Question New build, lots of problems

[Azrael444]

So a couple days ago I bought some components to upgrade my pc and since then it's taken me for a [CENSORED] ride. Right now I have:

• a Ryzen 3 3200g with its stock cooler (no overclocking)
• an Asrock A320m-hdv r4.0 (yeah I know)
• two 4gb 2666mhz sticks, PNY brand
• a no-name Sentey BXP65-LS PSU that claims to be 650w (yeah I know, but like. It's an APU. This sentey used to power a 3rd gen i5 with a 2gb R5 250x a couple years ago, so I didn't think it would cause any trouble)
• two HDD's, 1tb and 500gb

and I think that's it. Now, since its inception this computer has had

• multiple BSOD's (KERNEL_MODE_HEAP_CORRUPTION, THREAD_STUCK_IN_DEVICE_DRIVER_M, probably some others that I can't remember. It's BSOD'd like 5-10 times in the last 10 days, always when gaming/ doing something intensive),

• had some artifacting and crashed while playing Fortnite (that's on me really, I have a small desktop with no airflow and for some reason I decided to check if a notoriously hot cpu would run fine with its stock cooler on a reduced space. The answer was no)

• had a black screen once, again when playing Fortnite. Had to shut it down with the case button. It's worth mentioning that the very first day of building the pc, I actually played like 3 or 4 fortnite matches without issue, and they were pretty long matches where I ended up in the top 10 almost every time.

• has done the thing multiple times where it the screen freezes and the speakers start blaring a buzzing sound. This is particularly bad because it's done this not only on modern games, but also in [CENSORED] Black Mesa, a dx9 game that I used to play with no issues on a 1gb 9800gt (a 12 year old gpu). I'm pretty sure it's not a temperature issue, as I had Afterburner open the whole time and the temps were sitting on 52 c before it crashed.

Now to be fair, I didn't do a clean install of Windows after changing the mobo and cpu, which is a really big mistake I guess. I'm gonna do it right after posting this but honestly idrk if it will actually fix everything. But now onto the things I have actually done, I have:

• installed the AMD chipset drivers, the AMD Radeon Software that also downloaded some drivers, a couple of drivers that came with the mobo on a little cd, etc
• updated Windows 10 to the 2004 version (took ages)
• used DDU to erase old Nvidia and AMD files
• changed the thermal paste (on my week old processor. Yes I'm desperate)
• benched and stressed the cpu with CPU-Z, nothing really happened
• benched the cpu with prime95, it ran for a while until the cpu temp got to 85 c or so (I was checking with afterburner). Then a minute passed and I noticed the computer's clock hadn't changed, I tried to move the mouse and it wouldn't move. The screen was frozen and the ctrl+shift+win+b "reset graphics" thing didn't work, so I shut it off.
• ran the windows 10 memory test, it didn't find anything bad. I'll run memtest86 tomorrow if I find my USB drive
• ran chkdsk /r, it said Windows had fixed some things but that it required no further action
• played a couple games. Played far too many hours of Among Us with no issue but that game can run on mobile so it's not exactly a benchmarking tool. Was able to run Civ 6 in dx12 and dx11 for hours consistently for 2 or 3 days with no issue (haven't tried that game again tho), played like 10 mins of JC3 on low settings and besides the awful framerate and one instance where I thought it was gonna freeze, nothing bad happened.
• 
  
• Played a couple fortnite matches back when it allowed me to (both dx11 and dx12, wanted to see which one performed better), played one (1) match of Heroes and Generals that crashed mid-game, and then another that had some weird artifacting/ vignetting (like a purple ring around the screen). The framerate was fine for the 5 or so minutes I played, but the visual glitch was freaking me out so I clicked "exit to main menu" and then BAM blue screen. I pressed the little case button instead of holding down the big one, which I assume means that it rebooted instead of totally shutting down, and when it turned back on the screen was filled with red and white artifacts. Used the big button after that.
• 
  
• I also played Black Mesa right after building the pc, must have been like 30-40 minutes and no crashes or crazy high temps. My little brother played Roblox two times and both times it BSOD'd (the culprit was apparently dxgkrnl.sys) and he claims that Arma 3 also crashed, but to a black screen instead of a blue one. So far I haven't had any crashes during normal internet usage or when playing HD videos (I've seen some people having those issues, but nothing here) nor while doing digital art.

