Zotac - GTX 460 SLI, up in smoke!

[dma0991]

I'm wondering if it's the PSU at fault, then why is the OP able to get that same PSU to power a different system with a different video card without issue?

Powering it up is one thing. I am questioning why the safety feature on the PSU did not kick in when it happened. Had the safety feature kicked in sooner or instantly then this problem might not happen.

Maybe the PSU was fine before and after but it had a faulty OCP/OVP that caused to not detect a possible problem. It is just my assumption to what happened, it may be right or it may be wrong.

 

[SithSolo1]

Got any photos of the PCB side of the top card and the fan side of the bottom one?

 

[Pontius Dilate]

To be sure there is financial incentive involved here on behalf of the business.

They don't want to needlessly replace video cards that may well be part of a warranty fraud scheme itself. (remember Zap's thread on warranty fraud at BFG?)

There is a larger financial incentive here though IMO. If they accept these cards for warranty repair then they are legally signing on to recognizing that their cards can burst into flames. That has to be reported to the government, a recall would undoubtedly be required, as well as some rather expensive re-engineering of the card to ensure such does not happen on any future products.

This could snowball into a multi-million dollar liability for Zotac. If they steadfastly refuse ownership of the problem then they don't have to report anything to the government.

Best thing this guy could do for himself and his fellow citizens at this point is go to the CPSC website and report the video cards as unsafe products. Work on this from the angle that Zotac is trying to deny the existence of a very real issue.

http://www.cpsc.gov/talk.html

https://www.saferproducts.gov/CPSRMSPublic/Incidents/ReportIncident.aspx

This will start the wheels moving that need to be moving.

I totally agree. If the OP is convinced that his cards spontaneously started a fire inside his case and that nothing else is to blame he should definitely report the manufacturer in good faith to the CPSC.

 

[KaosFaction]

Aigo, I'm not the OP - I'm talking about an entirely different situation with an entirely different card. I think you conflated the OP's case with the anecdote I shared. I could be wrong, though, as I only understand maybe half of what you ever post anyway.

Anyway, if all of the other parts (miraculously) work in other systems, it's clear to me what went wrong with the OP's system. Zotac cards.

Or poor install (as what keeps being suggested throughout this thread).

 

[Seero]

Please treat the following a FUD, or read really carefully and try to understand what I am trying to say. (I know it is extremely hard, but try.)

We all assume that the material used by the cover is non-flamable, but the pictures showed that it is. Normally, plastics are flammable due to its chemistry compound. Again i'm not a chemist but I believe there are no ways to make plastic "fire resistant", only "fire retardant". However, making it fire retardant means it is hard to cut. If, and I mean 'IF', manufacturer chooses to cut corners and use fire resistant coating instead, then when plastic melted, it may become flammable.

It doesn't take a lot of heat to melt the cover, only about 90-120 degree. If there is a source which keep a single drop of plastic in its liquid state, the chemistry compound can change as things may evaporate, turning that single drop of plastic a fuel for fire. If that is the case, then you actually have a mini bomb sitting at your PC waiting for a chance to ignite. It is hidden underneath the video card so it is really hard to detect. It is sitting on the hot source which is keeping it in liquid state or constantly changing its states (that sounds fun to a non-chemist like me).

My theory is flawed because it takes some 300+ degree to ignite plastic which is not really possible from the heat off HS, but my theory does favor the above assumption, keeping a drop of fuel on top of the NB HS. My guess is, the whiter spot on the NB HS was the spot covered by melted plastics, and eventually ignited, burning the plastic on top of it, creating the dent at the cover. The temp of the ignition was so high that melted plastic turn into fuel, yet the resistant coating from new plastic prevented it from burning repeatly. Yes, the question is what triggered the ignition.

If what I said is plausible, then the entire case may not be user fault after all.If there is a drop of fuel sitting on the HS surrounding electronics, then a single spark or discharge can indeed ignite the fuel I.e. a discharge from video card towards the NB HS. The point where that drop of fuel seats must be the point which the HS is closest to the cover and therefore the point where a discharge will occur, just like how a lighter works. In OP's case clearly something was ignited. In fact, the temperature of the smoke what so high that it caused the shroud near the fan to melt as it sucks air in. Evidence of extreme temp is clear. Look at the PCIe pins off the 2nd card, only temperature over 300 degree can cause that kind of damages, and it is from the ashes resulting from the plastic burn.

