Zotac - GTX 460 SLI, up in smoke!

[Pontius Dilate]

nVidia's reference fan for the GTX 460 wasn't centrifugal either. nVidia also certifies their partners' cards for SLI with non-centrifugal fans and bridges which force sandwiched configurations. My point is that if either nVidia or their board partners (or AMD for that matter) thought that there was even a slight chance of the cards catching fire in an ordinary side-by-side configuration they never would've designed them to be used that in that way.

 

[mogman]

Happy Easter all, I apologize for the delay in responding, I've been out of town visiting friends for the holiday.

Thanks for all the comments thus far, I've posted a brief update with further details based on your questiosn along with some initial Zotac e-mail communication.

 

[Seero]

I just want to make some observations here.

Regarding the matter of placing the cards in the 1st & 3rd PCIe x16 slots, those are the only full x16 electrical slots on the board. The 2nd & 4th PCIe x16 slots are actually x8 electrical. For full x16 links to both cards in SLI you have to use the 1st & 3rd slots. This is true for many other motherboards. Also for 3-way SLI the cards have to be sandwiched. nVidia-supplied 3-way SLI bridges require that the cards are sandwiched, and some 2-way SLI bridges require that the cards are sandwiched. Therefore while we could say that this is not an ideal configuration (especially for the top card) it is certainly a fully approved and expected configuration by the hardware designers and manufacturers.

I don't see how the southbridge heatsink could possibly get hot enough to help start a fire on the cards or the board. That just doesn't sound possible to me. The board would have shut down due to electrical problems before that happened I would think.

I'm no forensic scientist, but it looks to me like flame shot out from the lower x16 PCIe slot and burned the back of PCB of the lower card and the heatsink shroud of the upper card. A sudden massive short in the lower card's PCIe slot which caused an intense electrical fire before the computer shut down? It's not at all clear that the card itself is specifically to blame. The OP could just as easily be requesting a new board from Gigabyte.

I'm too no forensic scientist, but I personally found this case extremely interesting.

I see what you are saying and agree. Those HS shouldn't be hot enough to ignite as they are attached to a chip with some thermal production. However, that doesn't mean they can't emit heat beyond the temp which will trigger the production. For example, have a heat gun can heat up things that can exceed the temp of the gun itself. Assuming the air flow is dead, heat could have radiated off video HS onto the cover from inside, while outside the cover is touching the NB HS, which again become to junction point of 2 source of heat. Yes, even than it shouldn't ignite, but possibly hot enough to melt stuffs, i.e. plastic in around that area, which is the setup of the incident. By looking at the clip of 4th PCIe, it is clear that de-shape(melting) is the result of pure heat, not fire, as there are no burn marks. Yet, the 4th PCIe was so far from the point in question, if the 4th PCIe de-shape was due to the point of ignition, then we can see how hot it must be at the point of ignition.

Heat may not be the caused the ignition. As you suggested it, may probably be a power scourge coming from the 2nd PCIe slot. Look at the 2nd PCIe, it is de-shaped so bad that the socket itself isn't straight. It is possible that those pin were touching each other within the connector, which caused a short so intense that it actually ignite the cover.

However, if that is the case, then the 2nd PCIe socket should be melted too. Look at the melted cover, there is a dent at the location away from the PCIe slot. Coincidentally, that dent is on top of the part where the NB HS white spot. My guess is, the card actually melted before the ignition through time, and the part melted onto the HS actually trapped heat, and therefore ignited. I'm not a chemist either so i don't know the ignition point of that type of material, all I know is that type of material produce huge amount of smoke which actually extinguish the fire upon ignition.

Of course, my assumption is the dent is where the ignition occur, but that doesn't explain the mark underneath the fan.

 

[Seero]

Happy Easter all, I apologize for the delay in responding, I've been out of town visiting friends for the holiday.

Thanks for all the comments thus far, I've posted a brief update with further details based on your questiosn along with some initial Zotac e-mail communication.

Happy Easter.

 

[aigomorla]

ummm...

his bad psu caused bad spike voltage, which caused his card to catch fire.

The reason why Zotac probably isnt honoring the RMA, is because damage like that is more traced to the PSU.

So u cant expect Zotac to replace what was working, and died due to a psu that ended up taking the entire system with it.

 

[mogman]

ummm...

his bad psu caused bad spike voltage, which caused his card to catch fire.

The reason why Zotac probably isnt honoring the RMA, is because damage like that is more traced to the PSU.

So u cant expect Zotac to replace what was working, and died due to a psu that ended up taking the entire system with it.

Just finished testing the original Corsair HX-850 power supply with a spare LGA 775 system, powers on a full system with a 2nd spare 8800GT.

The only failed/malfunctioning components from my system are now the two Zotac cards!

