Gaming rig...help with client computer with vid card issues.

Whelp, Rob (Skye) wins the case of beer.

The new Seasonic Gold arrived today, swapped it with the original ThermalTake Toughpower that was in there.
Booted it up with just the onboard...booted up fine.
Powered down, installed the nVidia 1050ti in the top slot...booted up fine...ran full desktop resolution fine.
Several reboots, several full power downs and fresh power ups, two rounds of Microsoft updates....every time...running fine like a champ.

So yup..faulty power supply.

They were VERY close to the water when the big surge came in from hurricane milton last fall...while I had 4.5 feet of water on top of my yard, they were on a peninsula closer to sea level then I was...and had over 6. They did barely have time to ...get important things out as the surge was coming in...he, being a big gamer, and runs of lot of custom half life servers...did grab his tower which was up on a desk...as water was coming into their home, when they evacuated. While it did not have water intrusion in the chassis, there was that visible evidence of a wee tiny splash on the back of the top vid card. Not even 1/4 of the volume of fluid you'd see from spitting. But..you know...sea air....has salt in it...and that salty moisture can settle. But from what I can see...it's not salt water that splashed on it, likely soda or some drink that may have toppled over on his desk and a drip when over the edge and down the back of the tower a tiny bit.
 
@YeOldeStonecat, honestly I don't think it was liquid damage.

You've got 3 DC rails in a PSU, 12v, 5.5v, and 3.3v. Each must operate within 10% tolerance. 3.3-.33 = 2.97v. Tolerance is ONE THIRD of a volt... that isn't much.

In my experience, as soon as the 3.3v rail becomes a 3v rail... board nope out of booting. DC rails will drop voltage under load, and will not directly appear when tested with a volt meter in all cases.

Here's where things get fun...

PCIe uses the 3.3v rail for signalling. So does the CPU. So if that rail gets weak, and you put 2 GPUs and a CPU on it... well splat. So when you started mentioning changes in behavior as you shifted the GPUs around, I'm going to assume the 3.3v rail is FUBAR. And that, happens just due to time and age. It's the smallest voltage, with the tightest tolerance, age widens tolerances, ergo... this is the normal way for a PSU to go!

If water / salt kills a supply, that will typically cause an arc fault on the 12v rail, if THAT happens... machine is junk. Anything connected to the 12v will fry, which is pretty much everything. I've seen such machines become rather impressively long lived blow torches, it's spectacular! But also quite devastating.
 
Back when I was building gaming rigs for people, I do remember as vid cards got more and more powerful.....the importance of a stable and capable 3.3v fail. The higher end vid cards demanded it.
 
Back
Top