Hi All, 
 
I just RMAed my EVGA 650w G2. I had a previous thread on the forum showing it was faulty. 
 
Anyway, the replacement arrived two days ago, it took a while to get working, and it involved resetting my CMOS.  Anyway, I got it working, the next day, I started it up, and downloaded an update for a game, and then it suddenly black screened from the loading screen.  I was a bit frustrated at this point, anyway, I ran a series of tests including the included psu tester, which it passed. But still, it felt exactly like the previous issue, with the faulty psu as the ethernet cable lights were on, and there was a slight move of the cpu fan when I pressed the power button.  I therefore, unplugged everything from the psu (but not from the mobo) apart from the 20 pin, on the theory that it might be cpu related and pushed the power button.  My graphics card immediately caught fire. Although the fire is out now, I'm still slightly in shock.  I don't know what caused the fire, mobo, psu or Gpu, or if it was my fault. Could someone let me know whether I should RMA the RMA, or if this is my fault.  If anyone could let me know which components are likely to be safe for reuse.  I am tempted to run a monitor directly from the motherboard but I am not sure if it safe. 
 
Picture of my GPU now: https:// cdn. discordapp. com/attachments/600461647072591909/741326090978590800/image0.jpg 
 
Best,

This kinda makes me think you might of had a different part of your system causing the problem even with the original power supply that was RMA'd  So if you use your power supply tester the power supply you have right now is working ok?

Hello black_falcon120
 
 
Sorry to see your system is having all the issues.
 
1) PSU does have safety circuits that can shut it down or stop it from starting ... if an issues occurs ... so, I do not know the root cause of the failure you mention in your post.  --> Was the failure to start a symptom of the GPU having an issue ?
 
 I'd call EVGA and get input from them as to your Next step in the process.
 
Testing is the only way to determine if a component is damaged after a close inspection for obvious issues
 
2) If It was mine, Id do a careful inspection of: MB, Case, GPU "fingers" that go into the MB PCIe slot
 
3) If only the GPU is damaged as shown in your picture --> & None of the PCIe "fingers" that plug into the MB shown sign of arcing or other damage.  No arcing on front or back of MB.........  then you might try the CPU's iGPU & see if your PC can safely boot.  I'd also unplug everything from the Case except the start button
 
--> You are taking a risk of causing More damage to your PC, if you miss damage or it is the PSU
 
Have a list of your Hardware handy when contacting EVGA, post it in this thread could be helpful
 
MB brand, Model,  Version & BIOS info
 
CPU
 
RAM
 
PSU:  EVGA 650w G2
 
Did you build the PC ?
 
What is your general experience level ?  first PC build, .................. built many PC this last year & previously (gives us an idea of what detail you may need)
 
 

Going to Try & move this to the Warranty section
 

sparetimepc
This kinda makes me think you might of had a different part of your system causing the problem even with the original power supply that was RMA'd  So if you use your power supply tester the power supply you have right now is working ok?

Possibly, but the original was definitely faulty, as it failed the psu tester.  The problem went from the cpu and case fans would spin up for 10 sec, to only the cpu fan would move perhaps 2mm.  
 
Just checked, the psu still works, but I can't see how one 20 pin connector is enough to cause literal flames to come out of back of my GPU's pcb. 

Cool GTX
Hello black_falcon120
 
 
Sorry to see your system is having all the issues.
 
1) PSU does have safety circuits that can shut it down or stop it from starting ... if an issues occurs ... so, I do not know the root cause of the failure you mention in your post.  --> Was the failure to start a symptom of the GPU having an issue ?
 
 I'd call EVGA and get input from them as to your Next step in the process.
 
