Random shutdowns after 2 weeks on new build

[Wu Tang Wyatt]

Just made my first build a couple weeks ago and everything has been running fine until just about an hour ago.
Tonight I had some friends over to check out the rig, I got a gtx 970 going to 32" and 40" vizios. I was playing dead space, running dolphin emulater at 5k render and even played the supposedly broken arkham knight on max settings with no frame cap to positive results. The whole time I was running a game on one screen and either spotify or chrome on the other TV.

After my friends leave I turn off the computer and go take a shower, do laundry, in other words left the room with the PC off for a couple hours. When I came back and booted up I had only just launched a SNES emulator when my PC shutdown. When it shut down the the led on the case power button would blink til I switched off the PSU. I booted back up and the PC stayed on the desktop with no programs running for about a minute before shutting down in the same fashion. I've tried different outlets, disabling one of the monitor and more but the problem persists. The computer boots fine before quickly shutting off so I don't suspect faults with anything internal other than the PSU.

One very weird thing to note: when I was running tests disconnecting things, after a shutdown I leaned back to switch off the PSU and lightly bumped the case which made the PC begin to start up again almost as if something loose had been knocked back into place.


That being said though, how could a corsair power supply suddenly die only a couple hours after exemplary performance? Could it be stressed to death that quickly? because I admit I skimped a little on the PSU even though that's not recommended but I can explain why if anyone cares.

My build
Gtx 970
Amd fx8310+cooler master 212 evo
MSI 970g43a
2x4gb g skill ripjaws ram
250gb Samsung evo ssd
175gb Seagate barracuda (had laying around)
215gb 2.5" Samsung HDD (from broken laptop)
DIYPC Zondda case
Corsair cx430w

Now I know 500w is the rec for 970s but manu recs are overshot and my load on part picker was only around 350w. I also read people on partpicker with very similar builds running their 970s fine on 430s. I was in that category myself for the past couple weeks. The 500w was also twice the price of the 430w as the 430 was on sale for $15 after MIR. I didn't want to pay double for 70w more and at $15 would be replaceable if need be.

So my question is what could be causing a reputable PSU to go from 100% to sputtering in just a couple hours of non use? Or is it possible I'm experiencing a power issue external to the PC? What do we make of the bump to start?

PC is currently unplugged from everything and has been over 30 min I'll probably give it another try in case the issue was external and temporary.

Sorry for all the text, just trying to be thorough and paint the complete picture.

 

[jimmysmitty]

I don't see anything on Frys website from $50-60 that is as good as the XFX or Seasonic linked in this thread. There are some Corsairs but they are above your price range.

 

[logainofhades]

I know CWT has made some good units too, but for PSUs, I say "if it's not broke don't fix it," which means I limit my selections pretty much to Seasonic, maybe Super Flower, maybe Delta.

I have to agree, as my gen 1 TX 750w is still alive and kicking, 6yrs later. :lol: At the time I bought it, it was one of the best PSU's for the money.

 

[ILBRelic]

"How could a corsair supply die after only a couple hours of exemplary performance"

What is exemplary for a power supply? they work or they don't.

This is all based on my understanding, and probably wrong in several ways but hopefully someone smarter can correct any inaccuracies. Here goes:

I imagine during that few hours of Dead Space a discrete component being forced to deliver too much current went through significant thermal stress resulting in corrosion or expansion leading to wild changes in impedance, loosing of contacts, or destruction of IC's. A few hours can be a lot if we're talking an initial rate of 2 degrees a minute. The fact that it started a second time is dumb luck and as soon as it heated up again it shut back down. The chances of it being external to the PC is next to none and easily testable (plug something else in, use a multimeter, etc).

Bumping of case leading to spastic power-on lends evidence to loosed contact but I'm just theorizing.

Either way it is NOT safe to use this PSU anymore (never was really), and may have already damaged your GPU (that 6 pin to molex should still be on the 12 V rail but I'm guessing that's what you overloaded) or MB in a way that wont present itself for some time. A bad PSU can deliver more current than your mobo or gpu are designed to handle or behave in semi-transient state which causes a domino like effect wherein thermal dissipation is impossible. This results in permanent changes to electrical characteristics of components, which in turn prevents or impairs normal operation.

You spent $350 on a GPU and $15 on a PSU, you reap what you sow. I would have shot for a 960 or ebay'ed a 4gb gtx 670 and spend a lot more on the PSU and MB. Also displays have nothing to do with your PSU directly, so no worries there. (will marginally increase your gpu's power consumption on idle )

Most people leaving reviews on places like newegg or amazon don't actually know or care to leave anything other than it "works fine for me". Reading those reviews is not research. Be wary when places like pcmag or cnet review a device too, real world results between identical models vary too widely.

 

[Wu Tang Wyatt]

Thanks for the info guys. Can anyone endorse this antec vp they got at Fry's? It's listed on here as tier 3 which seems acceptable
http://www.frys.com/product/7947110?site=sr:SEARCH:MAIN_RSLT_PG

 

[Wu Tang Wyatt]

I'd really appreciate an answer as I plan on buying this today

 

[Onus]

I would trust the Antec VP-550F; it's made by Delta.

 

[tiny voices]

Much better than the CX430 and actually has the right connectors you need.