Question Multiple BSODS after clean install and different parts.

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Jun 22, 2020
27
0
30
Hello,

I've had different BSOD's for the past year. Tried everything from changing parts to formatting PCs and installing only a few drivers or all of them.

The thing that helped the most was windows updates, sometimes I'd go without a BSOD for 1/4 weeks but then I updated it and they started again.

Last week I started getting a "USB overcurrent detected, system will shut down in 15 seconds" though, so I figured it was time to change the motherboard and so I did.

First system:

RYZEN 3600
Asus TUF B450M
GSKILL 2x8GB 3200MHz
SSD CORSAIR MP510 240GB
HD TOSHIBA HDWD120 2 TB
SAPPHIRE RX 580 4GB
CORSAIR CX550W
LOGITECH G603 MOUSE, G512 KEYBOARD, G533 HEADSET

Tried a lot of things like I've stated before, ran memtests, tried no XMP, changed ram to HyperX Predator, nothing helped.

Decided to get a new MB and cleaned every single part of the PC.

Tried a clean install and as soon as windows started I got the first BSOD. Installed chipset, video, audio and lan drivers, updated windows and kept getting 1/2 BSODs a day. The funny thing is they never happen while I'm gaming or running any intensive task. Only when browsing the web, listening to music or working on excel.

Second system:

RYZEN 3600
GYGABYTE AORUS B550 ELITE
HyperX Predator 16GB (2x8GB) 3200MHz
SSD CORSAIR MP510 240GB
HD TOSHIBA HDWD120 2 TB
SAPPHIRE RX 580 4GB
CORSAIR CX550W
LOGITECH G603 MOUSE, G512 KEYBOARD, G533 HEADSET

Any advices? I'll post the latest dumps:

4 DUMPS

https://1drv.ms/u/s!AlPs5aawCA1aiLRnu_TJDHVL9qaLYA?e=2lI9d4








 
You are on BIOS version F11d, or no?

You have checked for updated drivers for your keyboard, mouse and headset, recently?

If it's not a BIOS issue, and it's not a driver issue and none of the dumps seem to be the same thing each time, then I'd say it's either the power supply or the motherboard, but I'd certainly double check all wiring and cable connections, try different power and data cables where possible and also make sure you are plugged directly into the wall socket. No power strip for the PSU.
 
A lot of years ago I had a hard time trying to figure out what was wrong with a PC. The computer in question may work well, or it would BSOD, freeze, etc., it was very strange. To my surprise everything was fixed once I changed the keyboard (using the same USB port on the motherboard).

My advice may sound stupid considering the amazing feedback you got from the mods/ex-mods, but since you tried a lot of things already, uninstall the logitech software (and then if you know how, clean the system using CCleaner portable https://www.ccleaner.com/es-es/ccleaner/download/portable), shutdown the PC, unplug all logitech gear and get a standard usb mouse and keyboard with cable if posible.

If it works without issues then either 1 of those 3 items, or the software is broken.

If it does not work then it may be a bad PSU, a wiring issue, or a power strip, as Darkbreeze pointed out in the preceding post.
 

gardenman

Splendid
Moderator
I ran the dump file through the debugger and got the following information: https://jsfiddle.net/9ubvngpf/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

File information:111320-7015-01.dmp (Nov 13 2020 - 14:09:59)
Bugcheck:BAD_OBJECT_HEADER (189)
Probably caused by:memory_corruption (Process: NGUIdle.exe)
Uptime:0 Day(s), 3 Hour(s), 23 Min(s), and 48 Sec(s)

Comment: When this crash happened, BIOS version was F10.

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.
 
Don't use automatic utilities for BIOS or driver updating...................ever.

These are things you SHOULD absolutely do manually. You find and update BIOS updates manually, and you do the same thing for any motherboard or peripheral related drivers. Using any kind of automatic utility to do these things usually results in somebody having to come HERE for help, at which time we tell them DON'T DO THAT. LOL.
 
Jun 22, 2020
27
0
30
Don't use automatic utilities for BIOS or driver updating...................ever.

These are things you SHOULD absolutely do manually. You find and update BIOS updates manually, and you do the same thing for any motherboard or peripheral related drivers. Using any kind of automatic utility to do these things usually results in somebody having to come HERE for help, at which time we tell them DON'T DO THAT. LOL.

