Troubleshooting PC issues, reviewing memtest86 outcomes, looking for guidance.
Troubleshooting PC issues, reviewing memtest86 outcomes, looking for guidance.
Hello! I'm trying to fix my PC and figure out which part might be faulty. I'm stuck and unsure what conclusion to draw or what further tests I should perform. Any guidance or advice would be greatly appreciated.
What is the BIOS version on your current motherboard? There are multiple updates available that improve memory and USB support for your system. Additionally, please specify the exact DIMM slots you plan to install in—beginning with the one nearest to the CPU, numbered 1 through 4, with slot 4 being the farthest from the board's edge.
When sharing a troubleshooting thread, it's standard to provide complete system details. List the specifications as follows:
CPU:
CPU cooler:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS:
Monitor:
Include the PSU age along with its make and model. BIOS version for your motherboard at this time.
Are the RAM slots A2 and B2 filled? Where did you obtain the installer for your OS? Regarding the OS, have you installed it in offline mode on a Windows platform?
What is the BIOS version on your current motherboard? There are multiple updates available that improve memory and USB support for your system. Additionally, please specify the exact DIMM slots you plan to install in—beginning with the one nearest to the CPU, numbered 1 through 4, with slot 4 being the farthest from the board's edge.
Thanks for your responses!
Regarding the missing details, it seems intentional. I tend to explain too much, add unnecessary specifics, and sometimes overdescribe when I should have kept it brief (only 5 details were needed). I focused on what’s important from my perspective, adding more later if needed.
I don’t think the PSU information was necessary, so I omitted everything about it until you asked.
Then, here’s a summary: the last time I updated the BIOS coincided with a motherboard purchase, where I experienced timing issues and voltage changes. I was concerned about BIOS updates affecting performance.
Building my PCs made me realize the risks of flashing BIOS without proper testing. After troubleshooting for five days, I finally decided to update.
Now I’m ready to adjust RAM settings and run more tests.
So have you upgraded the BIOS to the most recent stable release, or not? What version is currently installed?
If you've updated the BIOS, it might be wise to perform a hard reset and restart from the beginning, as certain configurations can become "stuck" and fail to update even after an update.
BIOS Hard Reset steps
Turn off the device, disconnect the power supply, and detach the PSU cable from either the wall outlet or the power source.
Take out the motherboard's CMOS battery for roughly three to five minutes. In some instances, removing the graphics card might be needed to access the CMOS battery.
While the battery is removed, keep pressing the power button on the case for 15-30 seconds to drain any remaining charge in the CMOS circuit. Once the time is up, reinsert the battery, making sure it's oriented correctly as it was removed.
If you had to remove the graphics card, you can reinstall it now, but be sure to reconnect any power cables and your display cable.
Reconnect the PSU cable, power on the unit, and wait for the POST screen to appear with options to set up CMOS/BIOS. Enter the BIOS setup program and adjust boot settings for Windows or legacy systems, depending on where your operating system resides.
Save your changes and exit. If the system successfully boots, you can proceed further—such as restoring memory profiles, fan settings, or other custom configurations that may have been lost during the reset.
In certain situations, after resetting the BIOS, it might be necessary to load default or optimal values in the BIOS to ensure hardware tables are properly reset within the boot manager.
It's also important to note that if the issue involves a video signal problem, consider trying a different display type, such as HDMI, especially if you're experiencing no visual output or signal errors.
If a display is unavailable, switching to another monitor can be beneficial, as some systems may not function correctly with displayport setups.
Trying an alternative monitor is advisable when a hard reset is required, particularly if the problem stems from a video signal deficiency.
Good news, looks like problem is solved at last ! And yes, everything was related to BIOS. Here's a ton of extra informations about my whole PC problem, because oh my god i know so well how it can be frustrating to search online, find an exact description of your problem, to end up with a "Every thing is fixed, thank you !" with no explainations. So here's probably too many details and informations... most likely ending with small joke attempt getting completely out of hand slowly shifting to pure madness.
About problem description: It's crystal clear now. That ugly description "Many different very random glitches/crashes everywhere" gets an official description "System instability".
Informations about the BIOS: Before updating yesterday, RAM was set to 6400mhz. The last time i set it was 6 months ago, at 6000mhz (ram sticks specs). However, i have no idea "when" or "how" it ended up at 6400mhz. Did it jump to 6400mhz by itself exactly when the problem started ? Or after all my memtesting, when i put back the sticks in dual channel ? No idea. All i know is: It was set to 6000mhz 6 months ago, and last night, it was 6400mhz.
