Cold bug???

MichaelM · 10 Dec 2011 at 16:20

Hey, my system has been acting up a little with the winter weather closing in recently. I have been plugging in the computer each day and I would have to give it a couple of minutes before pressing the power button before it would do anything.
I'm making this post now because today I have been waiting for it to come on for at least half an hour and it has finally worked now.
About 5 minutes before it came on I took off the side panel and put an electric radiator against it...

Today I took the effort of checking all power leads and using the motherboards onboard power button but in the end a bit of heat and the external power button seemed to do the job as normal.

Stuff to note -
I have had absolutely no stability issues with the system when operating.
It boots into windows with no issues when it does have power. No failed overclock.
Only the CPU is under water cooling.
When it fails to turn on, nothing is on. No fans, no motherboard lights, no PSU noise, nothing.
I have used the PSU in my last system without such issue. CPU, mobo, ram, watercooling, fans and SSD/HDD are the only new parts since this cropped up.
520 Watts is the most I have recorded being drawn from the plug so it is never over the watt rating.

I've heard of cold bugs but that is usually with minus temperatures and CPU related so motherboard and fans should power anyway?
Any ideas? I'll be happy to get a list of possible issues and troubleshoot them each day.

Plec · 10 Dec 2011 at 16:54

I would save your clocked profile and clear the CMOS and run the system at stock.

If the problem is eliminated you will know that the clock may need a little more vCore for the initial boot (or other tweak). My old 775 setup had a similar problem and it was BIOS tweak that rectified the problem.

If that doesn't work you may want to look at your memory settings - or try the system with just 2 sticks for a while.

What USB HW ahve you got attached?

Rik1254 · 10 Dec 2011 at 17:02

Probably isnt the case, but looking from an automotive perspective, the PSU is having trouble charging the capacitors to give the initial pump of power to turn on.
Id say this is more apparent because when you said when it doesnt power on, nothing responds, which tells me that there is no/ not enough power.

Do you have another PSU you could try?

MichaelM · 10 Dec 2011 at 18:07

Plec said:
What USB HW ahve you got attached?

Keyboard and mouse, that's it.

Rik1254 said:
Probably isnt the case, but looking from an automotive perspective, the PSU is having trouble charging the capacitors to give the initial pump of power to turn on.
Id say this is more apparent because when you said when it doesnt power on, nothing responds, which tells me that there is no/ not enough power.

Do you have another PSU you could try?

Yea, this is why I would leave it a few minutes first. Room temperature does seems to be a coincidence at least. I know lower temperatures increases conductivity in materials so should help with the flow but in the case of a battery or capacitor it means the energy is transferred out easier so its harder to keep a charge. Not that I think any of the hardware is outside manufacturers specified operating conditions. I'm afraid I do not have another atx compatible PSU.

Tomorrow I will remove two of the four ram sticks, if that doesnt work I will remove a GPU and again if that does not work I will remove the PWM connection from the mother header followed by the fans. That should be more than enough stuff removed to solve such an issue, yes?

Thinking about it now it might be the fans. 9 PWM fans daisy chained with power from the molex connectors and all using the CPU header for the control signal via 4 akasa PWM fan splitters. Each splitter is connected to the others cpu header. Would it make any difference to have the first splitters control signal connected to a fan instead?

Plec · 10 Dec 2011 at 18:56

If you suspect the PSU - and it's certainly a possibilty - i would remove all superfluous connections apart from CPU/HS, gfx and 1 stick of memory memory (detatch hdd, optical, sound, fans etc..) and check to see if boots first time. It's an easy test and if it booted you would know it's eitehr the PSU or a component you detatched at fault.

Did you eliminate the clock as a factor?

MichaelM · 11 Dec 2011 at 15:33

Got it to power on the first time today... Before I attempted to remove any hardware I looked at the extension cord which had the PC, monitor, speakers, router, laptop and an external HDD for the laptop. I removed all but the monitor and PC from it and it powered on the first time with no issues.

School boy error

MichaelM · 12 Dec 2011 at 18:29

Looks like I spoke too soon, its not starting up today either.

So I tried -
Different sockets on the extension (power meter showing 240v on socket)
Plugging the PSU directly into the wall socket (power meter showing 240v on socket)
Plugging it into 4 other wall sockets
Removing two sticks of ram
Then removing 1 GPU's 8+6 pins
Checked for damage on both sides of the 24 pin
Removing all but the 24 pin (modular PSU saves my sanity here)
Having only 24+8 pin
Using the motherboard power button

No luck so far with getting it on today, I stuck it all back together and will give it a few minutes and i'm sure it will boot up as normal.
Looks to me like a PSU RMA is the next plan of action, yes?
Would they send me out a replacement for me to try first so I don't have any down time?

Plec · 12 Dec 2011 at 18:41

Unfortunately, they wont send you out a replacement to use while you wait as they need to ascertain that it's faulty (if it is at fault.).

Have you tried doing a basic PSU test?

Below are links to Huddy's site which has a step by step guide on how to test your PSU using a volt/multimeter if you're interested:

'Use a multimeter to check your PSU' and one to 'How to check if your PSU is faulty'

It's not a definitive test but it may help establish that there's a problem helping the rma process.

MichaelM · 13 Dec 2011 at 19:08

I just gave it a go with my multimeter. I think its an issue related to the PS_ON (Green/com's) wire because I can't short it to test the PSU so no voltage can be read. Which would also explain why the computer won't do anything either, the PSU thinks the 24 pin is not connected as the circuit is not completed.
240v is coming through the kettle lead and I went through every wire in the 24 pin cable to make sure there are no breaks or bad connections. Whatever is going wrong seems to be inside the PSU.

Perhaps there is a switching relay controlling this communications circuit and it is malfunctioning part?

Cheers for the links and taking the time to reply