Those molex adaptors were always shit, not really designed for PCI power if you ask me but whatever.

Does the PSU/GPU still work as far as you're aware?

I'm going to assume this is a single-rail PSU as a multi-rail PSU would have tripped the OCP before any of that damage occurred (or maybe the original owner just kept pushing it past it's limit lmfao.) I also noticed you have adapters on the cable. Adapters are really a gimmick. You should honestly consider upgrading your PSU before it ends up killing your GPU. I know you mentioned you turned the settings down, but that's not going to help the case. If your PSU is shitting out like this, high and low settings aren't going to stop it from eventually taking hardware with it once it decides to die.

Someone correct me if I'm wrong here with a guess or any information I me be wrong with here.

I believe video cards have a BIOS programming that allows them to regulate and dictate how much voltage that is going to the card from the PCIe lanes any GPU connectors such as the 6/8 pin. It's possible whoever owned the PC before you tweaked that setting that allows it to pull more than it really should; that or, the GPU failed or had a quick hiccup where the programming decided it needed more voltage and this resulted in the issue. That's a long shot and I may have explained it rather poorly, but it's a possibility.

Yes the graphics card still works also would you be able to link me to a wire I should use?

I was wondering if it was the psu because it doesn't seem to be made from a know maker and ive tried searching it online but had no luck.

Mrv8gaming4K wrote Yes the graphics card still works also would you be able to link me to a wire I should use?

I was wondering if it was the psu because it doesn't seem to be made from a know maker and ive tried searching it online but had no luck.

I would upgrade the PSU regardless. It you're being forced to use Molex-to-6pin adapters, then it's time to change the PSU. I would be really careful about using the PSU though. Just because nothing hasn't happened after it melted your cable; doesn't mean something isn't going to happen. Getting a recent PSU, be that semi-modular or full-modular or even non-modular, will not force you to use Molex-to-6pin adapters as it will have an 8-pin connector that is 6+2. 6 pins are connected together like your adapter is, however there is an extra 2 pins that barely hang off the edge. This is for people who have GPUs that need 8-pins; however, you can connect the 6 and let the extra 2 slightly hang over. It will not cause you problems.

TaigaAisaka wrote
I believe video cards have a BIOS programming that allows them to regulate and dictate how much voltage that is going to the card from the PCIe lanes any GPU connectors such as the 6/8 pin. It's possible whoever owned the PC before you tweaked that setting that allows it to pull more than it really should; that or, the GPU failed or had a quick hiccup where the programming decided it needed more voltage and this resulted in the issue. That's a long shot and I may have explained it rather poorly, but it's a possibility.

I'm an mecp who works in automotive but electrical theory is still semi relevant. a 14 gauge wire should accommodate about 15 amps at 12v. So unless this card is drawing like 200 watts on 12v, that shouldn't be anything close to enough to melt like that.
If the voltage to the GPU is too low, the current flowing through the GPU will increase. Vice versa.
I would also suspect maybe an inadequate motherboard. Im not sure on the path electricity flows in a PC but what leaves the PSU must come back... not sure if that happens via solely the 6 & 8 pins or if it actually cycles from those 6 & 8's through the card to the PCIe on the mobo, back to the main ATX plug....


What I would do
Re-seating all electrical pigtails and the GPU.
Id also try a new adapter, maybe on a different 4 pin strand if possible. The cheapest thing to try. After booting, grip the new adapter. If you feel heat, something is up. Shouldn't have to wait long.
Then maybe try the card in another PC.
I would try a PSU.
Or just buy a decent office PC on craiglist or kijiji. throw your card in there.
Likely the cheapest easiest solution lol, after verifying its not the card.

I just diagnose and fix car electrical stuff and PC's are just a hobby so this is just my $0.02

Jodabruh wrote

TaigaAisaka wrote
I believe video cards have a BIOS programming that allows them to regulate and dictate how much voltage that is going to the card from the PCIe lanes any GPU connectors such as the 6/8 pin. It's possible whoever owned the PC before you tweaked that setting that allows it to pull more than it really should; that or, the GPU failed or had a quick hiccup where the programming decided it needed more voltage and this resulted in the issue. That's a long shot and I may have explained it rather poorly, but it's a possibility.

