Kernel Power 41 with new mobo and CPU

ambrosekeen · Feb 13, 2018

First of all system specs

- I7 8700k CPU
- Aorus z370 gaming 7
- Aorus Extreme 1080ti GPU
- Corsair - Dominator Platinum 32gb (4x8gb) DDR4
- Thermaltake - Toughpower Grand RGB 750W 80+ Gold
- Windows 10 Pro
- 2x Samsung 850 Evo SSD
- Full custom water cooling so temps are not an issue

So the motherboard and CPU are brand new (upgraded from i7 6700k and gigabyte z170 gaming 5 mobo). And the night of my new upgrade I start to get random PC resets. Complete resets with no BSOD and straight reboot. Looked into it and I'm getting kernel Power 41 errors. First occurred 4 times in a single game of League of Legends. Then not again for the rest of the night while normal browsing/trouble shooting. Next day once playing PUBG, more troubleshooting. Then it survives all night while mining BTC (thought I fixed it with driver reinstall and re installing windows etc) to happen randomly while browsing the web, so not fixed. I'm thinking it's PSU but it's only 3 months old, Ive done ram tests and came back clear. Now running on 1 stick of 8bg until a reset to see if any stick doesn't cause it (very hard to replicate the resets as it's random, shortest is 5 minutes a part and longest has been a day and a half)

2nd issue is I just finished building my BTC mining rig. Specs as followed

- Gigabyte z170x gaming 5 (old mobo from pc previously no issues)
- Intel Celeron G3900 2.8ghz
- Corsair HX 1000W 80+ Platinum
- 3x EVGA 1070 SC black edition
- 1x EVGA 1070 FTW
- 1x Corsair 8gb dominator platinum
- standard ssd for OS and windows 10

Now the issue with my mining rig funny enough is the same. After multiple tweeks to get all GPU's working it's running into the same kernel error 41's which yells random power outage. Stuffed if I know what it is or is my luck that bad that I got a dodgy PSU which is brand new and very high quality??

Any help is appreciated or recommended tests to run. I'm running all tests i can think of when I have the time but to have these two issues at the same time tells me bad luck or I've done something wrong.

First system was a simple change of motherboard and CPU with a fresh windows install. All components are compatible with each other unless I have that mistaken?

And 2nd system is all brand new GPU's and brand new PSU with 2nd hand mobo (previously no issues). This rig has ran all night and day with no reset to randomly reset during the next night twice. On some occasions it fails to transmit a display to my monitor which causes me to reset because no matter what card I plug monitor into with hdmi or displaycable it won't recognise. Hard reset fixes this so it's screaming psu or faulty plugs somewhere.

Hope someone can help. Sorry for lengthy post

Darkbreeze · Feb 13, 2018

I see no mention of the single, maybe only, important consideration in all of this, which are the PSU model numbers?

How old it is is not what matters.

How many "watts" it "says" it can handle is not what matters.

What kind of 80plus rating it has, you guessed it, is not what matters.

What platform it is build with and who it was built for, what the SPECIFIC model number as listed on the PSU label, that is the only thing that matters and the only thing that can help us determine any kind of preliminary indication of quality or lack thereof.

ambrosekeen · Feb 13, 2018

Darkbreeze :

Thanks for the reply. The following are what I could find for model numbers:

Main PC
TPG-0750F-R

Mining Rig
Model RPS0076
Part number CP-9020139 / 75-002706

Darkbreeze · Feb 13, 2018

Are both these rigs in the same room, having similar problems?

Are they sharing a power strip or outlet?

That PSU in the main PC is a Thermaltake Tough power Grand, and is a fairly decent unit, for a Thermaltake. I'm not a fan of Thermaltake, especially their power supplies, as most of them are better used as paper weights. That model is ok though. Notice I said "model", not THAT unit. So, even with a great power supply, there is of course always a chance something is wrong with it. How old is that unit?

The other unit is a Corsair HX series 1000w based on that info. Also a VERY good unit for the most part. I doubt the PSUs are the problem unless they are fairly old and have been run hard for a while, which is possible.

Since both units are having issues, I'm betting there is something common to both systems, like a power strip or faulty power outlet, that is causing the issue. Are you using an extension cord or anything else that might be relevant to both units?

Darkbreeze · Feb 13, 2018

Are both these rigs in the same room, having similar problems?

Are they sharing a power strip or outlet?

That PSU in the main PC is a Thermaltake Tough power Grand, and is a fairly decent unit, for a Thermaltake. I'm not a fan of Thermaltake, especially their power supplies, as most of them are better used as paper weights. That model is ok though. Notice I said "model", not THAT unit. So, even with a great power supply, there is of course always a chance something is wrong with it. How old is that unit?

The other unit is a Corsair HX series 1000w based on that info. Also a VERY good unit for the most part. I doubt the PSUs are the problem unless they are fairly old and have been run hard for a while, which is possible.

Since both units are having issues, I'm betting there is something common to both systems, like a power strip or faulty power outlet, that is causing the issue. Are you using an extension cord or anything else that might be relevant to both units?

Are these units plugged directly into a wall socket? If not, I would try that, even try a DIFFERENT wall socket on another wall or room if possible, to see if the problem follows it or goes away.

ambrosekeen · Feb 13, 2018

Darkbreeze :

Both systems are at different ends of the house. My main PC with the TT PSU I only had the issue once I upgraded my mobo and CPU with a fresh install and the PSU is only about 6 months old.

