Random Shutdowns.

[Yakusoku]

So I have been having this problem for quite a few months now. Its really irritating to have my desktop turn off when I'm 3d Modeling & Photoshop (Especially), Gaming, or just Browsing the internet. At times it just shuts down even when I'm on  standby (Leaving for an hour and having just a browser on).

 

Parts

Spoiler

i7 4790k

Asus Hero VII

980ti SC EVGA

2x8 16g DDR3 Corsair Veng Pro 2133

RM850 +80 Gold

H60 Corsair Water Coller

Windows 10 Pro

 

 

How does one tell which part causes the problem. Hardware or Software? Anyone have any tips to diagnose this kind of problem?

 

Here's what I have noticed though, perhaps it would help if anything else for info needed.

• Tried having the PSU's main power directly pluged to the wall instead of a extension (A heavy duty one). No Change
• Noticed when it shuts down, the two monitors I have still is turned on. Extension wire would be fine? Also would the so called "Dirty Power" from the outlet not be the cause?
• I've heared that RM from Corsair isnt as reliable as the others, could it be just plain power supply? (I dont have an extra one or anyone I can borrow a PSU so I cant tell)
• It doesnt matter what I do with the desktop it will shutdown regardless of any kind of heavy or low to even no loads I put it through. It will shut down.
• Happens At least once a day. Two or more at times. 4-5 days in a week, since there are days it doesnt shut down at all.

 

P.S

Spoiler

 I've tried to RMA the RM850 PSU to Corsair they gladly take it back, the only problem is I am reliant on my desktop for work. I cannot have one day (Even on the weekends) to not use it. Deadlines for my work are very deadly, so I need my Desktop up and running. I wanted to know if its like EVGA where they would first send you the replacement unit then you ship them yours so you dont get any downtime.

 

Not sure where to put this post since I have no idea which is causing the problem.

Please and Thank You!

[NeoFrux]

When you're having these shutdowns, you say it's happening most often during games? Does the computer go through a shutdown routine, or does it just turn off as though the plug was pulled from the wall? You might want to check the temperatures via Open Hardware Monitor while running a stress test like Prime95 and Furmark to see where you're at. Note, if you see your temperatures rising much further than 80C (the GPU might hover around 80-85), stop the stress tests. You don't want to burn the thing down. The upside is that if that happens, you'll know that the problem is heat-related.

As for "dirty power", no, the power supply should be able to handle filtering that kind of stuff before delivering power to the rest of the computer. Dirty power will, however, reduce the lifespan of the power supply, which could cause a number of other problems.

I'm not sure about Corsair's RMA policy lately, I've only ever had to use it once or twice years ago; So far so good otherwise. Then it wasn't really a choice: They'd ship you a replacement upon receipt of the RMA.

[Yakusoku]

14 minutes ago, Runefox said:

When you're having these shutdowns, you say it's happening most often during games? Does the computer go through a shutdown routine, or does it just turn off as though the plug was pulled from the wall? You might want to check the temperatures via Open Hardware Monitor while running a stress test like Prime95 and Furmark to see where you're at. Note, if you see your temperatures rising much further than 80C (the GPU might hover around 80-85), stop the stress tests. You don't want to burn the thing down. The upside is that if that happens, you'll know that the problem is heat-related.

As I have said in my first post, even at "Idle" little no programs are running it will still shutdown. Wether or not its being pushed gaming, rendering or just plain browsing it doesnt seem to be the cause.

[NeoFrux]

Just now, Yakusoku said:

As I have said in my first post, even at "Idle" little no programs are running it will still shutdown. Wether or not its being pushed gaming, rendering or just plain browsing it doesnt seem to be the cause.

Well, what I was referring to was your emphasis "(especially)" on the 3D modeling & Photoshop side of it. I misread it as coming after the comma at first, but yeah. If there's no rhyme or reason to it, there's still some merit to trying a stress test, if only to get the power supply to ramp up and more likely shut down if it can't handle the load. Checking the temperatures either way is prudent at any rate.

[Yakusoku]

Would just simply rendering in Maya or 3DSMax be enough? These would represent real world scenarios.

 

If it is my CPU would go jump to 68-73 (No shutdowns when I render). As for GPU FFXIV (HIghest Settings Possible + Gsync On), at a 3 hour Raid, Dungeon run. I do it almost every day. Temps are the same 69-76 not reaching 80's. (Note: Using 2 Monitors a 1080p Samsung HDTV and a 1440p ROG Swift 144hz)

 

The thing that is problematic is its "Random". It doesnt matter what I do it will just shutdown randomly. Is there a way to find out maybe an "Error" reader to why the system shut down? It would register something when a System just suddenly turns off right?

 

[Xenesis]

I've had experienced random shutdowns during gaming and idling, the type of shutdown where the plug has been pulled. So I did a little experimenting over a week to see what the cause was. My temps during gaming was fine eg. CPU - 45C and GPU will hover around - 68C. Ran memtest and HD tune, no problems, RAM and SSDs+HDDs were fine. So I ran 3D mark... This is where I started to find random shut downs. 3D mark would load the first graphics test, and will instantly crash. So just by randomness, I thought I'll check all my cables, just by plugging out and replugging literally everything. Ran 3D Mark again and then I was finally able to manage to successfully run 3D mark with no issues.

