Is this a psu fault

niceplace

Weaksauce
Joined
May 20, 2005
Messages
125
I have a problem with my pc which started yesterday, i had shut the pc down and just as it was near closing down i remembered the way you do i had something else to do on it. So i pressed the space key to boot up again and nothing happened, i dont think it had got all the way through the shutdown procedure and i am wondering would that cause the proplem. I tried turning the psu off via its switch and letting the power discharge then turning back on again and pressing the power button on the pc, the fans would start for a few seconds then stop, i tried this a few times but got the same result. I then took the side and the front of the pc case off convinced it was the power switch at fault, i cut the two wires going to the switch and stripped some insulation from the ends so i could touch them together to see if bypassing the switch would get it to boot. But no it just wouldnt start. After repeatedly trying this i got it to boot but the monitor just had a black screen. After more of the same messing about power supply on off touching the wires together i once again got it to boot and i entered the bios to see if anything had been altered there but it was ok.
I then exited bios and to add insult to injury i was then presented with the black windows didnt shut down properly screen i thought i would take the option to go into safe mode as this sometimes cures things. I then restarted the pc and on it came i tried shutting down a few times the starting up and all was ok. But on trying to boot this morning i got exactly the same problem no boot.
The pc has a Hiper power supply in it dated 2001 with a 450w output this supply came with a power supply tester which i connected up to the supply and the 3v led was flashing rapidly as was the 5v one, does this indicate a problem. I have managed to get the pc booted and i am just going to leave it in standby in the hopes it starts from that state without all the hassel.
I really need to get this sorted quickly as i use the pc a lot so can anyone point me in the direction of what the fault might be.

Some info on the pc
I built it last year it has a Asus P8Z77-V mobo and a Intel Core i5 3570 3.40GHz processor.
 
2001 is a long time ago mate. It's probably just about that time.
 
If it is the psu it would be a relief as i was thinking could it be the mobo, all the hassel that would be. I am hoping someone may have come across the problem before and can say what it is. Another thought was the psu underpowered for the job and its finally had enough.

Forgot to say before the psu tester instructions are scant to say the least, it would have been helpful had they a few scenario,s and how faults would show on the LED,s.
 
Sounds like bad caps in the PSU. Also, why wouldn't you just unplug the power button and short the pins to turn it on? Depending on WHICH Hiper PSU that is, it could be high quality, or... lllloooooow quality. Buy a decent PSU. If your tester is flashing on 3V and 5V I would say it does indicate a problem. If it was normal, all the different voltages would flash the same, or all not flash.

http://www.newegg.com/Product/Product.aspx?Item=N82E16817182132 Check out this PSU. Gets good reviews, and is on sale.

http://www.hardwaresecrets.com/article/Rosewill-HIVE-650-W-Power-Supply-Review/1460
 
After my last post i put the pc into standby started it up about 4hours later and no problem coming out of standby.

Thermonuclear

After the above do you think it could still be a capacitor problem. When i bought the psu it was considered quite a high end model.

I am not quite sure what you mean by

"why wouldn't you just unplug the power button and short the pins to turn it on"

If you are talking about the power button on the front of the pc the wires were soldered to it also the back of it was sealed in solid type resin.
 
There is an aspect of "turning on" power supplies that requires a connection to keep the power supply running. I have not paid attention to the details, but a "fix" for testing is to use a paper clip to short the "control" wire to ground.
This does not appear to apply here.
The problems with using a 2001 power supply, if there is a problem, comes in two ways. The first is heat effects the performance of capacitors in a power supply over time. Depending on capacitor quality and internal design of the power supply, temperatures can affect caps. They may be fine to useless.
You can use a DIMM (Digital Multimeter) to check voltages. This is relatively simple. Checking ripple requires an oscilloscope and knowledge. Not usually possible.
The second is power usage has changed from the 3.3 and 5 Volt to 12 Volt.
Old power supplies are like: 3.3V 28A, 5V 34 A, 12 V 21A or similar. (this is from a Raidmax 450 using an old design, likely poorly executed) 2001 3.3 and 5V Amps would be higher.
Watts = volts x amps.
(lol Raidmax 450 is 12V 21A = 250 watt rating in modern terms, if it is not lying)
Modern power supplies will have like (430 Watt Corsair[more like a 380]) 3.3V 20A, 5V 20A, 12V 32A. The combined 3.3V and 5V is further limited to about 120 watts max.
A Seasonic 450 watt (SSR 450M Gold ) 3.3V 20A 5V 20A 12V 37A.
A 2001 power supply is likely about 50-60 % efficient, a CX430 is 80+ Bronze about 82%, Gold about 86%. or 88%. look it up, if you like.
Also a 12 year old fan may be wearing out, which could lead to higher temperatures which could fry power supply.