I don't think anyone would have even remotely imargine that the cover of the video card can turn into fuel. Even if there is a chance it is a big deal. There is a big difference between "not possible" and "highly unlikely". By looking at those pics, it really isn't as simple as a fried chip or a pop out resistor. It can kill, especially running SLI/Crossfire 24/7 unattended + poor cooling. Chance of winning a lottery is low, but someone will win. If the above is plausible, then we all should really place a fire distinguisher right beside computers.

Again, I have no proof, data nor the expertise to make this as a claim. At best this can only be a plausible case. If OP wants to get to the bottom of this, please follow idontcare's advise.

 

[Texashiker]

Unfortunately we cannot provide any type of RMA since the product is severely damaged physically.

So their card overheated, burnt up, and they do not want to replace it.

Forget that, I will never buy a Zotac card.

At the very least, I would go to amazon and newegg and a very negative review?

 

[SickBeast]

OP perhaps your best bet as of now is to go after Gigabyte.

After looking at the photos again, it actually looks like the graphics cards "smothered" the chipset heatsink.

It looks to me like you had some bad luck. I can't stand it when the motherboard manufacturers insist on putting the chipsets anywhere near the PCI-E slots. It's stupid. If they RMA your board and pay for the cards, I would not use them together again in the future unless you can find a way to give that chipset heatsink the space it needs to dissipate the heat.

Good luck. Either Gigabyte or Zotac should give you new parts. You deserve to have all three components RMA'ed IMO.

 

[tweakboy]

This is what happens when your vendor blows goat.

I bet the PCB is red on this card ?

 

[Seero]

.. I can't stand it when the motherboard manufacturers insist on putting the chipsets anywhere near the PCI-E slots. It's stupid....

I agree. Also, having 2x16 PCIe so close is also a fail, assuming PCIe 1 &3 is indeed the 2 x16 slots. Plus the missing Bridge fan.

 

[Seero]

Nvm.

 

[exar333]

Powering it up is one thing. I am questioning why the safety feature on the PSU did not kick in when it happened. Had the safety feature kicked in sooner or instantly then this problem might not happen.

Maybe the PSU was fine before and after but it had a faulty OCP/OVP that caused to not detect a possible problem. It is just my assumption to what happened, it may be right or it may be wrong.

This is a good idea; it may strengthen the OP's argument with Zotac if he/she can send in their PSU to get certified. If the OCP/OVP is functional and is proven, maybe that reinforces the GPU as faulty. I would want the PSU looked-at anyways, to give me a good piece of mind before using it again and having something similar happen. I wouldn't trust it otherwise...

 

[ArchAngel777]

Zotac does not have to admit anything by doing the right thing and replacing this guys card. They can still maintain "This is not a defect and there is no problem, but in the interest of customer service, we will RMA the card". It isn't that hard and if Zotac wants to gain more of a foothold here in the US, it is going to need to take care of this guy.

So far all we have are opinions. No one even knows this guy and some people are trying to blame him. Maybe it is his fault, but you can't prove it. So leave it alone. Zotac has more at stake here than this guy.

I personally don't have a problem with Zotac, but their response to this guy is affecting my opinion of them. So, again, in the interest of their business, they are going to need to take care of this.

 

[OCGuy]

So their card overheated, burnt up, and they do not want to replace it.

Forget that, I will never buy a Zotac card.

At the very least, I would go to amazon and newegg and a very negative review?

And you know this is what happened how?

For all they know, the OP is like one of the other posters in this thread who feels entitled to run hardware outside of specification, and be given new hardware on a whim.

 

[SickBeast]

Zotac cards are known to run hot. A pair of them combined with the Gigabyte motherboard with a stupid northbridge placement seems to have created a perfect storm for the OP.

IMO Zotac should RMA the cards and Gigabyte should RMA the motherboard.

 

[Pontius Dilate]

Zotac cards are known to run hot. A pair of them combined with the Gigabyte motherboard with a stupid northbridge placement seems to have created a perfect storm for the OP.

IMO Zotac should RMA the cards and Gigabyte should RMA the motherboard.

Any evidence for your first statement? And it's the southbridge. All motherboards have the southbridge located in that area. Also, I would bet anything you like that the southbridge heatsink did not cause the fire.

 

[Vesku]

If you can't get an amicable resolution from Zotac, I would consider small claims for something like this. Be prepared to show that it was a defect in the card and not another component.

 

[SickBeast]

Any evidence for your first statement? And it's the southbridge. All motherboards have the southbridge located in that area. Also, I would bet anything you like that the southbridge heatsink did not cause the fire.

http://www.anandtech.com/show/3810/nvidias-geforce-gtx-460-part-2-the-vendor-cards/6

According to the AT review, it ran 11C hotter than a reference 768mb card, and almost as hot as a pair of 460s in SLI. I also read in a Guru3d article that the Zotac card gets close to 80C when overclocked. That's hot. Now imagine two of them.

