[axl2468]

So this problem just began happening just recently - my PC would go through one of these 3 events at random (may happen once a day, may take a few days before it happens again):

 

1. Freezes for a minute or two, then the PC turns off (no restart).

2. Freezes for a bit, then a BSOD error appears.

3. Freezes for way too long, and I'm forced to forcefully shut down the PC.

 

Here are my specs:

CPU - i5-2400

Motherboard - ECS H61H2-MV

RAM - 8GB Kingston DD3-1600 (2x)

GPU - Sapphire RX 580 Nitro+ 4GB

SSD - Crucial BX500

 

I have a gut feeling that it happens more frequently when I leave it idle for a while, but I don't think it has an effect on anything and is probably just placebo.

 

I don't think any of my hardware changes have caused this issue since I've been using the same setup for some months before the problems started happening.

 

Here's a list of what I tried to do:

1. Replaced my secondary hard drive (which was getting too old anyways) with ones known to work (from another PC).

2. Checked my SSD (my main drive) health, nothing out of the ordinary according to CrystalDiskInfo, Hard Disk Sentinel, Speccy, and Crucial's own software.

3. Reinstalled Windows from a clean slate three times.

4. Reseated almost everything in my PC except for the processor.

5. Changed SATA cables for the SSD.

6. Used Intel's diagnostic tool.

7. Used GpuMemTest - no problems found.

8. Used Window's memory diagnostic tool and OCCT's memory test (granted, this one was the free 1 hour only version) - no problems found.

9. Tried the Windows Driver Verifier to determine if it's a driver causing it - that wasn't much help at all.

10. Uninstalled all Windows updates - no change.

11. Installed all Windows updates available - no change.

12. Did an OCCT stress test to check whether it was power-related.

 

I've also looked at the memory dump files for the BSODs - BlueScreenView says that all the BSODs (minus the ones caused by activating Driver Verifier) have the bug check DPC_WATCHDOG_VIOLATION and the Crash Address is at ntoskrnl.exe. I also checked the dumps using WinDbg, and they all point to something called "ataport".

 

I've attached the related dump files for reference:

 dumps.zip   818.13KB   62 downloads

 

I'm lost as to whether this is a software problem or a hardware problem. I am very confident that this isn't malware though, as I already tried it after doing a clean format to all drives in the PC. I'm also doubtful that it's a GPU problem nor a power problem, since I can play games on my PC just fine (though I've yet to try it under a high-intensive game, I'll try it with something like Doom Eternal)

 

I haven't tried out using Linux yet to see if it's a weird Windows thing as I don't really have the time to do that right now. Is there a way that I've missed to find out what the cause is? Thanks.

[Advertisements]

While we look at the crash dmps;

 

Provide the brand and model name or number of the power supply (PSU).

 

Download then run Speccy (free) and post the resultant url for us, details here, this will provide us with information about your computer hardware + any software that you have installed that may explain the present issue/s. 

 

To publish a Speccy profile to the Web:

1. In Speccy, click File, and then click Publish Snapshot.
2. In the Publish Snapshot dialog box, click Yes to enable Speccy to proceed.
3. Speccy publishes the profile and displays a second Publish Snapshot. You can open the URL in your default browser, copy it to the clipboard, or close the dialog box.

Edit to add, two of the crashes were caused by ataport.SYS drivers and the intermediate crash was caused by atapi.sys drivers, what they are and where you go to update them here and here

[phillpower2]

While we look at the crash dmps;

 

Provide the brand and model name or number of the power supply (PSU).

 

Download then run Speccy (free) and post the resultant url for us, details here, this will provide us with information about your computer hardware + any software that you have installed that may explain the present issue/s. 

 

To publish a Speccy profile to the Web:

1. In Speccy, click File, and then click Publish Snapshot.
2. In the Publish Snapshot dialog box, click Yes to enable Speccy to proceed.
3. Speccy publishes the profile and displays a second Publish Snapshot. You can open the URL in your default browser, copy it to the clipboard, or close the dialog box.

Edit to add, two of the crashes were caused by ataport.SYS drivers and the intermediate crash was caused by atapi.sys drivers, what they are and where you go to update them here and here

[axl2468]

Sorry for the late reply - I reset my BIOS to default and that apparently resulted in my GPU not being able to output any display. After bashing my head against the wall for more than an hour, I found out that it was because I needed to turn on CSM in the BIOS settings (this also fixed the issue of the GPU not displaying anything until I've logged into Windows and the PC not displaying anything when in Safe Mode - which is weird considering how it didn't behave like that before)

 

My PSU is an FSP Hyper K 500 (not the best one, I know - but at the time EVGA PSU's weren't available at my local store).

