Repeated power supply failures

Toggle sidebar Toggle sidebar
B

bdgrs0

Reputable
Oct 28, 2015
7
0
4,510
So I've recently moved into student accommodation, bringing my 4-year old PC with me.
Within 4 weeks, the power supply that had lasted me all this time died, taking both my hard drives with it. I reseated every component, checked for shorts, assumed the PSU was just old as heck, and replaced it. I decided that I might as well upgrade, and so I also replaced my mobo and GPU. The only components left from my old PC were the case, CPU and RAM.

So the replacement PSU died 2 days ago, and I have no idea what I should do. 2 PSU deaths in a week obviously means something is up, and I was hearing static in my headphones before it sparked and stopped working, so I'm assuming there's a short somewhere in the case.

Do I just buy a new case? Is there any method for testing for shorts?
 
Sort by date Sort by votes
Old PSU was EVGA NEX750G Gold, newer one was Corsair CX750M. Since I assumed it was either a power surge or "old age" that killed my first PSU, I used a different power socket for my replacement.
 
Upvote 0 Downvote



I'll admit I didn't do much research into buying them, I'll make sure to do so with the next.
I am indeed using a surge protector, so I'm not really tempted to buy a replacement PSU since I still have no idea what's caused these 2 to die.
 
Upvote 0 Downvote

What sort of protector? A cheap $10 affair that simply has a few MOVs in it and that's it? A $30-40 unit like an APC SurgeArrest Performance or TrippLite isobar which have actual power filtering components? Something fancy like an SurgeX or ZeroSurge?
 
Upvote 1 Downvote

MOVs, the only thing you will find inside most surge suppressors under $50, don't remove noise. They clamp excessive over-voltage to something that will hopefully be survivable to downstream equipment.

A MOV with a continuous operating voltage of 130V won't conduct any meaningful amount of current until 200V, just above the normal line voltage peak to avoid wearing out during normal operation and wasting power. Beyond 200V, current rises steeply but the real clamping only begins around the 300V mark.

If power at your location has considerable noise in the 200-300V region and the noise is killing your power supplies, a basic surge protector won't save you. If you got a really cheap MOV-only protector, it might not even start clamping before 470V.

With a better surge protectors that have an LC power filter, most high frequency junk and line transients get snuffed before they even reach the MOVs. I named the SurgeArrest Performance and isobar because those can be found for under $40 and I know for sure they have LC filters.

Here's an example of what a $40-50 isobar gets you:
http://www.tomshardware.com/picturestory/684-tripp-lite-isobar-tear-down.html

And here's what you get with the better entry-level protectors:
http://www.tomshardware.com/picturestory/681-cheap-power-surge-protector.html

As far as power line surges are concerned, I would much prefer having the isobar over any of the $10-15 models... they are nowhere near the same league.
 
Upvote 0 Downvote
So along with replacing my PSU with one of higher quality, I'll also get a much better surge protector. Would it still be recommended to replace the case? I was getting static in my headphones, which suggested to me that there was a short in the case, as I hadn't had it before.
 
Upvote 0 Downvote
I seriously doubt the case had anything to do with your noise, unless there is an unintended short to ground somewhere due to an extraneous stud, rear IO plate ground finger stuck inside a connector or something else of the sort. Whatever case you get, the motherboard will still be grounded to it through the motherboard mounting studs, the rear IO shield, the power supply ground, HDD/ODD/SSD ground through the SATA cable, etc.
 
Upvote 0 Downvote
So much speculation and too few facts. For example, ATX specs are blunt about this. Short all PSU outputs together. No damage must result. In fact, specs even define how thick that wire must be to make a short circuit as big as necessary. Why do you know a short causes damage?

If a short exists, nothing in a computer is damaged. Because all current goes through the short - no currents in motherboard, disk drives, etc.

I do not see anything that even says why you know a PSU, disk drive, etc is damaged. That makes useful replies even more difficult. To obtain an informed reply, your conclusons must include details on why you know. Using a meter with a few minutes of labor would have defined what part is defective ... and why. Then reasons for 'damage' could have been identified rather than only keep replacing its victims.

Basic electrical knowledge demonstrates why shorts would not explain your symptoms. Your reasoning apparently was, "I had damage. Shorts do damage. So I must have had a short." That syllogism is invalid.