 

Parts

Spoiler

i7 6700K

16GB Corsair 2400 DDR4

EVGA 750W G2

MSI Gaming Z710 M5

R9 390 Sapphire Nitro-X

Windows 10 Home

 

Can't guarantee this will work for you but worth a shot I guess? 

[Yakusoku]

11 hours ago, Xenesis said:

I've had experienced random shutdowns during gaming and idling, the type of shutdown where the plug has been pulled. So I did a little experimenting over a week to see what the cause was. My temps during gaming was fine eg. CPU - 45C and GPU will hover around - 68C. Ran memtest and HD tune, no problems, RAM and SSDs+HDDs were fine. So I ran 3D mark... This is where I started to find random shut downs. 3D mark would load the first graphics test, and will instantly crash. So just by randomness, I thought I'll check all my cables, just by plugging out and replugging literally everything. Ran 3D Mark again and then I was finally able to manage to successfully run 3D mark with no issues.

 

Parts

  Reveal hidden contents

i7 6700K

16GB Corsair 2400 DDR4

EVGA 750W G2

MSI Gaming Z710 M5

R9 390 Sapphire Nitro-X

Windows 10 Home

 

Can't guarantee this will work for you but worth a shot I guess? 

I'll give this a try. At the same time it would be a good time to clean the entire thing as I will unplug everything anyway.

 

Anyone else have an idea to how to diagnose this kind of problem? Perhaps if its not hardware maybe software is a possibility?

[0x1e]

17 minutes ago, Yakusoku said:

I'll give this a try. At the same time it would be a good time to clean the entire thing as I will unplug everything anyway.

 

Anyone else have an idea to how to diagnose this kind of problem? Perhaps if its not hardware maybe software is a possibility?

If it just powers off with no BSOD, freeze or anything else but instead just powers off then it's very unlikely to be a software issue.

I would also believe it's not a safety feature kicking in.

 

Instead it's faulty hardware, or incorrect set up.

Try running Memtest86. If it passes then knock your ram to it's rated timings and use 1333Mhz just to see if it stops shutting down.

If the Memtest86 fails then you can try RMA'ing the ram.

 

Your computer is no overclocked at all apart from the RAM?

[Yakusoku]

13 hours ago, 0x1e said:

If it just powers off with no BSOD, freeze or anything else but instead just powers off then it's very unlikely to be a software issue.

I would also believe it's not a safety feature kicking in.

 

Instead it's faulty hardware, or incorrect set up.

Try running Memtest86. If it passes then knock your ram to it's rated timings and use 1333Mhz just to see if it stops shutting down.

If the Memtest86 fails then you can try RMA'ing the ram.

 

Your computer is no overclocked at all apart from the RAM?

My computer is not overclocked not even my ram. Its speed is 2133mhz, but upon checking cpu-z the "NB Frequency" is at 3997.6 and changes to 3996.2

Is the Frequency the same as the speed its rated for? Because I dont remember changing its original speed.

 

PS: Test Pass no errors with Memtest86

[0x1e]

10 hours ago, Yakusoku said:

My computer is not overclocked not even my ram. Its speed is 2133mhz, but upon checking cpu-z the "NB Frequency" is at 3997.6 and changes to 3996.2

Is the Frequency the same as the speed its rated for? Because I dont remember changing its original speed.

 

PS: Test Pass no errors with Memtest86

Anything above 1333Mhz for DDR3 is classed as overclocked.

Your ram is able to hit 2133 but that doesn't mean your IMC (memory controller in the CPU) is able to keep up with that speed.

With a 4790k it's unlikely it can't go to 2133 and above, but I think it's best to eliminate it from the equation first as it's one of the easiest.

 

If your ram is rated for 2133 @ 12-12-12-36 @ 1.5 volts (random numbers used)

Then use

1333 @ 12-12-12-36 @ 1.5V

[Yakusoku]

4 hours ago, 0x1e said:

Anything above 1333Mhz for DDR3 is classed as overclocked.

Your ram is able to hit 2133 but that doesn't mean your IMC (memory controller in the CPU) is able to keep up with that speed.

With a 4790k it's unlikely it can't go to 2133 and above, but I think it's best to eliminate it from the equation first as it's one of the easiest.

 

If your ram is rated for 2133 @ 12-12-12-36 @ 1.5 volts (random numbers used)

Then use

1333 @ 12-12-12-36 @ 1.5V

Ok set the Frequencey down to 1333, do I have to run memtest for this? Or just do a 2-3 day test and see if I still get random shutdowns?

[0x1e]

17 hours ago, Yakusoku said:

Ok set the Frequencey down to 1333, do I have to run memtest for this? Or just do a 2-3 day test and see if I still get random shutdowns?

If memtest passed at 2133 then it should pass 1333. So it would waste time.

I say just use your machine like normal. If the machine crashes on the loads that you state then the problem should appear easily if it's still there.

[Yakusoku]

9 hours ago, 0x1e said:

If memtest passed at 2133 then it should pass 1333. So it would waste time.

I say just use your machine like normal. If the machine crashes on the loads that you state then the problem should appear easily if it's still there.

2133 Passed no errors.