Solved Event 41, Kernel-Power

[Morgud]

OK, ran several of those Tests.

I ran Kaspersky and Windows Defender Offline, Full System Scans both and both didn't show up with anything suspicious.

I ran Western Digital Data LifeGuard Diagnostics - DLGDIAG for Windows, both Quick and Extendet Tests on all Drives, didn't come up with anything suspicious again. Only in one of my HDs the SMART Status fails because its spin up time is to slow. But that's a secondary drive.

I didn't run Stress Tests, but i did run the Memtest. It ran 13:25:40 hours, 7 Pass and i got 11 Errors, plus in Pass 1 Test 7 i got a failing address. Somehow i got the feeling, that again i have either Bad Ram or the Motherboard got Bad. Both are not even 10 month in use, sad.

I will first check Ram#1 in Slot#1 using Memtest, then i will check Ram#2 in Slot#2 to see which one produces the errors.
If only one Ram Stick produces Errors, i will Test the Good Ram Stick in the slot where the Bad Ram Stick sat, to cross check the slot. Afterwards i test the Bad Ram Stick in the God Ram Stick's Slot.
Does that sound like a reasonable plan?

 

[Arc]

Yes, go ahead

 

[Morgud]

Just ran Ram#1 in Slot#1, ran 7:30 hours, Pass 7, i got 1 error(in T7 Pass6). I will try to do the second check tomorrow, Ram#2 in Slot#2.

 

[Arc]

Check RAM 1 is slot 2 now. Observe if it shows the same error there or not.

 

[Morgud]

Yes, if it doesn't then there is something wrong with the Motherboard, right? But sadly i don't have the time to let it run tonight, as we have some family guests today, which unfortunately sleep in the room where our PCs are. The noise could make them not sleep good, so i let it run tomorrow for about 7 hours, that should be enough for 7 Pass, of course if the error shows up earlier then not 7 hours. But it shown up at pass 6, so i have to let it at least run 7 pass.

 

[Arc]

If RAM 1 doesnt give any error in slot 2, and RAM 2 behaves as same as RAM1, then we may consider it as a Slot 1 related issue.

Do the rest of the tests whenever you feel like.

 

[Morgud]

I will get to it, latest starting this Monday, as then the Family moved out of the house, and i have allot time to do a through checking.
It is never good, to check stuff when half the family that life elsewhere come for Grandpa's 90th Birhtday.

 

[Morgud]

Just let Ram#1 run in Slot#2 for about 5 hours, pass4 and no error. But i want to use the PC, so i let it run overnight again til Pass 7. Last time, the error came up in Pass 6, so in about 12hours i post the result. I think the Motherboard is broke.

 

[Morgud]

So, ran Ram#1 in Slot#2 overnight, 11 hours Pass 10 no error. Ram#1 did show one error in Slot#1 Pass7. I'm checking Ram#2 in Slot#1 today's night, but looks like the Slot or something on the Motherboard broke.

So at this point, it has to be either Ram or Motherboard right? No chance the HD's can cause this? As they PASS the Quick and Extendet Test of "Western Digital Data LifeGuard Diagnostics" all, only one failed SMART cable check.

 

[Morgud]

Hi Pauly,

Now i am through with the tests, but i am a little confused with the results.
Ram#1 had 1 error in Slot#1, with a failing Address. In Slot#2 Ram#1 ran through fine for 11 hours, 10 Pass.

Yesterday i let Ram#2 run in Slot#1, and expected a Error, which came up after 10 minutes even before i gone to bed. So i ran Ram #2 in Slot#2 overnight, expecting no Error. But Ram#2 came up with 9 errors, after 11 Pass 12 hours.
The interesting part is, that Ram#2 had a Failing Address in Slot#1 and also in Slot#2, which was the same Failing Address. The only difference was the Good and the Bad address and it was both in Test 6 Pass 0.
000aa1cd004-2721.8MB Good/00000400 Bad/00010400 ErrBits/00010000
000aa1cd004-2721.8MB Good/00000100 Bad/00010100 ErrBits/00010000

So does that mean, i have a problem with the Motherboard AND one of the RAM bars? Should i check Ram#1 in Slot#2 again? Is it possible, that Ram#1 ran fine in Slot#2 for 12 hours, but a second check shows errors?