 

[3DVagabond]

On Friday, 4/22 at 1:09pm EST, I received the following update:

From: Hansel von Muller <hansel@zotacusa.com>
TO: Chris Messer <c*****@gmail.com>
CC: Andrew <andrew@zotacusa.com>, Luis Molina <LuisM@zotacusa.com>

Date:Fri, Apr 22, 2011 at 1:09 PM
subject: RE: RMA Request & Follow Up
mailed-by: zotacusa.com

I have looked into this issue and had it also reviewed by my superiors.

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

I apologize for the inconvenience

Hansel von Muller
Tech Support Lead
ZOTAC USA
877-59-ZOTAC (96822) ext. 605

What are they trying to say the damage is not due to the failure? The failure was due to the damage? That's the only way they could deny it on the grounds they are claiming. Strange response.

 

[MustangSVT]

Certainly is an interesting case. There was a flame inside computer.. I would imagine things would crash/lockup waay before anything reaches ignition level. Also from the detailed investigation by Seero (GJ btw.), motherboard seems to be the culprit.

Is there a reason you are suspecting videocard instead? is the mobo out of warranty?

 

[KaosFaction]

What are they trying to say the damage is not due to the failure? The failure was due to the damage? That's the only way they could deny it on the grounds they are claiming. Strange response.

They are claiming:
"Warranty will be voided if physical damage occurs to the PCB, chipset, or components caused by improper installation, damage during modification or installation of aftermarket attachment(s), or water or chemical damage of any kind."

As per their warranty. I think they believe it wasn't the cards that caused the damage, but damage had been done to the cards (Read Seero's analysis).

Has OP contacted Gigabyte yet?

 

[thilanliyan]

Call me cynical but companies will do whatever they can to get out of honouring a warranty.

His cards have physical damage. Regardless of whether it was caused by the cards malfunctioning, they will claim there is physical damage and so can't honour the warranty.

Finding the actual cause of the fault itself is so difficult in this case, weakening the OPs case even more.

 

[Pontius Dilate]

I don't think Zotac could possibly make any determination as to fault without at least the motherboard, both cards and the power supply. I can understand Zotac denying warranty service under the circumstances, but I would at least extend the offer to Mogman to send in the cards/motherboard/power supply for examination with the understanding that they aren't guaranteeing replacement and will return the items with their diagnosis. Everybody is afraid of lawsuits however.

Mogman, is there any chance you could take pictures of the motherboard and both cards in good lighting with something other than a cell phone? Also, what do you think happened?

 

[Idontcare]

Call me cynical but companies will do whatever they can to get out of honouring a warranty.

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.

 

[Gigantopithecus]

This is precisely why when friends and acquaintances ask me about saving a few bucks by going with a Zotac video card instead of something from EVGA or XFX, I insist they spend the extra money on a product that won't burst into flames (this isn't the first time I've seen a Zotac card ignite). Their rebate practices are deceptive to the point of borderline illegality, their RMA unit will do everything to avoid replacing dead cards, and they make **** products.

Caveat emptor.


Profanity is not allowed in the technical forums.

Idontcare
Super Mod

 

[nanaki333]

This is precisely why when friends and acquaintances ask me about saving a few bucks by going with a Zotac video card instead of something from EVGA or XFX, I insist they spend the extra money on a product that won't burst into flames (this isn't the first time I've seen a Zotac card ignite). Their rebate practices are deceptive to the point of borderline illegality, their RMA unit will do everything to avoid replacing dead cards, and they make **** products.

Caveat emptor.


Profanity is not allowed in the technical forums.

Idontcare
Super Mod

where else have you seen zotac cards ignite? links?

 

[yacoub]

They are claiming:
"Warranty will be voided if physical damage occurs to the PCB, chipset, or components caused by improper installation, damage during modification or installation of aftermarket attachment(s), or water or chemical damage of any kind."

As per their warranty. I think they believe it wasn't the cards that caused the damage, but damage had been done to the cards (Read Seero's analysis).

Has OP contacted Gigabyte yet?

Curious to hear the answer to that as well.

 

[slag]

I did see that, however the thing that sticks out as weird to me (and maybe it's just me), is, he said:

"I came home from work on Tuesday, 4/19, to find my system shutdown and a strong burning smell in the air."

So...you left your system on all day while you were at work.....why???

Something doesn't add up...but again, could be reading too much into it...

??

I think a large majority of people leave their computers on 24x7. When I need to use it, i don't want to wait the 30 seconds to a minute for it to be ready. I move the mouse, the screen comes on, and I use it.

 

[nanaki333]

i just read where you have the original motherboard and cpu working just fine. zotac denied your rma due to physical damage... well no crap there's physical damage! they caught fire! the psu could still be the culprit, but i don't know... sounds more like the card.