I'm an mecp who works in automotive but electrical theory is still semi relevant. a 14 gauge wire should accommodate about 15 amps at 12v. So unless this card is drawing like 200 watts on 12v, that shouldn't be anything close to enough to melt like that.
If the voltage to the GPU is too low, the current flowing through the GPU will increase. Vice versa.
I would also suspect maybe an inadequate motherboard. Im not sure on the path electricity flows in a PC but what leaves the PSU must come back... not sure if that happens via solely the 6 & 8 pins or if it actually cycles from those 6 & 8's through the card to the PCIe on the mobo, back to the main ATX plug....


What I would do
Re-seating all electrical pigtails and the GPU.
Id also try a new adapter, maybe on a different 4 pin strand if possible. The cheapest thing to try. After booting, grip the new adapter. If you feel heat, something is up. Shouldn't have to wait long.
Then maybe try the card in another PC.
I would try a PSU.
Or just buy a decent office PC on craiglist or kijiji. throw your card in there.
Likely the cheapest easiest solution lol, after verifying its not the card.

I just diagnose and fix car electrical stuff and PC's are just a hobby so this is just my $0.02

I believe his GPU pulls an estimated 170w, before any overclocking calculations (if any.) OP didn't list his motherboard, but looking from the pictures, it does look rather old as it's using the green paint on the board when almost all recent ones have been black; plus paired with the blue and green DIMM slots. Very well a possibility that his motherboard isn't able to support a certain threshold of a power draw. Looking at the Molex cables he has; he has 2-3 Yellow cables and 1-2 Black cables. The main focus here is yellow as it's +12VDC. Adding onto that, if even one pin isn't crimped properly in his Molex cable, he could very well be hitting a higher AMP threshold. I don't know the exact number for Molex, but I know SATA cables that aren't crimped properly; which, are mostly shitty Chinese knock-offs, they can reach about 20-ish AMPS and that will mostly likely take your HDD or SSD with it if it short circuits. That one is an extreme long shot and I really doubt that is the issue at hand, but who knows. This also goes all back to the PC's previous owner. He could have tweaked around in BIOS or even the GPU settings and OP wouldn't know the difference unless he looked in there himself. The cable melted before OP got it and is just now noticing it, because that's a smell you wouldn't be able to miss. Whatever the previous owner did is unknown, but it's very easy to tweak voltage and power draw your PC can intake; GPU included. I'm guessing the previous owner did that - while using the Molex adapter - smelled that something was melting, panicked, turned it off/reset the settings he did either in BIOS or something like MSI Afterburner and just let it be.

I would disagree about him trying the adapters on a different connector. Craig already mentioned those Molex adapters are shit, and he's right. They are utter shit. I can find countless posts about people having wires melting when using Molex adapters, hell, I can find things that have been burnt from a fire caused by those adapters. They're absolute utter shit and I wouldn't doubt if the person who sold it to OP got some really cheap ones. He's already "playing with fire" by being forced to use the Molex adapter even after the cable melted; his PSU could shit out and it could take hardware with it.

Buying an office PC also isn't really a solution. The way you mention to slap a video card, makes it sound like that would become his new PC. Pre-built computers have been known to have shit quality PSUs as well as extreme budget motherboards. He also has to account for the case because his GPU may not fit in the case. Those couple hundred dollars he would spend on a pre-built office PC could otherwise be used to get a new PSU, and possibly get a Ryzen 3 CPU and AM4 motherboard or get a Pentium g4560 and an LGA 1151 motherboard. I'd say the PSU is the most important right now, because he needs to get a PSU with adequate connectors as he should not have to use any adapters to power his PC. I'm even skeptical about extenders being used in PCs but I would trust an extender over a Molex adapter any day of the week.