Updated manually to F11d.

One thing that should be noted is that the BSOD's only happen when idle or low usage.
 

Colif

Win 11 Master
Moderator
Not going to argue with DB but the updaters made by motherboard makers at least stick within the lines and only update drivers for their devices, unlike Driver Booster that lies about how many outdated drivers you have and then dumps the wrong drivers on boards.

I wouldn't use any tool except those built into bios to update them. Some boards have internet inside bios now so you can get it to look for an update for itself.

technically, just cause its got idle in its name, its not really idle... - https://store.steampowered.com/app/1147690/NGU_IDLE/

as for error... um. stop showing me errors I haven't seen. We checked ram already right?
suggested fixes - update win 10, you on 20H1, I don't know if 20H2 would help
boot into safe mode, if it doesn't happen it could be drivers
run sfc/scannow
run av scan
i feel like aliens wrote this website - https://www.howto-connect.com/fix-bad_object_header-bsod-error-windows-10/

it looks like a memory error. it often happens in association with other errors of a certain type, some of which you have already had.
 
Jun 22, 2020
27
0
30
Not going to argue with DB but the updaters made by motherboard makers at least stick within the lines and only update drivers for their devices, unlike Driver Booster that lies about how many outdated drivers you have and then dumps the wrong drivers on boards.

I wouldn't use any tool except those built into bios to update them. Some boards have internet inside bios now so you can get it to look for an update for itself.

technically, just cause its got idle in its name, its not really idle... - https://store.steampowered.com/app/1147690/NGU_IDLE/

as for error... um. stop showing me errors I haven't seen. We checked ram already right?
suggested fixes - update win 10, you on 20H1, I don't know if 20H2 would help
boot into safe mode, if it doesn't happen it could be drivers
run sfc/scannow
run av scan
i feel like aliens wrote this website - https://www.howto-connect.com/fix-bad_object_header-bsod-error-windows-10/

it looks like a memory error. it often happens in association with other errors of a certain type, some of which you have already had.


as for error... um. stop showing me errors I haven't seen. We checked ram already right?

Yes, ran memtest and this is the second memory and mobo already. I sold the first memory kit and the guy who bought it has no issues.

suggested fixes - update win 10, you on 20H1, I don't know if 20H2 would help

Already did that. Some updates made it so I did not get a single BSOD for 3-5 days.

boot into safe mode, if it doesn't happen it could be drivers

Tried it, still happens.

run sfc/scannow

done it several times

run av scan

done it several times too

I'm starting to think it has something to do with the CPU.
 
Your problem is beginning to sound a LOT like this one, and I'm wondering if it isn't simply a known bug or issue related to ALL or many systems using a specific hardware combination. IDK really, but it sure sound similar.

Try disabling ALL of the C-states in the BIOS and see if you still have issues.

Read from this point forward, and then go back and read the full thread if you desire.

https://forums.tomshardware.com/threads/i-dont-understand-this.3658464/post-22044753
 
Jun 22, 2020
27
0
30
I read somewhere that the IDLE game mentioned uses a lot of CPU resources. Try a CPU test. Maybe Prime95. Darkbreeze mentioned it before and may know more about it than myself.

It uses between 3/5%. I'veused prime95 and other stress tests before. I've run so many stress tests for HOURS and HOURS and not a single BSOD. Lately I've been playing wow and it maxes out my GPU and about 25/30% of the CPU while also doing other tasks. ALSO not a single BSOD.
 
Jun 22, 2020
27
0
30
Your problem is beginning to sound a LOT like this one, and I'm wondering if it isn't simply a known bug or issue related to ALL or many systems using a specific hardware combination. IDK really, but it sure sound similar.

Try disabling ALL of the C-states in the BIOS and see if you still have issues.

Read from this point forward, and then go back and read the full thread if you desire.

https://forums.tomshardware.com/threads/i-dont-understand-this.3658464/post-22044753


Could you guide me through it? I've read two instances regarding similar issues on the AMD forum and the solution was to bump dram voltage and vcore voltage.
 

gardenman

Splendid
Moderator
I ran the new dump file through the debugger and got the following information: https://jsfiddle.net/e4ktxpc6/show This link is for anyone wanting to help. You do not have to view it. It is safe to "run the fiddle" as the page asks.