Honestly at this point I might just [CENSORED] around and flash the BIOS. Other than that and a clean windows install I have no idea what to do. Help please?

 

[Phillip Corcoran]

Definitely a clean Windows install --- we always advise that after motherboard change, even for Windows 10.

Your issues may not all be down to not re-installing Windows but there's a good chance most of them are -- only one way to find out.

 

[jyraya25]

1/-do a clean reinstall of windows and troubleshoot it test your games and stuff i think everything should work allright
IF NOT :
then your HDD your RAM and your PSU should be revised one of them may be the cause of your probleme i had the same probleme the HDD was faulty

 

[Azrael444]

Definitely a clean Windows install --- we always advise that after motherboard change, even for Windows 10.

Your issues may not all be down to not re-installing Windows but there's a good chance most of them are -- only one way to find out.

So I did a system reset which downloaded a Win10 copy from the cloud, but kept some of my files and apps. Is that good enough or should I go for a full drive wipe? Anyway it didn't fix anything. I downloaded the drivers from the Asrock website instead of using the cd, but from what I've seen they're the same version and everything, installed the amd chipset drivers, blablabla. Booted up Civ 6 (which was on my other HDD so it was saved from the purge) and BSOD'd after 10 mins, around 60c temps (PFN LIST CORRUPT, then upon rebooting it showed UNEXPECTED STORE EXCEPTION). Windows update does have some updates that I could install, but I don't think it'll do anything, but I'll do it anyway later. Also worth mentioning is that the old Realtek HD Audio Manager seems to be gone, but replaced by an app? my mic is ridiculously quiet tho, but that's another issue. So I ran chkdsk in my other drive, it ran for a while and then had an "unspecified error" so I ran it again, and it found some defective clusters (I think that would be the translation) but again told me that "no further action is required". Yet the event viewer shows 6 errors with source "disk", all saying "Device/Harddisk1/DR1 has a bad block". It also shows 8 ApplicationHang errors from MicrosoftEdgeCP.exe, which is weird because Edge never crashed. After the chkdsk was over I ran Civ6 again, played for 10 minutes again, and BSOD'd again, this time coded "SYSTEM SERVICE EXCEPTION" caused by fltmgr.sys. Event viewer shows 5 AppModel-Runtime errors code 65 right before the "Critical kernel-power" shutdown (the appmodel errors say something like "error with 0x57 when recovering hte Appmodel runtime state for windows package") and if I scroll down there are five more of those errors right before the other BSOD (and if I scroll even further down there are 25 DeviceSetupManager errors, from devnodes PCI and HDAUDIO. So I guess that's what the quiet mic is all about).
At this point I have no idea what's going on anymore. Is it... the harddrive? Which one? Is it the ram? Why does it happen only when gaming?
I've read a bit about memtest and driver verifier and seatools and hdtuner so I might as well run those, I also haven't flashed my bios (yet) so that's another step I can take. I'm just… lost

 

[Azrael444]

I may have fixed it- I ran memtest86 on my little Goku-shaped USB. Dual channel test failed almost immediately with thousands of errors, so I started running the sticks one at a time. Stick 1 on slot 1 passed, stick 2 on slot 1 passed, stick 1 on slot 2 passed, stick 2 on slot 2 failed catastrophically. Dual channeling with the sticks shifted (i.e. stick 1 on slot 2, stick 2 on slot 1) gave the same results as the first test, with slightly less errors than the first time but still thousands. But yet I had the brilliant idea of booting the computer anyways, with both sticks in, and running an old game to see if it would crash. I still had Black Ops 1 (released in 2010) so I played that, and not even 5 minutes in when I was about to respawn for the third time (veteran difficulty, don't judge me) it artifacted and the screen froze, so I removed stick 2 and tried to play it again, and this time it actually worked (managed to play the ending mission for like 5 minutes because the game was almost over, and a bit of Zombies afterwards). I don't wanna cry victory yet, but I really hope it's fixed now.

