Author Options:

how do i make a circuit that switches batteries when one is dead? Answered

i mean that when one battery is out of juice, the circuit switches the power from that battery to a charged one. does anyone have an idea?


. You can do it with a single SPDT relay. Wire the coil of the relay to the main battery so that it is activated when the battery is charged. Connect the common (C) of the relay to your load, NO to the main battery, and NC to the reserve battery. Wire the battery negatives together.
. You may have to do a little searching to find a relay that will repeatably drop out at the proper voltage.
. But, as orksecurity points out, you don't really need anything - just wire up your batteries in parallel.

Much more elegant than mine, Nacho! But I think you still need a second latching relay to keep the second battery active once the first one's been replaced. Unless you treat the second as a pure "reserve."

. There's also the possibility that the primary battery could recover enough to re-energize the relay, so a latch is a good idea. To save power, I'd probably go with a mechanical latch.
. There's more than one way to skin a cat.

Alternatively, of course, you could just hook the two batteries in parallel and draw equally from both of them, which may yield about the same useful life...

I had no idea what sort of circuit or system he was dealing with. Do multiple lead acid batteries behave well under that sort of configuration?

reading what orksecurity said, that sounds like it'd work too, but mim kinda scaredto try that, since these batteries (NI-CD @ either 7.2 or 9.6v) get hot when they are being used already, and wouldn't this make them hot too?

i think i'll try kelsetmh's way first, and if i can't get it to work, i'll use orksecurity's.

i plan on having this a very low budget car, i have a full gearbox, a chassis, a body, and plenty of lights and so far i have only spent $10! (on the chassis and body w/ the gearbox)

Sounds like a cool project! If the batteries are getting warm already, that may mean that you're drawing more current than they're designed for. Putting two of them in parallel gives you twice the current capacity at the same voltage, so they should behave better.

I think I'd try that first, before embarking on a relatively complex circuit project.

oops. i spelled kelseymh's name wrong. my bad :)

Suppose that one battery delivers more current than the other, so that it (initially) discharges faster. Then it will drop to a lower voltage than the other, and will start drawing current, rather than delivering it.

Now that I think about it, I guess that system is negative-feedback, so that even if the batteries start out unbalanced, they will balance themselves automatically. Harumph. You engineers always know best, don't you....

Yep; self-balancing.

And as the djinn said to the camel, "harumph yourself."

Harumph. You engineers always know best, don't you....


Use a pair of relays. Relays draw very little current to operate.

The idea is that you wire both relays' control lines to the first battery; as long as that battery supplies power, it will hold the relays in a "connected" state (see below). When the battery dies, the relays will "fail" back to their "normal" state.

Use one relay in a NORMALLY OPEN (NO) configuration. With that relay, wire the output lines so they allow current from the first battery through to the circuit. When the battery dies, that relay will fail OPEN and the dead battery will be disconnected.

Use the other relay in a NORMALLY CLOSED (NC) configuration. With that relay, wire the output lines so they allow current from the second battery through.
When the first battery dies, this relay will fail CLOSED, and the second battery will begin supplying current.

With this simple system, as soon as you plug in a new battery in the first slot, both relays will be tripped and the second battery will go back to being a backup, whether it fully drained or not. If you want a latching switch, so the second battery remains active until it dies in turn, then you would need to use a third, latching relay in NC configuration, as an intermediary.

You wire the this third relay with its control lines coming from the output of the second NC relay above (in parallel with the rest of the load). Its output lines would then intercept the positive line from the first battery going to the first relay.

So what happens is that (1) the first battery supplies power until it dies. Then (2) the second battery switches into the circuit, supplying power. The second battery also (3) takes control, so that it is the only one supplying power, even if you replace the first battery with a fresh one. As soon as the second battery dies, (4) the circuit switches back to the first battery.

thank you :) i am intending on using this for a RC car that i am working on, and with my other car, i find that one battery dies too soon, so i figured that maybe two batteries would last double the time...

i do have a couple questions though,
1- are these two of the same relay or are they different? where can i get them? i have a 6v relay from a cable converter with 4 pins, will this work?

2- could you draw up a schematic? im sorry but all that sounds kinda confusing...

thanks a LOT for the help though! if i ever manage to debug the drivetrain in this car, this circuit will be put in it and when a slideshow is posted you will get credit! :)

The Wikipedia article on relays has a nice figure showing the schematic diagrams.

Most common relays are SPDT with five pins, and operate in both NC and NO. Two of the pins (the ends of the coil) are the input or control pins.

The other three are labelled (on the data sheet, if not on the actual relay box itself) common, NC, and NO. In my description above, you'd always connect one wire (usually the "negative") to common, and the other wire to NC or NO, depending on which way you wanted the relay to function.

I'll try to draw up a schematic for you; in the mean time, given the diagram for the SPDT from Wikipedia, try doing one from my description yourself (consider it a homework problem :-).

Thanks! I made it up while I was writing the comment. I'll be very interested to see how well it works, and whether Zack247 needs to use diodes for back-EMF suppression, or any of those other real-world complications.

Ooops, I probably shouldn't have said that out loud...

My only concern would be a nice fat cap for when the switchover occurs, depending on how fault tolerant the end circuit is...again, great stuff!