Mosfet as switch or DC SSR ?

Yes. Though you stated it backwards. The larger voltage disconnected battery will have its negative terminal go below ground, forward-biasing the drain-body diode of the "off" NMOS.

Yes. See...

and

for how to do bidirectional switches.

I first did this on a custom LiIon charge/discharge controller chip.

...Jim Thompson

You could also do high-side with PMOS, but derive the gate supply from the highest battery voltage, thus insuring no "back" paths. ...Jim Thompson

Nope! Chalk that up to me thinking I could control the body placement as I do in custom chips. I'm doing a short-circuit fault detector right now... it got me confused :-) ...Jim Thompson

in.

nd

will

rse

How would you detect the charging cycle being over in order to change batteries?

Other than a simple float charger, I don't think I'd be electronically swapping batteries. Note that most chargers have an overvoltage sense. That is, if they see too high of a voltage (one sign of a defective battery), they don't charge the battery. Thus the on-resistance has to be low enough not to trigger the overvoltage sense circuit.

lead(reverse

Bodies? Oh, oh, so you's guys do that, too?

\ \How would you detect the charging cycle being over in order to change \batteries? \ \Other than a simple float charger, I don't think I'd be electronically \swapping batteries. Note that most chargers have an overvoltage sense. \That is, if they see too high of a voltage (one sign of a defective \battery), they don't charge the battery. Thus the on-resistance has to \be low enough not to trigger the overvoltage sense circuit.

I have a couple different ideas on using this. First is for my lead acid batteries in the lawn mower, 2 tractors, and seldom driven car I have. This would electronically swap the batteries connection to my battery tender, perhaps a battery swap every day so each battery would get charged every 4 or 5 days. Since a battery tender can be disconnected from one battery and connected to another without problem it shouldn't be a problem. This wouldn't be too bad with relays either but if I can do it with mosfets I would prefer it.

In the future I would like to make a multi-chemistry charger that can handle multiple battery packs mixed or matched.

RogerN

Just get a full-float charger (13.8V, current limit at say, 10A) and connect them all simultaneously, in parallel.

Have Fun! Rich

Great way to start a fire if one batery fails with a shorted cell.

I would maybe try it with fuses at each battery but I would still like to have a switch at each battery. That way each battery could be brought to nearly the same voltage before being connected in parallel.

I'm interested in the "one at a time" method because someday I want to be able to use it to process multiple batteries simultaneously. Port 1 might be a lithium Ion, port 2 might be 12V lead acid, port 3 might be 4.8V NimH pack, port 4 might be a 11.1V Li-Po pack.

RogerN

On a sunny day (Sun, 12 Dec 2010 09:20:03 -0600) it happened "RogerN" wrote in :

So what is wrong with diodes, diodes do not protect against reverse connection, but they do isolate batteries from each other, preventing one charging the other. You just need to add that .7 (Si) or .5 (Schottky) voltage drop.

In that case, you need more than just "one at a time". Each different type - LI, PbSo4, NiMH - needs a different charge rate/ charge termination. It would probably be better to do this:

DC Supply + ---+----------+-----------+----}}-----+ | | | | LI PbS04 NiMh Whatever Charge Charge Charge Charge Controler Controller Controller Controller | | | | Batt Batt Batt Batt | | | | Gnd -----------+----------+-----------+----}}-----+

If you want to use a micro, then a single charge controller circuit whose rate/termination is selected by the uc in conjunction with the particular battery type selected would be fine. The idea here is that, while there are many ways to skin this cat, you do need to control things differently depending on which battery is selected.

Ed