Asus P8P67 Pro caught on fire

[Allen]

As per my thread earlier today (which has been closed by Scan, something rather displeasing as I never got an answer to my question), does anyone know if I stand any chance of contacting Asus and getting a response from them?

Basically, it seems that when I built my PC back in April/May, I plugged the front USB port into the FireWire connection on the motherboard. Since the ports have never worked, I have not used them so I would assume that there was no power draw through. Despite this, and the fact the PC has been working absolutely perfectly for 6 months, the motherboard decided to randomly catch on fire, melting some components, and burning the case and the graphics card. All this whilst the PC was idling, I was at work when it happened, luckily the fire damage was isolated to within the case!

Now I am totally accepting that it was my fault the cable was plugged into the wrong connection, but I fail to believe that a) manufacturers would allow identical cables to be plugged in without some kind of power protection if the result was that the motherboard could catch on fire, and b) that since absolutely nothing was plugged in to the front USB ports there would be any power draw on the line which could cause this problem to occur in the first place.

So does anyone know if I have a leg to stand on? Does anyone understand the principles of motherboard electronics enough to provide an explanation why this could happen and if indeed it is 100% my fault or another component failure where it is just a coincidence that it happened in the same area of the motherboard as the FireWire connector?

[Guy]

I don't think you have a leg to stand on as well it was your own fault you plugged it into the wrong port. BUT no harm in contacting ASUS.

[Allen]

I don't think you have a leg to stand on as well it was your own fault you plugged it into the wrong port. BUT no harm in contacting ASUS.

But do you say that knowing with 100% certainty that plugging in a USB cable to a FW connector, even with nothing plugged in to the USB ports, can cause a fire? It's far beyond what I would call "motherboard damage" as the photo's quite clearly indicate.

If manufacturers knew this, don't you think they would make it more obvious than just making one black and one blue and a little note in the middle of a huge manual (that let's be honest, hardly anyone reads every word)?

[dangel]

On the up side, you saved yourself the hassle of "fault not found" when you do return something to scan One of the reasons I've abandoned them.

Checking pics on the web - yes the connectors are mechanically the same (which is horrible design no doubt) but the colour coding is sensible and there's a warning in the manual. As to why the cap blew - probably because the connector happened to short the wrong two pins together (earths can be linked on the cable). I feel very sorry for you - perhaps Asus will too You may well get a good gesture out of them but as to the letter of the law they're pretty much blameless.

I don't disagree (at all) with the idea that this can happen (if indeed it was the cause) is a major design flaw - not only that i'd say it's a safety issue too as things exploding carries fire risk/injury. Obviously the heat generated here was fairly hefty too.

I found a small reference here: http://forums.steampowered.com/forum....php?t=2071048 to the risk..

[kalniel]

Yep give them a bell and express your alarm at the severity of the resulting damage. Damage to the same component is one thing, fire risk is quite another, and not something that can be excused with a small disclaimer (just watch a few episodes of WatchDog).

[dangel]

Yup, on that you might have some sort of case - but at least give them a chance to do the right thing by you - not all companies have bad customer service and ASUS do trade on quality etc (personally i'm a big fan of them but this issue does cause alarm). Please do keep us posted on this one!

[Allen]

I will give them a call tomorrow, unfortunately I don't have the serial number with me at work and their lines close at 5pm.

[Agent]

Allen - the darkest part of the burn is directly under where (what looks like) a voltage regulator between the two slots as I mentioned in the other thread. Could you get another pic of this, maybe with macro mode?

I'm not saying it's the cause, but it looks like that might have been the main point of failure. I'm not convinced the fire actually started at the firewire header just yet

[Allen]

I can't find a macro setting on my phone. There's a close up, but no matter what I tried, I would not focus.

I will see what Asus say.

[peterb]

I had a look at both the picyures in the Scan thread and the manual for the mobo.

Without physically examining the board, it is hard to say what actually happened, but,from the wiring and pin out connections on pp 23-26 of the manual, it looks as if plugging in a USB cable to the IEEE 1334 socket would put a short cct on the 12 rail that is part of the IEE1334 specification.

The device Agent identified as a regulator looks more like a 1 ohm resistor (from the picture he linked to in the Scan thread) which has overheated and dropped from the board.

