Two dead motherboards lasting 3 months each

[TheOni]

I built my own PC back in Decemember, and it worked absolutely fine up until March, and I found the motherboard had died, returned it to manufacturer fine, got a replacement, which installed all nice and had been working with no issues then Friday last week I went to turn the PC on an no power, after a few diagnoses I found that the motherboard had died again, rang the supplier and have arranged to send it back and am now awaiting the replacement.

It's just really bothering me that both motherboards lasted roughly the same amount of time before they died?

The motherboard in question is "Gigabyte Z97X-SOC Force Intel Z97 (Socket 1150) DDR3 ATX Motherboard" which from all I've been reading is really hard wearing and pretty reliable as far as I can tell.

Basically when the motherboard is dead, you flick the power switch for the PSU to on, the power LED on the motherboard flicks on for a split second and then goes off again.

The only thing in common both times is that I had been away for a day or two and came back and the PC wouldn't start up (everyone else assures me they hadn't touched it, and I'm sure they know better). I would imagine it's more common that the PC suddenly dies while you are using it rather than during down time?

The first time it did happen, I wondered if maybe there was a powercut while I was out and caused a small surge just enough to blow the mobo, but because of that I bought a surge protector so I can be pretty damn sure it's not that this time.

When I asked the supplier to have a look at my PC specs if there was some kind of conflict or incompatibility that might be causing the mobo to die they couldn't really find anything.

Specs are:

MSI GeForce GTX 980 Gaming Edition 4096MB GDDR5 PCI-Express Graphics Card
Gigabyte Z97X-SOC Force Intel Z97 (Socket 1150) DDR3 ATX Motherboard
Intel Core i5-4670 3.40GHz (Haswell) Socket LGA1150 Processor
Cooler Master HAF X Gaming Tower Case - Black (RC-942)
Avexir Blitz 1.1 Original 8GB (2x4GB) DDR3 PC3-17000C9 2133MHz Dual Channel Memory Kit (AVD3U21330904G-2BZ1)
Corsair Builder Series CX 750W V2 '80 Plus Bronze' Power Supply (CP-9020015-UK)

So yeah, I'm really not sure what to think, I'm sure a replacement will work again but if I am doing something that's slowly killing the mobo over three months I'd like to try and avoid it if possible. (of course there is the chance that I've just been really unfortunate and had 2 faulty mobos)

NB:-

The only thought I've had that could be a possibility is the heat sink I'm using? it's a "Noctua NH-U12S Ultra-Quiet Slim CPU Cooler with NF-F12 Fan", don't know if maybe the base unit for it is causing a short or something maybe? but then would it have sat there for 3 months without doing anything first?

 

[NerdIT]

Hey there TheOni.

Sounds like you have a bad/failing power supply (PSU) or INSUFFICIENT WATTS on the PSU. Either of these will cause system instability, crashes, limited functionality and then possible burnout of motherboard and/or components.

750W is on the lower end of what I consider sufficient for your build.

You can buy a PSU tester online for under $20. Simple to use.

I'd bet my last dollar that's your problem

Hope that helps !

Jeremy Dixon
A+ Certified Technician
Applied Science degrees in Networking Technology and Systems Security

 

[TheOni]

Hey there TheOni.

Sounds like you have a bad/failing power supply (PSU) or INSUFFICIENT WATTS on the PSU. Either of these will cause system instability, crashes, limited functionality and then possible burnout of motherboard and/or components.

750W is on the lower end of what I consider sufficient for your build.

You can buy a PSU tester online for under $20. Simple to use.

I'd bet my last dollar that's your problem

Hope that helps !

Jeremy Dixon
A+ Certified Technician
Applied Science degrees in Networking Technology and Systems Security

Hey thanks for the reply,

I've just ordered a PSU tester and will let you know if it comes back with a bad result.

So 750W is not so good for my build do you think it's worth getting a more powerful PSU? Or do you think it's more that the PSU is faulty and not supplying as much power as it should do?

