Question Two weeks of irregular BSOD after PSU was killed from a possible surge ?

[XxLiveLivelyxX]

Hello!

I will keep this short as possible.

Spoiler: System Info

mobo: ASRock B450m Pro4
2x Gskill RipJaw 8GB Sticks of Ram DDR4
GPU: Sapphire Radeon Rx580 Lite oc
CPU: AMD Ryzen 5 2600
(Not sure about HHD [From the previous build])
(Just put this in last week) 750w tx/m Corsair PSU

Spoiler: Main Thread

I built my own PC and played games on it for about 2 or 3 weeks on mid settings (out of preference, not limitation). I had a 550w Corsair PSU at the time.
My son asked to see the game in Epic settings, and so I set it that way. After about 5 or 10 minutes of playing my ([Not sure what the rating is in J] surge protector [it has On/Off/Reset],) went (SNAP) and everything attached to that power bar was off...

My PC then periodically began crashing and displaying BSOD's -- Research suggested that maybe (power surge caused hardware problem) or that maybe it corrupt some files in the system32 folder, and needed a fresh format. I formated... Everything was going great. Then I was getting blue screens again after 2 or 3 days.

I acted quickly and bought a 650w PSU, formated my pc again, used it for a day or two and again while playing, (SNAP) The surge protector went off again!

Some friends suggested I invest in a real surge protector and upon looking into that, I bought a 4500 Joule Surge Protector, and only plugged my PC and External Hard-drive into it.
My research suggested I was overloading the 15Amp outlet with approximately 1300w or more of power. Granted( they're rated for 1800w, - 20%) so about 1400w. Damn was I cutting it close....

So now I have a 750w PSU in there plugged into its own wall outlet on the 1300J surge Protector. After formating again, I am still getting random BSOD.
One error suggested it could be a memory issue, but others suggested the GPU (or CPU since there has been an AMD Crash [AMDKMDAG.SYS]) But honestly, it could be anything that caused that file not to run.. right?

I thought I uninstalled my GPU drivers and ran on the basic fallback driver for a few days, but today... BSOD again.. which led to seeing that windows installed the drivers behind the scenes (Testing again).

I moved my 2 sticks of Ram into the other 2 available slots, and the BSOD suppressed for a couple of hours... I've been getting error codes with uncaught errors..

Spoiler: Conclusion

One thing I've been meaning to try is to update my Bios, cause I cant seem to get onboard to run graphics.
Since I updated my new PC when the pc sleeps, sometimes it won't wake up. and this seems to happen before the BSOD twice in the last 2 or 3 weeks. How do I safely update my Bios?

Maybe I fried the GPU? It still works when I do a stress test on it, and the same with CPU. They were only 5-minute long stress tests, but My CPU Heat just kept going up... and I stopped it at 79Celcius.

I hope that someone can help me solve this, as I have no more hair to pull out of my already bald head... At least my sense of humor didn't surge.

There is no TLDR. Thanks for reading!

EDIT: Windows seemed to have installed th RX580 driver behind the scenes or something, cause I just checked in the device manager and had to roll it back :/


EDIT: (THE NEXT DAY) I am running unity and watching discord stream and its working fine for now on basic windows fallback adaptor.

Spoiler: zgpu stuff

Here are my ZGPU stats if they might be helpful (Windows installed the drivers on its own):
During Render Test:

After Test:

 

[rgd1101]

on ups or surge protector? could be a short in the pc

 

[XxLiveLivelyxX]

Thank you so much for replying.
It wasn't a ups.

I didn't know much about the electrical side of things before this happened.. but i believe the power bar it was plugged into was a surge protector, as it has an on off and reset label.

How can I fix a short? What are some potential causes? Could it be cable management? Its messy in there..

 

[XxLiveLivelyxX]

alright, I have re-done my cable management in hopes that it will help. I did have the case's audio wires running under the gpu into its slot on the mobo... Though it was working fine before the Surge Protector went SNAP... but anyways, I ran it behind the case this time.

If after a few days nothing still, I will try and add the GPU drivers again, and see if it changes after that.

Any other suggestions, Ideas or whatever lay them on me. I'm listening

NOTE: No BSOD Since I posted the original thread.

 

[XxLiveLivelyxX]

After doing some research, It could easily be the AMD Drivers for my RX580... Cause since yesterday, after removing the drivers I have not yet faced any bluescreens... I did just experience a black screen and 3 beeps, and then my display returned.. What could this be caused by?

 

[rgd1101]

brand and model of the psu?
what the gpu temp on idle and on load?

 

[XxLiveLivelyxX]

The second psu was an evga 650w standard psu.
I put the specs under the zgpu spoiler tab in the original post. I hope that's what you're looking for. The test wasn't done for too long, as I didn't want to stress anything unnecessarily.

 

[XxLiveLivelyxX]

What do you guys make of this? This is with the basic display driver, and has been like this all day... My guess is it can't actually read this info, and is displaying something false?

Wont let me display image please follow link.
https://gyazo.com/31f6276b756650fea3bf0c0702bccb5d

Note: No BSOD since I uninstalled the amd drivers for my gpu and no beeps today either.

@madmatt30 Im sorry to bother you sir, but you've offered some good advice in the past I am hoping you could take a moment to add your 2 cents, if you have the time of course.

 

[XxLiveLivelyxX]

So after 2 days of no blue screens, I figured I'd try something new. I had realized that after formating (and even when I built my pc) I had not installed the mobo drivers..

