Possible PSU problem? Solved(kinda)

Grandmaster Fap! · 7 Dec 2018 at 21:05

So i've been running my new Ryzen system with no problems but ive started playing Shadow of the TombRaider and Ashen and both games have caused me random crashes, frozen screen and have to hard reset.

Ive run various gfx and mem benchmarks and all pass so im wondering, is this a PSU problem or something in Win10 be it a driver issue?

As an example ive just played through Lords of the fallen with no issue and this pushed my Cpu and gfx hard.

EDIT: Ive changed everything except the PSU which is a Coolermaster 600W, never been an issue but ive switched from intel to a 2700x keeping my 1070GTX, surely a 600w would be enough for a 2700x and 1070?

EsaT · 8 Dec 2018 at 17:05

Coolermaster has plenty of variety from cheapos to proper PSU.
So what's the particular model number.

And does Event Viewer show any errors with time mark matching those crashes etc?

Grandmaster Fap! · 8 Dec 2018 at 17:16

Been running stress tests all day, Cinebench, Realbench, AIDA, Prime, ran Memtest all night. Even running the Tomb Raider bench repeatedly. No fails.

PSU is CM Silent Pro 600w, Event Viewer shows Kernel Power 41, lt also shows an error, something to do with RealTek sound driver around the same time as the freeze. Ive uninstalled the driver and am just letting windows power it while i test some more.

EsaT · 9 Dec 2018 at 21:12

Basing on age of that model and cheap second tier capacitors in bad place for cooling, it's well possible that it reached end of its designed life.
https://www.hardwaresecrets.com/cooler-master-silent-pro-m-600-w-power-supply-review/2/
Wish Enhance had spent as much effort to capacitor quality and their "work environment" as to those oversized "Mine are the biggest!" heatsinks blocking airflow from reaching secondary capacitors.

Also it's basically at most 500W PSU by any standard from current decade.
https://www.hardwaresecrets.com/cooler-master-silent-pro-m-600-w-power-supply-review/6/
Gotta love that last line in bottom of sticker...

AS SEALED STICK WAS REMOVED, LOST OR DAMAGED, IT SHALL BE OUT OF WARRANTY VALIDITY

Seasonic Focus Plus Gold 650W seesm to be out of stock, but there's Antec with same insides.
https://www.overclockers.co.uk/antec-hcg-gold-650w-80-plus-gold-modular-power-supply-ca-24w-an.html

Though modern 550W PSU would be better than that CM as new.

Grandmaster Fap! · 10 Dec 2018 at 17:30

Ordered a new PSU, dont know what else to try as i reinstalled Win10, play game and boom crash. I can run realbench all day long pushing cpu and gpu way harder than these games can and its fine, pulling my hair out. Lamenting the day i swapped from Intel to Ryzen, nothing but hard work.

EsaT · 10 Dec 2018 at 17:57

From its description there are more stressing tests than Realbench.

My CPU isn’t being fully loaded – is there a problem?

Your CPU is not the only indicator of system performance. Synthetic benchmarks often isolate one part of the system and load it fully to express ‘maximum performance’, but this is not how your PC works day-to-day in applications.

So heavy games can well have more demanding load spikes.

Grandmaster Fap! · 10 Dec 2018 at 18:05

The annoying thing is i can play a game for an hour or just 15 mins and the comp locks, its like waiting to step on a landmine

Hopefully the PSU comes tomorrow and i can put this problem to bed!

Grandmaster Fap! · 11 Dec 2018 at 17:03

Sooo ive got a Seasonic Prime 750w, all plugged in and within 3 mins of running Realbench it locks and now i cant hold the pwr button to reset, i have to flick the switch on the PSU.

Boot up and run benches again, all fine, play games, all fine.....its....random and doing my nut in.

Kurgen · 13 Dec 2018 at 09:19

maybe its a fault with the mobo or ram, have you tried dropping to a single stick, and alternating them

Grandmaster Fap! · 13 Dec 2018 at 12:33

@Kurgen Not tried that yet mate as ive run memtest overnight with no issue so thought they should be ok. Just trying some other things at the moment and will try that if no success.

FuF · 13 Dec 2018 at 23:44

Possibly unrelated, but I used to have exactly the same problem as you. Eventually (after many months of struggling doing anything software/hardware related to fix it) I found that my heavy GPU (the aging 980 GTX) was sagging a bit and leaving the connected lane. Tied it with a cable tie against the case and did not have the problem again. Make sure that everything is well connected since you've just upgraded your system. A random power cable (both mobo or GPU) or RAM not inserted properly could be causing it. And funny enough, most of the time I was having the issue, benches for GPU or CPU would pass just fine. Always crashed on games and used to give the Kernel error 41 in the event viewer as well.

Grandmaster Fap! · 14 Dec 2018 at 09:33

@FuF Thanks for that mate, im starting to think this may have been the problem as id taken the stock fans off my GPU and fitted some 120s i had laying around. All was well for a week but then this problem started, i fitted the original fans back on it just for the hell of it and 'touch wood' ive not had a restart in a day and half....we shall see!

EsaT · 14 Dec 2018 at 20:14

Grandmaster Fap! said:
@FuF Thanks for that mate, im starting to think this may have been the problem as id taken the stock fans off my GPU and fitted some 120s i had laying around. All was well for a week but then this problem started, i fitted the original fans back on it just for the hell of it and 'touch wood' ive not had a restart in a day and half....we shall see!

Graphics cards have many other components than GPU chip itself which need cooling.
So when changing anything in its cooling it's hard to be sure that nothing starts running hotter than it should.
Somekind crash of graphics card or errors in its communication could certainly crash rest of the PC.

Grandmaster Fap! · 14 Dec 2018 at 20:36

EsaT said:
Graphics cards have many other components than GPU chip itself which need cooling.
So when changing anything in its cooling it's hard to be sure that nothing starts running hotter than it should.
Somekind crash of graphics card or errors in its communication could certainly crash rest of the PC.

Yeah all seems to be well at the moment, be it the replaced fans (EDIT: just to be clear i left the heat sink on the card and it covered ram and vrms)werent cooling something properly or sagging GPU or the fact i connected the fans to a mobo header and had Argus monitor control them based on GPU temp. It could have been anything or all of the above :/

FuF · 16 Dec 2018 at 20:47

Glad you've got it sorted! Hopefully some undisturbed gaming to come during Christmas now

joedeath332 · 17 Dec 2018 at 23:59

Grandmaster Fap! said:
Ordered a new PSU, dont know what else to try as i reinstalled Win10, play game and boom crash. I can run realbench all day long pushing cpu and gpu way harder than these games can and its fine, pulling my hair out. Lamenting the day i swapped from Intel to Ryzen, nothing but hard work.

Sorry you are having bad luck with your Ryzen I too went from intel to ryzen 2700x, however mine was a very positive switch and I am more than happy with the performance per buck.
Hope your new psu fixes your issue.

Edit just read psu did nothing, but hopefully stock gpu fans back on it will fix it.

Grandmaster Fap! · 18 Dec 2018 at 09:36

Thanks joedeath332, all seems to be well still, still not sure if it was software or hardware though.

***Possible PSU problem?*** Solved(kinda)

Possible PSU problem? Solved(kinda)