I still say it was a combination of the southbridge and the GPUs that caused this problem. The heat probably got to be too much and it caused the GPU HSF to melt and spontaneously combust.

Even at 950mhz on the core with 1.2v, my MSI Cyclone GTX 460 never exceeds 65C under load. The Zotac hitting 73C at stock speeds is bad IMO.

 

[Seero]

http://www.anandtech.com/show/3810/nvidias-geforce-gtx-460-part-2-the-vendor-cards/6

According to the AT review, it ran 11C hotter than a reference 768mb card, and almost as hot as a pair of 460s in SLI. I also read in a Guru3d article that the Zotac card gets close to 80C when overclocked. That's hot. Now imagine two of them.

I still say it was a combination of the southbridge and the GPUs that caused this problem. The heat probably got to be too much and it caused the GPU HSF to melt and spontaneously combust.

Even at 950mhz on the core with 1.2v, my MSI Cyclone GTX 460 never exceeds 65C under load. The Zotac hitting 73C at stock speeds is bad IMO.

That temp is still cooler than 8800GT. Imagine it is a pair of 480. They are hotter and longer then 460. There may be even more factors we didn't see. Or maybe we are barking the wrong tree. Who knows?

I am sure of one thing though, I don't want to be the next OP. I really want to find out what exactly went wrong so I can prevent it.

 

[Gulkorr]

to rule out the power supply i would call corsair and tell them what happend and would like to have the power supply tested to rule it out. and you want a letter sent back with the same power supply and you sent in, so that may, say, hey i sent this in to corsair they tested it nothing it wrong with it. so it is the card's what the deal now.

 

[yacoub]

to rule out the power supply i would call corsair and tell them what happend and would like to have the power supply tested to rule it out. and you want a letter sent back with the same power supply and you sent in, so that may, say, hey i sent this in to corsair they tested it nothing it wrong with it. so it is the card's what the deal now.

If you really want to resolve this properly, this would be a good first step since Corsair support seems to be well regarded and it's likely you'll get an honest assessment from them of the PSU, which will allow you to rule that out or identify it as the culprit part.

 

[Pontius Dilate]

http://www.anandtech.com/show/3810/nvidias-geforce-gtx-460-part-2-the-vendor-cards/6

According to the AT review, it ran 11C hotter than a reference 768mb card, and almost as hot as a pair of 460s in SLI. I also read in a Guru3d article that the Zotac card gets close to 80C when overclocked. That's hot. Now imagine two of them.

I still say it was a combination of the southbridge and the GPUs that caused this problem. The heat probably got to be too much and it caused the GPU HSF to melt and spontaneously combust.

Even at 950mhz on the core with 1.2v, my MSI Cyclone GTX 460 never exceeds 65C under load. The Zotac hitting 73C at stock speeds is bad IMO.

Well first of all, the Zotac tested your linked review is a different card with a completely different HSF than the OP's. Secondly, see all those cards running hotter than the GTX 460? They didn't spontaneously combust in Anandtech's overclocked rig running furmark.

Plus, the OP said his rig was idling. The cards weren't doing any work at all, and neither was the southbridge. The event you are describing is so unlikely as to have a practically zero chance of occurring. The OP had a short or surge in the lower PCIe socket which caused a brief, intense electrical fire before his rig shut down. The card may still be to blame somehow, but the cards simply did not spontaneously combust due to overheating.

 

[aigomorla]

Please treat the following a FUD, or read really carefully and try to understand what I am trying to say. (I know it is extremely hard, but try.)

no no no no no no no no !!!

Its really simple what he did, and he's gonna have a hard time proving to me that he didnt from the damage which is shown.

IF YOU PLUG IN YOUR EPS PIN inside a PCI-E Slot, you will mix all your grounds up.

What happens when you have a 12V+ and you feed it with a bad ground?

Simple... its like crossing a battery.

So what happens to the wire when you do that? It heats up til no tomorrow.

You see why zotac wont honor RMA?

Its either the PSU shot out a bad ground on the PCI-E lane.. or the OP forced the EPS into the PCI-E Plug.

 

[Seero]

Well first of all, the Zotac tested your linked review is a different card with a completely different HSF than the OP's. Secondly, see all those cards running hotter than the GTX 460? They didn't spontaneously combust in Anandtech's overclocked rig running furmark.

