My PC keeps turning off...

[Emily Young]

The Desk PC 2.0 has a problem, and Linus is about to go for vacation. Is Anthony a bad enough dude to diagnose and repair his workstation before he comes back?

 

 

Buy AMD Threadripper 3970X CPU
On Amazon: https://geni.us/Tb18Mc
On Newegg: https://geni.us/yB3h
On B&H: https://geni.us/dxaiVMR

 

Buy LG CX 48-inch 4K OLED TV
On Amazon: https://geni.us/TOEzA
On Best Buy: https://geni.us/u7mVGR
On Newegg: https://geni.us/uvVjyDS

[SignatureSigner]

Just needs a good kick.

[Skiiwee29]

@GabenJr noticed when you disabled C-states, it didn't show you when you saved that you made changes. Did you have to go re-disable it after?

 

[komar]

Video on how to find bad capacitors or coil wines next please.

Please Anthony, 3 out of my 4 alt boards are with such problems

[Velcade]

Can't wait to watch this when I get home.  I updated the bios on a PC last night and I've noticed it's been restarting randomly.  Hopefully a similar fix. 

[RejZoR]

You can limit the scope of how far C-States are allowed to go (to which C state level).

[creesch]

Good troubleshooting in general. However I would have done a few things first personally: 

 

1. Reset the bios for good measure, although in most cases flashing the bios does reset a bunch of settings.
2. Check for newer chipset drivers, certainly if it appears that powermanagement is involved these might actually contain fixes already. 

Other than that it is a solid video going that goes through the basics of troubleshooting which is the most important part

[sub68]

my life as a IT support in a nutshell

[Radium_Angel]

3 hours ago, komar said:

Video on how to find bad capacitors

They are pretty easy to spot.

 

[kblocks]

Just wanted to say, that this is a true Tech Tip! Maybe I haven't been paying much attention, but I feel like a lot of the videos aren't really tech tips (not that the channel is supposed to be limited to those).

 

Pretty cool though. Thanks!

[BaneWilliams2]

Does anyone know what chair Anthony uses? It looks like a gaming style chair, and I bet it's damn comfortable - I'd like to pick one up for myself.

[NemesisPrime_691]

So, Linus updated the BIOS and mentions that the system has been restarting since then, and he does not try to downgrade it to the previous, perfectly working version? 

 

I am all for these tech tips of diagnosing the issues and trying to fix them, but disabling C states is a sort of weird solution since you are basically disabling a feature that was working previously without any problems.

 

I would've tried to downgrade the BIOS first and if that did not work, then go through these steps. Just a quicker solution. Most of the motherboard manufacturers must also have some sort of a BIOS downgrade (flashback?) feature as well right? Why not use it then?

[komar]

15 hours ago, Radium_Angel said:

They are pretty easy to spot.

 

Mine are not sadly. The board takes 19V laptop charger and makes 12,5 and 3.3 and now its power cycling whenever I try to boot it.

[Velcade]

So I've updated the BIOS on 4 machines recently; two Intel two AMD, all ASUS motherboards.  I've noticed that the two AMD builds have been having restart issues and the two intel builds have been fine.  Looking at the C-states on each board the AMD boards default to enabled and the Intel boards default to auto.

 

Once I disable it on the AMD boards I don't have the random restarts anymore.  Strange distinction between the two.  Why not leave it auto for both.  Could this be some 'trickery' on AMD's part to claim their CPUs use less power than intel? 

[derr12]

6 hours ago, NemesisPrime_691 said:

So, Linus updated the BIOS and mentions that the system has been restarting since then, and he does not try to downgrade it to the previous, perfectly working version? 

 

I am all for these tech tips of diagnosing the issues and trying to fix them, but disabling C states is a sort of weird solution since you are basically disabling a feature that was working previously without any problems.

 

I would've tried to downgrade the BIOS first and if that did not work, then go through these steps. Just a quicker solution. Most of the motherboard manufacturers must also have some sort of a BIOS downgrade (flashback?) feature as well right? Why not use it then?

Mentioned he needed to for windows 11,  There are some security things that early bios versions lack in a lot of cases. so he would have to downgrade windows as well if he wanted to do that. 

