[SOLVED] Shutdown when Gaming (CPU or PSU or RAM) - Need help?

[RalphJoy22]

Greetings to everyone,

I am hoping to receive some help from experienced members on here regarding the issue I am facing. My desktop is kinda old, almost 4-5 years now and I don't usually use it for graphics intensive gaming either.

(I use to game heavily back in the day - not anymore. Only light gaming)

But once in a while, I do love playing games like Apex Legends or Rocket League for a bit. But starting last week I have noticed something quite unusual. My CPU idle temp is showing 63-70'c via Speccy - not sure if that is accurate. But when I do start games after like 5-15 minutes, my PC shuts down.

No error. No BSOD. Nothing. Just complete shutdown.
I've had many issues in the past with PSU's - so I am not sure if that's the case.

Regards.

 

[TechyInAZ]

Hmm, that is cooler masters very low end brand. It would be worth a shot trying to get something better like a Seasonic GOLD series or anything brand of PSU that has Superflower components (like EVGA G2 and G3).

Could be too that your thermal paste is getting hard. Given that you built your rig 5 years ago...thermal paste starts to loose it's liquid form after 4 years. Once it hardens, heat transfer becomes a lot more difficult.

 

[anotherdrew]

Your CPU does seem to running hot and that could be the result of a bad reading or the pump on your water cooler not working efficiently anymore. I would start by checking the idle temp in BIOS and then troubleshooting the cooler. Overheating can cause shutdowns like you describe.

 

[RalphJoy22]

Hmm, that is cooler masters very low end brand. It would be worth a shot trying to get something better like a Seasonic GOLD series or anything brand of PSU that has Superflower components (like EVGA G2 and G3).

Could be too that your thermal paste is getting hard. Given that you built your rig 5 years ago...thermal paste starts to loose it's liquid form after 4 years. Once it hardens, heat transfer becomes a lot more difficult.

Thank you for your response. Indeed it was a budget build and I didn't really go for HQ components back then. I wouldn't mind getting a PSU if it is indeed failing, any way to figure that out? Though I am sure, thermal paste was re-added 2 years ago, while I was cleaning it.

Your CPU does seem to running hot and that could be the result of a bad reading or the pump on your water cooler not working efficiently anymore. I would start by checking the idle temp in BIOS and then troubleshooting the cooler. Overheating can cause shutdowns like you describe.

Thank you for your response, so I loaded into the BIOS and I reduced the system performance and the idle CPU temp is at 59-61' currently. But inside the desktop, it shows differently like 70+

 

[anotherdrew]

Thank you for your response, so I loaded into the BIOS and I reduced the system performance and the idle CPU temp is at 59-61' currently. But inside the desktop, it shows differently like 70+

60 C in BIOS still seems high. I know the FX 8350 tends to run on the warm side, but that still seems too warm. Are you running software that monitors the pump speed? Have you checked if the radiator is full of dust?

 

[RalphJoy22]

60 C in BIOS still seems high. I know the FX 8350 tends to run on the warm side, but that still seems too warm. Are you running software that monitors the pump speed? Have you checked if the radiator is full of dust?

The radiator seems to be clean albeit a bit dusty. Well, I don't have any software to monitor the pump speed sadly. I don't even remember if Antec provided a CD or not for its companion software.

I had a liquid cooler before this one (It was an Antec one as well and within 2 years, it went bust). Now, if I remember correctly - I think the current one is also nearing the 3-year mark.

Also, I was monitoring the temps while playing a game of RL right now and the temp was hovering between 80-85'. Just got shut down again.

 

[anotherdrew]

The Grid software in the link below should be able to show pump speed.

http://store.antec.com/cpu-liquid-cooling/kuhler-h2o-950.html

 

[mdd1963]

Often the jump to a liquid cooler for AMD's FX8350 then deprived the mainboard's VRM of needed cooling from benefit of the CPU fan's blowing air downward....

Try opening the case, and posiitioning a small fan pointed across the mainboard's voltage regulator circuitry....; if things improve, you might have stumbled onto the reason....

 

[RalphJoy22]

The Grid software in the link below should be able to show pump speed.

http://store.antec.com/cpu-liquid-cooling/kuhler-h2o-950.html

Thank you for this, I will install it and see how that goes.

Often the jump to a liquid cooler for AMD's FX8350 then deprived the mainboard's VRM of needed cooling from benefit of the CPU fan's blowing air downward....