Furthermore, a properly designed PSU cannot damage other components (ie disk drives). A minimally acceptable PSUs come with a long list of numeric specifications. Using a meter, some requested instructions, and minutes of labor means numbers. Those numbers mean a next reply can answer all your questions. Without those facts, then every reply can only be wild speculation. Without those numbers, people who really know this stuff cannot provide a useful reply.

BTW, destructive surges occur maybe once every seven years. Protection inside a PSU is typically superior to anything that might be inside a power strip protector. Read its near zero spec numbers. A UPS typically claims ten times less protection. Again, useful answers do not ignore numbers.
 
Upvote 0 Downvote
Thanks for the reply.
The first PSU, when plugged into my motherboard, with only CPU and RAM attached, simply doesn't work. I power it on, and the fans in the case spin incredibly briefly, and then there's nothing. Then it doesn't do anything when powered on until I unplug it and plug it back in.

If a short doesn't cause damage, and a PSU failure can't damage components, what would explain my 2 dead drives? They're dead in the sense that, no matter what PSU, cables I use, they simply don't spin up.
Since I was afraid that turning it on might cause damage, I haven't actually tried my 2nd one since it turned off. There was a big spark, and a very strong smell of ozone, and the PC powered down and didn't turn back on when I tried right after. Is it worth trying again?

I have a 9V multimeter with me here, how would I go about checking the power supply?
 
Upvote 0 Downvote

If you need to completely power down the PSU before it will (try) to power up again, it likely means the over-current/power protection got tripped and locked the PSU out to prevent you from accidentally turning it back on and subject your PC's components to whatever caused the PSU to shut down repeatedly by making it far more inconvenient than spam-clicking the power button.

The ATX specifications say that the PSU must gracefully handle a short from any of its outputs to ground and from 5VSB to any rail. It does not say anything about shorts between the board and ground or supply rails not damaging the motherboard and its components. With the amount of energy stored in capacitors across the system, it is quite possible that any sort of momentary short will fry components faster than the PSU can react or even without the PSU noticing at all.
 
Upvote 0 Downvote

Restore everything as when the computer worked. AC power cord connected to a receptacle. Computer not on. Set a digital meter to 20 VDC. Attach its black probe to the chassis (bare metal; not paint).

Locate a purple wire (pin 9) from PSU to where it attaches to the motherboard. Use a red probe to touch that wire inside a nylon connector that attaches to motherboard. If necessary, make that connection using a needle or paper clip. It should read somewhere around 5 volts. Record that number to three digits.

Next, do same with a green wire (pin 16). Then press computer's Power On button. Monitor how meter changes and what it eventually settles to. First number should be something well above 2.6. Second number should be something near to zero, Actual numbers and time to change (behavior) are relevant.

Repeat same to a gray wire (pin 8). Note a higher starting voltage, a lower final voltage, and its behavior. Report those three digit numbers and behavior.

Setup computer to execute as much software as possible. IOW it should be outputting sound loudly, while searching the disk, while playing complex graphics (ie a move or game), while powering a USB device, while accessing the internet, etc. Having it access many peripherals simultaneously is important. If it cannot power up, then monitor any one red (pin 4,21-23), orange (pin 1,2,12 or 13), and yellow (pin 10 or 11) wire for what each does as and after its power button is pressed.

Report all three digit numbers from those six wires. Next reply will identify or exonerate suspects.

BTW, if wires are not colored, then a PSU may not be ATX Standard. See www.smpspowersupply.com for color and pinouts.


 
Upvote 0 Downvote
I'm not sufficiently sure I understand. Was the meter at 20 VDC and showing zeros? Did you connect its black lead to the chassis as described? Make contact with the purple wire using the red lead? The number remained at zero or was some other number displayed?

It was not nothing. A number had to be there. Please do not assume 0 is nothing. Only report what was observed. You may not know why. But that detail is important - only state what is observed.

Step back and confirm the meter works. Put the meter into an Ohms setting - maybe 200 ohms or 2k ohms setting. It will probably display something nonsensical. Touch red and black leads together. It should change to a number that is near zero. What is that number?


Some power supplies have a power switch on the back. Did you try it with that switch both in an off and on position? Or does that switch exist?

Do you know the power cord is connected to a live AC receptacle? IOW will that receptacle also power a light bulb?

Once these questions are answered, then we an move on to what probably exists and why.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom