PC won't turn on after power outtage & storm

[pepi93]

I'm from Toronto where we have been out of power for the past 3 days. I have anyway...

Just got my power back on and viola, my PC won't turn on...at all, no sounds, nothing...

During the storm I wasn't using my PC, I turned it off, kept it plugged in the power bar, APC Home/Office Surge Arrest...but I flipped the switch which I later realized was the actual surge protection

I did read that flipping the switch off doesn't really matter but who knows...I also read that if damage occurs it's usually the PSU that fries and it doesn't go beyond that.

I didn't unplug the PC completely...next time I will.

The storm wasn't bad, not much lightning, perhaps one or two bolts...I've never had an issue since living here and many systems...I live in a condo and not a house...

Everything else that was also plugged into the power bar works fine...monitor & speakers.

What else can I share?

1) the motherboard LED works and lights up
2) I did a quick test on the PSU with a paperclick, shorting the green/black cables and the PSU powered up fine (fan started to spin etc)
3) Main SSD also works as I ripped it out and tossed it in my laptop to recover some files I needed quick.
4) did a full voltage test of the PSU, all values are normal
5) checked voltage of CMOS battery, reading over 3V

I keep pressing the power on button and nothing happens...

Here's what I'm running

OCZ GameXtream 700W PSU (older kind)
ASUS P5B Deluxe
C2D CPU with aftermarket Thermalright HS
6GB of TeamXtreem RAM, 2 x 2gb modules, 1x 1gb modules
Gigabyte 550GTX ti GPU
OCZ SSD as main
WD Raptor as back up one
Seagate Barracuda as back up 2
Asus Xonar DX sound
several fans and a Lian Li case

Here are my questions...

What do I do next in my troubleshooting and how?

Thanks.

 

[westom]

Just got my power back on and viola, my PC won't turn on...at all, no sounds, nothing...

You are probably ready to learn about two completely different devices that, unfortunately, share a same name. The APC did what its manufacturer said it would do. It does not claim to protect from another and typically destructive transient.

However the other device that does have an always required low impedance connection to earth does protect from that other and typically destructive transient.

No protector does protection. Best protection for some incoming utility cables is a connection from that cable to single point earth ground. Cable TV and satellite dish with that wire have some of the best protection possible.

However other incoming utility wires (ie AC electric, telephone) cannot connect directly to earth. So the telephone already makes the same connection via an installed for free 'whole house' type protector.

Your AC electric probably has no such earthing. The other device (also called a protector) makes that connection to single point earth ground. These 'whole house' protectors are sold by more responsible companies including ABB, Leviton, Syscom, General Electric, Ditek, Intermatic, Polyphaser, and Siemens. A Cutler-Hammer solution sold in Lowes and Home Depot for as little as $50. In every case, the effective solution always have a dedicated and low impedance (ie 'less than 10 foot') connection to the one key protection device: single point earth ground.

No protector does protection. Some of the best protection is only a wire to what does the protection - single point ground. Other protectors do the next best thing. Make the same short connection via a 'whole house' protector. Because even the APC needs that protection.

Protectors are simple science. 'Art' is the earthing. Single point earth ground (every word has electrical significance) should have most of your attention. Even a sharp wire bend to earth compromises protection. It's an art. Effective protection means even direct lightning strikes. And nobody knows a surge existed. Even the protector remains functional.

Because a protector is only as effective as its earth ground.

 

[westom]

What do I do next in my troubleshooting and how?

None of those can report a good supply. A paperclip test can only identify some PSU defects. Measuring a PSU without a full computer load can also report a defective supply as good.

Remember, a power system has many parts; not just a PSU. To have an immediate answer requires three digit numbers from a multimeter.

What many call 'good' voltages are not always so. Spec numbers are not the complete answer. But measuring six key wires where the PSU connects to the motherboard (and disconnecting nothing) will identify the problem. Without doubt.