 

 

[Naveen Prashanth]

On 1/13/2022 at 11:30 PM, GabenJr said:

The Desk PC 2.0 has a problem, and Linus is about to go for vacation. Is Anthony a bad enough dude to diagnose and repair his workstation before he comes back?

 

 

Buy AMD Threadripper 3970X CPU
On Amazon: https://geni.us/Tb18Mc
On Newegg: https://geni.us/yB3h
On B&H: https://geni.us/dxaiVMR

 

Buy LG CX 48-inch 4K OLED TV
On Amazon: https://geni.us/TOEzA
On Best Buy: https://geni.us/u7mVGR
On Newegg: https://geni.us/uvVjyDS

You can use AMIBCP v5.0.1 app to dump your bios file and afuwin to edit your bios and rebuild (mod) it and update that bios to check if enabling/disabling any hidden settings will help you (also changing access to supervisor will make it visible in the bios menu directly)

 

[GeniusMonkey]

I ran into almost exactly the same issue with my Threadripper 3970X workstation when I switched from a PCIe 3.0 to PCIe 4.0 graphics card - random reboots when mostly idle with no real information in the event log. Turning off global C states in the BIOS prevented the issue from occurring, similar to what Anothony saw. The other setting that prevented the issue from occurring was forcing PCIe 3.0 in the BIOS. Turning off global C states is not ideal since it can impact the maximum single core Turbo and may slightly impact power consumption.

 

After a few frustrating days a trying to figure out what was going on, I finally tracked down the root cause of the issue: a memory overclock that was just on the edge of stability. Disabling XMP/DOCP and any manual memory timing adjustments prevented the issue from occurring. I was then able to refine this to just slightly increasing Trc for the memory to fix the issue. I hypothesize that incorrect values were being read from memory when coming out of the package C6 state (which saves core state to memory), resulting in badness.

 

It is plausible that a BIOS update could change some of the memory timings not included in the XMP profile resulting in the issue.

 

[NemesisPrime_691]

On 1/14/2022 at 11:47 PM, derr12 said:

Mentioned he needed to for windows 11,  There are some security things that early bios versions lack in a lot of cases. so he would have to downgrade windows as well if he wanted to do that. 

 

 

Hmm, maybe I missed that in the video. Makes sense then.

[TrialMacameau]

On 1/14/2022 at 6:19 AM, NemesisPrime_691 said:

So, Linus updated the BIOS and mentions that the system has been restarting since then, and he does not try to downgrade it to the previous, perfectly working version? 

 

I am all for these tech tips of diagnosing the issues and trying to fix them, but disabling C states is a sort of weird solution since you are basically disabling a feature that was working previously without any problems.

 

I would've tried to downgrade the BIOS first and if that did not work, then go through these steps. Just a quicker solution. Most of the motherboard manufacturers must also have some sort of a BIOS downgrade (flashback?) feature as well right? Why not use it then?

Only to find out that some board doesn't allow flashing a BIOS version that's lower than the one currently installed ( in the exception of MSI boards )

 

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

 

In my case after watching this video, ever since I built my PC about a year and a half ago, I was having some troubling issues with my system that caused a black screen of death with nothing on and simply hard reset itself. In the Event Viewer, Windows classified it as a WHEA-Logger type of system crash related to the CPU itself. I have an image on Twitter for it.

 

In the image this is the WHEA-Logger caused by the CPU

 

It says :

 

A fatal hardware error has occurred.

 

Reported by component: Processor Core

Error Source: Machine Check Exception

Error Type: Cache Hierarchy Error

Processor APIC ID: 1

 

By disabling C-State and putting the PSU current usage to Typical it might of fixed it but I need to test it further to make sure it stays like that. And I game pretty much all day long so I can test this with some heavy games that play and work well with this test. I'll report back after 72 hours give or take.

[PotatoParker]

i did everything in the video but its still shuts off randomly\.  but i do hear like a click sound everytime it shuts off

[noob0saur]

Recently i got my PC turning off while gaming, i checked everything from temps to power limits & even added a lot more voltage to my 5Ghz overclock, finally it wasn't any of these, guess what? it was a defective PCI-E extension cable...