Complex Troubleshooting Problem (GPU or PSU, most likely)

boothr2

Reputable
Jul 15, 2014
20
0
4,510
Hi everyone, before I describe my issue, I want to thank everyone on this forum for taking the time to read this, as well as for teaching me many things about building and troubleshooting over the past couple months, you guys are awesome.

<Walloftext>

So about 3 weeks ago, I finally put together my first build. The first 2 weeks were pretty smooth with no issues. About a week ago, my computer crashed while playing CS:GO (just went right to black screen, with a "no display" message on the monitor). The funny thing is, upon taking a look inside the case, only the CPU fan and the rear case fan (I have 2 case fans) were shut down along with the monitor display, while all other fans (front case fan, PSU fan, 3 GPU fans) kept running after the shutdown. I just figured it was an isolated incident and restarted with no problem.

Yesterday, the same thing happened twice, and another 2 times today, ALL DURING GAMING. So after some reading I decided to run Prime95 to stress test the CPU and RAM, and Furmark to stress test the GPU.

I ran Prime95 for an hour, and my CPU only reached a max temp. of 51 C, while my motherboard reached a max temp. of 66 C.

I ran Furmark for 20 minutes, and my GPU reached a max temp. of 73 C, while the motherboard reached 61 C. Although I never see these temperatures while gaming, these are still safe operating temperatures from what I have read. However, not during, but RIGHT AFTER I closed Furmark, my display, CPU fan, and rear case fan shut down like mentioned before. Also, I cannot restart my computer right after it shuts down now, I have to wait about 20 minutes after I turn off the PSU to get the CPU fan and rear case to spin and the display to appear on the monitor.

Here are my specifications...
Motherboard: Gigabyte GA-78LMT-USB3 microATX
CPU: AMD FX-6300 3.5GHz 6 core
GPU: PowerColor Radeon R9 270X 2GB TurboDuo
RAM: 2x 4GB A-DATA XPG DDR3-1600 sticks
PSU: Corsair CX500W
HDD: 1 TB Western Digital
SDD: 120 GB Samsung EVO
Case: Rosewill Line-M microATX mini

At first, due to a small case and lack of space to tie off cables, I thought I may have had a cooling issue, but seeing my temperatures of the CPU and GPU at 100% load have dismissed this thought. The other possibility, from what I have read, is a faulty GPU or PSU, or just insufficient wattage (I thought 500W would be enough for this build). If one of those is faulty, I'd much rather it be the PSU (cheaper, plus I bought it at MicroCenter).

Thank you all for reading, and I hope we can find out what is going on here. I also rely on this computer for data analysis through Mathematica and Igor, so it is important that I diagnose the problem here, so it does not interfere with my school work (but I'd also like to run my games without issues :)) All input is much appreciated.

</Walloftext>

EDIT: I hope this is an appropriate section for this post, I did not see a forum dedicated to troubleshooting, and I'm unsure exactly which component is acting up, so I posted under Systems.
 
Solution
Your setup doesn't have any other red flags other than the PSU. While PSU is most likely it is not certain until you change it to verify.

I searched prices and right now these are best quality unit for the price as of today:

Good:

PCPartPicker part list / Price breakdown by merchant

Power Supply: SeaSonic S12II 620W 80+ Bronze Certified ATX Power Supply ($71.30 @ Newegg)
Total: $71.30
Prices include shipping, taxes, and discounts when available

Better:

PCPartPicker part list / Price breakdown by merchant

Power Supply: XFX 650W 80+ Bronze Certified Semi-Modular ATX Power Supply ($69.99 @ SuperBiiz)
Total: $69.99
Prices include shipping, taxes...

boothr2

Reputable
Jul 15, 2014
20
0
4,510


LOL. What would you suggest? I'd like a suggestion based off of my problem ;) Thanks for the quick reply, I hope you can help me.
 


I'll look at it more in depth tomorrow. I'm about to wrap up here.

