Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Solar Cell Measurement Circuit

Status
Not open for further replies.

atpalmer

New Member
I am a building a solar cell test station that will automatically generate an IV curve of a single solar Cell. I plan on using a circuit to choose among 5 resistors to generate an IV plot. I am only showing one resistor for simplicity. I will also record Isc and Voc. I need help choosing a transistor that will not significantly affect the measurement. If not possible than I will use the transistors to power a mechanical relay. This is a low budget project so ultimate accuracy is not essential.The data acquisition device is a 12 bit minidaq (MiniDaq A/D and I/O board for the IBM PC)
 

Attachments

  • SolarDaq.gif
    SolarDaq.gif
    8.5 KB · Views: 630
Are you familiar with an R-2R ladder network? You could use one as a digitally controlled Load (sort of poor man's DAC) where you could switch in binary-weighted load current steps...

You could also use a voltage-controlled (or DAC controlled) current sink with which you characterize your solar panel by selecting the load current, and reading the panel voltage.
 
Isn't the IV curve almost always the same? Only specs I ever see for solar cells is open circuit voltage in full sunlight and short circuit current. The IV curves are effectively irrelevant aren't they? You simply try to store as much of the current as possible while it's providing anything.
 
I am a building a solar cell test station that will automatically generate an IV curve of a single solar Cell. I plan on using a circuit to choose among 5 resistors to generate an IV plot. I am only showing one resistor for simplicity. I will also record Isc and Voc.
Will your design pull out the internal component values for the equivalent circuit shown in this link?
Solar cell - Wikipedia, the free encyclopedia
 
Thank you for your input.
Are you familiar with an R-2R ladder network?
I read the wiki on R-24 but don't quite get it. I plan on using 4 resistors in paralell that I can switch to get 16 equivalent resistances. I have a working analog to digital converter(minidaq) to record the cell voltage and current(via voltage accros the shunt).

Isn't the IV curve almost always the same?
No, my cells are from a few years of factory samples so they include the bad batches as well as good. Isc and Voc can't completely define the cell.

Will your design pull out the internal component values for the equivalent circuit shown in this link?
Solar cell - Wikipedia, the free encyclopedia
I will use the data to find the maximum power point, shunt and series resistance, fill factor etc...

My primary question is if I can use the transistor as switch without significantly affecting the measurements. I have never used a transistor before. my background is in mechanical engineering.
 
First of all, thank you for your help and tips.
I uploaded a more complee schematic.
Are you familiar with an R-2R ladder network? You could use one as a digitally controlled Load (sort of poor man's DAC) where you could switch in binary-weighted load current steps...

You could also use a voltage-controlled (or DAC controlled) current sink with which you characterize your solar panel by selecting the load current, and reading the panel voltage.
I'm not familiar with r-2r. I plan on using 5 resistors in parallel to get a 32 different resistive loads. I don't have any powersupply for current sink. I am measuring the cells under a fixed lighting source.

Isn't the IV curve almost always the same? Only specs I ever see for solar cells is open circuit voltage in full sunlight and short circuit current. The IV curves are effectively irrelevant aren't they? You simply try to store as much of the current as possible while it's providing anything.
No, these cells are from a mixed lot of good and bad cells. This schematic is designed to test the cells so I can build a panel with only the good cells.

Will your design pull out the internal component values for the equivalent circuit shown in this link?
Solar cell - Wikipedia, the free encyclopedia
Yes once I have the IV curve I can calculate the maximum power point, fill factor, shunt resistance, series resistance etc...

My main concern is whether or not The transistor can be used as switches without affecting the voltage measured at such low voltages (.020V-.900V) and around 40-200 ma.
 

Attachments

  • Solar Test.gif
    Solar Test.gif
    20.8 KB · Views: 291
Here is my suggestion:
 

Attachments

  • SolarTester.jpg
    SolarTester.jpg
    363.8 KB · Views: 391
Sorry mike but i don't get it. How could I measure the cell performance under load if there is a 1 megaohm resistor in series. The current at 1 megaohm would be 9 micro amps. I am meauring a singe solar cell, not a panel.
 
That's 1 milliΩ (0.001 Ω)! It is there just so I could show the plot of I(R6), the hard to read purple trace.

Spice uses Meg for megΩ
 
Last edited:
Like Mike said note the lower case m milli vs the upper case M for mega, this is standard notation anywhere not just with LTSpice as far as I know.
 
Like Mike said note the lower case m milli vs the upper case M for mega, this is standard notation anywhere not just with LTSpice as far as I know.

Actually, Spice uses "m" or "M" for milliΩ;
and "meg", "Meg", "mEg", "MEg", "meG", "MeG", "mEG", or "MEG" for megΩ :D
 
It'll use M for Milli? Okay that's not so good =\ Good thing I usually use meg
 
It's a standard SPICE thing. SPICE is case insensitive so they needed to come up with different ways to describe mill and mega, than m and M.
 
Current plot is inverted

MikeMl,
First thanks for a wonderful circuit. However, the plot/curve of current I(R6) does not match yours...It is kind of inverted for my case. What am I doing wrong? See attached for more information.
Regards
 

Attachments

  • solar.png
    solar.png
    163 KB · Views: 151
Last edited:
I still dont see the point in this test, all you ever test with solar cells is the open circuit voltage and the current with the cell shorted across the outputs.

The bit in the middle dont tell you a darn thing.
How will you guarantee the light on the panel will always be identical for the test.......you can not, so the test is useless.

You can plot the curve out from, volts at zero current (open circuit) to full current at zero volts (closed circuit)

Pete.
 
Since I did this professionally starting in 1980, the first key is to measure the current through the device and the voltage across the device. the next problem is a steady light source. GE ELH bulbs are not too bad of a spectral match. A little diffecient in the blue fo Silicon based cells. You need to operate the light source on DC.

An integrating converter will be best. The lower the sense resistor, the better.

You absolutely have to have a 4-quadrant power supply, One that can source and sink current, preferably with some sort of current limit. I made one with an OP amp and 2n3904/2n3906 drivers, a few resistors and a couple diodes. You can now easily get OP amps that can source/sink 100 mA.
 
I would do it with a single high power n-CH MOSFET with a low value current sense resistor in source. You can make a feedback control that creates a constant current sink. The current would be set by voltage control on the feed back op amp reference voltage input. For a single cell measurement, you should be able to drag the voltage down to at least 0.1 vdc.

Scan it fast enough and you won't build up much heat in the MOSFET.

If you don't control the temp of the PV panel the whole measurement plot will change the longer the panel is exposed to illumination, as PV panel heats up. You should have a temp sensor mounted to the panel to record the temp during the scan.
 
Last edited:
Regarding the lighting setup for measuring the solar cells, I'm not sure I agree with all the dire warnings about getting the right color spectra, etc. After all, all the O.P. is trying to do is to separate good cells from bad. It doesn't matter if the light source they use isn't as close to sunlight as possible, so long as the light output is consistent so they're always comparing apples to apples.

In other words, once they characterize a known good cell, they can use those measurements to sort the rest of the cells.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top