Question PC is randomly powering off ?

[TheDarkOne198]

I have been getting some random powering off,it has happened twice just today but has not happened in a while before today. It will just power off,leaving the CPU debug light on,on the motherboard and the HDD light on the front of my case is solid on. To me this feels like it could be CPU overheating but everything is running cool enough according to HW Info and Ryzen Master and it has done it while idle and the CPU is not being stressed. Even now,all 16 cores of my 3950X are hovering between 39 and 45 C,depending on the core. I was using PBO earlier when it did it both times but its never run hot under my NH-D15 except on PBO and doing something like a cinebench run and with that it reports some of the cores hitting 90C or higher. Something that really stresses it. Other stresses like CPU-Z or the CPU test in 3D Mark do not get it that hot. Is this indicative of overheating? It really does not seem like it should be but by what it looks like,it must be,IMO. Setting to default,with cinebench running,all the cores are sitting around 65-72,depending on the core. Whats does anyone think? Could something else cause the powering off and the CPU light on,with a solid HDD light?

EDIT: It could also be my GPU overheating. I just noticed in nicehash that fan speed was 0% on my 6750XT even though it it should not be. I think some kind of shenanigans might have been going on with the fan control in the driver or something. I usually let MSI afterburner control the fan with its default fan tune profile but if for some reason my Merc 319's fans were not spinning,it could have easily overheated even sitting idle,I would think. Turning Afterburners fan control off and on a couple of times seemed to have fixed it. IDK if that was really the cause or if it would cause the CPU light to be on like that.

 

[Aeacus]

Full system specs, including PSU make and model (or part number) is? Also, how old the PSU is, and was the PSU bought new or used/refurbished?

 

[TheDarkOne198]

Full system specs, including PSU make and model (or part number) is? Also, how old the PSU is, and was the PSU bought new or used/refurbished?

MSI X570-A Pro
Ryzen 9 3950X
32 GB of 3600 MHz DDR4 (16GBx2)
Radeon RX 6750XT
1TB 970 EVO plus (System drive)
2TB WD Blue SSD
1TB WD Blue HDD
Super Flower Leadex SE Platinum 1200W (SF-1200F14MP), a little over 1-year-old and bought new.

It IS using the 8+4 EPS power because the board has the +4 and I have read it can help provide cleaner power to the CPU. As I said, It might be an overheating issue but there are so many times it runs fine. I just did another cinebench run for a few minutes and the hottest core almost hit 74C which is perfectly fine...but other times I have seen them hit 80C and nothing is different. What can cause them to run hotter sometimes but not others, doing the same things? Could the +4 EPS power cause things to go weird?

 

[sonofjesse]

Are you mining while it shutsdown or what?

 

[Aeacus]

Could the +4 EPS power cause things to go weird?

Highly unlikely.

I have been getting some random powering off

For most of the times, it comes down to two things: PSU or main electricity grid.

To test/fix both, you need: 2nd, good quality PSU, to test if you get the same shut down issues as well.
If you don't, issue was with your Super Flower unit (while good quality unit, there are lemons even among the best). If you do get same shut downs even with 2nd, good quality PSU, issue isn't with PSU.

For main electricity grid issues, you need an UPS, which keeps your PC running when blackouts/brownouts happen. I suggest line-interactive, true/pure sine wave UPS, at least 1200W capacity.
If shut down issues go away when your PC is backed up by an UPS, great, you solved your issue. But if shut down issues remain, even after using UPS and have tested with 2nd PSU, then issue could be either high thermals, or MoBo VRM.

Fixing high thermals is obvious. Fixing MoBo VRM = MoBo replacement.


This above, is when issue is with hardware. But if issue is with software, best option is clean Win install. Still, it could be that the issue is with the mining software you're using.
Here, and regards with mining software support, my help ends. My moral compass won't allow to help any miners.

 

[TheDarkOne198]

Highly unlikely.



For most of the times, it comes down to two things: PSU or main electricity grid.

To test/fix both, you need: 2nd, good quality PSU, to test if you get the same shut down issues as well.
If you don't, issue was with your Super Flower unit (while good quality unit, there are lemons even among the best). If you do get same shut downs even with 2nd, good quality PSU, issue isn't with PSU.

For main electricity grid issues, you need an UPS, which keeps your PC running when blackouts/brownouts happen. I suggest line-interactive, true/pure sine wave UPS, at least 1200W capacity.
If shut down issues go away when your PC is backed up by an UPS, great, you solved your issue. But if shut down issues remain, even after using UPS and have tested with 2nd PSU, then issue could be either high thermals, or MoBo VRM.

Fixing high thermals is obvious. Fixing MoBo VRM = MoBo replacement.