The marks on the case look like beads of molten solder that may have dropped from the board as the resistor overheated, and the marks that look like burn marks are the residue of the casing material that vapourised from the heat of the overloaded resistor. I'm not convinced that actual combustion took place, but there seems little doubt that something overheated. A good PSU will have shut down all rails in the event of the catastrophic failure of the device, preventing further damage, and I suspect that is what actually happened as the damage is limited to that area of the board.

The unknown, without looking at a cct diagram of the board is whether that component is linked to the IEEE 1334 connector circuitry. It seems reasonable to suppose that it is, and IF plugging in the USB cable caused a short cct on the 12 volt rail, and IF that resistor is pert of that 12 volt rail, AND there were no other current limiting devices in the short cct supply, and it is a 1 Ohm resistor then that resistor would potentially be dissipating 144 watts, so no wonder the casing melted and vaporised, and the solder on wires melted. But there is a lot of supposition there and one question is why it took six months to show up under those conditions

One answer could be that the resistor isn't the only current limiting component but as a result of continual overheating (because of the incorrectly connected USB cables) it gradually changed value/degraded to the point where it failed catastrophically giving the results shown in the photo.

The manual clearly states on pp vii that all cables should be correctly connected before using the product, as well as the warnings on pp23-26 that incorrect connection of the firewire or USB headers will damage the motherboard.

The lessons identified really are:

Always read the manual thoroughly before starting the build

Check and recheck the connections (apart from the different colour coded header, the mobo screen print labels it as IEEE1334)

Always use a good quality PSU that will provide a shut down capability in the event of a downstream fault /overload condition (which I think is what happened here)

Sadly not much help to the OP, who has my sympathy, but maybe prevent a similar accident to others

[Allen]

Thanks for the reply Peter, I was hoping for someone who really knew what they were talking about to be able to answer me.

My PSU is an "XFX P1-650X-NLG9 XXX 650W Modular", I would've thought that their PSU's would be pretty good quality.

Also, you say that you don't think combustion took place, well the fact my flat still smells of electrical fire 2 days after the event makes me think otherwise. Plus, I did attach a picture of the graphics card which is about 3-4 inches away from the area this happened (albeit above) that has been charred by whatever happened. The picture may not express it, but it looks like the kind of burning that is done to metal similar to when you hold a flame under a spoon to warm the brandy before pouring it over the Xmas pudding (I added that ref so you didn't think I was a junkie, lol).

I am going to call Asus soon to see what they say, although I am now beginning to feel that I will just have to, sadly, take this one on the chin. It's just a real shame that this has happened a mere 2 weeks before I move which is costing me in the region of about £2k and I just don't have anything spare to replace it, so I am left with some rather expensive mantle ware until I get it sorted.

[Allen]

BTW, if anyone wants to see, I have more photos here.

[Allen]

OK, they don't deal with component support over the phone, so I logged a ticket with them and expect an update in 48 hours.

[peterb]

Thank you for the link to the photos - very interesting, and it looks as if that component I identified as a resistor is still in place. It is also clear that the heat source was in the area of the SPDIF and IEEE 1334 connectors, The photos also show charring of the plastic connectors for memory and expansion cards.

PVC insulated cables and some of the hard plastics do give off very acrid and distinctive fumes when heated, and some of those (particularly PVC) give off toxic and acid fumes. PVC gives off Hydrogen Chloride, which dissolves in water to give hydrochloric acid, and there are also cyanide compounds. Looking at the GPU, the discolouration looks to be superficial and will probably clean off, and you should clean it off because of the risk of corrosion. Use plenty of Iso-propyl alcohol to wash the deposits off any contacts or metal work.

The discolouration of metal areas may be due to the acid effect of the fumes as well as carbon and soot given off by the charring process (similar to the carbon deposits left by a flame on the back of a spoon, as you mentioned earlier).

It would be interesting to see the back of the board if there is any damage there, but having seen those photos, I would revise my opinion about the resistor, and suggest that it might have been one of the circuit tracks that overheated, as a result of the plug error. But certainly having a good PSU probably stopped a bad situation from becoming worse.