 

[Pauly]

Hmm that is unusual it is possible to have issue with both but very unlikely to be safe I would run all tests again I know it's a pain in the ass but nice to be sure
And remember just because a stick passed a test now doesn't mean it will pass next time and vice versa one of the joys of electronics I guess, also this may sound stupid but have you marked the ram sticks to ensure they don't get mixed, only other thing you could try is putting good stick in good slot and just run with it for couple days/weeks to see if that proves your system may slow your pc down a bit but may give some insight into cause of crashes

 

[Morgud]

It sure is weird, that Ram#1 has 1 error in Slot#1, and 0 in Slot#2, while Ram#2 has 3 in Slot#1 in 10 minutes while having 9 in Slot#2.
One thing i could imagine, is buying a new Ram Set(2x4gig) just like the one i have, and do all tests again, or at least use one new Ram that's hopefully error-free out of the package and test the Motherboard this way.

A side note, i do have a 850wats PSU and 3 Harddrives plus a DvDrom in mys system. The only two things that changed end of November (my birthday)was getting a new Monitor from my wife that i attached using HDMI and a new External Harddrive (a Western Digital My Book 3TB). But the external Harddrive has its own power supply. So i doubt that there was a power surge of any kind because of the external Harddrive.
One weird thing that's started beginning December(that's when i started to have that issue)was that i started the PC and the CMOS of the Motherboard was not correct, so it asked me to either use set it manually or use a saved setting.

I've already had once a Motherboard that was failing, some of it's capacitors where broke, one could see that visually (the liquid was dried out, or it was bloating)on the Motherboard. Back then i remember, i had irregular BSOD's, the error message's i got aimed toward all kind of driver's on my system. Then i ran a Memtest back then, just to find out something wrong with the Ram.

After serval tests, that where like this, leading to no 100% conclusion, i checked my Motherboard and saw visually blown and dried out capacitors. I got this idea in some forum, where someone told me that those can break if there are power fluctuations, so i checked. But that Motherboard where like 3 to 4 years old. This is not even one year old, they should be still working.

However, i did check this board, before i posted in this forum, i opened my PC up in mid December, after quite a few HardHangs, to clean for dust and check the board. Nothing.

Also the FIRST thing i did, was mark the Ram Sticks with a sticker each, which a number for 1 and 2. All tests would be useless if i don't ^^. Right now i have Ram#1 in Slot#2 ... the Ram#2 Stick lies here on my Desk. Since Ram#1 in Slot#2 didn't produce errors ... i should be the only one without bugs.

It's quite possibly, that i had a few power surge here, some lamps are acting weird if the Air Condition in this house starts up(i life here only since February last year). So, maybe something fried the board, and one ram chip along with the slot, wondrous electronics.
However the second PC and the Notebook in here seems unaffected.

So, i guess, the only true solve would be, to get another Motherboard, and another Ram set. Then i could cross check both. The problem is, i didn't have any more Hangs since i am checking the system with you ... all i could do is get New Ram and kick them into the Slots and do a 12hour Check. If that shows me no error, do another 12hours one, if that shows no error i could guess it's the Ram.
If there is a error, then i could exchange the Motherboard, kick in the old Ram and do the check, if there is then no error it was the Motherboard, if there is ... Its the Ram AND Motherboard.

Frustrating.
I will do a new test, Ram#1 in Slot#2 tonight. It's the only configuration that ran errorfree until now. I just want to Play around.
I will probably write down my CPU and Ram Specifications, and buy a new Motherboard and Ram at Fry's. That allow me to at least check NEW Ram Sticks on this Motherboard in 4 configurations. If all 4 run fine(the old Ram as you remember, 3 of the 4 settings had errors)it's the Ram. If that not work, i probably have to reinstall Windows 7, as i will use the new Motherboard, hopefully then without problems.
Need to go there, to get a new GraphicCard for my Wife's Birthday anyway ^^. Lets see where this error is.