Now that i know the problem was linked to the BIOS, i remembered one "event" that happened approximately at the time the problem started. MSI Center asked for an auto-update, i said yes, update either simply failed, app crashed, system restart, or BSOD, can't remember exactly, end of story, forgot MSI Center over existed until today. I have no idea if it means anything though. (Only reason MSI Center is installed is to turn off ram/gpu LEDs. So i didn't care about that failed update.. you horrible app, you had your chance to update, you failed, too bad, you're on your own now.).
(MSI PRO B650-P WIFI , so you don't have to scroll up)
Bios version: It was 7D78v16. Updated to 7D78v19 last night.
(Kingston Fury Beast 2x16GB KF560C36BBEAK2-32)
BIOS version / RAM behavior: With the previous BIOS version (7D78v16), any EXPO attempts lead to either "frozen" before POST, endless restart before POST, restart between post and boot. Same behavior when setting it to specs manually (6000mhz 1.35V 36-38-38). I ended up setting it to 6000mhz, auto voltage, default timings, everything was very stable until 2 month ago when everything started going wonky. (around that MSI Center failed-update event). Today, BIOS version (7D78v19), Enabled EXPO (6000mhz 1.35V 36-38-38), Memtest86 2-pass without error, system perfectly stable all day. Oh, and just to be perfectly clear, all those freeze/restart around POST were NOT related to the normal "Motherboard and RAM learning about each other to be best-friend-forever".
Extra informations that may or may not be important at all:
- From the first time i installed that motherboard 6 months ago to this day, the only stuff i touched in the BIOS was: EXPO enable/disable, RAM frequency, RAM voltages, RAM timings (the first 3 timings like 36-38-38) and BOOT order. That's all. (Okay, human brain can have unstability issue as well, so it could still be possible that i touched something without remembing it. But the only reason would be to activate/desactivate "something i don't understand" that was required for one specific app or a Windows 10 feature to work properly. One good exemple of that would be enabling virtualisation for bluestacks. But i would remember that one, obviously.
- After installing the mobo-cpu-ram, i installed Ryzen Master to make sure temperatures matched those from MSI Afterburner. It did, never opened the app again, it's not auto-opening with windows, but i never uninstall it. I just tried to open it to check the exact version number i had, but i get a "ryzen master does not support current processor" error. Installed the app around july 10th 2023. I'll not investigate that further. No idea if it's meaningful. (CPU: 7700x)
- The only "burn test" i ever did was a few minutes of 7-zip performance test, which is very mild. So i have no idea what kind of stability i would get with a proper burn test, and i'm not planning to run one unless it's necessary. I like my parts to be handled gently and i do the same with my PCs. (Reason for 7-zip performance test: Personal curiosity. Monitored temperature and CPU clock speed. I wanted to observe the behavior of a CPU designed to give "best possible clock speed at 95'C".
I almost never visit this forum unless i need help of advices (it's so selfish, i'm ashamed), but i remember any important detail or information that i forgot to mention, i'll come back here to edit my post. Also, if the problem come back, or anything particular happens that i have reason to believe could be linked to that problem, even 3 years from now, i'll come back here and edit this post. (That's a promise to you, random future troubleshooter.)
So if you don't see any clear edits, it means i ran my PC same as usual for a whole week without any suspicious unstability, and then it passed a 10-Pass memtest86 with zéro error. (That statement won't be valid for the next 7 days and i don't like that. Also, you don't feel like scrolling up to check post date. Today's Dec 8th 2023, the 10-pass memtest is planned for Dec 15th 2023. That should do it.)
Future desperate random troubleshooter stumbling randomly here, i hope this post helped you, and if time-travel is invented... (Damn, i imagined an awesome joke earlier but forgot the ending). <----- Yes, this is the actual joke. You should be rolling on the floor in pure hysteria right now, laughing so hard that the idea of passing out from asphyxia crossed your mind at least a couple time. If you're not, there's probably a sense of humour compatibily patch somewhere.
Can be skipped, it's just how my post started:
Oh and by the way, it could be confusing to everyone (including future troubleshooters.) as to "why" i posted as many informations and details, which burried the main reason that made me completely stucked after days of troubleshooting and "could possibly" also happen to someone else... not to mention the extra unnecessary insanity. That's just me. One small acceptable joke attempt, poof, 10 pages of non-sence that i don't even delete before posting. It happens, i can't help it. (Close to being pathological i said. This one is not a joke.)
