Battery protection

[Bosman]

Some are lower in AH rating than others. My assumption is that if I just do them in normal parallel (I am setting up a 12v system) then when the smaller ones go down the bigger one may be drained faster as it will try to charge the small ones. Is this logic sound?

If it is so, will the diodes I suggest in the pic help to protect them sufficiently?

Input much appreciated

[Ungrounded Lightning Rod]

You're burning an extra 10% of your power in diode drops.

You can cut that in half (with six more diodes) by hooking the + side of the bridge directly to one of the batteries and replicating the + side of the bridge a couple additional times for the other two batteries, i.e. instead of one blocking diode from the bridge + to the battery use three blocking/rectifying diodes, one from each AC wire to the + side of the battery.

However this still won't necessarily solve your problem.  While the diodes will keep the high-volt batteries from discharging into the low-volt ones, the batteries "clamp" the charging voltage from the genny.  The low battery/batteries will hold it down, keeping the high batteries/battery from achieving full charge as early/often as the low ones do.  The non-right-angle voltage/current curve of the diodes will produce SOME improvement over direct connection of the batteries.  But it won't be adequate to balance them.

Try bringing your batteries to best condition by:

 - Watering them

 - putting an equalizing charge on them

 - rewatering them (carefully) to exactly the full point.

then:

 - tying them togehter

 - running them for a couple days (to clear surface charge), and

 - measure their state of charge in all cells with a good hydrometer, once when near-full and once when partially discharged.

If the hydrometer says all the cells are about equal specific gravity both at near-full and partially-discharged (and their cells didn't drift out of equalization over a few days), they're tracking well enough that you should be able to leave them in parallel. If one of 'em has come un-equalized it is defective (high-resistance short) and should be recycled and replaced - then rerun the test with the ones that are left.  If they are significantly out-of-match you'll have to go back to a balancing arrangement if you want to keep them all in service.

[ghurd]

Not sound logic.

The smaller batteries will not get lower SOC than the larger batteries.

The small and large batteries will have the same SOC.

They are in parallel and will be at the same voltage.

The small batteries will not "drain faster" than the large batteries.

Notice ULR said high and low volt.  Nothing about rated AH.

A bad battery will make the system not work so well, mostly regardless of its rated AH.

G-

[Nothing40]

That is the same type of setup I've been using for a couple years now.Despite the extra losses from the diodes drops,it works very well. I used fast recovery/schottky diodes to limit the voltage drop.

My batteries are all different ages and sizes,I've even mixed some Nicd's in with my SLA's. Everything lives together happily.

Here's a rough sketch I put together showing the idea,same circuit as yours;

http://i199.photobucket.com/albums/aa154/Nothing40/SolarStuff/batt-setup.jpg

[Ungrounded Lightning Rod]

Thanks, ghurd.  I missed that he was concerned about the charging rate of different sized batteries.

Bosman:  The issues with paralleling batteries are all with voltage versus state-of-charge.  So what matters is if the chemistry of the cells is a very close match, not whether the cells are of different current capacity.  (Think of the cells of the battery with a higher capacity as being composed of more than one paralleled cell each of lower capacity and you should get it.)

If two batteries have VERY CLOSELY matched chemistry the same voltage will charge them at the same fraction of their capacity per unit time, and will peter out at the same state of charge.  So you're fine.

What matters is that the state-of-charge vs. voltage curve is very steep.  So even a small difference in battery chemistry - plate composition/alloy ratio, electrolyte dilution, temperature, aging, level of sulfation, etc. - can make a significant difference in the state of charge for a given voltage.  If those mismatch, the low voltage battery will fully charge while the high voltage battery is still not fully charged, the high-voltage battery will discharge into the lower voltage battery, etc.  This leads to either the low voltage battery being overcharged and damaged or the higher voltage battery being undercharged and sulfating.

So when you're setting up a parallel system fresh with new batteries yo'd like to use batteries that are not just the same brand and model but from the same manufacturing lot.  That way they start out about the same, age about the same, and swing temperature in about the same way, which minimizes the chance they become unbalanced and start killing each other.  (You also try to mount them so they experience the same temperatures and cooling air, because temperature also affects voltage vs. state-of-charge.)

But you CAN parallel batteries of the same type but different makes, models, sizes, etc. IF they're closely matched on the state-of-charge vs. voltage curve.  Then even if you get them to match initially you'll have to check them more often and more carefully, in case they age or outgass differently, to keep them in balance and catch them early when age makes some of them no longer good partners for the others.