24-pin on Seasonic PSU -> MSI 790FX Mobo melted

P

prochobo

Limp Gawd
Joined
Sep 28, 2005
Messages
321
The PSU is a Seasonic X-1050 and mobo is an MSI 790FX-GD70

I was running 3x 5870's which were mining until yesterday, when I smelled burning. This has happened before and I'm on my second X-1050 (got my RMA two days ago). The 12v connector to the mobo just melted. This is the third time this has happened. The first time, the 12v wire burned out and they sent me a replacement cable. I should also note that the modular connection on the PSU itself also got partially melted. The second time, I got sent an RMA PSU and another replacement cable. Now this is the third occurrence.

Is it because I'm pulling too much power from my video cards? You would think an X-1050 could take this, but Seasonic says it might be my motherboard.

gIAsrweh.jpg

8SeuVJ4h.jpg

rnIvKj0h.jpg
 
do the video cards have any kind of power connection headers, and if so are they being used?
 
Does the board have an auxiliary molex power connector for multi card setups (I cant see one in the picture on MSI's website)? This could be the problem - the cards are drawing too much power through the 24-pin connector and burn it out.
 
The 790FX-Gd70 doesn't have an auxiliary 12v input for the graphics cards, that's your problem. The cards are drawing too much power through the 24 pin connector (up to 75 watts each).
 
The problem is likely compounded because of the melted plastic and charring on the pins of the 24-pin connector on the motherboard. More resistance to the flow of current equals more heat generated, which is why the problem is likely occurring even faster than the first time.

It could have been as simple as a poor connection the first time, which caused the heat and melting and then made things way worse from that point forward.

But the people above me are spot on, and this is why most boards capable of supporting triple GPUs have an extra 12v connection somewhere.
 
Also, there have been quite a few people who had similar problems with that board and certain CPUs, something about the 1090T and 1100T popping the VRMs under high overclocking conditions.
 
EVGA had a power adapter that plugged into an empty PCIe x1 slot to land more current on the PCIe bus, you might try something like that
 
This is also common if you get to much dirty power flowing through your circuit, your psu can't move fast enough to regulate all the power sags and spikes.
 
Blue Falcon said:
Also, there have been quite a few people who had similar problems with that board and certain CPUs, something about the 1090T and 1100T popping the VRMs under high overclocking conditions.
Click to expand...

That was a problem plaguing a lot of MSI boards, but it is not at all relevant to what's happening here, as the power circuitry for the GPUs and for the CPU are completely separate.
 
You need powered PCI-e risers if you're going to be mining with power hungry video cards 24/7. It's not necessarily a defect but more like a design oversight- you'll need to reduce the power draw through the 12v +4 connector somehow otherwise you'll keep burning molex connectors.
 
You must log in or register to reply here.
Back
Top