Complete Desktop Failure After Power Surge. Solutions?

Last night, we had a power surge. My desktop is plugged into a surge-protected power strip. However, after the power surge, the computer would not boot up again when I tried to turn it on this morning.

Here's what I've observed:

* Power strip still works - all other items attached to it are fine
* Amber indicator light on mobo is on, as it always is when the PC is off.
* No signs of physical damage on mobo (i.e. burns, popped transistors, etc)

I did not have time before work this morning to begin the troubleshooting process, but I've narrowed it down to 2 possibilities:

1) Best Case Scenario - power supply is fried, not bringing in enough voltage to power up the unit. Simple solution - replace the power supply.

2) Worst Case Scenario - mobo is fried. There is no solution - entire rig needs to be replaced.

I'm hoping it's not Scenario 2, especially since there is an amber indicator light on the mobo. If the mobo was fried, I imagine nothing on it would work.

Any suggestions?

Discussion is locked
Reply to: Complete Desktop Failure After Power Surge. Solutions?
PLEASE NOTE: Do not post advertisements, offensive materials, profanity, or personal attacks. Please remember to be considerate of other members. If you are new to the CNET Forums, please read our CNET Forums FAQ. All submitted content is subject to our Terms of Use.
Reporting: Complete Desktop Failure After Power Surge. Solutions?
This post has been flagged and will be reviewed by our staff. Thank you for helping us maintain CNET's great community.
Sorry, there was a problem flagging this post. Please try again now or at a later time.
If you believe this post is offensive or violates the CNET Forums' Usage policies, you can report it below (this will not automatically remove the post). Once reported, our moderators will be notified and the post will be reviewed.
- Collapse -
I keep this video handy to show how it's done.

This isn't me but he does what I do in the shop, office and home. I also have the usual DVM, Oscilloscope at the office so I can really show my stuff there. But here's a video on dealing with the usual dead PC.

- Collapse -
Thanks! This video is very informative!

Thanks! You also replied to my issue just over a year ago with my previous PC. I can't believe I'm here again with a rig that's just barely over a year old. I do still have the power supply from the previous computer, but i am not sure if that would be a good indicator, as that PSU is now 8 years old, and may have contributed to the failure of my last PC (I never actually was able to diagnose the true reason for the previous PC failure). It may not even be compatible.

I'll check and see. If it IS the Power Supply, I will not only purchase a new one, but I will also get a more powerful surge protector. I'm hoping that it's simply the PSU and not the mobo. Just because I didn't see any physical damage on the mobo doesn't mean none occurred.

I'll keep you posted. Thanks again!

- Collapse -
If you get bored, there's a nugget to be found

At about 50 minutes in with UBCD. You don't have to have XP to use UBCD as he did. UBCD is a SwissKnife to do all sorts of things. It's something I use along with a few Live OS CD/DVD/USB sticks.

- Collapse -
Strange update.

Strange update - I had disabled the power supply for the entire day. After getting all settled in tonight, I just flipped the PSU back to On, and tried booting up the PC. It came on no problem. All fans inside are running, nothing appears to be wrong. Computer has been on about 10 minutes, fan speed is gradually increasing as the components heat up from general use.

I'm always leary of computers "Self-fixing", as it likely means it's merely something that will happen again later. Thoughts?

- Collapse -
Over on Tomshardware

You can find a tar pit on discussions about heat sink compound and the need to bake it in for a few hours before declaring the compound work good or bad.

I'm not there to check your PC but I've lost count of how many just need a good pass with the canned air and a little twist of the heatsinks (bubbles in compound), checking the sinks are on right, and the usual unplug and plug in of all connections.

Engineering joke follows: Why do they call them "Connectors?"

- Collapse -
Clr Safety Lockout

Hardware has a safety lockout feature. It tripped, it is reset by disconnecting the AC power cord for a few seconds. Is that what you did - reset the safety lockout?

- Collapse -
How deep is the problem(s)

When it comes to electrical issues, such as surges the problem can be multiple or include more than one problem. That said, you could actually try fixing it and have the new or good part become bad because yet another part sucked it dry. That's the problem with electrical issues. So, as stated you should at least replace the PSU with a known good one of same specs or BETTER. Alas, once its connected it may work or will it. I say, because often enough some users report after a short use it becomes bad or won't boot again. Thus, you find you have another problem or issue at work. Which for tech is to test items alone on a work test PC and see how it behaves. Thus taking one step at a time and then return known test good parts back into the PC and/or replace as this is the best time to do if there are any "iffy" results and not take a chance with that part.

You stated the surge protector is still working, that may not be the case. Because in many typical surge protector of cheaper make a simple component is only a few cents in value and its good for one decent hit. Breaking apart the surge protector you may find a smoked hole in it or broken or at least discolored, can't be trusted. Replace the surge protector as well. Don't take anything for granted. Remember, a surge can be God knows huge and the PC is rated for 3-12vDC tops, so take it from there.

tada -----Willy Happy

- Collapse -
Additonal Facts

First, most damage has no visual indication. It is rare for a defective electronic part to identified by visual damage.

Even worse, protector parts must never fail in a manner than makes visual inspection possible. That type failure is called catastrophic. Defined unacceptable by protector part manufacturers. Typically occurs because a protector was grossly undersized to fail on a first or second surge. And was on the verge of creating a potential fire. Yes, that sentence should have everyone's attention.