 

[westom]

You can buy a PSU tester online for under $20. Simple to use.

A power supply tester can only report a few and most obvious defects. It cannot report many anomalies that might explain these symptoms. For about same money, a digital meter reports all those and other anomlies. Numbers are always necessary for a definitive answer.

A PSU is only one component of a power system. Any of those many other components might explain this anomaly. Only a digital meter will report numbers from the entire system. Numbers that often only report something useful to the fewer who really know computer hardware. Without those numbers, the fewer who really know this stuff can only be silent.

Heat sink will not cause failures (except with AMD processors). In fact, heat is a diagnostic tool to identify defective semiconductors that will run a computer for weeks, three monhts, or a year. And then cause failures later at lower
temperatures. Many want to cure a symptom (more fans) rather than understand the diagnostic warning provided by operation at higher temperatures.

To obtain a useful reply requires a meter, about a minute of labor, and some requested insturctions. Then the entire power system can be exonerated or the defect identified. Without those numbers, your only alternative is to keep replacing good parts until something stops failing.

Best evidence comes from the dead body. An autopsy to determine what on the motherboard was defective goes a long way to identifying the actual defect.

Finally, if your computer needs 750 watts, then it is also so hot as to toast bread. We tell computer assemblers that they need double the wattage rather than teach how to select a supply. If the supply is inaccuratedly selected, numbers from the meter will also identify that potential defect.

 

[TheOni]

Update

I received the PSU tester and have just ran it on the power supply.

I took pictures of the results it displayed for each power connection and collated them here

http://imgur.com/a/tJ187

(I should point out that the one that does not have a connection ID at the bottom of the screen is the 8pin CPU one, all the others are labelled 24p, PCI, etc.)

So, I'm no expert but the results (as far as this tester shows) are sufficient, for the power it is supposed to be supplying. It didn't blink red as it said it would if the voltage was too high or low at least.

From looking at the PSU tier list on this I see that the PSU I do have is only bronze certified and quite low on the tier list, I have earmarked the Corsair AX860 as a possible replacement if it is necessary/recommended, I'm hoping that the better quality component may prove better for the longevity of my motherboard, but I can't say I'm sure with my own knowledge really.

Here's hoping someone is able to tell something from the results at least.

 

[Tommy LeQuickee]

If different motherboards are dying in your system then your PSU is either cheap or borked.

Get a good high quality PSU. I use the SeaSonic 1200w Platinum 80+, highly recommended.

 

[westom]

Here's hoping someone is able to tell something from the results at least.

A defective supply (obvious when under near maximum load) will often report good numbers without load. If that power tester is loading a 700+ watt supply, then it must be as hot as a bread toaster.

That PSU may also be providing defective voltage numbers when fully connected to and loaded by all other computer parts.

Second, useful numbers must be to three digits. 3.3 and 5.0 are two digit numbers.

Third, apparently the PSU status monitor is not reported by that tester. The same status monitor used by the power controller to decide when a computer powers off or remains powered on.

 

[TheOni]

If different motherboards are dying in your system then your PSU is either cheap or borked.

Get a good high quality PSU. I use the SeaSonic 1200w Platinum 80+, highly recommended.

It was two of the same model motherboard if that makes any difference.

I have a feeling that I will be just upgrading to a higher grade PSU as you say, unfortunately I get the feeling this is one of those things that won't really be confirmed one way or the other without just sucking it up and buying a new one and seeing if the same thing happens again in 3+ months or not.

 

[TheOni]

A defective supply (obvious when under near maximum load) will often report good numbers without load. If that power tester is loading a 700+ watt supply, then it must be as hot as a bread toaster.

That PSU may also be providing defective voltage numbers when fully connected to and loaded by all other computer parts.

Second, useful numbers must be to three digits. 3.3 and 5.0 are two digit numbers.

Third, apparently the PSU status monitor is not reported by that tester. The same status monitor used by the power controller to decide when a computer powers off or remains powered on.

