STRANGE situation

[Cramig88]

Ok so, I just had the strangest thing happen. Does anybody know what happened?

 

System configuration

asus crosshair viii dark hero Wifi x570

5900x

4x 16gb teamforce argb/white CL14

xfx 6700 ti xt something (12g vram)

xfx 6600 8gb

Samsung 980 pro 1tb NVMe

lian li dybamic o11 XL, white

(monitors and cooling system irrelevant)

 

Troubleshooting steps you already did

Break down the pc one by one, eliminating all causes of the problem. My daughters pc has a 5900x also, as a last resort, swapped the Cpu. 
flashed cmos, used usb to flashback, removed all the ram/components. Basically left it as a test bench by the time I removed everything - even the NVMe.

first guess was power supply, swapped with other pc, a 1000w seasonic.  Even swapped all the power lines as the seasonic only works with its own cables.

left cpu fan in (tried others too).  Jumped the psu, just to check - works. 24 pin and 8 for cpu, motherboard RGB turns on, start button included.

hit start, no LED readout. Part after part, diff minimal testvench configurations. No LED POST, each attempt. Probably 5 hours or more of different configs. One dimm, diff locations, diff dimm, diff locations. 2 dimm, 3 dimm, everything.

no post - just lights up. Molex pump works when 24pin is jumped.

 

Any additional information that may be relevant.

here’s what happened. Pc was on for 48h or so, downloading and doing junk. Temps were normal 33c and 33-34 on both gpus (watercooled, monoblocks) under no load.

plugged in a usb-c flash I used the day before. Suddenly shuts off. Boom, lights out. I flip the psu to off, pull the cord as a precaution.

half an hour later I plug it back in, push the start button on MOBO, no post readout on the LED. Nothing.  At this point start removing parts/fans/5 and 12v RGB/front panel headers and so on. All the way down to swapping for an identical model cpu.

no post when power is applied as I said through various configs. Dead.

fearing the worst, goto Amazon and do a RMA exchange.

Leave it without power, just a cpu, fan, 24p/8p, that’s all. No NVMe, no ram. Barebones and swapped Cpu (I know it works)

10 hours later return to this nightmare still fearing the worst. Plug in psu, flip the power, board lights up. Low expectations. Hit start, it starts POST with LED display and all. Didn’t have that 10 hours ago on same “configuration”

now, 2 dimms, NVMe installed, 6700, only fan is Cpu obviously, and a molex for the pump. Pump is straight dc from molex, so no PWM or dc connected.

obviously had to reset fTPM, everything boots into windows 10.

 

so how is it that a usb-c shorted out the pc after using it 8 hours prior?  Bootable usb with hiren on it (no chance of virus/malware) and it poops while in Windows.  Then after hours of trying, no post readout - dead.

leave it alone after all my attempts mentioned, 10+ hours later, it randomly decides to post.

I should also mention I stared at it for 4 hours or so feeling defeated. No power applied, unplugged, everything. Barebones as I said. Before I went to bed and gave up, still wouldn’t post, board would just light up - start did nothing.

then, it just works, 10 hours later. Cpu and fan header, starts posting. LED readout, no ram. One dimm, LED, no vga. Plug in 6700, posts. No boot drive. Plug in NVMe. Bios displays on monitor!  says new processor. Clear fTPM. Ok. Restarts, bios says configure detected NVMe. Go into bios, save changes and exit (you have not made changes, save and restart?) yes.

boots straight into windows

 

what changed, leaving it unplugged for hours, then trying again way later?

what happened?  How could a 5w usb-c, kill the pc?  Yes initiallly it had a game max 1050w PSU, which was my first guess (jumped it works fine). Still, the both psu’s are overpowered for the setup. Intentionally. Then it just works?

 

please any help on this one would be great. I don’t know what happened or what changed after leaving it on a failed “attempt”. As I said, plugged it in, hit start, worked hours later just fine.

I’m completely puzzled

Thanks guys

[BiotechBen]

The first mention of GameMax is in tier C(cheap/low-end systems)

Your gm1050 is in tier D(very cheap/igpu systems) for the safety of your hardware, I would recommend changing it.

https://linustechtips.com/topic/1116640-psu-tier-list-rev-148/

[Cramig88]

Ok looking at the list is very overwhelming. Single rack, double rack?

what would you recommend as a HIGH quality PSU?  I would like to return this “game max” (obviously) and get a superior PSU. 
as mentioned I have a o11 dynamic XL which is a “full size” mid tower atx case. No I am not using the hot swappable 3.5” caddy’s, I actually removed them completely to accommodate Better wire management and put my psu on the top since the psu cables for my other build (thermaltake t71) were a STRETCH reaching the 8 pin cpu headers.