Testing is the only way to determine if a component is damaged after a close inspection for obvious issues
 
2) If It was mine, Id do a careful inspection of: MB, Case, GPU "fingers" that go into the MB PCIe slot
 
3) If only the GPU is damaged as shown in your picture --> & None of the PCIe "fingers" that plug into the MB shown sign of arcing or other damage.  No arcing on front or back of MB.........  then you might try the CPU's iGPU & see if your PC can safely boot.  I'd also unplug everything from the Case except the start button
 
--> You are taking a risk of causing More damage to your PC, if you miss damage or it is the PSU
 
Have a list of your Hardware handy when contacting EVGA, post it in this thread could be helpful
 
MB brand, Model,  Version & BIOS info
 
CPU
 
RAM
 
PSU:  EVGA 650w G2
 
Did you build the PC ?
 
What is your general experience level ?  first PC build, .................. built many PC this last year & previously (gives us an idea of what detail you may need)
 
 

Thank you, my specs are: 
 
i7 6700
16gb RAM 
I did not build the pc, the full specs are here: https://support.hp.com/gb-en/document/c04818319 
 
The only alterations to it I made, was adding a Samsung SSD and adding the EVGA PSU which replaced a 400w if I remember correctly and started sparking. Just to clarify, the only EVGA thing in the case is the PSU, and the 980TI is a founders edition.
 
I have some skill, especially after my issues, I have: replaced a psu, taken out the GPU and RAM multiple times, but I have not built a whole PC. 
 
Looking at the GPU, the fingers appear fine, apart from one strand of the 'trace' appears to have disappeared on the edge, but I don't think that's a big problem as it did boot yesterday... 
Thank you for your help. 

Cool GTX
Going to Try & move this to the Warranty section
 

At the moment, I'm also interested in what is safe for me to do, as well as the warranty issue, will I get that help there? 

sorry I'm having trouble posting on the Forums today - other members have mention the same "Forums issues"
 
 
Did you Replace EVERY wire when you changed the PSU ---> Using Only the cables that came with that PSU ?
 
You see, not Every PSU has the same outputs & the cable set is "custom" to that PSU design ... so the correct Voltage & ground connections happen @ the device end
 
Note: yes one of the contact "fingers" on the card edge that goes into the PCIe bus on the MB is shorter that the rest of them -- this is normal

Cool GTX
sorry I'm having trouble posting on the Forums today - other members have mention the same "Forums issues"
 
 
Did you Replace EVERY wire when you changed the PSU ---> Using Only the cables that came with that PSU ?
 
You see, not Every PSU has the same outputs & the cable set is "custom" to that PSU design ... so the correct Voltage & ground connections happen @ the device end
 
Note: yes one of the contact "fingers" on the card edge that goes into the PCIe bus on the MB is shorter that the rest of them -- this is normal

The only PSU cables I have came with the PSU originally, when EVGA replaced it, they were very specific that they were only replacing the pass, not any of its accompanying cables.  So they are still the same and they are being used with the exact same model (EVGA G2 650w).
 
thank you for clarifying, I'm still slightly in shock over seeing a literal flame on my GPU...
 
Also, I had a look at the motherboard, and there appears to be maybe only a small burn mark on the motherboard, it is very small, and I can't tell if it is a burn mark or not.
 
edit: here's the image https://cdn.discordapp.com/attachments/600461647072591909/741367744666271815/image0.jpg 
 
The burn mark is to the right of the lower cut in the metal thing. 

It's not unheard of for those video cards (GTX 980Ti) to short out or even catch on fire/burn up..Try/get another card and everything will hopefully be fine..In the meantime just connect your monitor to the motherboard and use the  Integrated Processor Graphics the i7-6700 has.

black_falcon120

sparetimepc
This kinda makes me think you might of had a different part of your system causing the problem even with the original power supply that was RMA'd  So if you use your power supply tester the power supply you have right now is working ok?

Possibly, but the original was definitely faulty, as it failed the psu tester.  The problem went from the cpu and case fans would spin up for 10 sec, to only the cpu fan would move perhaps 2mm.  
 
Just checked, the psu still works, but I can't see how one 20 pin connector is enough to cause literal flames to come out of back of my GPU's pcb. 

Yeah i meant i was just wondering if another component took out your power supply that had to be sent in, or was it a brand new one that was faulty from the get go? Yeah this part i circled does not look right to me. Those traces are leading directly to the PCIe slot.