PC Power-on Issue

[Sudy]

Hi guys, I don't think we have as many PC builders around here as back in the day, but Mrs. Nym's machine is having an issue and I wondered if this seems familiar to anyone.

Got an old Alienware Aurora R7 beefcake system from 2018. Never had any issues with it beside an overactive/noisy CPU fan, but I was never able to identify a major temp issue and no stability issues ever arose so I ignored it. I think the PSU's fan died well over a year ago, but the system runs case-open in a well ventilated area and has never had an evident operational issue, so again I just left it.

Yesterday the system stopped powering on. Checked the cable and the surge protector, seemed fine. But the chassis power button was just unresponsive. I assumed it was a dead PSU, so I started poking around. I discovered what is evidently a self-test button on the back of the PSU. When I depress it, the PSU runs! And if I press the power button, the system boots per normal. But when I tried to power it up again after releasing the button, nothing. However I can press the self-test, boot up the system, and release the self-test and it does stay on, and at first glance seems to function normally.

So my initial guess was that something power-related went wrong and some kind of safety mechanism was tripped. My concern is that operating in this manner may be dangerous/lead to subsequent issues.

But then I powered down and tried again, and the system did boot per normal via the chassis power button, even though I wasn't depressing the PSU self-test! So now I'm thinking, go about business per normal until further problems/warning signs arise. But I'm thinking the system should be powered down to be safe when we're asleep/out of the room for long periods. Probably also a good idea to replace the PSU on account of the dead fan alone, but of course I don't want to if I don't need to.

Any ideas or advice? Thanks as always!

[Isgrimnur]

Unplug everything.

And PSUs can degrade.

[Daehawk]

Id worry about that PSU becoming a dust and fire hazard...and this is me..Daehawk saying that.

[Sudy]

Thanks,

Yeah, when I did the successful PSU self-test followed by a full boot it was with only the PSU plugged in. What I don't get is why it didn't power up regularly after I'd reconnected everything the first time, but now it is.

But I've definitely had PSUs die before. In fact they always seem to be the first component to go in my systems.

Yeah, good point that it's probably safest just to replace the PSU on account of the dead fan alone. But since it's run without it for so long without issue (and I only even noticed it through physical inspection), I didn't necessarily think that was related to the current issue at least. But degradation is definitely a possibility/concern. I wish PSU fans were easier/safer to replace.

[Sudy]

And then I couldn't turn it on again. But... then I could. So definitely an unresolved issue lol.

I did some deep dust/cat hair removal and got a lot of junk out of the case. The system has always had that really loud fan and suspiciously high CPU temp but I've never been able to find a problem other than assuming the system builder did a poor job of putting the heatsink on. But it's really behaving right now, even though there wasn't too much dust in the CPU fan to begin with. I've always wondered if the sensor was just wrong. Anyway, I don't think this is the cause of the power issue as the temps are what I always remember them being.

[Daehawk]

Fans aren't that hard to get to. Ive never replaced one fully but I have taken a new PSU and opened it to reverse the flow of the fan one time.

[gilraen]

The self-check is supposed to have diagnostic lights on the PSU - did you check those?

[Rumpy]

I had a power issue several years ago, and it was almost embarrassing. It turned out to be a faulty power switch on the case itself. The plastic stem that usually makes contact had broken due to flimsy structural construction, and the solution was to use caulking to fill in the gap to provide support to the button itself. It worked.

[hitbyambulance]

i have a PSU where self-test works fine, but it refuses to boot under _any_ load. i suspect the problem is a short in one of the modular cable connectors (i received this PSU second-hand, and one of the 'matrix grills' on the PSU where a lead goes in - could well be the motherboard one - is missing, and possibly there are connectors now touching when a cable is plugged in)