I see what you are saying, I doubt it is putting a full load on the power supply (although it did get a little warm).

Do you think there is a reasonable way I can actually test the power supply under a full load?

 

[westom]

Do you think there is a reasonable way I can actually test the power supply under a full load?

Best test involves equipment that costs $thousands. Best solution for a layman was posted.

To obtain a useful reply requires a meter, about a minute of labor, and some requested instructions. Then an entire power system can be exonerated or the defect identified. Without those numbers, your only alternative is to keep replacing good parts until something stops failing.

Again, a PSU is only one component of the power system.

 

[Tommy LeQuickee]

If different motherboards are dying in your system then your PSU is either cheap or borked.

Get a good high quality PSU. I use the SeaSonic 1200w Platinum 80+, highly recommended.

It was two of the same model motherboard if that makes any difference.

I have a feeling that I will be just upgrading to a higher grade PSU as you say, unfortunately I get the feeling this is one of those things that won't really be confirmed one way or the other without just sucking it up and buying a new one and seeing if the same thing happens again in 3+ months or not.

Never skimp on PSU's. It's one of the few system parts you can't compromise on when it comes to quality

You can buy a high quality PSU (Seasonic 1200w platinum+ is quite cheap these days I picked mine up for £170, so a 850w maybe £120-£130ish perfectly reasonable for such quality) and guarantee nothing will go wrong (7 year warranty)

Or you can buy a cheap PSU for £60 but spend 10x as much on that replacing motherboards and other components all the time because it keeps blowing them up, shorting them out, overvolting, undervolting.. not good.

 

[NerdIT]

@westom -good advice.

You could take everything off you board, test, add a piece back, test again, log, repeat.

Probably a dead board or PSU.

Rosewill has always done me well, been using them for 10+ years. I currently use :

Rosewill Photon-1200, PHOTON Series 1200W Full Modular Power Supply, 80 PLUS Gold Certified, Single +12V Rail, Intel 4th Gen CPU Ready
http://www.newegg.com/Product/Product.aspx?Item=N82E16817182320

Enough room for tri-sli and anything I would ever need, Strong single 12v rail.

 

[westom]

Or you can buy a cheap PSU for £60 but spend 10x as much on that replacing motherboards and other components all the time because it keeps blowing them up, shorting them out, overvolting, undervolting.. not good.

No minimally sufficent supply will cause that damage. However, it must state in specifications that those required functions exist. Supplies marketed to computer assemblers can 'forget' to include some required functions. If specs do not state that function exists, then suspect the worst. A £60 supply should be more than sufficient. But it must include a full page of specifications.

First indication of an inferior supply is no numeric specs. You need not even know what those specs mean. But no specs mean the 2% who really know this stuff are silenced. Without numeric specs, the best informed cannot say anything useful.

Specification requirement applies to all manufacturers.

 

[TheOni]

Best test involves equipment that costs $thousands. Best solution for a layman was posted.

To obtain a useful reply requires a meter, about a minute of labor, and some requested instructions. Then an entire power system can be exonerated or the defect identified. Without those numbers, your only alternative is to keep replacing good parts until something stops failing.

Again, a PSU is only one component of the power system.

It looks as though testing it under load with a digital meter requires me to have a working motherboard to connect the PSU to (I'm still waiting for the replacement mobo from the supplier).

Honestly if there's a chance that the PSU I'm using is damaging my mobo I'd rather not risk connecting it up again if I can avoid it, I will be getting a new higher grade PSU for sure, but I think I will still try and test it when it arrives.

I'm hoping when I get the replacement motherboard I will get some information from the supplier what they actually found in their inspections, as that may shed more light on what is causing this.

@westom -good advice.

You could take everything off you board, test, add a piece back, test again, log, repeat.

Probably a dead board or PSU.