Measure the green wire. It should be about 5 volts. But all three digits are important. Report the actual number.

Repeat same for the green and gray wires both before and as power is applied. Provide those numbers and the power on behavior.

Finally report each of any one red, orange, and yellow wire as the power switch is pressed. Again, report behavior and its final three digit number.

 

[AnitaPeterson]

1) Buy a motherboard tester. You're in Toronto, so head towards the nearest Canada Computers store, and get something like this: http://www.canadacomputers.com/index.php?cPath=33_439
They're cheap, and if you're serious about building your own machines, you'll need one anyway.

2) Get another PSU and try it in your computer (connect only the basic minimum at first - like motherboard, CPU, videocard and the boot device).

 

[pepi93]

I did the multimeter test...all values are as they should be,

yellow at 11.85V
orange at 3.36V
red at 5.20V
blue at -11.49V
pink at 5.20V
grey at 5.20V (jumped pretty quick from 0, not sure if it's supposed to be a slow increase from 0-5V)

These values are for all oranges, yellows, reds etc...

I did this test while shorting the green/ground to keep the PSU powered up.

I also checked the CMOS battery, clocking in at over 3V

thanks for the replies so far.
going to try it with a different psu.

can't see any physical damage to mobo

 

[pepi93]

1) Buy a motherboard tester. You're in Toronto, so head towards the nearest Canada Computers store, and get something like this: http://www.canadacomputers.com/index.php?cPath=33_439
They're cheap, and if you're serious about building your own machines, you'll need one anyway.

2) Get another PSU and try it in your computer (connect only the basic minimum at first - like motherboard, CPU, videocard and the boot device).

those are all PSU testers unless I'm missing something? I posted my voltages that I got with multimeter for my PSU

 

[mfenn]

What do I do next in my troubleshooting and how?

Strip the PC down to the bare minimum, that means mobo, CPU, one DIMM, GPU, and PSU all outside the case. See if it boots. If it does, add components one at a time until it fails.

If it doesn't boot, then try different DIMMs and DIMMs slots. If that doesn't work, get a spare PSU (you should have one anyway). If that doesn't work, my bet is the motherboard.

 

[VirtualLarry]

During the storm I wasn't using my PC, I turned it off, kept it plugged in the power bar, APC Home/Office Surge Arrest...but I flipped the switch which I later realized was the actual surge protection.
I did read that flipping the switch off doesn't really matter but who knows...I also read that if damage occurs it's usually the PSU that fries and it doesn't go beyond that.

I didn't unplug the PC completely...next time I will.

TBH, most power/surge strips have an on/off switch, that effectively works the same way as unplugging it. I've never seen one where there was a switch to keep the power flowing, but disable the surge protection.

So I think your concerns are slightly unfounded.

Although, was the computer wired into any phone, ethernet, coax, etc. wiring? Surges don't just come in on AC wiring.

Anyways, I think the PSU just quit. I lost power once, and my HTPC that was running 24/7 for several years, didn't power back up. Well, not without unplugging it and replugging it and hitting the power button a few times. It finally refused to power up anymore when I had it on the testbench.

Note that this PSU had been on a UPS for several years too, it wasn't directly connected to the mains.

A new PSU, and everything powered back up just fine.

 

[VirtualLarry]

I did the multimeter test...all values are as they should be,

yellow at 11.85V

While technically in-spec, if that is measured with a multimeter under no load, then that seems low to me. It suggests that it would go lower (and perhaps out of spec) under load.

 

[westom]

I did this test while shorting the green/ground to keep the PSU powered up.

Shorting the green to ground complete compromised the entire test. Making measurements useless. Instructions said to only press the power button. If no voltage appears, then only report that number. Only relevant is what those numbers are when the power button is pressed. And no wires were disconnected or jumpered.

No number was provided for the purple wire - the first and a critical three digit number. Please follow directions as specifically stated. Next was to measure the green wire AS the power switch is pressed. To record its behavior when the power switch is pressed and what its final voltage is. Same for the gray wire.