Try opening the case, and posiitioning a small fan pointed across the mainboard's voltage regulator circuitry....; if things improve, you might have stumbled onto the reason....

I am not much familiar with what you just said but I am guessing that, if putting a small fan aiming towards the MB brings down the temp - we are looking at a busted liquid cooler, is that it? (lol - sorry still learning about these things)

 

[RalphJoy22]

The Grid software in the link below should be able to show pump speed.

http://store.antec.com/cpu-liquid-cooling/kuhler-h2o-950.html

Right, so I tried installing but it doesn't detect the pump like at all.

 

[anotherdrew]

Right, so I tried installing but it doesn't detect the pump like at all.

Sorry that didn't work. At this point it seems you might be at the next level of troubleshooting, which might require spending some money.

1. The PSU could be the issue. Best way to test it is with another PSU (there are test rigs ... if you knew a shop that had one, great ... otherwise they are cost prohibitive).
2. This could be a heat issue with the CPU or VRMs. Some fans might help, but there is no easy, conclusive test.
3. We have not even talked about memory. Fortunately, that is an easy test as you can run the machine with one stick in different positions or run memory testing software.

 

[mdd1963]

HWMonitor will show you the actual CPU core temps....and clock speeds. If core temps not reaching throttle temps, but throttling appears to be occurring, then the MB would be the one doing the throttling (if it is even occurring), potentially due to VRM temps and/or power limits...

Cooling off the VRM circuitry with a simple house fan and an open case would seem a simple thing to test....

If results the same, I'd not think VRM throttling an issue....

Naturally, a PSU can easily cause insta-power-off symptoms as well....at some point, it will also need to swapped (at least temporarily) to rule out as a potential cause...

 

[Dark Lord of Tech]

It could be the very poor PSU , but that board has no heatsinks on the VRM , so most likely the CPU is throttling and shutting down. That board shouldn't have a 125w CPU mounted on it. Very poor power phase and thermal design.

 

[RalphJoy22]

Sorry that didn't work. At this point it seems you might be at the next level of troubleshooting, which might require spending some money.

1. The PSU could be the issue. Best way to test it is with another PSU (there are test rigs ... if you knew a shop that had one, great ... otherwise they are cost prohibitive).
2. This could be a heat issue with the CPU or VRMs. Some fans might help, but there is no easy, conclusive test.
3. We have not even talked about memory. Fortunately, that is an easy test as you can run the machine with one stick in different positions or run memory testing software.

Right, I was thinking of an upgrade anyways. So, please help me decide as I am not much familiar with on how to proceed regarding the hardware specs.

HWMonitor will show you the actual CPU core temps....and clock speeds. If core temps not reaching throttle temps, but throttling appears to be occurring, then the MB would be the one doing the throttling (if it is even occurring), potentially due to VRM temps and/or power limits...

Cooling off the VRM circuitry with a simple house fan and an open case would seem a simple thing to test....

If results the same, I'd not think VRM throttling an issue....

Naturally, a PSU can easily cause insta-power-off symptoms as well....at some point, it will also need to swapped (at least temporarily) to rule out as a potential cause...

Well, I was thinking of upgrading it all together, so please help me alongside the others on how to proceed.

It could be the very poor PSU , but that board has no heatsinks on the VRM , so most likely the CPU is throttling and shutting down. That board shouldn't have a 125w CPU mounted on it. Very poor power phase and thermal design.

Well, it has served me well all these years without any problems. So, are you suggesting I need to change the mobo? I don't know what VRM is or what it does.

---------------------------------------------------------------------------------

At this point, I am quite frustrated because I've not been able to use my PC properly. So, here is the thing. I am thinking of upgrading my desktop and all the parts within it but I won't be able to do it all in one-shot.

Now, I don't know with which parts I should start first but these are the things I've on the mind.

