Power Supply blew out my motherboard. Will EVGA cover it?

[arnavvr]

As the title states, I woke up, and my PC wouldn't turn on. Weird right? I tried clearing CMOS, reseating the RAM, and nothing seemed to work. And then I took a closer look. I noticed a definite burning smell, and saw that the 24pin connector had fried, and that the motherboard was unusable, and the power supply probably is too. My question is, is EVGA liable for the damage that their power supply caused to the rest of my system? It certainly isn't Gigabyte's fault, but the motherboard also isn't EVGA's product. 

 

What do manufacturers do in the event of a situation like this?

[NinJake]

Give them a call. EVGA has provided me with the best customer experience and I hope they continue to do so to others. Gigabyte is also a reputable brand. Hope all gets sorted.

 

My main question is what PSU and what Mobo are you using?

[Semper]

EVGA's warranty states that the warranty exists only for their unit, they don't warranty the failure of other components as a result.

You can attempt to see if they will cover damages, however I find it unlikely that they will. I was unsuccessful.

[CommanderAlex]

I don't think EVGA can do a lot, but you can give it a shot. This is from their warranty page on power supplies.

 

https://www.evga.com/warranty/power-supplies/

[Guest]

11 minutes ago, arnavvr said:

As the title states, I woke up, and my PC wouldn't turn on. Weird right? I tried clearing CMOS, reseating the RAM, and nothing seemed to work. And then I took a closer look. I noticed a definite burning smell, and saw that the 24pin connector had fried, and that the motherboard was unusable, and the power supply probably is too. My question is, is EVGA liable for the damage that their power supply caused to the rest of my system? It certainly isn't Gigabyte's fault, but the motherboard also isn't EVGA's product. 

 

What do manufacturers do in the event of a situation like this?

 

 

That's unusual.  That's a +3.3V and ground.

 

What are the specs of the PC you're powering with this and what EVGA power supply are you using?

 

[arnavvr]

Just now, jonnyGURU said:

That's unusual.  That's a +3.3V and ground.

 

What are the specs of the PC you're powering with this and what EVGA power supply are you using?

 

Intel Pentium G4400

Gigabyte Z170X-Gaming 7

2x Radeon RX480

EVGA 850G2

[Guest]

9 minutes ago, arnavvr said:

2x Radeon RX480

Interesting.

 

Are you mining crypto with this?

[arnavvr]

Just now, jonnyGURU said:

Interesting.

 

Are you mining crypto with this?

Yes.

[arnavvr]

4 hours ago, jonnyGURU said:

That's unusual.  That's a +3.3V and ground.

 

What are the specs of the PC you're powering with this and what EVGA power supply are you using?

 

Looks like -12V and ground but you're probably better than me at this  

[Guest]

5 hours ago, arnavvr said:

Yes.

Yeah.  That's why.

 

They're not going to cover that.  

[Guest]

14 hours ago, arnavvr said:

Looks like -12V and ground but you're probably better than me at this  

-12V is on the clip side of the 24-pin.  Sounds like you don't know what you're doing.  Maybe you should reconsider what you're doing here. 

 

[boghubodaghi]

Hmmm... 3.3V, huh? Why didn't the OCP kick in?

[Guest]

9 hours ago, boghubodaghi said:

Hmmm... 3.3V, huh? Why didn't the OCP kick in?

Because the board wasn't drawing too much current over all.  Just too much current on a single +3.3V lead.  That's why this isn't a PSU problem.  This is a motherboard problem.

 

My snarkiness aside (for the benefit of OP and others that may run into this).....

 

This board pre-dates GPU mining by 1 year.  It's a "gaming" motherboard.  While it's made for 3-way SLI/Crossfire with it's three x16 slots, you have to remember that 2x and 3x SLI/Crossfire does not scale 2x and 3x and does not require 2x and 3x the power.  It seems that the +3.3V pins on all of the PCIe slots are fed by the same wire off the 24-pin (would be good to get a set up and put a clamping ammeter on the +3.3V to confirm).  

 

Mix this with an equally old graphics card with an unusually high load on the +3.3V and multiple by two and you're going to burn up power supplies.

 

OP is actually lucky.  I would just buy a replacement 24-pin cable, a new motherboard, put it all back together again and try again.

 

[Mark Kaine]

On 3/2/2021 at 5:41 AM, jonnyGURU said:

Yeah.  That's why.

 

They're not going to cover that.  

So in theory, a PSU manufacturer would have to cover the damage done by their product *if* you can prove that it was actually the PSU's fault...? 

 

I mean, that's interesting with all the talk about 'protection' you'd think they're liable, but the difficult part would then be to actually prove that it was because of the PSU - which is probably difficult for the average consumer...?

 

Or are they just not liable at all, generally? 

 

 

 

[Guest]

5 hours ago, Mark Kaine said:

So in theory, a PSU manufacturer would have to cover the damage done by their product *if* you can prove that it was actually the PSU's fault...? 

 

I mean, that's interesting with all the talk about 'protection' you'd think they're liable, but the difficult part would then be to actually prove that it was because of the PSU - which is probably difficult for the average consumer...?

 

Or are they just not liable at all, generally? 

 

 

 

A lot of manufacturers will cover incidental damage.  

 

Do they HAVE TO?  No.  Not in all countries.  Some countries, you'd have to take them to court.

 

[Mark Kaine]

47 minutes ago, jonnyGURU said:

A lot of manufacturers will cover incidental damage.  

 

Do they HAVE TO?  No.  Not in all countries.  Some countries, you'd have to take them to court.

Interesting also, but yeah, in the end it will depend on laws, or 'goodwill' of the manufacturer. 

[Guest]

40 minutes ago, Mark Kaine said:

Interesting also, but yeah, in the end it will depend on laws, or 'goodwill' of the manufacturer. 