Only measure any one red, orange, and yellow wire as stated. Measuring other wires was not requested and provides no useful information. CMOS battery voltage would only report how many more years before a battery need be replaced. That number also irrelevant to this problem.

PSU tester is useless for so many reasons. In part, because at least one PSU wire is disconnected. Useful tester provides three digit numbers without disconnecting anything. Disconnecting can sometimes complicate the problem. And violates good diagnostic practice.

 

[piasabird]

If you are lucky it is the power supply. Try that first.

You do realize that if the motherboard is connected to a network device a surge can come through that also???

You should probably take it apart and test the motherboard outside of the case with no video card to see if it beeps. Might also disconnect the HD if one is attached. If the surge destroyed motherboard components it is hard to test that without spare parts.

 

[pepi93]

Shorting the green to ground complete compromised the entire test. Making measurements useless. Instructions said to only press the power button. If no voltage appears, then only report that number. Only relevant is what those numbers are when the power button is pressed. And no wires were disconnected or jumpered.

No number was provided for the purple wire - the first and a critical three digit number. Please follow directions as specifically stated. Next was to measure the green wire AS the power switch is pressed. To record its behavior when the power switch is pressed and what its final voltage is. Same for the gray wire.

Only measure any one red, orange, and yellow wire as stated. Measuring other wires was not requested and provides no useful information. CMOS battery voltage would only report how many more years before a battery need be replaced. That number also irrelevant to this problem.

PSU tester is useless for so many reasons. In part, because at least one PSU wire is disconnected. Useful tester provides three digit numbers without disconnecting anything. Disconnecting can sometimes complicate the problem. And violates good diagnostic practice.

Thanks for the post.

I've never done this so I just looked on youtube and did what it told me.

Could you please either direct me to a link where the instructions are given to do this properly or if it's not too much trouble give me step by step instructions on how to do it the way you suggest?

Thank you.

p.s. purple wire came in at 5.2V but as you said, the way I did it was useless so I guess this info is irrelevant.

 

[pepi93]

If you are lucky it is the power supply. Try that first.

You do realize that if the motherboard is connected to a network device a surge can come through that also???

You should probably take it apart and test the motherboard outside of the case with no video card to see if it beeps. Might also disconnect the HD if one is attached. If the surge destroyed motherboard components it is hard to test that without spare parts.

by network device I'm assuming you mean my broadband cable, in which case yes it was still connected to that...and that cable came from my modem directly into my pc.

 

[pepi93]

TBH, most power/surge strips have an on/off switch, that effectively works the same way as unplugging it. I've never seen one where there was a switch to keep the power flowing, but disable the surge protection.

So I think your concerns are slightly unfounded.

I appreciate you clearing this up.

 

[westom]

Thanks for the post.

Use YouTube to find where the wires are. But many describe things irrelevant to your problem. Do not use a paper clip. Do not disconnect any wires. Simply do this:
Measure the purple wire. (I called it green - my mistake) It should be about 5 volts. But all three digits are important. BTW that would be voltage on the purple wire when computer is powered off (but still connected to AC mains).

Repeat same for the green and gray wires both before and as power is applied by pressing the power on switch. Provide those numbers and the power on behavior.

Finally report each of any one red, orange, and yellow wire as the power switch is pressed. Again, report behavior and its final three digit number.

Most of your voltages appear to be on the high side. Maybe your meter is uncalibrated? Is the PSU fully connected to the motherboard? Implies its 12 volts might be defective - actually less than 11.7 volts. Yes, the ATX spec is a different number. But 11.7 is the number when we consider how a meter works and other problems. Videos also do not discuss that.

A wide variation in 5 volts high and 12 volts so low implies more problems. Maybe not the problem causing your power up problem. Another problem that mayl become obvious months or years later.