 

money is not a question, reliability is much more important. I mean prior to plugging in the usb flash, as mentioned used a day before, no issues. I had my iphone usb-c to thunderbolt (whatever they call it) plugged in. Phone wasn’t so that’s irrelevant I guess. Unplugged that, put in drive, gone

so crap PSU was the problem as I thought. Why would it have run POST hours later, when I left it exactly as it was?  Possible the capacitors/chokes/VRMs/etc “cleared”?  I still find it weird…

 

Corsair I saw on the list… recommendations?

thanks again

[BiotechBen]

10 hours ago, Cramig88 said:

Ok looking at the list is very overwhelming. Single rack, double rack?

what would you recommend as a HIGH quality PSU?  I would like to return this “game max” (obviously) and get a superior PSU. 
as mentioned I have a o11 dynamic XL which is a “full size” mid tower atx case. No I am not using the hot swappable 3.5” caddy’s, I actually removed them completely to accommodate Better wire management and put my psu on the top since the psu cables for my other build (thermaltake t71) were a STRETCH reaching the 8 pin cpu headers.

 

money is not a question, reliability is much more important. I mean prior to plugging in the usb flash, as mentioned used a day before, no issues. I had my iphone usb-c to thunderbolt (whatever they call it) plugged in. Phone wasn’t so that’s irrelevant I guess. Unplugged that, put in drive, gone

so crap PSU was the problem as I thought. Why would it have run POST hours later, when I left it exactly as it was?  Possible the capacitors/chokes/VRMs/etc “cleared”?  I still find it weird…

 

Corsair I saw on the list… recommendations?

thanks again

No sweat, walls of text can be quite intimidating.

I have a Seasonic Focus GM-550 and it's fantastic, I would say with a 6700xt and 5900X that either a seasonic prime/Focux GX 750W+ or a Corsair RM/HX  also with 750W+ and efficiency of gold of better.

(high quality, been making power supplies for about 40 yrs) 

https://www.amazon.com/CORSAIR-RM750x-Certified-Modular-Supply/dp/B079HGN5QS/ref=asc_df_B079HGN5QS/?tag=hyprod-20&linkCode=df0&hvadid=312727440900&hvpos=&hvnetw=g&hvrand=928372250071671050&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=9007020&hvtargid=pla-516264783994&psc=1

 

https://www.amazon.com/Seasonic-SSR-750FX-Modular-Warranty-Compact/dp/B073GWKDVM/ref=asc_df_B073GWKDVM/?tag=hyprod-20&linkCode=df0&hvadid=309779531175&hvpos=&hvnetw=g&hvrand=9770801813579376822&hvpone=&hvptwo=&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=9007050&hvtargid=pla-375280494182&psc=1&tag=&ref=&adgrpid=62412137260&hvpone=&hvptwo=&hvadid=309779531175&hvpos=&hvnetw=g&hvrand=9770801813579376822&hvqmt=&hvdev=m&hvdvcmdl=&hvlocint=&hvlocphy=9007050&hvtargid=pla-375280494182

[Frizz]

12 hours ago, Cramig88 said:

Ok so, I just had the strangest thing happen. Does anybody know what happened?

 

System configuration

asus crosshair viii dark hero Wifi x570

5900x

4x 16gb teamforce argb/white CL14

xfx 6700 ti xt something (12g vram)

xfx 6600 8gb

Samsung 980 pro 1tb NVMe

lian li dybamic o11 XL, white

(monitors and cooling system irrelevant)

 

Troubleshooting steps you already did

Break down the pc one by one, eliminating all causes of the problem. My daughters pc has a 5900x also, as a last resort, swapped the Cpu. 
flashed cmos, used usb to flashback, removed all the ram/components. Basically left it as a test bench by the time I removed everything - even the NVMe.

first guess was power supply, swapped with other pc, a 1000w seasonic.  Even swapped all the power lines as the seasonic only works with its own cables.

left cpu fan in (tried others too).  Jumped the psu, just to check - works. 24 pin and 8 for cpu, motherboard RGB turns on, start button included.

hit start, no LED readout. Part after part, diff minimal testvench configurations. No LED POST, each attempt. Probably 5 hours or more of different configs. One dimm, diff locations, diff dimm, diff locations. 2 dimm, 3 dimm, everything.

no post - just lights up. Molex pump works when 24pin is jumped.

 

Any additional information that may be relevant.

here’s what happened. Pc was on for 48h or so, downloading and doing junk. Temps were normal 33c and 33-34 on both gpus (watercooled, monoblocks) under no load.

plugged in a usb-c flash I used the day before. Suddenly shuts off. Boom, lights out. I flip the psu to off, pull the cord as a precaution.

half an hour later I plug it back in, push the start button on MOBO, no post readout on the LED. Nothing.  At this point start removing parts/fans/5 and 12v RGB/front panel headers and so on. All the way down to swapping for an identical model cpu.

no post when power is applied as I said through various configs. Dead.

fearing the worst, goto Amazon and do a RMA exchange.

Leave it without power, just a cpu, fan, 24p/8p, that’s all. No NVMe, no ram. Barebones and swapped Cpu (I know it works)

10 hours later return to this nightmare still fearing the worst. Plug in psu, flip the power, board lights up. Low expectations. Hit start, it starts POST with LED display and all. Didn’t have that 10 hours ago on same “configuration”

now, 2 dimms, NVMe installed, 6700, only fan is Cpu obviously, and a molex for the pump. Pump is straight dc from molex, so no PWM or dc connected.

obviously had to reset fTPM, everything boots into windows 10.

