[seth22]

im using mcp6241 pdip by microchip
http://docs-asia.electrocomponents.com/w...381350.pdf

it night time now, using focusing flash light, my v+ measures at 5.4mv, while v- at 10.3mv. outputs 80.2mv, im not sure where went wrong

[werk_ag]

... my cf is also rated at 1v 100mA under full sunlight

Hi seth22

A Rsh of 1.5 Ohm is fine for those pv cells.

Maybe, I like to make the things more simple than I should, or maybe I should keep in silent...
I can't say if the OP you chose is suitable for the task or not, this is up to you, I just can try to give some practical ideas.

You said some posts above that you have access to real pyranometer. Right?

If you want, try this:
- Instead of a fixed 8.2K (it seems 8.2K) resistor use 10K variable resistor.
- Keep the ADC readings with a 1.1V reference voltage.
- Add this to your code (supposing you are using an Arduino):

Code:

SolarRad= map(ADC, 0, 1023, 0, 1400);

where ADC is the reading from your ADC
- Check the Solar Radiation value you are getting from the real pyranometer, and adjust the 10K variable resistor until you get from SolarRad the same reading (or near). Continue checking the readings along the day, or plot a graph if you want. The 10K pot may need some small adjustments or even the used formula.

If in the dark, you get readings greater than 4 to 5 W/m2, maybe your OP is not very suitable for the task, or you are using too long cables between the OP and the ADC.

If this method seems inappropriate, just don't try it

[seth22]

hi werk

i now do not have access to the pyranometer as i'd returned it to my friend already, probably no where near to borrowing it in the next few months. maybe i'll try to redo with a lower resolution first (ignore op amp) and think about using a different op amp next time

[werk_ag]

In my humble opinion, trying to do it without an OP is useless. The voltage drop at Rsh is too low to be read. I believe, that at least for tests, it is preferable use any OP than none.

[seth22]

so i should try to make the gain higher to check if the graph goes higher?

[AllyCat]

Hi,

using focusing flash light, my v+ measures at 5.4mv, while v- at 10.3mv. outputs 80.2mv, im not sure where went wrong

A flashlight is probably not bright enough. The output of a "rail-to-rail" Op-Amp cannot go absolutely down to zero volts, but 80 mV does seem a little high. But it will depend on exactly what load it is connected to. The 10 mV at the - input is consistent with that output voltage, but the PV input has not reached that voltage, so the Op-Amp cannot "balance" and amplify anything.

Is it a LED or "wire filiament" bulb in the flashlight? A filiament lamp (particularly mains powered) is generally preferable for tests, because it puts out lots of IR, which the PV panel is (or should be) most sensitive to.

If you increase the Op-Amp gain by very much, then it (or the ADC) will overload when you use real sunlight.

Cheers, Alan.

[werk_ag]

hi werk

i now do not have access to the pyranometer as i'd returned it to my friend already, probably no where near to borrowing it in the next few months. maybe i'll try to redo with a lower resolution first (ignore op amp) and think about using a different op amp next time

If now you don't have access to the pyranometer data, you can try to find some nearby weather stations with reliable Solar Radiation data, and try to calibrate your device with those data. This should be done only in sunshine days, with a clear sky, otherwise in cloudy days the readings significantly change from place to place.

[seth22]

hi all, sorry for disturbing again. my area (or state) doesnt have any weather station. if i were to do calibration on the cf, without having access to pyranometer and weather station, how should i proceed?

[AllyCat]

Hi,

Basically, you need to wait for a day with "clear blue sky" and create a graph with a shape as shown in post #1 of this thread. Then you can compare the peak level with the "maximum solar radiation" curve shown for the specific date and your location, using software such as Cumulus (in post #1) and apply a calibration factor as required.

Cheers, Alan.

[seth22]

Hi,

do i reduce the gain if the data always hitting max limit, ie 1023 very frequently?