Need some advice on a horrible problem!

TerrorFirmer

I have the weirdest issue with my new build and it's actually frustrating to the point of wanting to throw it into a skip as I cannot find what is wrong.

It's an X370 K7, 2700X, 8GB TeamGroup 3000mhz, Lian Li 011, Seasonic 620W, Vega 64.

So the issue: In a nutshell, the PC turns off (in a strange way, more below) mainly when I plug in any powered USB device into the front panel - like an external drive powered from the mains.

This also happens if I just touch the live USB connector off the metal front of the case a few times or just make brief contact with the front USB ports without actually plugging in - the obvious suggestion is a short somewhere?

But here's where it get weirder - the PC doesn't auto shut down like a short would normally cause, the monitor just turns off instantly, but sound continues to work fine for about 20 seconds, then I finally get a buzzing sound and a restart.

It also seems to happens occasionally when I plug a powered USB device into the boards native USB ports though nowhere near as much, I can't replicate this reliably like I can with the front ones.

The fact that it's not an auto shut-down but a weird no signal, sound continuing for 20 seconds then lock-up, is really throwing me and I don't know where to begin.

I've tried a different PSU, ram, GPU, disassembling and re-building the entire PC multiple times, cannot find any bad contacts or wires, but still persists.

Even if I disconnect the front panel, touching the metal fascia with a powered USB connector sends it into the weird shut down cycle so it's like something is somehow being conducted somewhere!

If I leave both my external HDD's well out of range the PC can often work for hours otherwise. I played Overwatch for six hours the other day and no issues.

Where should I be looking from here?! I'm convinced it's the case but I cannot for the life of me work out how, why, or where and it's driving me insane.

game4it70

TerrorFirmer wrote: »

Even if I disconnect the front panel, touching the metal fascia with a powered USB connector sends it into the weird shut down cycle so it's like something is somehow being conducted somewhere!

It has to be a short somewhere.The obvious thing to do is breadboard it to confirm the case is causing it.
Is there a standoff in the wrong place by any chance?

TerrorFirmer

I'm not home at the moment but my next step is to build it outside the case but connected to the front panel to see what happens. I've rebuilt it at least twice using different standoffs and screws so I'm positive there's nothing up there.

The other thing that's really throwing me is the random way it shuts down (dead monitor, sound keeps going for 30 seconds, I can even still talk to people on teamspeak until it locks up) - usually with a short it would just shut down!

Venom

The first thing I'd check would be the motherboards standoff's as mentioned above and also the front panel connections wiring and that it's in the correct slot on the board.



Have you had a build in that case before or is it new?

game4it70

It is the weirdest problem i have heard of in a long time.

Have a look at the I/O sheild too.Sometimes the slots in them have small metal tabs that might be touching a part of the ports its not meant to.

TheBoyConor

I was going to say are you sure you have the wires for the front panel connected in the right order? My build had a case with all separate connectors for each individual wire and it was quite confusing and tedious to make sure they were connected correctly.

Would there be a stray screw, or a screw too long or too fat shorting something on the case to the board?

Are you sure that the case, PSU and board earths are all in connected correctly and sound? Are you also sure that the plug socket and the electrical circuit you are on in the house has a correctly functioing earth set up? You could also try to plug it in and run it in a different room or circuit or a different house even. What I am thinking is that possibly the earthing set up is some way gammy and current from a conductor in a USB device is touching your case and the case, instead of running the current to ground via the earth is shunting it into the MB and knocking it off.

That's a theory, but I might be totally wrong. Still, it is worth ruling out.

Inviere

I’d start by disassembling it and building it outside of the case, as advised above. Also triple check the front panel connectors, and make sure things like the ATX connector are firmly home

TerrorFirmer

TheBoyConor wrote: »

Are you sure that the case, PSU and board earths are all in connected correctly and sound? Are you also sure that the plug socket and the electrical circuit you are on in the house has a correctly functioing earth set up? You could also try to plug it in and run it in a different room or circuit or a different house even. What I am thinking is that possibly the earthing set up is some way gammy and current from a conductor in a USB device is touching your case and the case, instead of running the current to ground via the earth is shunting it into the MB and knocking it off.

That's a theory, but I might be totally wrong. Still, it is worth ruling out.

Obviously I'll have to have a good look once I'm home but something more along these vague lines is what I'm thinking (though i don't claim to know anything about this particular area).

I've built/re-built it, with all different parts even except the board and case, and even just single two standoffs/screws to temporarily hold it in place, so it's not anything immediately obvious like bad standoff/connectors not in place or the usual suspects.

Will update as the evening progresses!

TheBoyConor

Of course, there is always the possibility that the board is faulty. But obviously, rule everything else out first before declaring it shot.

If it comes to it, have you another machine that you can set the board up in if it comes to it so as to test it out?

TerrorFirmer

So if I build the PC outside the case but connected to front panel, it works fine and no issues plugging devices in or out.And touching a powered USB cable to the metal facing does nothing.

Only happens when in the case. Problem is I cannot narrow down what the bloody problem is.

It's only resting on 2 barebone standoffs (all others removed) with no IO shield and nothing from the front panel connected except the USB header and it still happens, I cannot for the life of me see where there could be a grounding issue or how a charge/current is passing through?

Painful.

Venom

Check on your motherboards manufacturers forums to see if anyone else is having any similar problems with that board, case, standoffs or USB issues.

Are you cable managing the various wires and connections in the new case while trouble shooting?

