Why does this NiMH solar charger work?

[aniseed]

I'm new to electronics and have been looking into building a solar cell charger for NiMH batteries. I got hold of a pair of small solar panels rated at 5v open circuit voltage, and 100mA short circuit current. I confirmed this under bright sunlight with a multimeter. Each panel will easily exceed these values if the sun is bright enough.

I connected one of them to a pair of AA NiMH batteries in series, thinking that perhaps I could trickle charge them under the sun. But there was hardly any current flowing into the batteries. So I added another panel in series with the first, as in the diagram below.

http://myhabitat.net/t/naive-charger-circuit.png

This time, a current of about 100mA flowed into the batteries. The voltage across the solar panels was at 2.6V. I left my 'charger' out in the sun for a few hours. Then I disconnected the solar panels, and discharged the batteries through a load of 50R while making occasional measurements of the voltage across the battery terminals. In short, the batteries started at 2.46V. After 5 hours they were down to 2.37V.

Does this mean that this charger circuit actually works? I thought that even a simple charger that just delivers a steady trickle charge would need some kind of extra circuitry in there somewhere.

 

[NascentOxygen]

I'm new to electronics and have been looking into building a solar cell charger for NiMH batteries. I got hold of a pair of small solar panels rated at 5v open circuit voltage, and 100mA short circuit current. I confirmed this under bright sunlight with a multimeter. Each panel will easily exceed these values if the sun is bright enough.

I connected one of them to a pair of AA NiMH batteries in series, thinking that perhaps I could trickle charge them under the sun. But there was hardly any current flowing into the batteries.

Hi aniseed, http://img96.imageshack.us/img96/5725/red5e5etimes5e5e45e5e25.gif

I don't have a good explanation for why one panel failed to charge a pair of NiMH cells. Have you ruled out the possibility that your first panel was faulty?

It sounds like your final arrangement worked okay. Is 100mA close to the recommended current for charging your cells? NiMH cells are robust and tolerant, they work well in garden solar lights, and they tolerate widely different charging currents as the sun's intensity on the solar panel changes throughout the day. You just need to make sure that the charging current does not exceed the manufacturer's maximum value or you may shorten the life of the cells.

 

[vk6kro]

It is obviously working, but the results seem a little odd.

Could you take the solar panels out in bright sun and put your 50 ohm resistor across each one and measure the voltage? Try it with any other similar resistors you have. 20 ohms would be interesting.
Try drawing a graph of voltage vs current for each panel if you can get a few readings.

This may tell you if one of them is faulty. If they give different voltages, one of them may be faulty.

Solar panels usually deliver a fairly constant voltage as the current is increased but then the voltage drops rapidly above a certain current and it is then unable to deliver much more current. It will continue to deliver much the same current even into a short circuit.

If you wanted to make a working circuit, you need to put a diode in series with the solar cells because solar cells will conduct in reverse when there is no sunlight and this may flatten your batteries overnight.

 

[aniseed]

Thank you for the replies and suggestions.

I will take more measurements of the panels under load, see if any of them is defective. I will also try to charge the batteries from just the one 5V panel again, see if I made a mistake the first time round.

At any rate, now I have confidence that even this simple circuit is capable of charging the batteries.

 

[aniseed]

I have tested my two 5V solar panels, using a 20R load as recommended by vk6kro and taken a few readings. The results are in the attached .xls file. The results show that each panel is working well and has a nice linear voltage-current characteristic.

I have gone back and connected just one panel to the pair of NiMH batteries on a sunny day and I can confirm that the batteries were getting about 100mA. I conclude that my initial observation of a single battery not working was a mistake. Apologies for the confusion.

In case anyone is interested, I found that connecting the two panels in series as in my initial post is not as effective as connecting them in parallel. In the latter configuration I get about double the current flowing into the batteries, while the voltage across the batteries remains about the same. If someone can provide an explanation for this I would be very interested.

FYI, Each NiMH cell is rated at 1300mAh. I think that charging at a rate 10% of that in an hour, i.e. at 130mA, would be suitably safe.

 

[vk6kro]

I think you may have misunderstood that test.

With the solar panel fixed in one position, facing the sun, you should be able to place various resistors across the solar panel output and graph the voltage that results.

I think you put a 20 ohm resistor across the solar panel and then rotated the solar panel. This then gives a measurement of your 20 ohm resistor, but doesn't test the solar panel.

At no load (open circuit) you will probably get 5 volts or so.

With a 50 ohm load, you might get 4.7 volts. This would mean a current of 94 mA. (4.7 volts / 50 ohms)

With a 20 ohm resistor you might get 3 volts (3 volts / 20 ohms = 150 mA)

and you might get 200 mA into a shorted output.

Then graph that with current along the horizontal scale and voltage along the vertical scale since current is the independent variable in this case. If you have more resistors, you will get a better idea of the solar panel characteristic shape.

Anyway, it looks like your panels are working properly.

 

[aniseed]

vk6kro, thanks for the clarification. You were right, I merely moved the panel around.

I re-did the test using different resistors, and got the behaviour that you described. Attached are the results, in case anyone is curious.

 

[vk6kro]

OK, that is looking better.

Just a couple of things you might like to look at.

You don't need to measure the current. You can calculate it. You will notice that column 2 divided by column 1 should equal column 3, but it doesn't because this is a measured value. See chart below where column 4 is the calculated value.
You know the values of the resistors and the multimeter has resistance of its own, even on the mA scale, which makes the reading inaccurate.
R...V...I...I
20 0.43 18.3 21.5
51 0.96 17.7 18.8
100 1.72 16.8 17.2
200 3.24 16 16.2


You need to find the "knee" in the graph. This is where an almost vertical line from about 5 volts suddenly turns left and goes almost horizontally towards the vertical axis. This is the point of maximum power from the panel.

This means measuring an open circuit voltage and a couple of lower current values, if you can.

So, maybe 500 ohms and 1000 ohms if you can.

The currents in the spreadsheet seem a lot lower than before. Was it less sunny than before?

 

[vk6kro]

This is an estimation of the load vs voltage graph for your solar panel:

http://dl.dropbox.com/u/4222062/Solar%20panel%20load%20graph.PNG

The black curve is obtained from your measurements. The purple bits are estimated from the measured part. Brighter sunlight will cause the knee to move to the right.

Maximum power is obtained near the "knee" which is at about 4.5 volt and 15 mA. This corresponds to a load of 300 ohms.

You can read off the current that will be delivered to a 2.6 volt battery by checking where the 2.6 volt line from the vertical axis intersects this curve. Looks like about 16 mA for these light conditions.

 

[aniseed]

You are correct, vk6kro. It was a lot less sunny when I took the measurements in my last post. The sun was setting at the time.

I have redone the test, this time including 510 and 1000 ohm resistors, and under sunny conditions. Attached are my results and graphs.

In short, the knees are at 4-4.5 V (for 45-55 ohm, yielding 80-90 mA).

vk6kro, thank you for taking the time to explain all of this to me. I now have a much better understanding of solar panels.

 

[vk6kro]

Glad it helped.

Incidentally, with Excel, the column which is furthest to the left will become the horizontal axis in a scatter graph.

So, it is worth starting a new column for voltage to the right of the one for current.
You can then either cut and paste the voltage data into the new column or just say "=(data in voltage column)". Then drag down to fill all the squares.
You can then force the graph to have current along the horizontal axis. In this case it makes more sense to do this because the different currents are causing the change in voltage.