here?

[JT118]

No unfortunately, but my daughter has become a Kiwi so we come out when we can. Below may be useful for the Davis sensor.


https://www.meteocercal.info/forum/Threa...avis+solar