 

[aigomorla]

What are they trying to say the damage is not due to the failure? The failure was due to the damage? That's the only way they could deny it on the grounds they are claiming. Strange response.

there trying to say the level of damage could not have been from just failed cards.

The damage was most obviously done by some type of power problem, which can be traced more at the PSU end then the video card end.

For the card to light up like that OP, it needs more power then what the PCI-E can provide.

That Giggy board doesnt even have AUX power imputs on the board, so its a physically impossibility that the board gave out that much power itself to fry the cards in the degree of damage it recieved.

This is why zotac wont honor the RMA.
They are saying its not there fault... its YOUR PSU's fault.

And i kinda see Zotac's option...

where else have you seen zotac cards ignite? links?

exactly...

Have you guys seen what it takes to ignite a video card?
Because I HAVE.

They dont do that anymore... they used to burn right though the PCB into something even more nastier then the OP showed.

But the OP's damage is traced at the power points.
Its OBVIOUSLY a PSU fault... which overspiked it up the butt, and then caused the PCB on the card to run more current then it was designed which in turn melted the card.


THIS TYPE OF DAMAGE CAN NOT HAPPEN ON JUST THE PCI-E LANE ALONE.
IT HAS TO BE TRACED TO THE 6/8 PIN POWER ON THE PSU.

^ this is Zotac's comment on it....


Now show me another Zotac up in smoke... then u can probably fight an RMA.

 

[Gigantopithecus]

where else have you seen zotac cards ignite? links?

Friend of mine who's local and swaps parts with me frequently had a Zotac 9800GT VRM blow on him and melt out part of the PCB. I'll see if he has pics and post them if he does.

 

[aigomorla]

Friend of mine who's local and swaps parts with me frequently had a Zotac 9800GT VRM blow on him and melt out part of the PCB. I'll see if he has pics and post them if he does.

your section is not the VRM.. its the PCI-E lane itself.

i was talking to a friend who works RMA.. infact even zap worked RMA.

To be honest, it looks like u might of messed up, plugged in the 8pin ATX instead of PCI-E... brought wrong ground wires in your card.. and your pci-e exploded...

as the board got super heated.. the fan you had up top gave more air igniting the cards.. which then you had a fire..

the moment card 1 or card 2 hit thermal shut down, it turned your system off..

That is why u came home with it off..

 

[aigomorla]

Just a closing comment from me..

OP... I still think u had about a 70-85&#37; oops factor at this moment.

Its very hard to prove that you didnt miss install the cards with that level of damage.

Now dont get me wrong, We had a problem like this on boards a while back.
LGA1156, and foxconn sockets melting under load.

But the only way to prove this, was after we saw quite a bit of people with the same problem.

The only way your going to get an RMA accepted, is to show the vendor, your not the only one with this type of damage.

And at this point, im gonna say, without using a waterblock, your level of damage is very difficult to get, in a properly setup system.

 

[dma0991]

This is really interesting as I have not seen such problems happen before. I guess I can still learn something new everyday.

From what I have experienced before, I made a connector for my side fan and the connector had exposed contacts. The exposed metal touched my casing and my PSU tripped immediately as in a complete shutdown before any real damage happened. The PSU was only able to turn ON after I had turned OFF the main for a few seconds. All of this happened in an instant and my PSU didn't wait till something was burning.

This leads me to believe that OP's main problem lies with his PSU which I am wondering why did it not shutdown immediately in case of a problem. From the pictures I can see that the problem started with the top card which burst into flames and causing burn marks and soot on the card below.

There can be a possibility that the temperature sensor on the top card to be damaged before the the flame even started (from a possible surge?) and delaying GPU overheat to shutdown. A few seconds later, the temp sensor on the card below detects overheating and shuts down the computer.

OP is pretty lucky to have the flames died off before the flame got bigger and burnt down the house. It is a tragedy nonetheless but I doubt that Zotac will take responsibility as the card might not be the cause of the problem in the first place.

 

[notty22]

I think a edit to the OP is confusing me, there is reference to 3 cards now. I'm pretty sure there was only 2 , is that correct ?
My first thoughts were there was some kind of dead short /over volt arcing that happened to the lower card, across the pci-e gold contacts, and the psu fed that, without tripping for a amount of time ,unknown.
If the PSU still works, I'd guess eventually a power off signal from the m/b is the only thing that powered down the psu, and stopped feeding the dead short.

 

[Gigantopithecus]

your section is not the VRM.. its the PCI-E lane itself.

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.

 

[Gigantopithecus]

This leads me to believe that OP's main problem lies with his PSU which I am wondering why did it not shutdown immediately in case of a problem.

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?