AVR Freaks

But that's not precise. The model is not a current source, it's a current source and a anti-parallel diode (d1) -- the voltage across the latter being the reason why a solar cell behaves as a current-limited voltage source.

Is it a flooded or sealed lead acid battery? Will it be cycled frequently? By 6V12A I assume you mean C=12 amp-hour capacity.

Flooded cells can be continuously charged at rates up to a couple C without damage as long as you as add water as needed to keep the electrolyte levels above the plates. SLAs will not lose water if the charge rate is less than C/3, 4 amps in your case, see e.g. http://www.powerstream.com/SLA.htm.

Those panels would put out the maximum 10 watt power around 18 volts, or ~550 milliamps, so that is the maximum current you would get with a direct connection to a 6 (or 12!) volt battery. For charging, looks like you could have up to 7 panels in parallel without any kind of charge controller.

But if you ever get to the full charge the voltage will keep rising and once above 7 volts (temperature dependent, see the link) the plates will corrode. If you don't want to monitor the voltage yourself then at a minimum you will want a charger that clamps at a lower "float" voltage.

Of course you would be wasting half the panel wattage with a direct connection, more efficient would be a buck converter to trade the excess voltage for more current. Most commercial ones will do a daily absorption charge before dropping to float, not so great if the battery is rarely discharged. ATTinyx5's driving a FET and bucking coil make great custom charge controllers.

Google Translate from ze Germans to English:

http://translate.google.com/translate?js=y&prev=_t&hl=en&ie=UTF-8&layout=1&eotf=1&u=http%3A%2F%2Fwww.mikrocontroller.net%2Ftopic%2F46903&sl=de&tl=en

Also, gchapman, I hope I didn't misinterpret your last comment. I wasn't intentionally aiming for The Pedantic Asshole of the Day award.

Well, if effective means efficient I think those are incompatible requirements. AGM batteries can last for 15 years if never discharged, but once the voltage goes below float you want to recharge as soon as possible to prevent capacity loss through sulfation. For solar charging that usually means by noon, after which the rest of the day's PV production is wasted.

I'd look at 2 15AH LiFePO4 cells in series charged to 6.9 volts. Costs less, capacity the same, chemistry just as safe, and letting the charge level hover around 50-90% over days and weeks will actually increase lifetime, compared to a full charge by noon and wasting the rest.

Well I was thinking of 15AH Headway cells but I can't find a current source for them, the factory made some production runs in February but things have been quiet since then. I got their 10AH cells for US$17 each a year ago and am very pleased with their performance, wouldn't hesitate to try them in an AVR application where they were recharged once a year.

If it is possible to use the battery for 3 months and then recharge over 9 months, efficiency would prefer a low self-discharge rate. With that requirement, first would be lithium ion, next rechargeable alkaline, then lead-acid in third place.