 

Here's the Speccy profile: http://speccy.pirifo...eCkx7mRJIMKGoX5

 

Also, I've done more things based on what I read in other similar posts in this forum:

- I've uninstalled the Realtek drivers, maybe Windows' own audio driver works fine for my system

- I've reinstalled my graphics drivers.

 

As for the ataport and atapi drivers, I think they're already updated (since according to the links that you added, you get them through Windows updates - my Windows has the latest updates now).

Edited by axl2468, 27 July 2022 - 06:56 AM.

[phillpower2]

No worries, we understand that you may have other things to take care of.

 

Strange behaviour with the GPU there but you at least have it sorted.

 

The PSU to put it bluntly is junk, it is not even powerful enough to support your GPU which AMD say here requires 500W, you have additional storage devices and an extra monitor hooked up so need something that is at least 550W and a minimum of Gold efficiency rated, your PSU claims only 456W which I would also question.

 

Also, I've done more things based on what I read in other similar posts in this forum:

 

 

It is not good practice to follow the advice of more than the one forum or to go off and do your own thing, I make a point of not assisting in either instance so as to avoid getting caught up in any mishaps.

[axl2468]

Oh I see, my bad. Besides power being a possible reason to the crashes, is there any other possible cause for the problem?

Also, I can't really afford a new PSU right now, so would you recommend underclocking + undervolting the GPU so it wouldn't place too much of a stress on the PSU? Maybe limit it to the lower end of the GPU's factory clock/voltage setting? (AFAIK AMD's driver software allows this)

[phillpower2]

The manufacturer of your GPU like AMD state a minimum of a 500W PSU, you can confirm this here

 

The only safe thing that you can do is remove the add on GPU from the PCI-E slot on the MB and hook your screen up to the appropriate video port on the MB, at the moment you are slowly but surely killing your add on GPU and possibly other hardware.

[axl2468]

Okay, I understand. I'll have to stick with integrated graphics for now it seems.

 

I'll let my PC run its course for a while, then I'll report if the crashing goes away. Thanks for the help!

[phillpower2]

Best course of action as an appropriate PSU will be far less expensive than having to replace a bad GPU or worse still the whole PC should the PSU go bang.

[axl2468]

Would this power supply be good as a replacement?

[phillpower2]

No, Corsair only recommend the CX range be used for the below, crazy price as well;

 

CX Series Modular power supply units are an excellent choice for basic system builds and desktop PC computer upgrades, offering high reliability, low noise, and the flexibility of modular cabling. 

 

 

That`s for the modular version and they are of better quality than the older hard wired models.

 

A much better option that has a ten year warranty, is $20 less expensive + gives you a free PSU tester EVGA SuperNOVA 550 GA

 

Important note, before making any purchase see if you can borrow an appropriate PSU to swap in for testing or alternatively get the PC checked by a local tech, the poor quality PSU may have damaged your hardware and the GPU being the most power hungry they tend to suffer first and the most. 

[axl2468]

Best Answer

Alright, so for an update:

 

Replaced my power supply with the one you gave me. Had it sent to a technician since the problems were still occurring, and it was determined that the motherboard was at fault. Replaced the motherboard - seems like the problem has stopped happening.

 

Thanks for the help!

[phillpower2]

Important note, before making any purchase see if you can borrow an appropriate PSU to swap in for testing or alternatively get the PC checked by a local tech, the poor quality PSU may have damaged your hardware and the GPU being the most power hungry they tend to suffer first and the most. 

 

 

You 100% needed to replace the original PSU in any event but can I ask whether or not you did either of the above before you purchased the said PSU and the issues that you say were still happening, was that with the new PSU and the GPU or the original PSU with no GPU in the board.

[axl2468]

I bought the PSU first since I was gonna get one in the first place. Issue still happened regardless of PSU and whether or not I had the GPU in the board (even tested with a fresh windows installation on a different hard drive with all possible setup combinations just to check) so that was my sign to bring it to a local repair shop. They had the pc tested with a different motherboard (in a fresh Windows installation, but the problem still occurred even under a fresh install anyways) and the issue stopped happening. Technician suspected that the problem had something to do with the SATA controller.

Edited by axl2468, 06 August 2022 - 08:37 AM.

[phillpower2]

Thanks for the additional info 

 

They had the pc tested with a different motherboard (in a fresh Windows installation, but the problem still occurred even under a fresh install anyways) 

 

 

Bit confused by the above tbh.

[axl2468]

Oh I just meant that they took my PC, and swapped the motherboard and the boot drive (with a fresh Windows install). That "but the problem still occurred even under a fresh install anyways" was just me trying to say that using the different boot drive probably wouldn't change the results anyways.