TLDR (short version): Wanted help to understand why "updating the bios" solution worked, but it seems like i managed to make sens of it unexpectedly by typing. (Normal people do that just by just thinking, looks like i required a keyboard to do it). If you could proof-check both my hypothesis at the bottom of the post, it would be much appreciated ! (Proof-Checking everything would be ideal, but only if you have much spare time.)
TLDR and a last comment after finishing the post: Started as an explaination about why i posted the previous post, then asking for personnal knowledge proof-checking because i don't like the fact that i don't understand both the problem as a whole and why the solution worked (Nothing made sense to me when i started the post), to unexpectedly find out (while typing) the tiny bit of "knowledge" i was missing to make everything make sense, then a few new assumptions that i'd like to be proof-checked, to end with 2 different hypothesis about the cause of my system instability. Both being solved by the exact solution mentioned at the beginning by Darkbreeze, which is updating the BIOS. (Thanks !).
If you have enough free time and would like to help me by proof-checking all of my assumptions, i'd really like it, but since i ended up with two hypothesis that finally make sense (to me), only proof-checking both hypothesis at the very bottom of the post would be good enough since they are compatible with all the assumptions i had. (I was SURE i got something wrong, but not so sure anymore).
Original post starting here.
Final note before posting: I won't edit everything, but if something seems wrong because i mentionned "Before-BOOT" , replace it with "Before Windows 10 started loading" and re-evaluate. Also, i'm still a bit confused about "when" exactly the "BOOT" event happens. Statement such as "PC is not Booting", does it mean Windows was trying to load but it failed/crashed/restarted while it was loading ? Or it means the "Windows loading" process can't start ?
One thing i'm sure of, my knowledge is either completely outdated or simply not good enough, which i believe could also be the same for someone else.
If anyone could take a few minutes to explain to me which concept i have completely wrong (based on my hypothesis bellow and the assumptions i'm making), it would be very appreciated.
First, the reason for my previous post is all about the 6000mhz ----> 6400mhz RAM in BIOS. If it was just a matter of "oh no, i didn't update bios, that explains everything and i
understand why.
", the post would be "Thanks a lot for reminding me about that, you just made my day".
Here's my hypothesis and my assumptions, please tell me where i'm wrong.
My personnal "After-Problem-Fixed" hypothesis:
The value changed by itself 2 months ago exactly when the "system instability" started. 400mhz over RAM specs, very justified system instability. Simply setting it back to 6000mhz would have fixed the problem. Updating BIOS fixed even more problem. (It would explain everything, that's the only thing i'm sure of. But to me it would be "impossible" to have happened)
Another hypothesis (not mine):
Updating the BIOS fixed everything. (It would also explain everything, but same here, from my knowledge it would be "impossible".
Probably wrong assumptions, plus questions:
Assumption #1:
Turning on PC in chronological order: Human press power button, PSU activates, Motherboards checks if everything is okay, shakes hand with RAM (If not, multiple reboot/freeze until they are best-friend), "POST", BIOS is loading, BIOS applies it's value to every components, inspection of devices one by one according to boot-order until it finds one with bootable data, "BOOT-ATTEMPT" , process the data succesfully, "BOOT", from that point you're in software (Any OS, Windows, Linux..).
Assumption #2:
Anything changing "on thier own" inside the BIOS without myself entering in the BIOS and without physical interraction with components: Values reverting to "previous state" or "default", Yes. EXPO turning on/off: A far fetched "maybe". Anything else, Impossible.
Assumption #3:
Values changing inside the BIOS itself after removing/replacing a part, such as playing with RAM Sticks for memtesting then putting them back the way they are designed to work: Same as last assumption. "previous state", "default", Yes. EXPO on/off: A far fetched "maybe". Anything else, Impossible.
Assumption #4:
Values inside BIOS cannot be modified by After-Boot software, but can be overridden by After-Boot software: Yes, but not without my input. And all those overrides are effective only in After-Boot.
Assumption #5:
Four reasons for BIOS update: Improve compatibility for components that were already existing when the first BIOS version was made. Adjustments to make future components compatible. Adding a new feature. Correcting human-made mistake made when designing previous BIOS versions.
Assumption #6:
If you stay at one specific BIOS version, and never touch any components, the "before-BOOT" system behavior will stay exactly the same, always.