By using a paper clip, then the power controller is unknown. Again, measure those voltages both as and after the power switch has been pressed. And ignore many of the YouTube videos that are only deceptive. And will not explain the analysis that can be provided from three digit volt numbers.

BTW, there should be a third digit to the 5.2 number. Is the meter on a 20 VDC scale?

 

[pepi93]

Use YouTube to find where the wires are. But many describe things irrelevant to your problem. Do not use a paper clip. Do not disconnect any wires. Simply do this:

Measure the purple wire. (I called it green - my mistake) It should be about 5 volts. But all three digits are important. BTW that would be voltage on the purple wire when computer is powered off (but still connected to AC mains).


Repeat same for the green and gray wires both before and as power is applied by pressing the power on switch. Provide those numbers and the power on behavior.

Finally report each of any one red, orange, and yellow wire as the power switch is pressed. Again, report behavior and its final three digit number.

Most of your voltages appear to be on the high side. Maybe your meter is uncalibrated? Is the PSU fully connected to the motherboard? Implies its 12 volts might be defective - actually less than 11.7 volts. Yes, the ATX spec is a different number. But 11.7 is the number when we consider how a meter works and other problems. Videos also do not discuss that.

A wide variation in 5 volts high and 12 volts so low implies more problems. Maybe not the problem causing your power up problem. Another problem that mayl become obvious months or years later.

By using a paper clip, then the power controller is unknown. Again, measure those voltages both as and after the power switch has been pressed. And ignore many of the YouTube videos that are only deceptive. And will not explain the analysis that can be provided from three digit volt numbers.

BTW, there should be a third digit to the 5.2 number. Is the meter on a 20 VDC scale?

Based on what you wrote you want me to keep the PSU plugged in the wall but PSU button is off...test each wire specified, both with PSU power switch on and off and record both numbers.

If I put 5.2 V it means 5.20, sorry for not clarifying.

Aside from how to test PSU wires while plugged in mobo I believe I understand what you're saying?

 

[pepi93]

I'm pretty sure my PSU is dead?

I did as you told me and also found some help online...here's what I have...I didn't test all the wires but I didn't really feel the need to.

purple, 0V off, 5.13V when power switch on
yellow 0V for both
green 3.50V on, 0V off
red and orange both 0 whether power switch is on or off..

should I keep testing or is it conclusive? I tried my best to dig that probe so it's touching the metal component of the pin...but from what I read if the purple doesn't have 5V in both the on or off position, the PSU is not delivering power to the motherboard?

Thanks for your help!

 

[westom]

should I keep testing or is it conclusive?

I now understand what the confusion is. Only some (few) computers have a power switch on the back. That switch should always be on. A power switch I was referring to is a power button on the front; pressed to turn on the computer.

Measure the purple wire when the computer is off but AC mains is connected (and that rear panel switch is always on). Then do other measurements when and after the front panel power button is pressed.

No voltage should be zero. It may be 0.11 or 0.09 or something like that. All three digits are important. Hopefully you are using a digital meter; not an analog one.

So far, numbers suggest a good PSU. But reperform that procedure to have voltages from all six wires. So that the conclusion is complete and without dobut. Then an explanation of what your computer is doing will follow.

 

[pepi93]

Thanks for all your help, the PSU is dead...

I just tried it with my friends PSU and motherboard powered up fine...I didn't have anything else connected as he forgot to give me his modular cables but I'm pretty sure I'm good to go...

I'm thinking of this PSU, what do you think? I want it to handle a Corei7 and a Z86 chipset or higher motherboard....not sure how much power RAM takes these days...
I plan on making these upgrades when there's good rebates
(back to school or boxing day, black friday specials)

I currently only run 1gpu but might eventually have 2...I doubt it though as I'm all about the free space in a case and I don't do any really high end gaming.

http://www.canadacomputers.com/product_info.php?cPath=33_443&item_id=053094

do I need to go higher?

http://www.canadacomputers.com/product_info.php?cPath=33_443&item_id=053095

different brand?