 

boothr2

Reputable
Jul 15, 2014
20
0
4,510


I understand, same here. Looking forward to what you (as well as others) have to say!
 

boothr2

Reputable
Jul 15, 2014
20
0
4,510


Anyone else? :( Can anyone else comment on the integrity of my Cx500W PSU for gaming with my listed graphics card? Thank you.
 

boothr2

Reputable
Jul 15, 2014
20
0
4,510


So do you think the PSU is actually the culprit here, with respect to the specific problem I'm having? And if so, can you recommend a specific PSU model better suited for my rig?
 
Your setup doesn't have any other red flags other than the PSU. While PSU is most likely it is not certain until you change it to verify.

I searched prices and right now these are best quality unit for the price as of today:

Good:

PCPartPicker part list / Price breakdown by merchant

Power Supply: SeaSonic S12II 620W 80+ Bronze Certified ATX Power Supply ($71.30 @ Newegg)
Total: $71.30
Prices include shipping, taxes, and discounts when available

Better:

PCPartPicker part list / Price breakdown by merchant

Power Supply: XFX 650W 80+ Bronze Certified Semi-Modular ATX Power Supply ($69.99 @ SuperBiiz)
Total: $69.99
Prices include shipping, taxes, and discounts when available

Best

PCPartPicker part list / Price breakdown by merchant

Power Supply: EVGA 750W 80+ Gold Certified Fully-Modular ATX Power Supply ($79.99 @ Newegg)
Total: $79.99
Prices include shipping, taxes, and discounts when available
 
Solution

boothr2

Reputable
Jul 15, 2014
20
0
4,510


Update: I just booted up a game of League of Legends and right after connecting, the shutdown occurred. This is the first time it happened immediately, so I will not try to run any more games. Does this support the hypothesis that the PSU is indeed insufficient? Thanks for all your patience! I have a Microcenter nearby so if you think I should replace my PSU with one of the ones listed, that's what I'll do.
 

boothr2

Reputable
Jul 15, 2014
20
0
4,510


Just bought a Bronze 750W EVGA PSU from microcenter (gold was too much money) and I'm stress testing the GPU now to see if it will crash like last time. Will post results!
 


The bronze EVGA's aren't made by the same OEM. There are more complaint instances. See how it goes.

 

boothr2

Reputable
Jul 15, 2014
20
0
4,510


So I just got done reassembling and testing everything. Yesterday, when I ran Furmark, my system crashed right after I shut the GPU stress test off. Today with then new 750 W PSU, I ran two tests (same length, 20 minutes) and no shutdown occured :) Interesting note, when I ran the GPU test with the 500 W PSU, my motherboard reached a temp. of 61 C; with the 750 W PSU, my motherboard only reached 41 C! Is it possible that my PSU had insufficient wattage to power the motherboard at a decent/full load, causing the shutdown? So far so good, I will run some games tomorrow and if there are no crashes, I will award you best answer. Thank you so much for your help, I will post back tomorrow with an update.

EDIT: During the stress tests on the new PSU, I noticed the PSU did not feel warm at all, where as the other one warmed up quickly. Is this a sign that the issue was actually with the PSU after all?
 


Everything you said is par for the course on the CX, and bad PSUs in general. I'm so glad the new PSU works like a charm.

One thing I noticed on a lot of threads around here about GPU problems is that often they have a CX PSU, and more than half the time it's not even putting out 12v on the 12v rail. But because the GPU doesn't change power draw, and voltage times current equals power, you end up with higher current and higher current makes lots of excess heat in I^2*R losses. That's why your mobo is running cooler now since the VRMs are not working as hard.

Like I said, it annoys the hell out of me that during reviews a lot of these units were given good ratings. So it's possible that Corsair weeded out the duds when they sent review samples out. Frankly there are a lot of problems around here with them.

Pass on the info :)

 

boothr2

Reputable
Jul 15, 2014
20
0
4,510


The physical chemistry grad student in me says this makes sense ;) So far so good, and hopefully the new PSU is the permanent solution; thank you so much for your help! By the way, I no longer hear any POST beeps on startup from the speaker on my motherboard (I noticed this before the shutdowns started happening). Should I be concerned that I don't hear the "OK" beep?