loading

will putting five volt regulators in parallel reduce the strain on them?

I'm using a peltier plate and some CPU fans to make a micro fridge for my car. for testing purposes, I'm going to need a pretty beefy power supply. because the plate draws four watts at twelve volts, I need a supply that can push out around three and a half to four amps. however, I don't want to go out and buy something, because I have a very limited amount of money. what I do have, however, is a large amount of leftover twelve volt regulators from an earlier project, rated to work with up to one amp. will connecting several of these in parallel (maybe five of them, just to be sure) be enough to reduce a thirteen and a half volt power supply with 3.9 amps of current to twelve volts? this is the closest I can get to a twelve volt supply with the materials I have, and I already have a heatsink for the regulators (I tore apart a dead computer power supply).

codongolev (author) 6 years ago
I did it despite warnings against because I had a surplus of regulators from an earlier project and didn't care about blowing them. it worked. I don't know why. maybe when they get hot they have more resistance or something, evening it out, or maybe it has something to do with the fact that the resistance is variable. whatever the case, I've had it running for a day and a half and it hasn't burned up yet.
kelseymh6 years ago
I don't think you can do that without some supporting circuitry to keep the current balanced. Current follows the path of least resistance, so whichever one of your regulators has the lowest input resistance is going to try to draw all the current, and get blown out. Repeat for the four remaining ones, and you'll eventually be a less than happy camper :-/

A Google search for current balancing circuit brings up a couple of useful schematics on the first page. They may be adaptable to your needs (treat each of your regulators as one of the "LED chains" in the example).
codongolev (author)  kelseymh6 years ago
I'm not sure... I've made plenty of flashlights with LEDs and it seems putting them in parallel would have the same effect due to slight differences in production, wouldn't it? besides, all the regulators are the exact same model, so even with slight variations, they should do the same thing as the LED example, right?

I might just try it... I mean, I bought a bunch of regulators because it was cheaper to buy a package of ten than just to get one. :P
I think the difference is that with LEDs, they fail due to over-voltage (bias-induced breakdown), not over-current. With the LEDs in parallel, they're all at exactly the same voltage (that of the batteries), and each one can draw whatever current they need in order to operate.

The regulators, however, take whatever current is pushed into them (technically, whatever the load pulls through), and just convert the voltage from input to output. Since you've got a single load on the parallelized output side of the regulators, the current path is determined by the relative resistances.

However, I did make an incorrect statement, and rather misleadingly :-( By Ampere's law, the current will divide across the parallel branches in inverse proportion to the resistances. It is only for the case of one parallel branch having much lower resistance than the others (e.g., you in a bathtub along with your favorite resistive appliance) that the current will seem to "switch paths."

So long as none of the currents exceed the regulator's rating, you could have a stable situation. If you're close to the limit, though, unbalanced branching could push one regulator over the top, causing a cascade failure.