 

[Morgud]

Yesterday, i ran Ram#1 in Slot#2, this configuration shown no error last run, i got 2 errors after 11 Pass(12:30 hours).
Originally, i ran a test 5 hours Pass 4 with Ram#1 in Slot#2 without any errors, then i interrupted the test(since i wanted to use the PC)and afterwards i let it run overnight 11 hours 10 Pass without error. So that means it ran 2 tests without errors, and the 3rd test it produced 2 errors, and one failing address in Pass 2.
Let me write down the Failing Addresses:
T7 P1 - 0018439a004 - 6211.6MB - Good ab265854 - Bad ab275854 - ErrBits 00010000 - 1st run, 2*Ram
T2 P2 - 0002a439094 - 676.2MB - Good 3a0324c0 - Bad 3a1324c0 - ErrBits 00100000 - Ram#1 Slot#2
T6 P0 - 00aa1cd004 - 2721.8MB - Good 00000400 - Bad 00010400 - ErrBits 00010000 - Ram#2 Slot#1
T6 P0 - 00aa1cd004 - 2721.8MB - Good 00000100 - Bad 00010100 - ErrBits 00010000 - Ram#2 Slot#2
T7 P2 - 0002a439094 - 676.2MB -Good bdc804c1 - Bad bdd84c1 - ErrBits 00100000 - Ram#1 Slot#2
There is no Failing Address with Ram#1 in Slot#2 since in the first run there was no error.

If you look carefully, you see that Ram#2 had in both Slots the very same Failing Address with the same MB value. So does Ram#1 when he produced its Failing Address in Slot#1, and when he produced the Failing Address in Slot#2 at the third run.
Now maybe its due to something wrong with both Ram Sticks. I mean, IF the Motherboard SLOT would have a problem, wouldn't both Ram Sticks need to produce the same Failing Address in the same Slot? Instead, the first Ram Stick produces its own Failing Address in either Slot, as does the second Ram Stick.
Weird isn't it?

If the Ram is error free, is it supposed to always pass the Memtest with zero errors? Or can even a perfectly fine Ram produce errors, if it's in a working Slot?

I'm thinking of getting me a brand new Motherboard, as this one has only one PCI slot and that makes it too cramped to add a good Sound Card ... the two slots Graphic Card would have not enough space if i add one. Plus, i want a USB 3.0, for faster transfer to my External HD.

So if perfectly working Ram in a working Motherboard should produce no errors within Memtest, then it would be easy to cross check if i get new Hardware.
I just have to plug in the new Ram into the old Motherboard, and do a check with both new Ram Sticks plugged in, if the Motherboard is fine it should not give me errors ... if it was the Ram.
Otherwise, i would need to exchange the Motherboard, and plug in the old Ram, and then there should be no error in Memtest right?
The last possibility would be, that both got damaged from whatever (power surge maybe?), and once i build in the new Motherboard and new Ram both should produce no errors in Memtest.

I would really appreciate your opinion on this Pauly, as i am unsure.

 

[Pauly]

TBH im not sure, if it were my machine i would be tempted to swap the RAM first, easy for me to say as i have other machines i could swap with but gotta be cheaper than new board if you are gonna go for new hardware, it would seem there is a problem with ram#2 because of the fault on identical position in either slot but ram#1 doesnt produce conclusive results, my only other thought could be memory controller
let me see if i can round up the troops and get some others opinions

 

[Morgud]

That would be nice if you could get other opinions.

The hardware is not even one full year old, as i bought the whole system last year in mid to late February.
So what i am going to do for sure would be, buy a new Motherboard and Ram, then first check the new Ram on this board using Memtest.
Now if you kindly ensure me that new Ram in top condition has to NOT produce ANY error then i will see quickly if it's the Board. If the new Ram indeed produces errors, as the old does then i consider it that some part of the old Board broke. Slot, Memory Controller whatever ... there are allot components on Motherboards that can break or glitch.

The next step would be to exchange the Motherboard, then i could test first the old Ram for errors in Memtest, and then the new Ram for errors.
It could still be possible, even thought it has little chance, that something happened that fried my Board AND Ram. Would make sense if a power surge occured, since the Ram sat on the Board the whole time.

After i figure out, what component is the problem, i could either return the unneeded New Ram/Board, or send in the old faulty Ram/Board.
Either way, i am not going to wait for 2-3 weeks, for exchange Ram/Board to be sent to me, as i use the PC allot.

