Author Options:

How to 'load' a CIS solar panel to maximize power output and efficiency? Answered

I have 4 decent CIS solar panels that, in series and in bright sun, will give ~16V, and can offer ~ 200mA shorted. However, if shorted, the voltage obviously sags to almost 0V, so basicly no power is delivered based on the fact that V*I=P at the other extreme, there is almost 4V open circuit, but essentially no current flow. Again, 0W of power delivery.

The voltage and current figures above are just memory and I did log the data at various resistances awhile ago. My intention is to charge 3 lithium batteries in series to 12.6V, that will then feed power to a modified car charger (basic buck circuit for downconverting 9V-18V to 5V)

Since the current is so low, can I maximize efficiency by connecting the cells directly to the battery pack without any additional circuitry other than a blocking diode? Or in order to speed up the charging do I need to add some complex control circuit that will intelligently match the impedance?

If the battery charging speed is based on current, does that mean a direct connection is best?


Got a pic.

They sound like garden light solar cells. If they are they are good for a charge rate of about 100 to 60 ma.

The load of charging a battery will give you a better current value.

The battery should be marked 750 mah or something like that.

At a rate of 75 ma and a little more it will take ten hours to charge the battery with just a diode.

Then you will need a shut off circuit when the battery is charged or you will need to monitor the charging.


I am not charging NiMh or NiCd batteries, I am charging lithium ion/polymer. Perhaps this one. This will feed a small 12V to 5V 1A buck converter extracted from a car charger like this one. It can charge with either a USB input (5V to 12V boost converter), or a separate 15V wall brick.

Even though the maximum current out of the cells is (I think) 200mA, but if I had a buck circuit which loaded the cells down to around 2.7V/cell, and converter had a constant current output, I think at low voltage outputs (when the battery is near dead), I could achieve slightly higher charging currents. (say, an output of 8.1V @ 800mA with an input of 10.8V @ 180mA) I do know how to charge lipo batteries properly.

I think your picks did not come through.

I play with a lot of solar light solar cells and your values match the garden lights.

What I mean is the way the solar cells act when charging a battery.

The voltage across a cell drops when it is charging a battery and rises when the battery is charged.

While charging the battery the current is 100 ma when the batteries are charged the current drops to 20 ma.

These solar cells are loaded all the time by the circuit.

The change in the voltage between charged and discharged is shown in these picks.



I do not have any pictures since I am designing and not prototyping. I will buy the necessary switch-mode charger if I know it will be more efficient in harnessing power from the panels.

My main question is, will the batteries charge faster by directly connecting the 4 cells in series into a reverse diode and after a shunt regulator, charge the batteries, or will incorporating a switch mode (no way in h3II I will use a linear device!)

here are the panels I am using. http://www.ebay.com/itm/4-X-CIS-Solar-Cells-3-2-V-DC-60mm-x-60mm-80-mA-25-Watt-Mini-Panels-/271020888710?pt=LH_DefaultDomain_0&hash=item3f1a1a9286 I was a bit off on the specs, as it is 4.5V per cell open circuit, (the 4 make 18V), and the current is in fact only 100mA shorted. I will redo my data logging of one of these cells at different resistive loads again so have some accurate figures to work with.

Wow to charge that battery you need 1.1 amp.

To charge that battery you need a lot of cells in parallel.

It might work better if you used the cells to charge a NiCad and then use the NiCad to charge the LiPo.

you dont need to charge it that fast, that is the maximum safe charging rate. I have decided my cells are not powerful enough to effectively charge the batteries. At 100mA (0.1A), it will take aprox. 22 hours, or about half a week in full sunlight. Realistically, without adjusting for the suns position, that can be upwards to a week, so it is only useful for keeping the battery charged, like a trickle charger.

Where I live we have months where the full sun average is 70 hours.

I like solar power but it is only useful for about 4 months where I live.

Wind power is much more reliable to the point there are wind farms all around the town I live in.

In the far back ground you can see many more industrial wind turbines.



Those wind turbines are an ancient design! None the less, electro-magnetic/mechanical solutions can be nearly 100% efficient, often higher than 80%, compared to the 20-30% efficiency of photoelectic systems. Since my phone charger has to be small and portable, I dont think that will be a good solution though.

Ancient technology for sure that is why I like this pic with the old windmill pump.

Yes I took that it is one of a series of published photos of mine.

Although the white wind turbine is only a couple years old it is not much different than the antiquated water pump next to it.

Well loaded in full sun they are 3.2 volt 80 ma .25 watt each.

All 4 in series 12.8 volts 80 ma 1 watt loaded in full sun. Should zener diode to 12.6 volts

They will charge 11.1 volt 800 mah LI batteries in twenty hours. ( 3 Cell Phone Batteries in series.)

It is a bit of a slow or pre charger for LI batteries of 11.1 volts 800 mah.

Shouldn't blow the batteries might take forever to charge them.


I suppose that due to the low power output, the solar phone charger I have in mind will be very impractical. I did buy a cheap solar charger, the the panel would not charge the pack beyond 40%, and that equated to 8% charge on my phone. (the unit I had was a complete china BS that was not worth the plastic it was made out of.)

you really need a proper circuit to look after lithium batteries

I agree, but since the solar cells are weak, I do not think they are capable of damaging the cells with overcurrent. Overvoltage may be a problem with a direct-drive, but can probably be eliminated with a simple shunt regulator (12.6V zener diode, etc).