[seth22]

hi alan,

thanks for the information. today was cloudy, need to wait another day for calibration. is 0.3 ohm low enough for my cf cell? or can i use several low R resistor and take multiple results?

also, i'm using arduino uno that has 5V reference, how can i make the reference into 3.3V? if my cf is connected across 10R, the irradiance value is it simply the value from logger > convert into voltage and divide by 10R and correlate to the I-W/m2 reference?
thanks

[AllyCat]

Hi,

0.3 ohm is certainly low enough to emulate a "short circuit" load for the PV panel. But it will give only 30 mV which will be useless if using a 5 volt (or even 3.3 v) ADC reference voltage. Werk uses a 1.1 volt reference even with an Op-Amp gain of around 7 - 10.

With a signal as low as 30 mV fsd (full scale) you may even have issues with the "input offset voltage" of a typical Op-Amp, producing a suppressed (or raised) zero level.

Generally, you can't calibrate using theoretical or data sheet values, because you would need to know (accurately) the conversion efficiency of the PV panel (at the precise operating conditions) and the gain of the Op-Amp, etc..

Cheers, Alan.

[seth22]

for the op amp, only rail-rail that matters? does cmos/fet opamp or frequency, voltage affects the ability or accuracy?

[AllyCat]

Hi,

No, I can't think of much. Obviously it must be specified for 5 volts rail to rail operation and the input bias (and output) must "include the earth/negative rail". Some FET types may have a relatively poor (i.e. large, say > 1mV) input offset voltage which may be signficant if you're using small input voltages (e.g. less than 100 mV). Some Op-Amps may need external "stabilising" components.

If in doubt, post a link to the device in question and we may be able to see if there are any "obvious" issues in the data sheet.

Cheers, Alan.

[seth22]

they are alot of them, i find this the cheapest http://my.rs-online.com/web/p/operationa...s/0403018/
this is the whole list of sorted op amp
my.rs-online

[AllyCat]

Hi,

That particular chip (MCP6241) has an input offset voltage of +/- 5mV which may produce a rather large offset (error) of the "zero" luminance level, with the positive gain configuration used by Werk. It would be less of an issue if using a "virtual earth" negative feeback configuration as described earlier.

Cheers, Alan.

[seth22]

tried mapping current from cf with irradiance from pyranometer, however the trendline did not cross the origin 0,0; used ~1.5R across cf, intercepts at around -50 W/m2 at 0A

next i tried to collect the voltage and current value from my cf, and the resistance is not 1.5R anymore, it becomes around 4 ohm for all data, for instance
28.8mA - 113.8mV (R=V/I=3.95R)
31.8mA - 126.2mV (R=V/I=3.97R)
does this has something to do with the series resistance of the cf?

[AllyCat]

Hi,

Are you making those measurements with one meter or two (concurrently)? In general you can't use one meter because the presence of the meter affects the measurement. If you are using two meters, ensure that the voltmeter is connected directly across the PV panel only, not the ammeter as well.

Generally, the best method is to ensure that you know the exact load resistance and then you can calculate the current from the voltage drop across that resistance.

Cheers, Alan.

[seth22]

so i've finished inserting low R across my all my cf (~1.5R) with op amp gain of 6.86 (approximating 140mV drop at 0.1A); i'd obtained the data from the logger

using assumption:
max irradiance 1400W/m2, Vdrop 140mV and adc Vref @ 1.1V
[spoiler][/spoiler]

was a very cloudy day today, clear sky reached somewhere after 11.45am. irradiance meter shows around 900-1000W/m2 after 11.45am but my logged data still doesnt show increase in voltage (except for the yellow line, but it spikes too fast and drops at 1pm when the irradiance is still high up ~1000?

the logged data seems off, any idea whether the yellow graph was the only correct one?

[AllyCat]

Hi,

You will need to make some "spot" measurements with a multimeter (volts range). The + and - inputs on each Op-Amp should always be at exactly the same voltage (probably around 100 mV with fairly bright sun) and the output about 7 times larger, say 700 mV, relative to ground.

What Op-Amp are you using? I don't recognise the particular pin connections shown.

Cheers, Alan.