Oh, btw, i forgot. While running the Memtest, i unplugged all my Harddrives and the DvDrom, as it didn't seem to make any sense to let them run unnecessary 12 hours without need to. I hope that didn't influence the Memtest.

 

[Dave76]

If you get one error during memtest86+ it means there is a RAM issue, it might be corrected by adjusting the settings.
If that same RAM card passes memtest86+ the next time, it doesn't mean the RAM is ok now, it means it has an intermittent fault, which is the worse kind to track down.
These errors can be caused by bad RAM, bad RAM settings, bad motherboard, BIOS issue, or even a problematic PSU.

Let's take a look at the RAM and settings, you might be able to adjust the settings to get the RAM stable.

Can you download CPUZ and post a snip of the CPU, Mainboard, Memory, and SPD tabs.
For posting the information , enter 'Snipping tool' in the Start button > Search box, hit enter. Select the area you want to post and save to a convenient place.
How to Post a Screenshot in Seven Forums
Go to your BIOS and tell us what the CPU, RAM and the CPU integrated memory controller voltages are.

Use Core Temp , to check the CPU temps when at idle, under load and/or testing.
Let us know the minimum and maximum temps you get.

 

[Morgud]

Hi Dave76,
Core Temp runs since a few days already, it tells me min 3C and max 29C.

I did update a long time ago my BIOS, because i got that Bulldozer issue, where the PC just freeze whenever you try to run certain games.

My PSU is a Thermaltake TR2 RX 850w one, it should be still be fine ... bought it back then with the whole System.

Here the Snips.

BTW, so normally a good Board and Ram should not create ANY Memtest86+ errors? Or can even good hardware get some errors? I just want to be sure, that the freezes are indeed either Ram or Board, and not another part in the PC here, since my HD's passed the WD test, and otherwise i not have anything in here other then the Graphic Card and CPU, the latter should work i hope.
Somehow i feel like back then, when i had trouble, lost. Back then i got out of that by buying a new Mainboard.

 

[Dave76]

Those are strange temps, 3°C is just above freezing.
Try another program:
Link >Speccy Free System information Utility

Power spikes can cause damage that is very hard to locate.
Do you have a voltage regulator or UPS?

Hardware issues are also hard to diagnose, sometimes a bad keyboard can take weeks to find and cause many different kinds of problems. You need to find the most likely and work from there.

Need these voltage settings, they are the ones you need to change.

Go to your BIOS and tell us what the CPU, RAM and the CPU integrated memory controller voltages are.

CPU integrated memory controller voltage is CPU-NB voltage in your BIOS.

Download your motherboard manual if you don't have it.
Link >MSI 760GM-P23 (FX) manual

 

[Morgud]

I just played a little FarCry3, when i went out of it, CoreTemp showed me 32C. 3C was idle on Desktop. Right now it shows me 48C max.

For the Power Spikes, my wife told me earlier, that why it only affected my PC and not hers. I just took a look, and she use her old Power Surge Protected Plug. What is a Voltage Regulator and UPS?

I suspect, that there where some Power Surge Incident starting December. I recall seeing fluctuations in the light in the House here. You know, i lived in Europe before, and there our electricity don't fluctuate as that lights would become some darker for moments.

So, in case there indeed was a Power Surge, affected could be everything in my PC, right?

CPU-NB = CPU Northbridge? I go and take a look.


Erm ... where in the Bios do i see that? I've looked around and could not see a menu where i see all Voltages listed. I do have some greyed out options, where i am supposed to be able to setup the Voltages of said elements.

 

[Dave76]

A CPU cooler can't lower the CPU temp below room temp, unless your room temp is 37°F it's not reading right.
Check it with Speccy, it's in the link above.

A power surge protection plug will work just fine.
A voltage regulator keeps a steady 110v and has surge protection, a UPS (Uninterrupted Power Supply) has surge protection and a battery backup if the power goes out, maybe 15 minutes depending on how big it is.
I'm afraid a power surge is possible to damage one or all of your PC components.

Yes, CPU-NB is CPU Northbridge, actually the memory controller was previously in the NB chip but now is part of the CPU, IMC (Integrated Memory Controller).

Download Your motherboard manual I linked above, it's on page 24.
It's under 'Cell Menu'.