Plus, the OP said his rig was idling. The cards weren't doing any work at all, and neither was the southbridge. The event you are describing is so unlikely as to have a practically zero chance of occurring. The OP had a short or surge in the lower PCIe socket which caused a brief, intense electrical fire before his rig shut down. The card may still be to blame somehow, but the cards simply did not spontaneously combust due to overheating.

This is not against you but just for the sake of possibilities. In theory, at idle, nothing should be hot, and nothing should ignite within a computer even if it is overclocked. Clearly, it did. Even if there was an electrical short, the PSU should shut down in milliseconds. The damage was clear that it can't be done with just a simple short unless the PSU shut down or it actually did but something else was burning after it has shut down.

Note that the mobo and psu can still function, indicating that the short can not be sever or it will kill both. Also, we should be able to locate the source if it is indeed a several short by locating the pins and sever damage around the pins that caused the short as the temperature of the pin must be higher than its surrounding. We are not able to locate such pin.

The damage seems to occur somewhere between the lower part of the 2 video cards. Knowing the cards are populated in slot 1 and 3, making socket 2 the possible area where shorts occurred. There are no questions about the possibility of a short from that slot, but can't be a short that cause the damage on the card because a) there is only one portion of the cover that is on top of the slot is burned, b) the burn of the cover is far sever than the slot itself, c) the cover itself is non-conductive, meaning it can't be the domain of the short. d) We should be able to locate the domain of the short as it should severly damaged, more so than the cover, yet either it isn't on the picture, or not visible through the picture.

If sever electrical short didn't occur, than what caused the damage? Fire was obvious as there weree ashes and smoke, but nothing seemed to be the fuel of the fire. There are no fuel, but there was a fire. How can that be possible? I checked those stickers but there ain't burned, but more like covered by ashes.

Like all fuel, its mass will reduce. In simple terms, it will be consumed. Nothing in those pictures were consumed other than the plastic cover as we can see the ash on the back of the 2nd card as well as the ashes at the 3rd PCIe slot. If it wasn't the fuel, than what kept burning it into ashes?

 

[Seero]

no no no no no no no no !!!

Its really simple what he did, and he's gonna have a hard time proving to me that he didnt from the damage which is shown.

IF YOU PLUG IN YOUR EPS PIN inside a PCI-E Slot, you will mix all your grounds up.

What happens when you have a 12V+ and you feed it with a bad ground?

Simple... its like crossing a battery.

So what happens to the wire when you do that? It heats up til no tomorrow.

You see why zotac wont honor RMA?

Its either the PSU shot out a bad ground on the PCI-E lane.. or the OP forced the EPS into the PCI-E Plug.

I will not challenge of the possibility of a faulty PSU, mis-configuration, or shorts. I read each of your post carefully, as well as all others because I too like to find out the cause, and thought about its possibilities. In short, it is indeed a possible cause.

For me, there are far more questions than answers I can get from the pictures. Yes, it is possible that the OP is lying about afk during the incident and it is somehow due to user. But, isn't it also possible that the OP speaks nothing but truth and his opinions? That is, someone how setup his/her computer which appears to work flawlessly for months until suddenly it decided to set itself on fire?

Actually, I should really stop trying about this.

 

[Gulkorr]

no no no no no no no no !!!

Its really simple what he did, and he's gonna have a hard time proving to me that he didnt from the damage which is shown.

IF YOU PLUG IN YOUR EPS PIN inside a PCI-E Slot, you will mix all your grounds up.

What happens when you have a 12V+ and you feed it with a bad ground?

Simple... its like crossing a battery.

So what happens to the wire when you do that? It heats up til no tomorrow.

You see why zotac wont honor RMA?

Its either the PSU shot out a bad ground on the PCI-E lane.. or the OP forced the EPS into the PCI-E Plug.

I first would like to add i am not here to offend anyone, i have been reading this and, i suggested that he call corsair and tell them what happend and would like to have the power supply tested to rule it out. and you want a letter sent back with the same power supply and you sent in, so that may, say, hey i sent this in to corsair they tested it nothing it wrong with it. so it is the card's what the deal now.

and the reason for this being is you could make them fit they make the connectors out of hard plastic. so unless your using brute force to the connecter it could of not made that mistake. but i am not ruling it out all the way.

I would like to add how could he make a 8 pin EPS go a in a 6 pin gpu connector.

because the ZOTAC AMP! GTX 460 has 2 6-pins no 8-pin

The corsair 850HX does not have a 6 pin EPS pin it has only a 8pin so how can you get a 8 pin EPS you would need alot of force. so the likely hood of him doing that are slim.

i am not saying anybody is right or no i am just following the the pictures