Rosewill has always done me well, been using them for 10+ years. I currently use :

Rosewill Photon-1200, PHOTON Series 1200W Full Modular Power Supply, 80 PLUS Gold Certified, Single +12V Rail, Intel 4th Gen CPU Ready
http://www.newegg.com/Product/Product.aspx?Item=N82E16817182320

Enough room for tri-sli and anything I would ever need, Strong single 12v rail.

I have picked out a platinum certified 850w PSU as a replacement, I can't imagine I would really need anything higher than that if I'm only running a single card set up? although of course open to suggestions.

 

[westom]

It looks as though testing it under load with a digital meter requires me to have a working motherboard to connect the PSU to (I'm still waiting for the replacement mobo from the supplier)..

Even a damaged motherboard consumes power IF a controller powers on the PSU. If 'missing and required' functions in that PSU permitted motherboard damage, then numbers from the meter (using requested instructions) would have also identified that. No other layman tool provides information to empower your best help.

BTW, if that power supply was providing too much (and destructive) voltage, then that voltage would have been even worse with no load and the power supply tester. Just another reason why that fear is unjustified.

 

[TheOni]

Even a damaged motherboard consumes power IF a controller powers on the PSU. If 'missing and required' functions in that PSU permitted motherboard damage, then numbers from the meter (using requested instructions) would have also identified that. No other layman tool provides information to empower your best help.

BTW, if that power supply was providing too much (and destructive) voltage, then that voltage would have been even worse with no load and the power supply tester. Just another reason why that fear is unjustified.

In hindsight yes it probably would have been a good idea to test the dead motherboard, but I remembered how it took a fair bit of time to get a replacement last time so I wanted to get it returned ASAP this time.

I don't know if it's of any use but I can tell you that when the dead motherboard was still in the case connected up, flicking the PSU power switch to on momentarily lit up the power LED on the motherboard.

I'm trying to get in contact with the supplier this weekend to A. chase up my return, and B. I will be asking them if they have a report from the inspection of the motherboard (either the original one or the one I have just sent back) hopefully maybe they will have some useful information that might tell us something a bit more definitive.

 

[westom]

I don't know if it's of any use but I can tell you that when the dead motherboard was still in the case connected up, flicking the PSU power switch to on momentarily lit up the power LED on the motherboard.

Power controller does that. It orders a PSU on. It then does numerous checks including stable power. It powers off immediately when any inputs are wrong - including a status indicator from the PSU.

Power cuts do not cause damage. Power cuts do not cause surges. And surges (even those powerful enough to destroy power strip protectors) are made irrelevant by what already exists inside the PSU. However that surge may bypass existing protection if using a power strip protector. Adjacent protectors can connect surges directly into a motherboard - bypassing protection inside a PSU.

Is the new motherboard good? Heat does not damage motherboards. Heat is an excellent diagnostic tool to find defective boards. Once you have the system running, then operate in a 100 degree F room. If software crashes, then suspect some part that should be happy even in 100 degrees is defective - and will probably get worse in months or a year causing failures at 70 degrees.

Of course, you may be suffering from coincidence. It happens. But not on your list of most likely suspects.

 

[NerdIT]

A digital PSU tester with readouts is what I was referring to - I have found them to be useful.

Also, you can always bust out a multi-meter, and check contact points at certain points on the mobo. There are plenty of guides on how to do this -it is tedious and time consuming, but you can pinpoint where the power starts to go screwy as it flows through your system.

3 dead board vs a bad PSU -I'd say PSU or crappiest luck ever

You mentioned the CPU cooler heatsink, I have seen mobos short out because the cooler fins were to wide and made contact with capacitors/DIMMS..So I would double check that -and check to see if anything -like a screw- isn't lodged between the back of your mobo and case -also seen this cause shorts.

Good luck!

P.S.
Afterthought, I have a 1200W PSU, and I run 3 GTX 980s, and like 12 fans, and a i7-5820k - My average load is about 675-750 Watts when under FULL load -and it does not get that hot..at all.. Are you ensuring the PSU has airflow??

