Author Options:

Home Electronic Project, Can Use Some Help :) Answered

Hey all I'm trying to hack together a few things to make a uninterrupted power supply unit. So far, I've got a 500W power inverter from 12V to 240V, and a car battery. Unfortunately, I also need a power supply capable of pumping enough amps into the inverter and battery to get the whole thing started. I've tried an old ATX power supply I had lying around (actually several :s and I figured if I can get at least a few amps out of it it'll power ONE computer at least) but as soon as the inverter is connected to the 12V line the power supply shuts down (even when it's not powering anything at all, just the inverter ATM). It's like some sort of safety mechanism, same thing happens if you short two wires across. I know there's nothing wrong with the inverter because I can plug it into a normal wall transformer (which fried as soon as I plugged something into it because the transformer only output 1A ), and it works fine powering a computer on the battery. Anyone know of a way of overriding this mechanism, or even better, a way of hacking together a cheap and nasty 240V -> 12V 50A power transformer? Maybe splicing two power supplies together, if it's possible? 'Shady


240V -> 12V 50A power transformer
Wow! It look like you need some serious power here, I think you will need to hack a microwave's transformer, try ask stasterisk, they might help you in some way.


10 years ago

I believe that this is a safety feature of ATX power supplies- out of the big tangle of wires, there is one pair that needs a resistance between them to convince the power supply that the motherboard is present and not on fire. You could always plug a motherboard into your power supply, that should keep it happy :) otherwise the "ATX PSU into bench power supply" Instructables seem to suggest connecting a 10 or 20 ohm resistor between one of the +5V rails and ground will work (it will get quite hot, dissipating up to half a watt). Point to note: my UPS at work seems to run mains straight through it when not charging or running off the batteries, it charges the batteries for a few seconds a day to keep them topped up and only kicks in the inverter when mains power goes down. This presumably needs some fairly sophisticated jiggerypokery to start up the inverter in a few milliseconds of the mains power going down, and of course your battery may reach equilibrium and just maintain a steady charge through trickle charging. I believe your average car battery charger only puts out about 6A, maybe enough to power a laptop or uber-low-consumption thin client but not a full desktop rig with advanced graphics card and all the trimmings. A computer PSU may be your best bet.

I believe your average car battery charger only puts out about 6A...
Hardly. There are low-amp "trickle chargers" meant to maintain charge in batteries that are rarely used, but there are much more powerful ones available. At around 10 pounds weight are ones that can briefly supply 100A for starting, heavier ones can do 150 and up.

Look for one that's automatic and has a "deep cycle battery" setting, because really the battery used should be deep cycle if you want it to last long, regular car batteries are about pumping out lots of amps very fast for starting. My small one does both, for 12V there are 2A and 15A automatic settings, 100A boost, and also a manual 6V setting. The large one is for regular only and has a 50A "fast charge" setting, however fast charging reduces battery life.

Since 240VAC is what's being asked for, I found this forum topic from Australia that goes deeper into the subject with recommendations.

And why the heck would you need 50A going to the battery? You'd have to be planning on a lot of time without power quite frequently to justify having to charge the battery that fast. 10 to 15A should be plenty for "normal" expected power outages.

forgesmith: Fair point, I was going on my old cheapo charger that only indicated up to 6A (but might have delivered more). Remember, however, that it sounds like his plan is to step down to 12V, trickle charge the battery and invert back up to 240V from that. If your computer PSU is 500W that's 2A at 240V but 40A at 12V, plus inverter losses. The 12V current is going to be powering the computer as well as charging the battery, so it will beed some hefty current. I'm a little skeptical about running the computer's mains power through 12V though. There are better ways of arranging this, that I guess depend on your inverter- will it, for instance, "feed through" mains while connected and swap to battery power when disconnected? Ultimately unless you already have the charger, battery and inverter you might be better of getting a refurbished/second hand UPS from the interwebs and maybe replacing the battery.

I'm a little skeptical about running the computer's mains power through 12V though.
Actually I'm pretty sure that's how at least some UPS' work, input is line voltage that charges the battery, output is from inverter running from the battery, and that's it. Zero changeover time, nothing's getting changed over anyway, use this setup for sensitive things where you can't have any current bumps. A battery is an excellent power conditioner for DC, the AC output quality is solely whatever the inverter can do.