I guess Seasonic is the staple?

 

[VirtualLarry]

Shorting the green to ground complete compromised the entire test. Making measurements useless. Instructions said to only press the power button.

Westom, that IS the "power button" on an ATX PSU. You short the green wire to ground to power-on an ATX supply. That's what the mobo does when you push the case's power button.

 

[VirtualLarry]

Thanks for all your help, the PSU is dead...
http://www.canadacomputers.com/product_info.php?cPath=33_443&item_id=053094
http://www.canadacomputers.com/product_info.php?cPath=33_443&item_id=053095

No need for something so high-powered.

Antec VP-450
http://www.canadacomputers.com/product_info.php?cPath=33_442&item_id=036731

Corsair CX500M
http://www.canadacomputers.com/product_info.php?cPath=33_442&item_id=056810

Last one has a $20 MIR.

That is, unless you plan on upgrading to a gargantuan or multiple video cards. For your current rig, either of those would be fine.

 

[pepi93]

No need for something so high-powered.

Antec VP-450
http://www.canadacomputers.com/product_info.php?cPath=33_442&item_id=036731

Corsair CX500M
http://www.canadacomputers.com/product_info.php?cPath=33_442&item_id=056810

Last one has a $20 MIR.

That is, unless you plan on upgrading to a gargantuan or multiple video cards. For your current rig, either of those would be fine.

My Plan is to upgrade to a Ci7, asus z87 pro or Gigabyte GA-Z87X-UD4H (for it's oc capability), probably about 8gb of ddr3 ram...might toss in another gpu but will stick with one for now...in any case, I'll get a new gpu at some point...who knows what they have by that time...Nvidia 800 series maybe...

Looking to upgrade latest by Christmas time.

I think I will go for the 750 PSU

 

[westom]

That's what the mobo does when you push the case's power button.

A 'power system' does many functions before it "shorts the green wire to ground". And then does more functions to decide if the green wire remains shorted. Then it eventually decides to let the CPU execute. PSU is only one component of a 'system' that does more than most realize.

OP shorted a green wire and his supply worked fine. Why? The power 'system' is more complex.

How to select a power supply: each DC voltage has a current number. A new supply must supply that same current for each DC voltage - or slightly more. If the system is designed by an engineer (ie Dell, HP, etc), then current numbers were sufficient for any upgrade.

How to determine if the new PSU is actually working properly. Normal is for a defective supply to boot and run a computer. And then the supply fails months later. Causing strange and confusing symptoms. But numbers from a multimeter can see that defect when the PSU is first installed. Identify a PSU defective before its warranty expires.

 

[pepi93]

How to determine if the new PSU is actually working properly. Normal is for a defective supply to boot and run a computer. And then the supply fails months later. Causing strange and confusing symptoms. But numbers from a multimeter can see that defect when the PSU is first installed. Identify a PSU defective before its warranty expires.

I think I would like to do this test with my new psu that I buy today. Could you give me more details on what I'm looking for? I plan to get the Corsair HX750 Gold..

Would be nice to know what is failing before I start to experience various symptoms because I have a dud...

Thanks in advance.

P.S. my old OCZ lasted me for about 6-8 years...I actually forgot when I bought it...not too shabby ..granted I don't have the pc on 24/7

 

[Torn Mind]

The Core i7 is hardly a power-hungry chip. 84 W TDP, and that is with the IGP enabled and the integrated voltage regulator occupying some of it.

Even the most powerful graphics cards(Titan) will not be going over 250 W TDP. the lower the tier of performance you go, the less high the TDP is. Generation number matters, but the second number from the left determines which performance tier you're getting.

And hitting the absolute maximum is not going to happen while gaming. You need Intel Burn Test+Furmark running at the same time. 450-550W is enough for a single GPU setup.