File information:111520-6968-01.dmp (Nov 15 2020 - 12:24:58)
Bugcheck:SYSTEM_THREAD_EXCEPTION_NOT_HANDLED_M (1000007E)
Probably caused by:memory_corruption (Process: MemCompression)
Uptime:0 Day(s), 0 Hour(s), 42 Min(s), and 04 Sec(s)

This information can be used by others to help you. Someone else will post with more information. Please wait for additional answers. Good luck.
 
Go into the BIOS. Find the DRAM voltage setting. Click into the field where it says 1.35v and change it to 1.36v. Hit enter. Go to the Exit tab, save settings, exit BIOS.

how_to_overclock_using_gigabyte_am4_motherboard_%284%29.jpg
 
Well, then my only guesses can be, if you are getting BSOD with XMP enabled, and raising voltage doesn't help, then it has to either be faulty memory, bent pin or two on the CPU, overly tight or unevenly tightened CPU cooler or bad motherboard. Of those, take your pick based on what you already know you've done or eliminated.
 
Jun 22, 2020
27
0
30
Well, then my only guesses can be, if you are getting BSOD with XMP enabled, and raising voltage doesn't help, then it has to either be faulty memory, bent pin or two on the CPU, overly tight or unevenly tightened CPU cooler or bad motherboard. Of those, take your pick based on what you already know you've done or eliminated.

I've already changed the MOBO and Memory. CPU pins are perfect and I've reseated it many times. Thermals are great. I've tried different default bios settings, a bit of overclocking and underclocking. Same issues, always BSOD when relatively low loads (max 10% cpu usage).

I've scoured the AMD forums regarding BSOD's involving the Ryzen 3600 and there are a lot of reports saying early chips were faulty due to TMSC's early 7nm process. I bought this close to its release so this might be it. I'm going to try and RMA the CPU and see what happens. Also, I got a 750w PSU for a GPU upgrade(possible RX6800 or RTX 3070). Since I also want to upgrade my memory I will buy another 16GB to get to 32GB and I'll be able to test another set of memory.

Will keep you guys updated!
 
I would not be against it being either of those suggestions. I think an RMA of the CPU as well as testing another memory kit is a good idea, HOWEVER, and this is a BIG however, I would absolutely not "assume" that purchasing a second memory kit will work out when you throw it in with the existing memory kit, even if it is the exact same model or part number. Chances are very good on Ryzen that it won't (Or any any platform, that it MIGHT not), plus the Ryzen platform is particularly hamstrung by the fact that in a number of cases, depending on the chipset and board model, that when using four DIMMs your maximum supported speed is going to drop to something less than what it is able to run at with only two DIMMs installed, if the memory is 3200mhz or higher.

It can be done, sometimes, but it is not likely going to be a plug and play event. You will probably have to put some work in to get it to do so, even if it was a four DIMM kit you purchased.

Right now, memory is pretty cheap. If you are going to buy new memory I'd suggest that you find a 2 x16gb memory kit that shows it is supported for that motherboard on the memory manufacturers compatibility list (Corsair memory finder, G.Skill memory configurator) and buy a single kit that has the full capacity you wish to run. It's less stressful on the CPU/Internal memory controller, and less stressful on the VRMs as well. I'm not saying it can't be done, or that some haven't done it, but it's usually been on a top shelf flagship motherboard, not a mid tier or budget model. And even then, it's usually required some hair pulling and foot stomping. Obviously, this is a decision that is totally up to you and there will likely be some with dissenting opinions who have been fortunate enough to make it work.
 
Jun 22, 2020
27
0
30
Hey guys, I'm waiting for AMD's RMA. I always set the CPU voltage to auto as everything else. I tried upping the voltage on the memory but still got BSOD's so this time I've tried setting CPU and VSOC voltage to normal and Dynamic Offset to around 0.04 v.

View: https://imgur.com/q9DFOHm


Turns out the PC is up for 48 hours and going without a single BSOD. In the meantime I've played games and also let it idle with that idle game that always gave me a BSOD. Nothing so far.

Upping the voltage is safe? Is it normal for CPU's to work like this or maybe it really is a faulty CPU?