 

so how is it that a usb-c shorted out the pc after using it 8 hours prior?  Bootable usb with hiren on it (no chance of virus/malware) and it poops while in Windows.  Then after hours of trying, no post readout - dead.

leave it alone after all my attempts mentioned, 10+ hours later, it randomly decides to post.

I should also mention I stared at it for 4 hours or so feeling defeated. No power applied, unplugged, everything. Barebones as I said. Before I went to bed and gave up, still wouldn’t post, board would just light up - start did nothing.

then, it just works, 10 hours later. Cpu and fan header, starts posting. LED readout, no ram. One dimm, LED, no vga. Plug in 6700, posts. No boot drive. Plug in NVMe. Bios displays on monitor!  says new processor. Clear fTPM. Ok. Restarts, bios says configure detected NVMe. Go into bios, save changes and exit (you have not made changes, save and restart?) yes.

boots straight into windows

 

what changed, leaving it unplugged for hours, then trying again way later?

what happened?  How could a 5w usb-c, kill the pc?  Yes initiallly it had a game max 1050w PSU, which was my first guess (jumped it works fine). Still, the both psu’s are overpowered for the setup. Intentionally. Then it just works?

 

please any help on this one would be great. I don’t know what happened or what changed after leaving it on a failed “attempt”. As I said, plugged it in, hit start, worked hours later just fine.

I’m completely puzzled

Thanks guys

Just to explain what I think happened, it sounds like the usb c drive or that port may have been shorted or faulty. If so and that PSU was faulty or junk my guess is over current protection kicked in and perhaps a fuse failed in the PSU (again just speculation).

 

Now as far as most PSUs they basically have resettable fuses/switches that clear almost instantly after a PSU power cycle. But if something went haywire it may have taken hours for all the capacitors to drain if you didn't hold the power button down with the PSU unplugged.

 

Again that's my best guess, as some PSUs need a full cycle and to be drained to properly reset

[Cramig88]

Wait, I can manually drain capacitors and such, by holding the power button on the pc, with everything unplugged (psu)?

I imagine a feature exists, I just didn’t see anything in my motherboard manual. After all, a 400$ motherboard (deff not bragging), should come equipped with “safeguards” against cheap PSUs and quality control “anomalies”.

I don’t think the usb-c (on the front panel, or drive) was faulty, I had just bought it, used it the day prior. Again only put hiren boot on there as I was trying to recover some old RAID volumes and break the associations so I could format them and use in this build on a new raid config.

usb-c jumpdrice goes in, lights go out. Panic, disconnect psu. First guess was that, so took daughters seasonic, swapped 2 psu cables. Lit up, dead to post. Tried the other psu with a jumper, didn’t work for hours then suddenly it turns on

 

so all in all, lesson learned. Never buy a cheap psu, even if it means RGB or whatever the case. That was the only selling point for me, not the reduced cost. The capacitors, MOSFETs, everything got confused when the pc suddenly wanted a whole 5 watts for the drive - because the psu delivered messy and/or unregulated power.

 

 

as I recall, I had a similar issue with my last build. I had it on a test bench (just the box), was trying to figure out why my nvmes were intermittent. As in, work when they want to. Bios wouldn’t even pick both up. Maybe one, none, the other, all randomly.

it was a seasonic 1000w gold 80+, I had MOBO power and cpu, Pc was off. Removed nvme from bottom slot, to try the top. As soon as I installed it in the top. Quite literally the lights flickered and I heard a zap out the power supply.

I had killed the PSU, but the motherboard saved itself (or psu sacrificed to save motherboard). But, this was out of ignorance. I had a Corsair waterblock on the nvme. The bottom is conductive. It got power, touched the board, zap. 
so I even exchanged it on Amazon, put thermal pads on the bottom of the blocks, never had a problem since. Although I don’t bother to take them out anymore. If they work, they work. Or one is shorted out, my mistake.
 

 

 ~

I really appreciate everyone’s help on this one.  I knew cheap psus are bad but I hope this helps others realize just how bad they can be. After all, a sudden demand for 5w “killed” my pc. In a panic I went through diagnostics, swapped the Cpu even.  All of that could have been prevented with a quality PSU. In fact, I would never have had the problem nor experience.

anybody considering a cheap psu needs to read this thread. It very well could happen to you, and if trying what I did… big risk swapping cpus. I could have potentially destroyed a single $4-500 USD processor or both, just troubleshooting.  All due to a 1050w PSU that I think costed 140$ or less…

 

I will take a look at a few of those PSUs mentioned, probably go with higher voltages though.  Anybody should have more than required at all times, not the minimum requirements or maybe 100w of grey area.

lol, I should have thought twice about the cheap psu from the start!  I only have like 10 USB ports on the back IO panel as is!