CPU:
Intel - Core i7-8700K 3.7 GHz 6-Core Processor

Cooler:
Corsair Hydro Series, H80i V2, 120mm Dual Thick Radiator, Dual 120mm PWM Fans, Advanced RGB Lighting And Fan Control With Software, Liquid CPU Cooler

Motherboard:
ASUS TUF H370-PRO GAMING Motherboard (Intel Socket 1151/8th Generation Core Series CPU/Max 64GB DDR4-2666Mhz Memory)

RAM:
G.Skill F4-2666C15S-16GVR Desktop Ram Ripjaws V Series - 16GB (16GBx1) DDR4 2666MHz

PSU:
Antec Hcg-1000-Extreme Smps 1000 Watt 80 Plus Gold Certification Fully Modular Psu With Active Pfc

GPU (Either one):
MSI GRAPHICS CARD PASCAL SERIES - GTX 1070 TI ARMOR 8GB GDDR5
Msi GeForce RTX 2060 Ventus XS OC 6GB GDDR6 192-Bit Gaming Graphics Card, Active Fan Control, Metal Backplate
GALAX GeForce RTX 2070 EX (1 Click OC) 8GB GDDR6 256-Bit Gaming Graphics Card Active Fan Control Metal Backplate RGB Effect

SSD:
Samsung 860 QVO 1TB Internal SSD (MZ-76Q1T0B)

HDD:
WESTERN DIGITAL DESKTOP HARD DRIVE 4TB BLUE (WD40EZRZ)

--------------------------------------------------------------------------

Now, I was thinking of changing the Cooler & PSU first and see if that solves the problem. At which point, if my PC starts to function normally - I'd be able to get the remaining over the consecutive weeks. So, something like ->

1) Cooler & PSU --> 2) Mobo, RAM & CPU --> 3) GPU --> 4) SSD & HDD

Just want to know from you guys, if that is the ideal way to proceed? And if all the parts I shortlisted above are compatible. Not much of a tech guy so would appreciate the fullest help.

Regards.

 

[anotherdrew]

I think the idea of changing the cooler first is a good idea. Personally, I would go with air cooling unless you planning to overclock or you have a small case with limited air flow.

THIS is what I use on my i5-8600k ... what ever you get, make sure it is AM3 and 1151 compatible.

For the PSU ... might I suggest you look at THIS and THIS. 1000 Watts is overkill if you are not planning to run 2 video cards

I currently have an RTX 2060 and it's great, but I only game 5-10 hours a week and it's mostly goofing around. When I'm working the computer is very quiet which is desirable to me.

I might do the SSD at the same time as I do the CPU/MB/Memory because it will require an OS reinstall (I assume your OSes are going on the SSD).

 

[RalphJoy22]

I think the idea of changing the cooler first is a good idea. Personally, I would go with air cooling unless you planning to overclock or you have a small case with limited air flow.

THIS is what I use on my i5-8600k ... what ever you get, make sure it is AM3 and 1151 compatible.

For the PSU ... might I suggest you look at THIS and THIS. 1000 Watts is overkill if you are not planning to run 2 video cards

I currently have an RTX 2060 and it's great, but I only game 5-10 hours a week and it's mostly goofing around. When I'm working the computer is very quiet which is desirable to me.

I might do the SSD at the same time as I do the CPU/MB/Memory because it will require an OS reinstall (I assume your OSes are going on the SSD).

Wow, thank you just the type of advice I was looking for.

Let's see. I am going for the liquid cooling because my case is a mid-tower which I'll upgrade to a full tower in the end after the major upgrades are done. Actually, the area where my desktop is has limited airflow - so, I'd prefer it has proper cooling.

Ah dang it, I am not from the US and those are expensive for me. The Seasonic one costs me $214 and the EVGA one will cost me $224. However, THIS will cost me only $127 - can that be used? Well, I was thinking of doing SLI in the future - would that be possible?

Well I believe, it's a solid upgrade over my 1060 right now. From 760 to 1060 and now 2060, possibly an SLI in the future. So, should be just fine for 1080p/1440p gaming.

As for the SSD, yea I guess I'll do it like this then:
1. Cooler & PSU --> 2. CPU, Mobo, RAM & SSD --> 3. GPU --> 4. HDD --> 5. Cabinet & Second GPU.

What do you think?

 

[anotherdrew]

The Corsair RMx series is very good. RMi is slightly better, but I wouldn't spend more than a $20 premium. THIS is a pretty good tier list. You will find other tier lists out there for more options. Don't hesitate to check reviews too.

1000 Watts is still a lot, especially with the nVidia cards. 300 W per card, 100 W CPU, 50 W for accessories (these are inflated round numbers ... the RTX 2070 has max draw around 230W) puts you at 750 W ... add 100 W headroom and you would be well served by an 850 W. Then again, prices when you get into that range don't change much so look for bang for the buck. Remember to look if you have enough connectors ... especially if you want to run SLI at some point.

