Gigabyte Motherboard went up in smoke... what could have caused it? And what could've died with it?

[haza12d]

I recently bought a used Gigabyte Motherboard from Craigslist a well as a CPU for what was supposed to be a test bench.

The motherboard was a Gigabyte Z97X Gaming and the CPU is a Pentium G3258

Everything looked fine on the board and on the CPU. No bent pins or signs of wear or damage.

After plugging in the power supply and turning the system on for the first time, the cpu fan spun, the leds on the board lit up... Aaaand there was pop and a puff of smoke and a litte flame that shot out from the left side of the board close to the heatsink to the left of the socket.

I immediately turned off the power supply and waited for the smoke to clear and started disassembling the system.

I had the board lying on an ESD Mat and I was wearing and Anti Static strap when I put it together to see if it worked before I threw it in the case so I'm kinda ruling out ESD.

My question is what could have caused the fireworks show in my dining room?

Could a bad CPU fry a board like that? Or was it the power supply? It is a refurb Corsair CX750M but after testing the PSU on another system, it seemed ok although the dead board did require an 8-pin for the CPU and the test board only needed a 4-Pin.

Could the CPU and RAM have been affected as well?

What the real bummer is that I looked up the warranty on the board and it had just expired on 10/30/2017!!! T_T

I'm hoping that Gigabyte would help me out and RMA it since the warranty just lapsed a few days ago.

Below are the list of components I was using and a picture of the damage.

CPU: Pentium G3258
Mobo: Gigabyte Z97X-Gaming 7
RAM: 8GB G.Skill RipJaws DDR3-1600
PSU: Corsair CX750M (Refurbished)

https://imgur.com/a/1a6UR

 

[DSzymborski]

I recently bought a used Gigabyte Motherboard from Craigslist a well as a CPU for what was supposed to be a test bench.

The motherboard was a Gigabyte Z97X Gaming and the CPU is a Pentium G3258

Everything looked fine on the board and on the CPU. No bent pins or signs of wear or damage.

After plugging in the power supply and turning the system on for the first time, the cpu fan spun, the leds on the board lit up... Aaaand there was pop and a puff of smoke and a litte flame that shot out from the left side of the board close to the heatsink to the left of the socket.

I immediately turned off the power supply and waited for the smoke to clear and started disassembling the system.

I had the board lying on an ESD Mat and I was wearing and Anti Static strap when I put it together to see if it worked before I threw it in the case so I'm kinda ruling out ESD.

My question is what could have caused the fireworks show in my dining room?

Could a bad CPU fry a board like that? Or was it the power supply? It is a refurb Corsair CX750M but after testing the PSU on another system, it seemed ok although the dead board did require an 8-pin for the CPU and the test board only needed a 4-Pin.

Could the CPU and RAM have been affected as well?

What the real bummer is that I looked up the warranty on the board and it had just expired on 10/30/2017!!! T_T

I'm hoping that Gigabyte would help me out and RMA it since the warranty just lapsed a few days ago.

Below are the list of components I was using and a picture of the damage.

CPU: Pentium G3258
Mobo: Gigabyte Z97X-Gaming 7
RAM: 8GB G.Skill RipJaws DDR3-1600
PSU: Corsair CX750M (Refurbished)

https://imgur.com/a/1a6UR

A bad or improperly connected power supply is the most likely cause. Whether they still honor the warranty is a bit of a crapshoot - not only do they not have to honor it past the date, they also don't *have to* honor it if you're not the original owner, though many of the larger companies don't really enforce it and frequently don't even ask for anything besides the serial number.

The date may be farther back, however, if you don't have the original receipt; in that case, they'll go by the serial number, which will give them the *manufacturer* date as the starting point, not the sale date.

 

[BFG-9000]

ESD mats are electrically conductive--that's the only way they can work. They have a black side that is very conductive from carbon black, and a colored side which generally measures about 1Mohm that is reasonably safe to power up low voltage devices on, but isn't as safe as say, cardboard. ESD mats are intended for conducting stray static away to ground during assembly or soldering, not for actually running things on.

If you used the ESD mat upside down, or a silvered antistatic bag, or a steel worktable top, then the maximum voltage present of 12v could be applied to some very sensitive areas with nearly unlimited current, more than enough to blow things up. Or more likely, something powered was grounded that shouldn't have been, particularly if you actually grounded that mat. I always cringe when I hear about people testing new systems atop antistatic bags, but if the conformal coating spray on the bottom has no holes, it usually insulates well enough. Unfortunately there are a lot of pointy things on the bottom of a board so a used one is likely to have any coating worn off there.

As the only things under those red VRM heatsinks are power Mosfets, the chances are good that one of them blew. Pop off the heatsink's two plastic pins and have a look underneath. Replace any dead-shorted + exploded ones and it may live again.

 

[haza12d]

I actually just used the esd mat to build it, not run it. It was on top of cardboard when the fireworks show began. Lol

 

[BFG-9000]

Then you may just have been unlucky that it failed right then, if nothing was shorted. On the bright side, when Mosfets fail from their usual dead-short failure mode they don't usually take out the CPU or anything else, so replacing it may get the board running again if you can't get a RMA.

Had a Mosfet blow during IntelBurnTest while overclocking a terrible 3-phase VRM board to ~135w and everything worked again after replacing just the one. I had just built the system and installed + setup all the special software which took all day, then was doing the final burn-in test before releasing the machine into Production, LoL. It's been running for the last 9 years ever since at the same overclock that blew it up. Was just a bad Mosfet.

 

[Satan-IR]

Yes just want to add to above that sometimes it's just a faulty bad component usually SMD like fets or caps that pass during QC and tests but under load or with some voltage or current stress give up.

You have to have the board tested by someone with the know-how. Sometimes replacing the faulty piece brings them back to life.

You mentioned fireworks and I'm guessing the damage must be visually identifiable. You must also have smelt the ozone and whatnot when it happened. The address you gave kind of points to VRM parts maybe a Mosfet or a capacitor.

 

[haza12d]

Thanks, ya'll! I'm still waiting to hear back from Gigabyte and see if they will RMA it and if they don't (Which I have a feeling that they probably won't) how much it would cost. I'll see if there's anyone local that can fix the board and see how much that costs. I'm probably looking to spend $50 for the repair, more than that and I don't think it'd be worth it.