There are power supplies that old running fine, mostly server ones. Glad yours is similar quality.
 
Last edited:
I meant the end of the power switch that connects to the motherboard. That generally is not adhered with anything, just relies on the pressure of the pins in the connector. http://3.bp.blogspot.com/-DLbJYcJBQU4/Tfdza0o8WuI/AAAAAAAAABA/SNYk2bmHi4w/s320/connectedzm9.jpg Like this. Normally this can just be unplugged, and then you can short the pins like this http://pics.bbzzdd.com/users/mechBgon/powertrip.jpg, to power the PC on. Yes, I still think it's likely capacitors. Like wtourist said, a PSU from 2001 is NOT the same as one now. Video cards and the like didn't pull the kind of amps they do now, so it wasn't normally needed. In February 2003 the ATX 2.0 spec was introduced, because it was more efficient to power a PC mostly with 12 volts. http://en.wikipedia.org/wiki/ATX#ATX12V_2.0 Your PSU is older than that, and will have less 12V and more 3.3v and 5v.
It's also older than the 80plus certification. Although that is voluntary and subject to debate, it means that your PSU is very likely inefficient as hell compared to a modern PSU, like wtourist said. Again, capacitors do wear out and have a finite life. Especially if you are running a newer PC with it that requires a newer ATX spec than it was designed for, you're probably lucky it ran as long as it did. The caps would be subject to a lot more stress than they were meant to handle and likely just wore out.
 
Thermonuclear,thanks for the in depth reply very informative. I see what you are saying about shorting the pins now.

I made a mistake with the date of the psu its actually December 2004 still quite an age though. The model No.is HPU-4S425 and it was rated at 77% efficiency. Here is a link to the specs page

http://hipergroup.com/products.php?lv=3&cate=1&type=23&pid=7


I was positive last night that i had left the pc in standby yet when i came to use it this morning it was in the off state and funny enough it started no problem. I will do the same tonight, leave it in standby that is and see if its still in that state tomorrow.

Whats got me baffled now is why was there no problem starting today, dont you just hate intermitent faults.

If i have to buy a new psu thats ok but as you can imagine i dont want to do that, fit it and find the problems still there.
 
Still likely capacitors. They tend to be intermittent until they are totally shot, and even then sometimes the unit will work. Really hard to pin down without seeing the caps. Being a power supply, I wouldn't do that unless you know exactly how to deal with electricity. A PSU can give you a nasty shock, even unplugged. http://hardforum.com/archive/index.php/t-773383.html As they say here, don't short the terminals on any caps and normally you will be ok. https://www.google.com/search?q=bad+capacitor&oe=utf-8&aq=t&rls=org.mozilla:en-US:eek:fficial&client=firefox-a&um=1&ie=UTF-8&hl=en&tbm=isch&source=og&sa=N&tab=wi&ei=GsX7Ud2cNY_G4APnQg&biw=1440&bih=779&sei=HcX7UY7KCbT64AP13IHABg Examples of bad capacitors. If they are REALLY shot, you'll know immediately. Once again be careful, and I take no responsibility if anything happens. I did warn you. :D EDIT: Also, power the PC down and unplug the cord from the PSU to the wall. Hold the power button on the front of the PC for a little bit (30 seconds,) and it should discharge any latent energy in the caps. Might want to do it a couple times.
 
Thanks once again, i am going to take a look at the psu innards and also the capacitors on the mobo, see if i can spot anything.

Oh and dont worry i am well versed in working with electrical items if a little rusty but you will not be held responsible for any mishaps :D
 
Great, good luck. Check any expansion cards also. GPU, soundcard, whatever. Good luck, and let us know!
 
Took the psu out and checked the capacitors, one small one (couldnt see the value) looked a bit dodgy with a slightly curved top. Trouble was it looked like it would need major surgery to remove it. So i decided to reinstall and see if i could get it to run just to give me some time to order a new one. Needless to say i couldnt get it to run at all apart from for a few seconds. I decided to check the power switch out and it looks as if that could have been the problem as after disconnecting the cables from it and touching the bare ends together the pc started up and has run ok since, wierd. So i need to sort a new switch out and hopefully thats the end of it.
 
Why did you cut the wires? You could just pull the connector off the motherboard and short the pins with a screwdriver.
 
Just a moment of madness, not thinking something out before wading in at the deep end.
 
Back
Top