Let's Make Robots!

solar trickle charger-solar engine

Hi, can anyone help?

I am trying to make my robot battery powered with trickle charge with two solar panels.

So far I have just tried 6v solar connected to battery and circuit with diode protection but this doesn’t work.

My next thought is to have two battery supplies, switched by relay when one is running low and the one not being used is trickle charged with the solar.

Trouble is I have been reading that the batteries wont charge until the solar voltage is above the total battery supply (5v). I live in the UK and on a gray day (which is most of the time) my solar panel only produces 3-4v.

So now I think I have to add a solar engine which apparently the solar panel charges a capacitor to the desired voltage (in this example 6.8v) and then releases it to a load (motor or light), but in my case, battery pack and then stops for the cap to charge again..
I and looking at the
"Chloroplast" solar engine. Circuit attached.

my previous thinking was that capacitors only charged up to a maximum of the same voltage that it was receiving i.e. 3-4v

My questions.
Is a solar engine the right way to go? Or does anyone have a better ideas.
is there a better or more suitable circuit as some of the components in the
"Chloroplast" solar engine
are US parts? Or does anyone know the European equivalent.
I thought Caps only charged up to a maximum of the voltage that it was receiving i.e. 3-4v. is this true or will it charge up to the 6.8v?

Thanks to anyone who can help.

solar_engine.gif10.9 KB

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Keep the voltage of the charger arround the same as your packs (or little above) Like oddbot said, 6v would be just ok for a 4.8volt pack. Slow loading a nimh pack and your looking at 100mah - 150mah.

I guess you need to check on what actually comes out of the solarpanel. You mention 3v a panel and 12v for 2 panels?

Probably your best off putting them in parallel, giving you maybe 4-5 volts (to charge a 4.8 nimh it'll work, maybe you wont be able to max it but hey). It would be charging the pack with 100 to 200 mah (max) which is fine.

Parallel and series, keeps confusing me to as an electronic noob. Check this:


I'm working on a thing like it (and some others), great way to get all you can get from the solarpanel, mount them on a bot that seeks the light :)

Considering you live in the UK and as you say the weather is grey you will probably need reflectors to focus more light onto your solar panels. Aluminium foil on cardboard would work. This would let you get more power from the solar panels you have.

could a person use a comparator powered by the running pack to test for a given voltage and then enable a regulation circuit when necessary?

thanks all for you help on this.

re comment from Chris the Carpenter. trickle charge, i'm not to worried about time as long as in charges eventually....i'll cross that bridge at at a later date if its an issue.

i chose to go parallel because series will give me around 12v 100ma the battery packs i was planning is 4.8v (4x1.2v aa) 2900mah.
and i thought that would be to much for a 4.8v battery. what do you think?

Jax, i chose a Chloroplast engine over a Miller solar engine only for the fact that it pulsed at 6.8v, all the miller ones i saw was 4.8v which i thought was a bit low to charg the batteries. thanks for the info though. cant really do bigger pannels as i am limited on space. this bot is going to live in a hamster ball about 5inches dia.

telefox, cheers. i'll get the meter out and also check on smaller voltage/higher amps solar panels.

i'm quite liking Chris The Carpenter's idea as it's simple,but when in series will give me around 12v 100ma the battery packs i was planning 4.8v (4x1.2v aa) 2900mah. is 11.4v after the diode to much?


Ah yes, 6.8V is way higher than a stock MN1381 is rated for. You can set it up to go off at higher voltages but it's bad for efficiency.

You can run the 11.4V and have a simple voltage regulator handle the sunny days but I think your very first step should be to get an ahmeter on your panels to figure out exactly what kind of current you are producing. I've shelved my solar projects for the season because of the lack of amps. Without proper illumination there is no worthwhile current generated. I see you're content with a long charge time but it might be longer than you expect.


You said a couple times that you wanted to trickle charge you batts. Then you went parallel to get some current, but you didn't get the volts you needed. Seems simple to me: Get the panels in series. Done. Do whatever you got to do to get the volts you need and in terms of the current, I say just use whatever you are getting. Worse thing that could happen is a lot of waiting for your bot to charge. Easy solution: Just charge at night when you are sleeping. This way you will not be wasting your time watching it charge. --Now that's thinkin' right there...

worth a thought. but still in low light it might not make the charge voltage i need to charge the batteries.

im actually want to have two battery packs, each consting of 4 x aa nimh batteries. whilst 4 (4.8v) powers the picaxe (robot) cuicuit. the other 4 will be recharging. when the active batteries runs low a relay will kick in and switch the solar and batteries.

i have been eyeing up the

Pololu Adjustable Boost Regulator 2.5-9.5V

do you think that will bo the job or do you think its a bad idea?

That Pololu boost converter is nice, but the efficiency suffers terribly as the current levels drop.

If your input current is 200mA at max. rated value with 2 panels in parallel, then the current you get when the voltage is only ~3V will likely be less than half that, so not even 100mA is going to the boost converter. The output of the converter will in turn be roughly half as much again, at which point the boost converter cannot run very well at all.

Grab a multimeter and see how much current you're getting from your panels at 3-4V, at least then we'll have some complete figure to work with. However I imagine you'll either need bigger panels to get the voltage up as JAX has suggested, or you'll need some more efficient smaller panels that can generate enough current to feed a voltage booster like the Pololu boost converter.

A couple points:

If you can't generate the rated voltage of the solar panel there surely won't be enough current to charge your batteries in a timely manner. The rated voltage of solar cells can be acheived with a good deal less than full sunlight. The rated current output will require the equivalent of full sunlight..

If you use a voltage booster you still won't get around your lack of current. If you can only make 3-4V on your 6V/200ma panel you are probably only generating a few milliamps at best. If you have a 1000mah pack you're facing hours and hours and hours of charging.

A solar engine can't make more voltage than the solar cell can generate. You can only produce bursts of higher current..

What made you favor the Chloroplast engine? A Miller solar engine is more efficient and cheaper (the last time I checked). IIRC the MAX8212 makes a better (and voltage programmable) SE as well but uses more a few more components.

That said, I think the solar engine is probably not a sound idea. If you can make the charging voltage necessary you should just skip the SE entirely and work with the current you have instead of losing some of it in the SE's components. It sounds like you already have a challenge getting enough voltage and IMO at least a C/10 recharge current out of your current solar panel selection.

My suggestion sucks, but it is to get bigger panels.

I have 2 x 6v 100ma panels so i thought to put them in parallel to get as much current out as poss. but i still need some voltage boost circuitry to guarantee 6v at the battery as the solar panel will produce a varying voltage depending on the weather and of course I lose the 0.6v at the diode. I am not informed enough in electronics to design this circuit myself but I’m sure someone must’ve already done this and can point me in the right direction?

running a picaxe circuit and motors which need 5 <> volts