So, I know for Corsair for example, and probably many others, in order to properly "budget" for damage claims, you simply get an insurance policy specifically to cover costs incurred by your own products.  I think it's called "General Liability Insurance".  Works just like car or home insurance.  You pay every month, hope you never need it, then you wreck your car, the washing machine floods or your power supply catches someone's motherboard on fire. 

 

[Mark Kaine]

20 hours ago, jonnyGURU said:

I think it's called "General Liability Insurance".  Works just like car or home insurance. 

yeah, I figured that's how it works... probably not cheap! 

 

 

would be good to have a list of manufacturers who provide this sort of help... this isn't a request or anything, just saying info like this is often difficult to find, as it happens with computer stuff often, I'm currently trying to resurrect an old AM2 Vista PC for example, and it's difficult, so many standards within the same generation seemingly, etc... or even just finding out which standards are supported with what cables for my new 'gsync compatible' monitor ain't as easy as you'd think... 

 

 

 

[FalconFour]

It's worth tossing the note in here, that it doesn't look like anything is "blown out" at all. The only smoke you let out is plastic smoke, not magic smoke. Replace the burned pins on the motherboard connector, and replace the burned pins on the PSU connector (preferably with a better connection), and put it back together, almost guaranteed it'll boot up and work fine. Or just permanently marry the PSU to the motherboard and solder it directly to those pads (3.3v?) if you're any good at soldering (please don't try this with one of those pencil-stick destruction derby RadioShack special irons - use a TS100 at 24v with a good D24 tip, or something).

 

Just popping in here to say your board is probably fine if you can just find another way to get those burned pins connected again (after removing/cleaning up the burn). ;) It was just a loose/bad connection that caused it - possibly over-current as well, but having the connection point burn, and not the wire itself, suggests a crappy contact to me.

[Guest]

19 minutes ago, FalconFour said:

It's worth tossing the note in here, that it doesn't look like anything is "blown out" at all. The only smoke you let out is plastic smoke, not magic smoke. Replace the burned pins on the motherboard connector, and replace the burned pins on the PSU connector (preferably with a better connection), and put it back together, almost guaranteed it'll boot up and work fine. Or just permanently marry the PSU to the motherboard and solder it directly to those pads (3.3v?) if you're any good at soldering (please don't try this with one of those pencil-stick destruction derby RadioShack special irons - use a TS100 at 24v with a good D24 tip, or something).

 

Just popping in here to say your board is probably fine if you can just find another way to get those burned pins connected again (after removing/cleaning up the burn). It was just a loose/bad connection that caused it - possibly over-current as well, but having the connection point burn, and not the wire itself, suggests a crappy contact to me.

It's just going to burn out again. 

 

[FalconFour]

14 minutes ago, jonnyGURU said:

It's just going to burn out again. 

 

Not if you hard-wire it, lol. Like I said, because the connector burned out and not the wire, it really seems like the interconnect was the weak/high-resistance link here. I see this [poop] all the time in the EV world. Connectors are seriously not treated with the respect/analysis they need. Often all this amperage just goes through a teeny-tiny little concentrated point (because the connection is damaged/weak/crappy and only holding on by a thread), it gets hotter as a result, further increasing resistance, things burn out, and "Oh golly gee how did we get here, must have been overcurrent"

[Guest]

52 minutes ago, FalconFour said:

Not if you hard-wire it, lol. Like I said, because the connector burned out and not the wire, it really seems like the interconnect was the weak/high-resistance link here. I see this [poop] all the time in the EV world. Connectors are seriously not treated with the respect/analysis they need. Often all this amperage just goes through a teeny-tiny little concentrated point (because the connection is damaged/weak/crappy and only holding on by a thread), it gets hotter as a result, further increasing resistance, things burn out, and "Oh golly gee how did we get here, must have been overcurrent"

I don't disagree with you.  Just the fact that it's a +3.3V and adjacent ground and that the board is being used to mine which means those cards are full bore 24/7 makes me think it's just current.  I mean.. think of how hot those pins had to get.  Over 70°C! 

 

I do like the idea of soldering the wires in, though.  It's an old board and it's not like it's worth anything.  I once did a mod where I soldered all of the power wires to the back of the motherboard.   So the board would power up, but people couldn't figure out how.  To add more mystery to it, I didn't use an ATX power supply.  I used one of those 5.25" bay PSUs hooked up to a pico-PSU.  Great conversation starter at LAN parties.  People would ask how it was powering, but I would never show them what I did. 

[legcramp]

The two RX 480s obviously pulled too much power from the PCI-E slot and this happened. 

[chocolatekarma]

Just as I thought...  most miners have no idea WTH they're doing.

[FalconFour]

21 hours ago, SGT-AMD said:

Instead of melting the connector, you will start burning traces.

That is another type of genie smoke you can never let out!

True, perhaps

 

But consider this: the motherboard is designed with those slots, and each slot has a current rating. It's not likely that the motherboard designer just YOLO'd it and thought "nobody would ever legitimately use these slots", nor that the GPU would YOLO it and be like "nobody will ever... use this GPU". They're designed to work together, by spec, and the ratings are built to maximums.

 

It seems extremely dubious to claim that the GPU must've been responsible for the burn for 3.3v reasons - GPUs get their power from the 12v 6- or 8-pin connectors, and have onboard conversion to power the bulk logic from that. 3.3v from the PCIe slot is just there to power the bus logic, which is basically a fixed figure (card is there -> card uses a fixed amount of power to exist and interact on the bus).

 

My money still holds on a bad connector. Besides, if they repair it this way, it's living a second life already. If the next thing that goes is a trace (still improbable bordering on impossible), then what're you out? One motherboard that was already destined for the bin before trying it