The argument is over the time involved. Let's say the PSU will be maxed at 500W out, add up the losses and call it 50A needed from the battery. The Cold Cranking Amps (CCA) of a battery is how many amps it will put out for at least 30 seconds at zero deg F. 400 sounds good for a new car battery, 400A * 30sec * (1 min / 60 sec) * (1 hr / 60 min) = 3.33Amp-hours (Ah) at maximum discharge. 3.33Ah / 50A * (60min / hr) = 4 min, however we're using less than 400A so at 50A I'd expect more actual Ah. Let's pick 10 minutes.

During a day you have 1440 minutes. If the UPS is used until dead once a day, that's 1430 minutes available for charging, if the power came right back on after it died. 50A * 10 min / 1430 min = 0.35A. Just a bit less than 2/5 of an amp, just over 4 watts, during about a full day of charging is all that's needed. Break out the solar panels. The amount needed for the battery is pretty minimal, it'd have to be discharged several times a day, or need to be used soon after it was discharged, to justify needing anything more than a few amps.

Since it's doubtful the PSU will be maxed out always, we don't need 50A@12VDC 24/7. Added up thru the day, let's say you need 50A for 4 hours. 50A * 4h / 24h = 8.33A. With the battery as storage unit, ignoring the UPS function and letting the charge vary, that's less than 10A needed. Add up computer and battery charging, for a system being used under "normal" conditions with "reasonable" expectations of UPS performance (if power's out more than a minute or two then you shut down), you can get by with only 15A.

If you think you'll need faster charging or your monster brute of a PC will eat up the kilowatt-hours, then you could consider more than 15A. But I sure can't see 50A needed. Plus, fast charging kills batteries, I'd limit it to 20-30A into the battery max. Now I can see where one might disagree on that, new vehicles can have 100A+ alternators. Well, back in the Old Days when Mavericks roamed the countryside, standard Ford alternators were around 35A, about the biggest system load was the blower motor on high, the batteries still got charged and cars still started just fine. Those large modern alternators are for powering the goodies. In any case alternator output is adjustable, check into it and you should find the battery gets charged with nowhere near that full alternator output.

(third comment this topic, hope I remembered to add everything this time...)
It's possible that there is an inrush of current into the inverter when it's hooked up causing the PSU to shut down. With a fully-charged battery in the system that shouldn't happen.

It is also likely this would never work with a PSU. For one thing, a car battery is really around 13.8VDC when fully charged. Inverters sense voltage, which drops as the charge is depleted. A PSU has a regulated 12V which the inverter (if it's a decent one) would sense as a battery that's nearly dead and would refuse to work with just a PSU. That voltage would also probably be too low to properly charge the battery. Also the battery will consume amperage until charged, without current regulation you'll overload a PSU immediately. And with the higher voltage and great amperage available you'd need diodes to keep the battery from feeding back into the PSU, likely killing it, which would reduce the actual voltage further (either diodes or dead).

Read what I posted here on a car charger, they're relatively inexpensive with practically zero hassle. I've yet to release the magic smoke from inside one, although Lord knows I've tried...

Oh, i've had this problem before! on psu startup, it sets the amperage limit, since you have nothing pluged in, it thinks something is breaking, so it shuts down... the way to get around this, put a diode letting power flow from the psu to the battery/inverter, but not vice verse. now, connect everything, INCLUDE the diode. switch on the inverter, then switch on the PSU. do it in that order.... a word of caution, use heavy gauge wires, i have had PSUs melt through speaker wires.... do me a favor, make sure you DON"T kill yourself, ok?

. If I understand what your saying, the inverter is working, but the PSU won't turn on. If so, the PSU is looking for a signal from the mobo that the voltages are OK. If the PSU is not plugged into a mobo, it doesn't get that signal and shuts itself down. . For an ATX PSU, jumper Pin 14 (green wire) to any of the ground pins 3,5,7,13,15,16,17 (all black) and the PSU should think everything is OK.