Second, nothing indicates a surge existed. Power loss is not a surge and does not create a surge. What would typically trip that safety lockout function is repetitive voltage drops to zero. A 120 volt protector ignores any voltage below 330 volts. Obviously a voltage drop below 120 volts is completely ignored by a protector. Obviously that is not a surge.

Third, any protector that must be replaced after a surge is grossly undersized and ineffective. But easily marketed to naive consumers who only believe the first thing they are told. And who routinely ignore numbers. A scam so easy to promote that Monster sold $100+ protectors that were electrically equivalent to one selling in Walmart for $10. Consumers who all but want to be scammed assume a higher price means better quality.

Properly sized protection (ie a 'whole house' solution for $1 per protected appliance) means multiple direct lightning strikes without damage to any appliance or that protector. That 'whole house' solution is properly sized, properly earthed, and based in how protection was done, successfully, even 100 years ago.

Fourth, we would routinely identify a surge's current path to find and replace each damaged semiconductor. Then nothing failed even years later. Disk drive will have an incoming surge path. But no outgoing path. So disk drives (like memory) are not damaged by a surge. Meanwhile ethernet, USB, and modem ports are easily damaged (or overstressed). Because those are an outgoing path for that surge.

Fixing it the first time involves learning how surges do damage and which protectors actually do protection. Also learning which protectors can compromise protection inside electronics. Learning which protector is not grossly undersized to fail and disconnect as fast as possible; leaving a surge still connected to appliances.

- Collapse -
Probably a Blown Power Supply

You have most likely blown your power supply. These are cheap, but somewhat complicated to install as you need to run power to EVERYTHING, your motherboard, your disk drives, your graphics card, your optical drives, pretty much everything, and route the cables appropriately.

I usually unplug one thing at a time from the old power supply and plug the new one into it, one-by-one, and only when everything is connected do I physically remove the old power supply and install the new one into the case.

I see below that it has miraculously healed itself. I have seen that happen before, and it is an aspect of some power supplies similar to a fuse that is reset automatically when power is disconnected to it and then reconnected. I would highly suggest, however, that you get yourself a UPS and plug your CPU and other critical hardware into it instead of just a surge protector.

- Collapse -
Miracle - a fuse that heals itself

No such thing as a fuse that resets itself. If that fuse blows in a PSU, it remains blown - must be replaced.

Many observe symptoms of that safety lockout feature. Then assume that is a fuse (it isn't) that somehow resets itself (it doesn't). Then assume the feature is inside a PSU (it is elsewhere in the power 'system').

Also bogus is the myth that a UPS protects hardware. If true, he can define what part inside a PSU is protected, why that part miraculously fixed itself, and provide UPS spec numbers that claim that protection. None were provided for one simple reason. Many make UPS and other recommendations without basic electrical knowledge. Bogus recommendations are quickly identified by no spec numbers. As if a blown fuse can miraculously heal itself.

Symptoms were classic of a safety lockout feature. To say more would required hard numbers.

- Collapse -
Here's that miracle.
- Collapse -
Miracle Still not Defined


A subjective citation that forgets to mention one always critical fact - the numbers. Show me. Show numbers that say it will perform that function? Power supply fuses must be rated 250 volts or higher - and other spec numbers. Were is this miracle fuse?

That citation does not list numbers appropriate for a power supply. If you have something to contribute, then identify this miracle fuse - with spec numbers. Still not defined is the miracle fuse.

OP's symptoms say nothing about a miracle healing fuse. Polyfuse is not in PSUs for good engineering reasons - called specification numbers. Same numbers also not found in that citation that only discusses low currents in 3.5 and 5 volt operations.

Defined were other functions that explain the OP's symptoms.

- Collapse -
Odd request.

Let me share we used polyfuses way down to 5 Volts. These are not new miracle parts and easy to find more detail. I'm unsure why you are demanding me to introduce you to a common component in an electronics today.

Later we used much higher values and voltages. My reply was only to open the door to one possible explanation. You can find more.

- Collapse -
More Background

I used polyfuses back when Raychem (if I remember correctly) sued and bankrupted another company for violating their patents. Back then, a coffee maker used these parts (only rated for 60 volts) in series as backup protection for 120 volt service. Kitchen fires resulted.

Polyfuse is still not sufficient to replace an AC line fuse - as spec numbers demonstrate. Resettable fuses are not used in PSUs.

Polyfuse does not explain the OP's symptoms. Otherwise that polyfuse must be capable of safely interrupting over 250 volts - without creating a fire.

Again, the point. OP's symptoms are consistent with operation of a safety lockout. That may or may not explain what happened. To say more required hard facts - such as numbers from a meter.

His system did not heal itself because it was powered off overnight. Worse, a defect may still exist. But again, that can only be determined by obtaining relevant numbers - which is probably not necessary.

- Collapse -
Welcome back.

To see more of how you post I read a few of your prior posts. You seem to come out swinging hard a lot. Here I want to share ideas, product, bits and pieces of inside the electronics we have today.

This may mean we'll have to dance a lot. Your move.

- Collapse -

I think you're probably correct as to your options. Having worked on other's desktops in the past, I keep at least one extra power supply unit for testing just that kind of situation. They aren't all that expensive. If I couldn't borrow one, I'd take a chance and buy one.

CNET Forums