This above, is when issue is with hardware. But if issue is with software, best option is clean Win install. Still, it could be that the issue is with the mining software you're using.
Here, and regards with mining software support, my help ends. My moral compass won't allow to help any miners.

To be fair,I was only mining for extra heat generation because its below freezing. I stopped for mining for profit when it stopped being profitable but its cheaper than running the heater,lol. I never did it for anything other than pocket change,anyway. No longer doing that and I just wanted the room warmer.

One time it did it earlier this year,I recall and it refused to turn back on for several minutes, either not powering on or again just giving the solid CPU light and the disk activity light was solid on. Eventually it actually just turned itself on. Its why I am leaning towards thermal issues. This feels like it was maybe too hot somewhere and needed to cool.

As for testing PSU and all that,I really do not have the means to do it so I either need to repaste and see if that helps or just hope everything just keeps working. It was not doing it with the previous board I had,which started having POST issues and I cannot afford a UPS right now. Or much else. I do not think it was a brown out though,lights did not flicker. As I said,I am leaning towards thermal issues. Possibly one core going too hot or like you said,VRM issues but I hope not....I cannot afford yet another motherboard and the windows installation is still pretty fresh as it is,too. I also regularly run SFC to keep things in order. Thanks anyway.

 

[TheDarkOne198]

Are you mining while it shutsdown or what?

I think so but I do not remember. I know it has happened while gaming before too. The mining was just because of it being so cold and I wanted a bit more heat,but it should be warming up a bit in the next few days so I will probably stop again. I wonder if it was power spikes but my hardware should not trip OCP or OVP on a 1200W Platinum unit but who knows.

 

[sonofjesse]

I would buy a space heater and NOT use the mining rig.

See if you have any more issues. Sounds like you was pushing something too hard and the power supply didn't like it.

 

[Aeacus]

I do not think it was a brown out though,lights did not flicker.

All PSUs must have hold-up time of at least 16ms (milliseconds), while your unit falls short of the target, of only having 13.6ms hold-up time,
review: https://www.techpowerup.com/review/super-flower-sf-1200f-14mp/5.html

Meaning that when power is lost for up to 13.6ms, your PSU can keep your PC running.

Do you have any idea how short of a time 13.6ms is? 1 second is 1000 milliseconds. Average human reaction time is 150-300 milliseconds. So, when there was a power cut, between 14 and 150 ms, you even didn't notice it, while it was more than enough for your PSU, to shut down your PC.
So, just because lights didn't flicker, doesn't mean there wasn't a power loss.

Btw, line-interactive UPS has transfer time of 2ms, from the moment of power loss to the battery power. Which is more than enough time for PSU to keep the PC running during transfer time, but completely un-noticeable to every human.

I have UPSes in use with my PCs and sometimes, for a split second, i've heard my UPS to switch to battery power and then back to main power. Just because i didn't notice a power cut, that was only some milliseconds long, doesn't mean there wasn't a power cut. There was, and my UPS also catched it, keeping my PC running without issues.

To be fair,I was only mining for extra heat generation because its below freezing. I stopped for mining for profit when it stopped being profitable but its cheaper than running the heater,lol.

Using mining rig solely as a heating unit, in hopes to heat up your room with decay heat, is very inefficient. You'd be wasting a far more electricity compared to the dedicated space heater. Also, it's the worst excuse i've ever heard from a miner, as of why to mine.
Btw, only way of mining rig as a heater being cheaper, compared to the dedicated space heater, is when you don't pay for electricity bill.

Oh, proper space heater costs as little as 36 bucks,
amazon: https://www.amazon.com/Overheat-Protection-Tip-Over-Portable-Thermostat/dp/B07W4SN7N7/

 

[TheDarkOne198]

All PSUs must have hold-up time of at least 16ms (milliseconds), while your unit falls short of the target, of only having 13.6ms hold-up time,
review: https://www.techpowerup.com/review/super-flower-sf-1200f-14mp/5.html

Meaning that when power is lost for up to 13.6ms, your PSU can keep your PC running.

Do you have any idea how short of a time 13.6ms is? 1 second is 1000 milliseconds. Average human reaction time is 150-300 milliseconds. So, when there was a power cut, between 14 and 150 ms, you even didn't notice it, while it was more than enough for your PSU, to shut down your PC.
So, just because lights didn't flicker, doesn't mean there wasn't a power loss.

Btw, line-interactive UPS has transfer time of 2ms, from the moment of power loss to the battery power. Which is more than enough time for PSU to keep the PC running during transfer time, but completely un-noticeable to every human.