 

[DSzymborski]

Honestly, you really need to do the full drive wipe. It can easily behave now and then you find another obnoxious similar BSoD gremlin weeks or months from now.

Anything worth doing is worth doing correctly.

 

[Azrael444]

Honestly, you really need to do the full drive wipe. It can easily behave now and then you find another obnoxious similar BSoD gremlin weeks or months from now.

Anything worth doing is worth doing correctly.

it most definitely didn't behave
Little rundown- ran memtest86, dual channel failed in both combinations, stick 1 in single channel passed in both slots, stick 2 passed in one slot and failed in the other. Played Black Ops 1, dual channel failed after dying a couple times. Both sticks worked by themselves though, no matter how many times I died. My brother played roblox under my strict supervision for an hour and a bit, crashed with the "bad" stick (stick two). Switched sticks and put in the "good" stick, played Black Ops again, didn't crash again. Brother played roblox again for about the same amount of time, crashed again. Both times it was a split second BSOD that I didn't even get to read, then it rebooted itself and there was no minidump to analyze. Downloaded aida64, stress tested for an hour with only one stick (the "bad" one I think) and it never crashed, but had some access violation errors at the end. Put in both sticks, ran aida64 for another hour, same thing. Cpu temps peaked at 78c, motherboard and hdd never got past 33c, no crashing but access violation errors at the end (they would keep appearing as long as I had the "statistics" screen open, so I didn't get to read the aida logs). Played Civ6 in the lowest settings for like an hour (both sticks of ram) and it never crashed, got bored of waiting for it to crash so I cranked the settings back up to high. Still didn't crash, so I played Black Mesa which a) wouldn't load a particular save (ok this might be the game's fault) and then b) after loading a different save and trying to load the first save from within the game instead of the main menu, blue artifacts appeared all over the screen so I hard-rebooted. Then (or maybe this was after upgrading the bios, I honestly can't remember) I loaded the save that did work, and after about 10 seconds the game froze with the dreaded "bzzzzzt" sound stuck in a loop. Hard reboot.
After scouring the internet for a little bit, I came across a forum topic of people with the same motherboard and APU and seemingly related issues, with one of them advising to downgrade the bios. I did that (downgraded from 3.70 to 3.30) and guess what, it also didn't work. There was freezing and bluescreens almost on boot (one of them was IRQL something, the other ones I don't even remember. They didn't write minidumps), it took me like 5 reboots to get the computer stable in order to download another bios and put it on the usb stick. Anyway once I got it stable I decided to reinstall AMD chipset and gpu drivers and then chill for a bit and eventually ended up opening a drawing website (edding's wall of fame) and plugging in my Wacom, and it BSOD'd immediately (I think it was KMODE EXCEPTION NOT HANDLED, courtesy of -surprisingly- amdkmdag.sys, an AMD driver. Not really going by memory here, just reading the minidumps and seeing at what time each of them was and then guessing which error appeared where in my timeline, so this one may be wrong). So I upgraded the bios again (this time to the latest version, 4.0) and booted up the computer. No freezing or BSOD's on boot but when I opened the wall of fame website it BSOD'd again (PAGE FAULT IN NONPAGED AREA, courtesy of both ntoskrnl.exe and ntkrnlmp.exe. Two equal errors at the exact same time, or so it says on WhoCrashed) (worth mentioning that on bios version 3.70, the one that came installed with the mobo, the website did work perfectly). Honestly this all seems so random and tiresome, I mean I still think it's the motherboard but I really don't know what to do here.