Sign in with
Sign up | Sign in
Your question
Solved

PC shuts off an instant after turning on

Last response: in Components
Share
December 14, 2009 1:23:07 AM

As soon as I power on my pc, the fans start up, leds all light up, then about 0.5 seconds later it powers right back off. Hiting the rear switch off then back on lets me repeat the process for no reason.
Is this a garunteed PSU fail or is it possible it's something else?

Happened when I left my pc running for a couple days straight, came home and the pc was off, but my speaker system and alarm clock were still on so it wasn't a power failure that sparked it :S
So much for my download streak -.-

More about : shuts instant turning

a b ) Power supply
December 14, 2009 2:35:17 AM

1. Unplug PSU.
2. Find and remove the CMOS battery (looks like a coin).
3. Reset CMOS (jumpers or switch - depends on mobo).
4. Wait 24 hours.
5. Insert CMOS battery (remove jumper if needed).
6. Plug in PSU and power on.
m
0
l
December 14, 2009 12:52:56 PM

Will update with your suggestion in 24 hours :) 

My board is giabyte's MA790XT-UD4P (am3) if thats of any interest, I will also start it up with 2 sticks of ram instead of all 4 in case of some abnormal glitch ive been given the honor of having.

Thanks
m
0
l
Related resources
a b ) Power supply
December 14, 2009 12:57:01 PM

Sounds exactly like the PSU overcurrent protection kicking in. That doesn't help much though as that could either be a bad PSU, bad component, or the whole computer is fried. Definitely start with viper's suggestion though. After that you might have to start taking out critical components and seeing if the board can ever even get to the BIOS error beep codes. I think for Gigabyte you can just short two jumpers or wait an hour after removing the battery, but no harm in waiting too.
m
0
l
a b ) Power supply
December 14, 2009 2:24:08 PM

I had to wait 24 hours on one of my Gigabyte boards. I waited 3 hours to start with and that wasn't long enough.

I forgot to add this step but discharge the capacitors before you power it all back up.
1. With the PSU still unplugged and the CMOS battery still out.
2. Hold the power button for 45 seconds to 1 minute.
m
0
l
December 14, 2009 4:13:57 PM

K I held the button down. I'm assuming it's the psu, its a thermaltake toughpower 850w modular, my system only sucks 380 watts MAX (electronic energy meter in wall socket) and as I understood from reading up on a PSU review once, that means it's running in a terribly inefficient state, especially when im idling at like 120 watts :S

If needed, specs are as follows
Phenom II x4 3.2ghz (stock)
4870 with like 5% factory OC
4x 2gig ocz at 1.7 volts & timings slower than stock (7-7-7-24)
2x hdd
2x dvd rw
x-fi gamer sound card
network card
some cheesy esata card that sucks
fan controller
6x 120mm led
2x 80mm led
1x 90mm led
custom inserted 140mm in PSU, some super scythe fan, the circle ones, I had to replace the stock fan cause it made annoying grinding sounds, plus this one blows more air with less sound

I seriously thought all that would suck more watts than it does, only to realize later I only actually needed a 650watt PSU at best since il never crossfire or sli
m
0
l
a b ) Power supply
December 14, 2009 4:47:13 PM

It isn't necessarily very inefficient at the low loads. I know my new PSU can hold about 80% even at very low loads. Optimum efficiency does occur higher up though.
m
0
l
December 15, 2009 7:11:23 PM

Alright i'm a few hours late to plug it back in, but I did, and the problems remains.
m
0
l
a b ) Power supply
December 15, 2009 8:03:40 PM

Could have been a surge. At this point, i think you just have to start removing components and see if you can ever get it to boot. Start with the basics, as in only CPU, GPU, one stick of ram, and all necessary fans.
m
0
l
December 15, 2009 8:53:12 PM

I am having the same kind of problem with an Enermax PSU. I determined that there is nothing wrong with my build as I installed a different PSU and it works fine.

The PSU oddly enough works fine on another build.

m
0
l
December 17, 2009 1:55:43 PM

LOL, that happened to my old 4850, I thought it fried, so I went out and bought a 4870 on sale, gave the 4850 to a friend and it worked fine on his pc :S retested it on mine and it still didn't work...

I ruled out my ram, video card, all sata devices, all pci cards cards and fans, it's not like I have a spare am3 board or quad core laying around so I couldn't rule those out, but wtv.

I went out and bought an Antec Truepower New 650W Modular Power Supply on sale at NCIX for 79.99 cad last night, so I will see friday morning if the psu is the culprit :) 

Thank you TheViper and EXT64 for your help :) 
m
0
l

Best solution

a b ) Power supply
December 17, 2009 3:33:21 PM

I hope it fixes your problem. I hope all of you reading this have learned a valuable lesson: you should have tons of computers (and thus spare parts) laying around your house like I do to make troubleshooting easier :lol: 
Share
December 18, 2009 9:45:35 PM

Well there we have it, it was the psu -,-
Installed the Antec and everything runs like a dream again =D

I think I discovered a possible culprit, my cheap @ss Sunbeam fan controller has a burn mark in the molex connector, one of the pins are black/orange and part of the white plastic turned yellow/orange :S
Odd as it was always fully pluged in and secured, wish I could sue, that was a pricy psu.

What I would do with a stinger missile launcher man oh man...
m
0
l
a b ) Power supply
December 19, 2009 2:11:11 AM

the associate said:
I think I discovered a possible culprit, my cheap @ss Sunbeam fan controller has a burn mark in the molex connector, one of the pins are black/orange and part of the white plastic turned yellow/orange

Even before the IBM PC existed - nothing on a computer power supply's output causes damage. In fact, one standard test for every supply is to short all outputs together and power on. Nothing must be damaged.

Intel even defines how thick that shorting wire must be. A common myth - the load can harm a supply. Only if a supply is sold on dollars and greater watts. Therefore is misssing essential functions that were standard even 40 years ago.

More often, failures are directly traceable to manufacturing defects that may not be obvious for months or years later (ie the famous electrolytic capacitor failures due to counterfeit materials).
m
0
l
!