I have UPSes in use with my PCs and sometimes, for a split second, i've heard my UPS to switch to battery power and then back to main power. Just because i didn't notice a power cut, that was only some milliseconds long, doesn't mean there wasn't a power cut. There was, and my UPS also catched it, keeping my PC running without issues.



Using mining rig solely as a heating unit, in hopes to heat up your room with decay heat, is very inefficient. You'd be wasting a far more electricity compared to the dedicated space heater. Also, it's the worst excuse i've ever heard from a miner, as of why to mine.
Btw, only way of mining rig as a heater being cheaper, compared to the dedicated space heater, is when you don't pay for electricity bill.

Oh, proper space heater costs as little as 36 bucks,
amazon: https://www.amazon.com/Overheat-Protection-Tip-Over-Portable-Thermostat/dp/B07W4SN7N7/

I actually don't pay for it,living with family and we do not have space heaters due to pets running around but do have gas central heating. I just figured little added heat so it does not run as much might actually save money. But I might be wrong. I always remember hearing miming was something one can do in the winter because of the heat it makes being partially reimbursed by the mining income,not that you can make much these days anyway. I only ever mined to make a little bit here and there with hardware I happened to have anyway and since the ethereum merge,that became pointless.

In any case,about the power cutting out,you have a good point. I wish I could afford a UPS though,if it was a brown out,there was not much I could do about it. It is winter time and I bet brown outs are more common. If it is just that,not much I can do about it.

 

[TheDarkOne198]

I would buy a space heater and NOT use the mining rig.

See if you have any more issues. Sounds like you was pushing something too hard and the power supply didn't like it.

Maybe and since it's warming back up a little I have stopped mining. As I said,it has happened while gaming too but only once or twice all year. I would not think my hardware could overcome a 1200W PSU as I have assumed everything in my PC,all at once should only be close to half of its capacity. I know transients are a thing though and maybe it is enough.

 

[Aeacus]

I actually don't pay for it,living with family and we do not have space heaters due to pets running around but do have gas central heating.

I figured as much. Since if you don't have to pay for electricity out of your own pocket, it is easy to waste electricity.

In a similar example:
Let's say you have garage and to heat it up, you keep the car engine running (decay heat from the car engine). But since you don't pay for the gasoline/diesel, you may "think" that doing so is cost effective, while in reality, it is extremely wasteful. Both the resource and money to pay for it.

Btw, i too have a pet and my doggo doesn't go poking into my space heater (switchable between 2000W and 1000W) when i turn it on, when need be. I too have central heating (hot water based), but sometimes, it isn't enough.

I always remember hearing miming was something one can do in the winter because of the heat it makes being partially reimbursed by the mining income

Mining with one rig and one GPU, has never been profitable. Electricity cost alone is what kills any profits small time miners hope to get. And then there is wear and tear on the hardware as well.

In a similar example:
What you have, is flower pot with some dirt in it, and what you're looking for, is competing with farmers with thousands of acres of land, in hopes to make a profit in growing food. That's how comparable single GPU mining is, to proper mining rigs with hundreds/thousands of GPUs.

 

[TheDarkOne198]

I figured as much. Since if you don't have to pay for electricity out of your own pocket, it is easy to waste electricity.

In a similar example:
Let's say you have garage and to heat it up, you keep the car engine running (decay heat from the car engine). But since you don't pay for the gasoline/diesel, you may "think" that doing so is cost effective, while in reality, it is extremely wasteful. Both the resource and money to pay for it.

Btw, i too have a pet and my doggo doesn't go poking into my space heater (switchable between 2000W and 1000W) when i turn it on, when need be. I too have central heating (hot water based), but sometimes, it isn't enough.



Mining with one rig and one GPU, has never been profitable. Electricity cost alone is what kills any profits small time miners hope to get. And then there is wear and tear on the hardware as well.

In a similar example:
What you have, is flower pot with some dirt in it, and what you're looking for, is competing with farmers with thousands of acres of land, in hopes to make a profit in growing food. That's how comparable single GPU mining is, to proper mining rigs with hundreds/thousands of GPUs.

It used to be somewhat profitable, enough to buy a game every now and then but that's no longer a thing. In any case, I have mostly stopped anyway. Now I just let something like MetaTrader 5 Strategy Tester make what it can and aside from the unusual powering off that I have seen a few times this year, which I am guessing is just a brownout,as you said...I get occasional BSODs. I had PBO on but I see both Page Fault in Non-Paged Area and IRQL Not Less or Equal and the other day, Unexpected Store Exception and the only other time I see that last one is in a particular game, and thought it was a bug in the game. It could be something else though. I got that first one earlier, just watching YouTube. Maybe MetaTrader overheated a core too much. All drivers are up to date and I have no idea what is causing them. Even WhoCrashed is not of any help.