Have you fiddled with any bios settings? Are you on the latest bios version or maybe try a reset back to default settings?

Can you pop the entire build in an old case to see if the same issue occurs?

killanena

I've seen the metal spokes oi shields touch off the motherboard do this before. It might be like a millimeter away and then you touching it or inserting a USB could be all the extra pressure it needs to make contact. That is of course if your using the rear USB slots.

TerrorFirmer

OK, getting somewhere now. When I replace the Vega 64 with an old RX460, the issue goes away - no shorting issues at all.

Even though the backplate for the Vega and RX460 are the same size and touch the case in the same way, somehow the Vega 64 being installed shorts out the case when I a) insert a powered USB device or b) simply touch the metal fascia of the case front with a live USB cable.

Any ideas?

ED E

Do you own a multimeter?

From quickly catching up on this thread (so excuse if I've missed something) it sounds like a 5v/12v rail on the Vega 64 is shorting to the frame/mounting. The whole case should be at ~0V(Ground/Earth) but if its not earthed properly then you could have it brought up to 5/12v which is a safety issue and also where the issues arise.

If you have a multimeter put one probe to the case (unpainted) and one pin to a grounded object like a radiator (unpainted bit) or the earth of a socket (avoid the live/neutral here!). That should be 0V. If its not, there's your fault.

TerrorFirmer

I'm confused by this - why would the Vega in isolation being doing this?

I've just tried a GTX1070 in the case and that's also working perfectly fine.

game4it70

When using vega card is the lead fully seated and not touching the small strips of metal that separate the slots on the case?

TerrorFirmer

game4it70 wrote: »

When using vega card is the lead fully seated and not touching the small strips of metal that separate the slots on the case?

Nope, definitely nothing that simple.

At least I've narrowed it down to the Vega 64, but I'm just utterly perplexed as to why it is conducting some sort of current or whatever via the front panel when the GTX1070 and other cards are not.

btkm8unsl0w5r4

Maybe the USB front port is damaged and conducting to the case.....sound like a super weird problem and guessing the solution will be something totally not intuitive....

TerrorFirmer

Gwen Thankful Beanbag wrote: »

Maybe the USB front port is damaged and conducting to the case.....sound like a super weird problem and guessing the solution will be something totally not intuitive....

Sadly not as I've tried 3 other cards and they don't do this. It only exclusively happens with the Vega 64.

Utterly bizarre and infuriating.

ED E

TerrorFirmer wrote: »

I'm confused by this - why would the Vega in isolation being doing this?

I've just tried a GTX1070 in the case and that's also working perfectly fine.

Failure, defect or damage.

TerrorFirmer

ED E wrote: »

Failure, defect or damage.

Just to put the cherry on the cake I accidentally just tore the serial number off the back of the card and wrecked it. :mad::mad::mad:

Joeface

Could you remove the bracket from the Vega. It could be kinked in some way causing the short.

TerrorFirmer

Joeface wrote: »

Could you remove the bracket from the Vega. It could be kinked in some way causing the short.

Yep tried that, still the same. It wasn't touching the case in any way shape or form and still the same issue.

Just checked OCUK there and they currently have 4 in b-grade so I think that probably says it all.

keffiyeh

Would've put money on it being the case. The fact that it's the gpu is bizarre. Does your gpu take an extra power connector/pin than the other cards by any chance?

Yeah I'd rebuild in a different case, to check.
The Vega could just be pulling on the MB just enough or dimensions are just different enough to have the board connecting with the case or I/O shield.

Of worse, seeing as it's a display issue first. The Vega is getting pushed against the case somehow and the power from the USB is shorting on the card?

TheBoyConor

Rent don't you run it built in the case but without the GPU card and just on the on board graphics and see what happens with the USB.

TerrorFirmer

Just an update on this, card is giving problems in every build, but because I accidentally removed the serial number sticker off the card (it got stuck to another sticker and torn off) I've been told zilch they can do. I now have a lovely €600 paperweight.

Any suggestions what to do with a dud Vega 64??

Inviere

TerrorFirmer wrote: »

Just an update on this, card is giving problems in every build, but because I accidentally removed the serial number sticker off the card (it got stuck to another sticker and torn off) I've been told zilch they can do. I now have a lovely €600 paperweight.

Any suggestions what to do with a dud Vega 64??

Oven reflow? Y'never know....:):o

K.O.Kiki

TerrorFirmer wrote: »

Just an update on this, card is giving problems in every build, but because I accidentally removed the serial number sticker off the card (it got stuck to another sticker and torn off) I've been told zilch they can do. I now have a lovely €600 paperweight.

Any suggestions what to do with a dud Vega 64??

http://matija.suklje.name/rooting-and-flashing-your-device-does-not-void-the-warranty-in-eu

You are still covered by EU law.
Keep escalating & threaten to bring this to court.

TerrorFirmer

K.O.Kiki wrote: »

http://matija.suklje.name/rooting-and-flashing-your-device-does-not-void-the-warranty-in-eu

You are still covered by EU law.
Keep escalating & threaten to bring this to court.

Thanks, very interesting. Though couldn't they realistically claim that without the serial sticker I can't prove I purchased from them?

Venom

TerrorFirmer wrote: »

Thanks, very interesting. Though couldn't they realistically claim that without the serial sticker I can't prove I purchased from them?

If your case has slots slots for vertical GPU mounting, maybe just getting a riser cable would get the GPU working in your new setup?