So i was changing my PSU fan according to the directions i've read around the web and this thread.

However, affter changing the fan in my Directron Lite-on 300W to an L1A, something's gone wrong... horribly wrong.

First i tried connecting the motherboard to the PSU so i could see if its becoming hot. My initial exuberance at still being alive was quickly cut short by the fact that my system wasn't getting past the fans spin, FF code flash, all power down.
So i set to troubleshooting - connecting and disconnecting every component in the system, including the RAM. About the only thing i didn't touch was the CPU and HSF.

Then i tried shorting the green and black in the timeless way shown to thousands of n00bs daily. No go. The fan spun for around 1-2 secs, then shut down. Then i had to disconnect power and remove the shorting wire from the ATX connector. I tried both black wires adjacent to the green wire (left and right).

Now i'm at a loss. What could it be? The connector was a standard 2-pin connector that i plugged the Panaflo into. The fan spins, indicating nothing's wrong with it or the connection. Could it be that the PSU is looking for more current draw on the fan connector??

Confused,
Mark

Specs:
Directron-Lite-on 300W PSU (stock fan Nidec Beta SL)
EP-8RDA+
1700+ with SK-7 and Enermax adj fan (tried both with and without the fan connected to motherboard and with a 3-4 converter and reset the CMOS by pulling jumper)
1GB Samsung PC3200

did you try putting the original fan back?

uh-oh.... i don't have another PSU. Now i'm scared.

gonna put the old fan back.

Is there no chance that you shorted something along the way? It sounds like some kind of short behavior. What about the pins in the 20-pin ATX connector to the mobo? Are they all secure? Is it possible that one of them is loose and not making contact?

A power supply that briefly turns on and immediately shuts off is detecting a short or other internal failure. It is shutting down to prevent itself and/or your hardware from being destroyed due to excessive current flow.

Did you take out any screws inside the case doing this project? You may have a wayward screw lying across a power and ground line on the motherboard. Or there could be some metal lying loose inside the power supply itself.

Try unplugging everything, wait a day for any stored power to dissipate, then pick up the computer and slowly turn the whole thing over in your arms. Hear anything rattling around inside the case?


Generally, power supplies are the most dangerous and potentially lethal component in the whole computer. They have covers for a reason. If the power supply was just plugged in, be careful inside the power supply and do not touch any exposed metal, INCLUDING the heat sinks. The capacitors may still be storing a charge, you could get a hell of a shock if you're not careful. If there is something loose in there, leave it alone for a day to let the power dissipate before you try to remove the loose item.

-Scalar

Well, i took the whole system out because my case is really small and i had the whole system on a wooden coffee table the entire time.

I was working on a carpeted surface though, so i hope that didn't kill it

I did this same fan swap recently and encountered the same problem. In my case, the PSU fan connects via a 3-pin MB header, except that the wires are reversed. I opened the PSU back up, cut the little tab off of the header, turned the connector around, plugged it back in, and it worked fine. If your old fan also had a 3-pin connector, compare it to the L1A and make sure the wires are going to the same pins. If they are, then this was of no use to you. Sorry If it does help, great!

Aha, that's probably it.

Just because the power supply may have a fan connector inside similar in shape to the motherboard fan connector, that absolutely does NOT mean it is wired or used the same way.

There is no standardization for connectors inside power supplies, because generally you shouldn't be opening it up anyway and the supply designers aren't trying to make it user-serviceable.


-Scalar

I just swapped my PSU fan myself. The connector for the PSU fan was the same as my panaflo. Great I thought. But then I checked the leads and noticed that they were backwards. So just like Jaik said, check to make sure the wires are correct.

Well, not always. It might be due to insufficient current draw on some of the rails. Happened to me with an Enermax 365 something or other, 350W.

I encountered this once with a system I was assembling. After installing motherboard, CPU (P4 1.8A) and RAM, without any of the drives connected, I tried to power it up for a quick check. Fan spun up, then turned off immediately.

After some swearing and cursing, I tried connecting the hard drive. Powered up without a problem, and has been working fine ever since. Unfortunately, still trying to cut down on the noise, which is what keeps me reading the articles and the forums.

Mark, please keep us informed. I hope it's just the reversed fan header. Good luck.

Well, if the fan header was reversed, would the PSU fan still spin up? The older fan was also 2-pin, so rpm-monitoring is out of the picture.

I made sure the fan was plugged in the right way because i noticed i had to bend the tab a little before plugging it in and i also had to cut away the little tab from the Panaflo's 2-pin tail.

I also tried it with a hard drive connected and a couple of fans (one to 5V and the other to 12V) and it wouldn't power up.

I'm at wits' end - can't think of any reason why it wouldn't power up. There wasn't really anything to short-circuit in there.

Hi, Mark,

You're right, it's unlikely the fan will be spinning if it is reversed.

Without another power supply it's rather hard trying to guess what's wrong. I went back and re-read your original post. What do you mean by "FF Code Flash"?

Does the same sequence of events happen each time you tried to power up your whole system? Or do you need to unplug and replug PSU from mains before you can get it to even spin the fan?

"FF Code Flash" means that the whole board apparently powers up, then shuts back down immediately, only having power long enough to "flash" the FF code on the diagnostic LED panel.

I have to replug PSU from mains to get any activity. I have also reset the CMOS a few times. Should i try to remove the battery, in case the reset isn't happening because of the power? (The CPU was at 7x210 at 1.4v)

Never used a diagnostic panel before. Is "FF" normal?

Having to unplug the PSU definitely sounds like a power supply problem. I had to do that too when I had too little load on my power supply.

I don't think you need to remove the battery to reset the CMOS if the motherboard has a jumper to reset CMOS.

Your CPU setup is pretty exotic (at least to me)... I wondered if it might be something with the motherboard that is causing the power supply to shut itself off.

I'm afraid I've reached the limit of my diagnostic abilities. At this point I would start swapping components starting with the PSU to see what's at fault. Is there one you can borrow? Otherwise the Fortron 300W with 120mm fan from Newegg for $32 shipped seems like a really good deal.

Well, the damn thing works fine with the stock fan, except that the stock fan is LOUD. We're talking Beta SL loud. I could use it on my heatsink if i wanted to, its that powerful. And i don't have a midrange fan that draws more current than the L1A. Can the PSU have current-sensing so that it shuts down when there isn't enough load on the fan connector? Is there a way to combine two L1A's so that they draw double the current? (It would be Parallel right?)

Interesting. Does it power up without the fan attached? If so, then you could just run the fan leads out to a fanmate1 plugged into a mobo header.

It might be worthwhile to see what amperage of fan is needed to get the PSU working. Maybe try a bunch with different ratings. Running 2 or more Panaflos in parallel is worthwhile, but at 12V, they won't be that quiet.

Have you humored us by trying the header the other way? Just troubleshooting to eliminate it as a option.

Yes i did try it with the polarity reversed. The fan didn't spin up at all, while the rest of the system spun up and then spun down. Now i'm at wits' end - it doesn't even work with another fan that draws 0.23A!!! How the fuck were they able to make a goddamn fan propreitary?

A photo of the connector is coming in a few mins.

Hello Mark:

Is the stock fan thermally controlled? If so, and it just runs too fast under a light load, you could try moving the thermistor! Or, you could stick a resistor in series on the positive feed to slow it down...

Hi Neil. As you can see in the pic i posted, the fan only has the 2-pin connector and no thermistor that i can see. Plus its rpm always seemed to be constant. There's no thermistor on the fan itself, and nothing else attached to it. Don't know how the PSU can differentiate between the stock fan and the L1A, other than current draw.