To digress - incidents like this do show how easy it is to make an error. A couple of years ago a colleague came up to me with those dreaded words "You know a bit about computers don't you, Peter?". Although I tried to persuade him that he was confusing me with my evil twin brother, he persisted with a tale of woe in that he had been fitting more memory to a friend's computer, but when he plugged it back in and switched it on, nothing happened - apart from clouds of smoke coming out of the case.

He brought it in, and opening it up, it was obvious that something had overheated, all the insulation had melted off the PSU leads, which were severely discoloured. We stripped it down and found that the floppy disk connector had been displaced by one pin, so one of the supply rails was shorting out. The PSU was a really cheap and nasty one, the sort characterised by two letters or three letters, sometimes with a play on the American pronunciation, so that went in the skip.

A replacement PSU was ordered (an FSP - from Scan - of course) and fitted, all cables checked, switched on and... nothing. A fan twitch and that was it. We checked again, and we had done exactly the same thing, but this time, the PSU just shut itself down and when we corrected the problem, all worked properly. Again lessons were check and double check, and use a good PSU. (I have posted this before, so I hope the moderators won't take me to task for double posting )

I would make a couple of other observations (and this is certainly NOT a criticism of the OP)

Mobos are not really consumer items in that they are a component in a system. You can't take it out of the box and just use it, it needs to be part of an assembly, so with that in mind, it is beholden on the user to make sure that the build is correct. And that is not to put anyone off from building there own machine, but in doing that, you are largely responsible for your own support and mistakes, and as we have seen, mistakes do happen.


Which brings me to the final point, although in this particular instance it wouldn't have helped, and that is Scansure.

A lot of attention was focussed on Scansure, when it was introduced, mainly because of the way it is marketed. I have been as vociferous as any on that subject, but I do not wish to see that debate re-opened here, it is done and dusted (and I'll remove any posts bringing that up again in this thread). But, what that negative attention did do was to ignore the benefits of the product itself, which can remove a lot of the risk of self build, not just for new constructors, but also more experienced ones who might be putting together a high end rig, costing a lot of money that they can't afford to write off, and in those instances, Scansure might provide a level of security that is well worth considering. Obviously you should alsways ensure that an insurance product is suitable for your needs first - but this is NOT a thread or post about Scansure.

So that brings to an end my waffle on what is a sad story, so I hope that allen is successful in his talks with Asus, that lessons can be identified to help others, and above all, that no-one has been put off self build because of this incident.

[kalniel]

Which brings me to the final point, although in this particular instance it wouldn't have helped, and that is Scansure.

A lot of attention was focussed on Scansure, when it was introduced, mainly because of the way it is marketed. I have been as vociferous as any on that subject, but I do not wish to see that debate re-opened here, it is done and dusted (and I'll remove any posts bringing that up again in this thread). But, what that negative attention did do was to ignore the benefits of the product itself, which can remove a lot of the risk of self build, not just for new constructors, but also more experienced ones who might be putting together a high end rig, costing a lot of money that they can't afford to write off, and in those instances, Scansure might provide a level of security that is well worth considering.

Scansure doesn't look like it provides much security, not just in this case:

EXCLUSIONS
This insurance does not cover:

1. Any claim arising from abuse, neglect or malicious damage.
2. The VAT element of any claim if You are registered for VAT.
3. Any breakdown or internal cause whatsoever or damage that occurs
during the delivery of Your Computer Component.
4. Any action which invalidates the manufacturer’s warranty.
5. Any claim valid under the manufacturer’s warranty or which are subject
to a manufacturer’s recall.
6. Any claim where no fault is found with Your Computer Component.
7. Any claim made outside of the 28 day Period of Insurance.
8. Theft ,loss, wear and tear, fire, explosion, flood, lightening, storms, or
other bad weather conditions,
9. Cosmetic damage such as damage to paintwork, dents or scratches.
10. Caused damage caused by You failing to follow the manufacturers
instructions
11. Any other costs that are caused by the event which led to Your claim,
unless specifically stated in this Evidence.

Point 7 obviously applies, but even worse are points 4 and 10, which seem to rule out covering accidental damage as well (accidental damage would invalidate warranty, incorrect assembly would not be following manufacturers instructions).

[Chris P]

Scansure does cover instalation damage and if this had happened during the 28 day period then we would have replaced all of the components included under the policy.

Best Regards