Assumption #7:
In this whole post, all my "impossible" become "possible" and every assumptions becomes invalid only those situations: Defective components.
Hypothesis i discarded:
One software update (let's say windows update) making the system instable with the already-installed BIOS version, making the "BIOS update fixed the problem" without involving the 6000mhz/6400mhz undefined event. It would make sense. However, it would break my
Assumption #5
since Memtest86 test are run "before-BOOT". I ran a 2-pass Memtest86 6-months ago with no error. This week, Memtest86 same errors every passes, always.
Comments and Questions about that previous discarded hypothesis:
Just realised the "statement" i just made "Memtest86 tests are run before-boot" contradict something in my
Assumption #1.
I'm a bit confused now, is it just me using the wrong terms with no impact on my logic, or did i just find what's wrong with my knowledge. Okay, so Memtest86 is on a storage device (USB stick). It contains bootable data. So basically, a boot-attempt is made, then "boot", then Memtest86 program starts running, which, from my own assumptions and personnal definition of "software", enters the same category of any OS such as Windows. I may have something here.
Assumptions about Memtest86:
I always assumed Memtest86 had to be run "before-boot" to stay at the hardware level. With the "just-added-to-personnal-knowledge" concept of "Booting into Memtest86"
,
which would define Memtest86 as "software", the only logical reason to run Memtest86 from a USB stick would be to prevent any disturbance from the operating system. Running any kind of Memtest86 inside windows would indeed be either unreliable or very limited in it's testing capabilities. Oh my god, if i'm not completely wrong about everything, my personnal knowledge is updated from "hopefully-right new assumtions".
Updated Assumptions about Memtest86:
Memtest86 is a "software". It "could" override BIOS value while running, but not change them in the BIOS Itself. All my "Before-Boot" assumptions about Memtest86 are burned to ashes. Memtest86 is run from a bootable device, not only to run "Outside-OS", but also to be the only "Booted-Software" running on the system. (Don't even know if it's possible to run multiple independant "Booted-Software" at the same time, but it would seem logical that it would interfere with Memtest86. It doesn't matter anyway, different topic completely.)
Extra Assumptions about Memtest86:
Different version of Memtest86 could give different results. That would explain why i had memtest86 errors this week and not 6-months ago. Different Memtest86 version giving different results. Never bothered checking version number, just downloaded the lastest one.
Two description of my instability problem and it's source. (Still at the "hypothesis" level)
- System instability (After booting into Windows 10). Most likely caused by an "undefined" update that happened 2 months ago. (Windows update, or other app update running inside windows). Compatibility issues between that "undefined" update and the installed BIOS version (7D78v16), which caused system instability inside windows 10. Possible solution: Install another BIOS version (Older or Newer). Solution chosen that solved the issue: BIOS updated to lastest stable version. (7D78v19).
Solution is exactly the one adviced by Darkbreeze (Thanks !). Doesn't explain that RAM frequency at 6400mhz though, but not worth investigating at this point.
-
Human-made error in the installed BIOS version (7D78v16) that could cause RAM frequency value inside BIOS to change unexpectedly to an "undefined" value that is not "previous" and not "default". Complete system instability caused by RAM frequency being set to a value higher than RAM Sticks Specification. Possible Solutions: 1- Manually changing RAM frequency value to specification inside the BIOS. 2- Install another BIOS version (Older or Newer). Solution chosen that solved the issue: BIOS updated to lastest stable version. (7D78v19).
Same solution, updating the BIOS. I could test that hypothesis by reverting to previous BIOS version, set RAM to 6400mhz manually, and observe if the system behavior inside Windows is exactly the same. But at this point it would be for personnal curiosity only, so not worth testing. Also, i may be wrong about that since i didn't attempt any overclocking for a very long time (Aside from setting RAM to specification), but 6000 ---> 6400 seems a bit too much out of specs for my instability issue that could "randomly" let the system stable for 24 hours.
One final question: If the system instability is strictly related to compatibility issues between the OS and BIOS version, is it possible for the instability to show up right after the operating system "BEGINS" loading ? Or it will show up only "AFTER" it is loaded ?
Memory compatibility issues and performance concerns usually aren't heavily influenced by the operating system. Generally, these problems appear to stem from how memory is set up or matched, and they persist whether you're using Windows or another platform. Clearly, applications that demand a lot of memory are more likely to encounter such difficulties, but since all systems rely on memory for operation, even basic tasks like booting might be affected. In severe cases, the system may fail to start at all or not load any files.