Ongoing BSODs, possible defective memory?

I've been having ongoing BSODs for a few months now. They started out rare, but have gotten much worse, to the point of this PC throwing multiple at me in the course of a day. They're to the point now that certain games cannot be played because they will 100% BSOD in 5-10 minutes or so.

At first, I thought it might be my processor's memory controller... but it doesn't make sense that it would be damaged and continue to work like it is right now. Then I thought it was the memory (and at the time had a thread here about this, I think), so I did the usual Memtest passes... and had no errors. Strange.

Then, for whatever reason, Windows decided it no longer wanted to create dump files for me to analyze despite the option being enabled. So now uploading those here are out of the question, as the minidump folder is empty regardless of what type of log I want it to create (small or full kernel dump).

As an example of how these BSODs have "progressed"... I play(ed) Team Fortress 2 quite a bit when this started happening in other games another user played. Before long, TF2 started throwing BSODs within 5-10 minutes also. I tinkered with the options and disabled the "Multicore Rendering" option... which helped, for a while. Now, even with that disabled, playing TF2 is out of the question.

Also now, BSODs will happen while casually browsing the internet and such. It seemingly depends on how long I have the browser open and how much memory it ends up using before I close it.

As for the BSODs themselves, the stop codes are all over the place. 116, C2, 7F, 50, numerous others. Specific errors range from BAD_POOL_CALLER, SYSTEM_SERVICE_EXCEPTION, to page faults in non-paged. (Again, making me think it's the memory due to their randomness.)

But again... memtest found no errors the last time I ran it over night, and just last night the Windows Memory Diagnostic found nothing, although I only let it run with the default settings, which consists of 2 passes and probably not-thorough-enough testing.

So that brings me here! What do you all think? I'd gladly upload some dump files to analyze, but... Windows won't write any for me. This hardware is pretty old (2009 build), so I know it's well beyond time to build a new rig, but if I can keep this one running, and maybe fix its problems, for now... that would be fantastic.

One thing I haven't tried is Driver Verifier. I know you all suggest it here frequently, but I'm not sure if it would be any use considering most of these BSODs point to either faulty memory, or a motherboard. (Or maybe PSU, considering how inconsistent they are.) I don't know how helpful it would be considering the sudden lack of dump files being written post-bugcheck.

Sorry for the long response! I tend to forget I make threads here sometimes when I'm busy with work and whatnot.

After making this thread, though, I opened this relic up and decided to reseat everything aside from the CPU... and it seemed to work, at first. Then, I had Steam crash upon exiting a game, and continue to crash when I tried launching it. Thankfully Steam creates its own dump files when it crashes, so I peeked at those and they all pointed towards memory/heap corruption. Unsure if related, but that makes me look at the RAM a bit more.

Then BSODs started happening again. The first one was a SYSTEM_SERVICE_EXCEPTION, then a PAGE_FAULT_IN_NONPAGED_AREA, so I went poking around the internet for possible causes of that, and came across a thread elsewhere that suggested disabling your page file and re-enabling it, which I did.

Much to my surprise, after the BSOD just a moment ago, Windows is creating dump files again! (I'll upload it with this post.)

This last one was MEMORY_MANAGEMENT, although it seems to reference a Windows 7 driver fault involving the... Audio Graph Isolation? whatever that is.

About the hardware tests: I ran all of those back when this trouble started, including the HDD tests many times (as they were difficult to even get working), and everything passed. The only thing I really suspect at this point is the memory, because almost all of the BSODs have referenced some kind of memory-related issue. Of course, after Windows quit writing dumps (corrupted page file, maybe? I don't know why it would quit when it'd been fine all this time), I couldn't analyze anything for myself to try and pin anything down.

It's still strange to me though that the memory would pass a night of memtest. Although I've read another thread from here where a guy ran memtest one night and it passed, then again when he left for work... came back, and it found errors. So I suppose it's possible.

I wouldn't be surprised if I had some kind of faulty driver causing issues, either, considering I've gone to great lengths to keep this GWX crap disabled. (Which gets recreated if I run SFC. Yay?)

At this point, I'm unsure if it's a memory issue (I've ordered new memory regardless, because we did have a module of this stuff fail pretty early on and had to RMA it a few years back), perhaps a faulty driver issue (thanks, Windows update!), or possibly another piece of hardware (motherboard? Only thing I can really suspect any longer).

Edit: If you'd still like me to perform all those hardware tests again, I can at a later point. Unfortunately, I work second shift, so I'd have to do the majority of them over a few nights and post the results as I completed them. I know memtest alone takes eons to finish.

Whoops, forgot about that bit. Here you go.

Actually had some kind of odd crash a bit ago; thought it was a BSOD at first (froze, stuttering sound, figured the blue screen was coming in just a second), but then the PC seemed to recover for about .5 seconds, and quickly froze again. No dump created that time, sadly. Nothing other than MS Antimalware errors in the event log also, which I doubt were related, but one of the errors did reference a user config file for the game I was playing (Elder Scrolls Online).

Not sure why; again, probably unrelated.

Sorry again for the very long responses. Work is pretty hectic and extra-draining lately.

Good news, however. I ran most of those diagnostics, and even took screenshots, and everything checked out... even the memory. But! That new memory I'd ordered prior to making this thread showed up, and I installed it the night afterwards when I got in from work.

Haven't had a single BSOD since then. I'm... really at a loss as to why extended periods (I'm talking 11+ passes here, not just a few hours, but 12+ hours) of memtest would show no faults with the memory that was installed, yet here I am with new memory and everything is okay.

I did see an ancient thread on here where a guy was having BSODs, ran memtest and it passed, then ran it again and it showed errors really early on, but I've ran memtest I don't know how many times throughout the past year since this started, and it always checks out.

The only thing I can think of that might have impacted the outcome is restoring the OC settings in the BIOS to default, despite nothing being set differently from when the machine was built. Everything in that section was set to auto, as it always has been. Yet when resetting those settings to default, and installing this memory, everything is fine. Also lowered CPU temps by a large margin, too, somehow (Back to sub-40c idle temps with a stock Intel fan). So I'm not sure what happened there, but that could have very likely been a contributing factor, or THE contributing factor.

In any even, I'm going to mark this as solved. Can't believe it took a year of troubleshooting and dealing with random BSODs to finally fix the problem... but I guess that's how it goes. And here I was looking in to building a new machine soon... which I still want to do, but now I can hold off on it a bit longer and save for better parts.

Anyway, thanks bunches.

Note: weird things to note with this particular instance that may help in future threads.

1. Windows quit writing dump files on its own, despite the setting being set for them. Disabling the page file, rebooting, enabling it, and rebooting again seemed to fix this. Unsure if related to the memory or not. I'd assume so since that's what gets written to the file.

2. Might have posters reset just their BIOS' OC settings to default, even if they're still stock settings, to see if that has any effect at all. Again, unsure if this was related or not, but it apparently had some form of impact on my particular instance.