Regarding SLI, with the RTX series nVidia is moving away from SLI support and SLI support in games is limited. The multi GPU setup is more for high end video editing and rendering applications.

Ventilation ... if you are seeing MB temperatures in excess of 40 C you need to think about more case fans. HWinfo (as mentioned above) is a great resource for monitoring temperatures. HWmonitor is a bit easier to read, but some people have had issues with it. Asus also offers monitoring software.

 

[RalphJoy22]

The Corsair RMx series is very good. RMi is slightly better, but I wouldn't spend more than a $20 premium. THIS is a pretty good tier list. You will find other tier lists out there for more options. Don't hesitate to check reviews too.

1000 Watts is still a lot, especially with the nVidia cards. 300 W per card, 100 W CPU, 50 W for accessories (these are inflated round numbers ... the RTX 2070 has max draw around 230W) puts you at 750 W ... add 100 W headroom and you would be well served by an 850 W. Then again, prices when you get into that range don't change much so look for bang for the buck. Remember to look if you have enough connectors ... especially if you want to run SLI at some point.

Regarding SLI, with the RTX series nVidia is moving away from SLI support and SLI support in games is limited. The multi GPU setup is more for high end video editing and rendering applications.

Ventilation ... if you are seeing MB temperatures in excess of 40 C you need to think about more case fans. HWinfo (as mentioned above) is a great resource for monitoring temperatures. HWmonitor is a bit easier to read, but some people have had issues with it. Asus also offers monitoring software.

Ah okay, well I guess I'll get that one because the price is very good compared to the others being provided. Regarding the connectors, is that dependent on the PSU?

In that case, I won't go SLI - probably will try getting 2070 in that case. I hope that will be sufficient for 1080/1440p gaming.

Yea, I never went for external case fans - but this time, I'll try and add them for proper airflow throughout the system.

 

[anotherdrew]

The 2070 will kick butt at 1080 and be very good for 1440. It also pairs nicely with the i7-8700k. You will have a computer that will keep you happy for 3-5 years easy. PSUs all come with a certain set of cables/connectors. If you wanted to run SLI, then you would want to make sure that your PSU came with 2 PCI-E cables each with a pair of 6+2 connectors. An 8 and a 6+2 normally will work, but you lose some flexibility (incase your graphics card has 2 6 pin plugs ... which is uncommon). The same situation applies if you wanted to have 10 SATA drives (most PSUs don't come with 10 SATA power plugs). Basically, it's just making sure that you get what you need and not finding out your choice is deficient when doing your build (something many of us learned the hard way).

Fans ... for a gaming computer I see 2 120mm fans blowing in (probably in the front) and 1 120 blowing out (back or top) as a minimum. Lots of fan choices out there ... some good, some just pretty ... but that is another post altogether.

Enjoy!

 

[RalphJoy22]

The 2070 will kick butt at 1080 and be very good for 1440. It also pairs nicely with the i7-8700k. You will have a computer that will keep you happy for 3-5 years easy. PSUs all come with a certain set of cables/connectors. If you wanted to run SLI, then you would want to make sure that your PSU came with 2 PCI-E cables each with a pair of 6+2 connectors. An 8 and a 6+2 normally will work, but you lose some flexibility (incase your graphics card has 2 6 pin plugs ... which is uncommon). The same situation applies if you wanted to have 10 SATA drives (most PSUs don't come with 10 SATA power plugs). Basically, it's just making sure that you get what you need and not finding out your choice is deficient when doing your build (something many of us learned the hard way).

Fans ... for a gaming computer I see 2 120mm fans blowing in (probably in the front) and 1 120 blowing out (back or top) as a minimum. Lots of fan choices out there ... some good, some just pretty ... but that is another post altogether.

Enjoy!

I see that makes sense - so I'll go with 2070 as it is more future-proof for the build.

So, I ordered the PSU I shared, Corsair RM 1000x and the Corsair H80i v2. So, PSU & Cooler is on the way and hopefully, with this, the existing problem of my desktop shutting down will be solved.

As I'll not be going with SLI - I believe I'll be fine. After a few more days of saving, I'll have enough for the CPU/Mobo/RAM/SSD.

Thank you for your help!!