So today I did just that.

Then I got an unreal engine 4 crash report.

While looking into the crash report to see if it could be something to do with the UE4 (On my laptop) I heard my pc booting and sure enough it was back at the ASrock boot screen... So I guess it Blue screened... Im so (insert Swear here) over this. Why is this happening!??? I don't have money to bring it to a repair shop, and can typically resolve this stuff on my own... but I've had enough. I hope someone can offer some sort of help..

 

[Uncodable]

After doing some research, It could easily be the AMD Drivers for my RX580... Cause since yesterday, after removing the drivers I have not yet faced any bluescreens... I did just experience a black screen and 3 beeps, and then my display returned.. What could this be caused by?

Hello XxLiveLivelyxX,

From personal experience, I've had bugs with AMD's device drivers before, in particular the ones for the RX 550. From what I've read, you allowed Windows to automatically install the drivers for your GPU. Now, you can usually get away with Windows automatically installing them, you should try installing the drivers from the official AMD website here: https://www.amd.com/en/support/graphics/radeon-500-series/radeon-rx-500-series/radeon-rx-580. Do not take my word for it, but there could very well be a chance that the surge protector causing the bug checks was a fluke. Still keep this noted, as just about anything can cause a bug check. Lastly, I would like for you to upload all of your crash dumps from the past week. Do NOT send the MEMORY.DMP dump file in 'C:\Windows', as this is a dump of the computer's memory as a whole. This can contain credentials, such as passwords. Rather, send the minidumps located in 'C:\Windows\minidump' or the directory that you set minidumps to be dropped at.

 

[XxLiveLivelyxX]

Hi, Thank you for your reply. I would like to send these files to you, as I have a couple of them. I cant seem to open them though.
I am trying to add you on Discord, but it says the spelling or number is wrong..

 

[XxLiveLivelyxX]

I called a repair shop to get some prices, and they suggested Crystal disk info to see if its the harddrive:


https://gyazo.com/4d64c262464b0b20ba1165172947284c

But I dont understand what I am reading..

 

[Uncodable]

Hi, Thank you for your reply. I would like to send these files to you, as I have a couple of them. I cant seem to open them though.
I am trying to add you on Discord, but it says the spelling or number is wrong..

I went to get food, sorry. I changed my username on Discord, it's uncodable#9099. Anyway, the files were made by Windows itself, so they are protected. This is what you have to do manually for each of the crash dumps (do these steps in order):

1. Traverse to the minidump directory (go to the folder)
2. Select and right click a mini dump file
3. Click on properties and go to the security tab on the top of the newly opened menu
4. Click on advanced, a new menu should have appeared
5. You want to give yourself ownership, so click on the change owner (will spawn a UAC prompt, respond to it accordingly)
6. In the text box, type in your profile's name (in my case, I would put in CrystalCode) and click on OK and apply
7. Close the permissions menu and reopen it to refresh it
8. Give yourself permissions to the file by clicking on "Add"
9. Click on select principle and type in your username, the exact same one you used for the ownership change, and hit OK
10. Give yourself full control over the file and then press OK, and apply. If you get a prompt saying that it is a protected Windows file, click on "yes."

 

[XxLiveLivelyxX]

I ended up copying the folder to my desktop, where I am now able to upload the folder to my server space.

I've added you on discord,

Should anyone else be interested in the dump files, ill be happy to send you a link - Just request them

Based on what I seen, it appears to be a Hard Drive issue, but Im not sure.

 

[Karadjgne]

Regardless of claims, a power strip with the on/off/reset is not a surge protector nor suppressor. It's a fire-hazard disguised as a safety product. It's nothing more than 3 strips of metal inside some cheap plastic with a switch rated breaker attached to the power leg. It's entire design is to overheat and trip if you add more than the rated amperage. It won't stop a surge unless that surge starts from inside your house and is absolutely massive voltage. A lightning strike will travel through your main electrical breaker box and trip those breakers long before it trips that power strip.

For instance, if you have 5A worth of stuff plugged into that strip, you can bet the breaker in the mains box has not only that 5A load, but the TV load, light bulbs, alarm clocks, and everything else that's on that circuit.

If that power bar is rated 15A,and you load it to 14A thinking you are safe, imagine what the breaker in the mains for that circuit is seeing, and overheating, and 2 seconds away from tripping if you plug your phone in to charge.

There are a few exceptions. Like the ones made by Monster Cable or APC, that have UL listed protections, honest Joule ratings, surge strike damage warranty upto $25k etc, but they cost $60+, not $3.99 from Walmart.

And you rely on that seriously cheap contact from a psu plug to those cheap metal buss strips to provide stable power. Once they heat up even a little, those bars weaken, losing tension and you get erratic voltage/amperage to the psu, which then gives your pc fits.

 

[XxLiveLivelyxX]

After doing some memory Diagnostics, I've noted that with 2 sticks of ram in the mobo (1 and 3) or (2 and 4) I get hardware problems detected. However, with only one or the other stick, There is no hardware problems found.. Seems like I've done in the mobo? whats happening?

 

[XxLiveLivelyxX]

Well, it seems to have been a problem with the motherboard / RAM as previously stated. I installed my gpu drivers today, played my favorite game on basically low settings for a good hour or so, did some programming. Watched some YouTube. And no blue screens.

I've contacted ASrock and their going to honor their 1 year warranty. Hopefully, this is the end of discussion.