Current limit on ATX molex lines?

[mathog]

I tried to run a fan (with a built in control board) with the 12V from a 4 pin ATX molex connector. The fan has a PWM input and that wasn't attached, so it should not have tried to start. Nevertheless, as soon as the molex was plugged into it the power supply shut down. Presumably the fan was drawing too much current. The fan works if plugged into its normal driver, as far as I can tell it is working normally. I have no idea how much current that fan normally draws, it comes out of a car. It is a squirrel cage about 6" in diameter and 1" thick.

How much current must that fan have pulled to cause the ATX to shut down?

I can cut up the ATX supply to get 12V from other places. Is it worth trying all the auxiliary connectors, or should I just cut up the main connector?

[SteveNielsen]

Re: Current limit on ATX molex lines?

Yesterday I had a similar experience with motor taken from a hard drive, which I assume now is PWM driven. I connected it to 12v from a drive connector on a small ATX PSU I use for bench experiments, and it did a short wiggle and stopped. Removing the 12v had a nice little spark to it. I was able to do this three times, trying different wire combos with the 3 leads on the drive until the PSU shut down on the 3rd attempt. Once allowed to cool the PSU is working as before so hopefully no permanent damage to it.

Does the PSU have the current or power ratings for the various output voltages printed on it? In my case the +12v output of the PSU I used is rated at 4 Amps.

[RJARRRPCGP]

Re: Current limit on ATX molex lines?

How much current must that fan have pulled to cause the ATX to shut down?

For a 500 W PSU, that would have to be close to 40 amps. lol

[stj]

Re: Current limit on ATX molex lines?

it's not the amount that causes the shutdown, it's the sudden spike.
i used to have this problem powering my dremel type tool from an ATX supply and sometimes turning the drill off would trip the psu.

[mathog]

Re: Current limit on ATX molex lines?

Does the PSU have the current or power ratings for the various output voltages printed on it? In my case the +12v output of the PSU I used is rated at 4 Amps.

It does and it is 16A total for +12V. That would be split across the 5 yellow lines coming out of the case. It doesn't say what the over current protection is on a single +12V line. In another forum somebody had measured the maximum constant current the fan draws at max speed, and it was under 2A. However, they did not measure the power on draw, so stj may be onto something about the initial spike being too high.

The ATX power supply is a Sparkle Power International FSP300-601U.

[mathog]

Re: Current limit on ATX molex lines?

Found this reference (but not yet the sources it references):

http://brazenartifice.wordpress.com/...2v-regulation/

I've read that some ATX power supplies will shut down completely if too much current is drawn from the 12 volt lines without at least an amp drawn from the 5 volt line too. From first-hand reports, that tends to happen part-way through long prints. I'm pleased to think that shouldn't happen to me.

That would fit what happens pretty well. Maybe I should just look for a 12VDC wall wart, those things should not be nearly as picky.

[SteveNielsen]

Re: Current limit on ATX molex lines?

It does and it is 16A total for +12V. That would be split across the 5 yellow lines coming out of the case. It doesn't say what the over current protection is on a single +12V line. In another forum somebody had measured the maximum constant current the fan draws at max speed, and it was under 2A. However, they did not measure the power on draw, so stj may be onto something about the initial spike being too high.

The ATX power supply is a Sparkle Power International FSP300-601U.

All the +12v yellow wires are connected to the same place in the PSU. There's no "splitting" of the current with only one device connected.

In my personal experience it was not the initial spike that shut the PSU down, if that were the case it would have happened immediately. That is not to say it could not happen or never happens that way, it just did not for me.

[mathog]

Re: Current limit on ATX molex lines?

All the +12v yellow wires are connected to the same place in the PSU. There's no "splitting" of the current with only one device connected.

Useful info, thanks.

This is making less and less sense. I had two junk 1.34A 12V fans. The connectors were cut off both of the junk fans, they were wired in parallel, and then attached to the same Molex connector. When the power supply started they both spun up and the power supply stayed on. This is without adding a 5V load resistor. That's 2.7A and not even a hiccup. Replaced one of the connections with a probe and touched it and released it a few times. Nice spark each time, and no complaints from the PS.

As far as we know the fan in the car only pulls 1.6A when it is going full bore. We don't know what it pulls when the car first comes on, that initial current spike could be quite large so long as it was of short duration. It isn't like the car battery is going to care about another couple of amps for a millisecond or two. (The wires connecting the fan in the car's harness are pretty small, maybe 20 or 22 gauge, so clearly there is not much current going through them normally.) The power supply must care about this startup current though, which is why it shuts off. The fan in the car is working normally when its +12V and pulse control come from the car.

[RJARRRPCGP]

Re: Current limit on ATX molex lines?

Useful info, thanks.

This is making less and less sense. I had two junk 1.34A 12V fans. The connectors were cut off both of the junk fans, they were wired in parallel, and then attached to the same Molex connector. When the power supply started they both spun up and the power supply stayed on. This is without adding a 5V load resistor. That's 2.7A and not even a hiccup. Replaced one of the connections with a probe and touched it and released it a few times. Nice spark each time, and no complaints from the PS.

As far as we know the fan in the car only pulls 1.6A when it is going full bore. We don't know what it pulls when the car first comes on, that initial current spike could be quite large so long as it was of short duration. It isn't like the car battery is going to care about another couple of amps for a millisecond or two. (The wires connecting the fan in the car's harness are pretty small, maybe 20 or 22 gauge, so clearly there is not much current going through them normally.) The power supply must care about this startup current though, which is why it shuts off. The fan in the car is working normally when its +12V and pulse control come from the car.

I dunno if there's a thing as finicky as the GFCI hooked up to the washer in the room off the kitchen.....

[SteveNielsen]

Re: Current limit on ATX molex lines?

The 1.34A 12V fans are not pulse driven though, are they?

I am not greatly knowledgeable about PWM motors but my guess is when applied a steady 12v across the phase coils of the motor it is unable to convert the energy to motion and drew too much current from the PSU. I think in a PWM motor the pulse durations and timing produce an effective operating current of much less than a steady voltage would.

BTW, You're making me read too much mathdog