The Corsair PSU is brand new out of box. I haven't tried swapping them over yet to see if that makes a difference as there is nothing imparticular that is causing the resets. ALTHOUGH today I noticed the Corsair system which is setup for mining bitcoin I noticed isn't actually having the kernel error, but instead I'm losing connection to the monitor fully besides power as it's still on but not displaying anything. BUT something is causing the mining to stop and I can't see what it is because the monitor doesn't display so I have to hard reset and it works again. Monitor is a 1 month old 2k Viewsonic only using it during the setup stage then returning to my main pc as 2ns monitor. I've got some time off as of Sunday so I plan on doing full tests and hardware swaps but I am also unsure what's tests to run if anyone can help with that would be great

Darkbreeze · Feb 13, 2018

Are you doing ANYTHING on the mining rig, aside from mining? No 3D apps or gaming on there, right?

Even so, you are likely a bit light on the PSU capacity because although we know that the demands of mining aren't as high as for gaming, still, you're talking about four 8 pin connectors that can draw upwards of 150w each, plus 300w in slot power, so that's 900w without even factoring in the demands of the CPU, memory, drives, fans or motherboard.

I don't have mining experience per se, so I can't speak to what the actual demands of the system are on that configuration and I've not seen any legitimate per card model mining power calculators that seemed trustworthy, but regardless it seems like a close thing just off card spec.

I've seen a lot of indications that bad riser cables are a problem too. Might just double check all those connections fully and maybe even swap a few things around just to see if it goes away. I've seen crazier stuff fix worse problems.

Are you, or can you even, use one of the cards that are mining for display also, or are you using the integrated graphics for that? Honestly, I'm not 100% sure how that works on mining rigs. Haven't tried it myself yet. I'm too busy trying to find a damn card but all the damn miners keep buying them up. (Looking at you here. LOL. JK)

ambrosekeen · Feb 14, 2018

Darkbreeze :

Mining rig is just mining. Only thing installed is drivers, windows and the software I mine with. Although that could be other issues like riser cables etc but most people use a 1000w for 6gpu's and I'm only using 4. I might try get some hardware to test how much it's all using. I'm using a card for display which is in the main PCIE 1x16 slot not using a riser so it's directly on the mobo.

I might try remove some GPU's and see if it's stable with just the one then progressively add more until it crashes and see what results I get. Because at this stage I'm lost for possible solutions without testing every single hardware I have. Is there a way of testing mobos without swapping them out? Because my new CPU is only compatible with the new mobo and code versa so if either of them are dodgy I can't find out which one it is

Darkbreeze · Feb 14, 2018

That's usually the dilemma with motherboard issues. Generally, without equipment that only high end operations have, it is almost impossible to actually test a motherboard (Or an electrical engineering education and a board specific wiring diagram to go off of) so you are left with pretty much eliminating everything else and convicting the board only by process of elimination unless in some rare cases you can actually see something physically wrong on the board like bulging or leaking, or burned caps or traces.

On the other system, even though it's a relatively decent PSU, it IS a Thermaltake, so I always have suspicions. Might really want to try a different PSU in there. Also, what memory tests did you run on that?

Loss of power resulting in instant off are almost, but not always because other things can be shorted and cause the issue or the power loss could be outside of the system, always due to a PSU or motherboard issue.

ambrosekeen · Feb 14, 2018

Darkbreeze :

That's usually the dilemma with motherboard issues. Generally, without equipment that only high end operations have, it is almost impossible to actually test a motherboard (Or an electrical engineering education and a board specific wiring diagram to go off of) so you are left with pretty much eliminating everything else and convicting the board only by process of elimination unless in some rare cases you can actually see something physically wrong on the board like bulging or leaking, or burned caps or traces.

On the other system, even though it's a relatively decent PSU, it IS a Thermaltake, so I always have suspicions. Might really want to try a different PSU in there. Also, what memory tests did you run on that?

Loss of power resulting in instant off are almost, but not always because other things can be shorted and cause the issue or the power loss could be outside of the system, always due to a PSU or motherboard issue.

I have only done the windows defrag memory test so far. Can you recommend others to do for me?

ambrosekeen · Feb 14, 2018

Darkbreeze :

Just got home from work and when I checked the logs my main gaming pc with the ThermalTake PCU had set three times with the kernel 41 error while I was at work. All these times it was at idle. When I get some time I'll try connect the Corsair PSU to this rig and see if that makes any difference

Darkbreeze · Feb 14, 2018

Make sure you have bios version F5h for your Z370 board installed. If you have the January version installed, I would look into whether or not that is able to be rolled back. Restarting and other issues are a common problem for Intel bios versions that were released in January as a result of the Spectre and Meltdown vulnerabilities. I suspect that if you have a newer version than the December release, this could well be the problem.

And if that is not the case, then update to the version that I listed. Should be done anyhow, even if you were having no issues.

You will have to reconfigure any custom bios settings after updating including any overclocks or memory settings.

If this doesn't help, then I'd try a different PSU. If that doesn't help, I'd try removing the graphics card, using the integrated graphics and see if the problem remains.

Search

Kernel Power 41 with new mobo and CPU

ambrosekeen

Prominent

Darkbreeze

Darkbreeze

Retired Mod

ambrosekeen

Prominent

Darkbreeze

Retired Mod

Darkbreeze

Retired Mod

ambrosekeen

Prominent

Darkbreeze

Retired Mod

ambrosekeen

Prominent

Darkbreeze

Retired Mod

ambrosekeen

Prominent

ambrosekeen

Prominent

Darkbreeze

Retired Mod

TRENDING THREADS

Latest posts

Moderators online

Share this page