 

[westom]

A digital PSU tester with readouts is what I was referring to - I have found them to be useful.

Yes it can detect some obvious defects - and not all defects. Obvious for many reasons including - PSU is only one component of a power system. An entire and interconnected system must be reported by numbers.

Most guides to using a digital meter are incomplete. Some are just wrong.

Apparently a power controller (another and completely different power system component) is turning off power. Why? That can only be answered without disconnecting any wires. And cannot be answered by a PSU tester.

Proper airflow? If a computer is working properly, then one 80 mm fan in a 100 degree F room causes no failures. If a computer is consuming 600+ watts, then heat from fans should be hot like a bread toaster. What computer is that hot? Another myth that power cuts cause hardware damage. A digital tester (that clearly cannot create an always necessary load) cannot report a PSU as good. It can identify some defects (identify a PSU as bad) and not report a PSU as good. Only useful numbers come from the system without any changes - without even one connector disconnected.

 

[snowangl67]

I built my own PC back in Decemember, and it worked absolutely fine up until March, and I found the motherboard had died, returned it to manufacturer fine, got a replacement, which installed all nice and had been working with no issues then Friday last week I went to turn the PC on an no power, after a few diagnoses I found that the motherboard had died again, rang the supplier and have arranged to send it back and am now awaiting the replacement.

It's just really bothering me that both motherboards lasted roughly the same amount of time before they died?

The motherboard in question is "Gigabyte Z97X-SOC Force Intel Z97 (Socket 1150) DDR3 ATX Motherboard" which from all I've been reading is really hard wearing and pretty reliable as far as I can tell.

Basically when the motherboard is dead, you flick the power switch for the PSU to on, the power LED on the motherboard flicks on for a split second and then goes off again.

The only thing in common both times is that I had been away for a day or two and came back and the PC wouldn't start up (everyone else assures me they hadn't touched it, and I'm sure they know better). I would imagine it's more common that the PC suddenly dies while you are using it rather than during down time?

The first time it did happen, I wondered if maybe there was a powercut while I was out and caused a small surge just enough to blow the mobo, but because of that I bought a surge protector so I can be pretty damn sure it's not that this time.

When I asked the supplier to have a look at my PC specs if there was some kind of conflict or incompatibility that might be causing the mobo to die they couldn't really find anything.

Specs are:

MSI GeForce GTX 980 Gaming Edition 4096MB GDDR5 PCI-Express Graphics Card
Gigabyte Z97X-SOC Force Intel Z97 (Socket 1150) DDR3 ATX Motherboard
Intel Core i5-4670 3.40GHz (Haswell) Socket LGA1150 Processor
Cooler Master HAF X Gaming Tower Case - Black (RC-942)
Avexir Blitz 1.1 Original 8GB (2x4GB) DDR3 PC3-17000C9 2133MHz Dual Channel Memory Kit (AVD3U21330904G-2BZ1)
Corsair Builder Series CX 750W V2 '80 Plus Bronze' Power Supply (CP-9020015-UK)

So yeah, I'm really not sure what to think, I'm sure a replacement will work again but if I am doing something that's slowly killing the mobo over three months I'd like to try and avoid it if possible. (of course there is the chance that I've just been really unfortunate and had 2 faulty mobos)

NB:-

The only thought I've had that could be a possibility is the heat sink I'm using? it's a "Noctua NH-U12S Ultra-Quiet Slim CPU Cooler with NF-F12 Fan", don't know if maybe the base unit for it is causing a short or something maybe? but then would it have sat there for 3 months without doing anything first?

Try flashing your bios mine did the same thing until I usb flashed the bios with the correct bios from gigabyte.com

 

[Nemesia]

Try flashing your bios mine did the same thing until I usb flashed the bios with the correct bios from gigabyte.com

This thread is 6 years old. Please look at the date before answering in a thread.

Last message "Aug 24, 2015".