BSOD help (may be related to GPU temps.)

[Bsbsod]

Ok, i'm going to take a break from this for a few hours. I'm growing a bit frustrated and that's just not a good feel for fixing things (in fact, that's when they tend to get broken). I will run on 2gb of memory now, an additional 2 tomorrow and so on and so forth until a crash occurs again, unless it happens today. in which case I may just break down and purchase new RAM just to test the theory and return it if its not the case.

I did the Furmark test and it did not cause any issues, I will try the GPU ram test as well and come back with results if anything goes wrong, if nothing happens I will be back tomorrow to provide additional info on the experience.

 

[writhziden]

Alright, sounds like a good idea. And I agree that taking a break when getting frustrated is a good idea. That way, you don't end up with a computer shaped hole in your windows, walls, or doors.

 

[Bsbsod]

So the night of sleep gave me some more ideas, one of which, ultimately, may be the fix for this whole thing. This morning, (after leaving the computer on all night) I decided I would add another 2gb stick, but while I was looking at the 3 sticks of Ram I have left I noticed something.

I don't know how this didn't hit me any sooner.

The ram is made by Corsair (XMS2-DHX 2gb 800Mhz) but then what's this? some numbers were different on the backplates. And then it hit me like a flying brick. The timings, hell even the voltages were different. I bought these in pairs, but had not installed them as such. I don't know how this ever worked for so long and only now has started causing BSODs. Since that realization I put another stick in (the matched one) and worked on the computer for hours without issues, pushing it to the very limit.

I figured if there was still any RAM issues it would rear its head when I installed the third and fourth matched sticks. Since any RMA that may need to happen (yes, corsair has a lifetime warranty) needs to happen in pairs, I figured I would step up to the full 8gb and give it a go, if any issues happen now I will at the very least know which pair I need to return.

I will keep you posted with any additional developments. I will be face-palming over not thinking of this earlier for many weeks to come.

 

[writhziden]

Mixing and matching RAM can cause blue screen crashes. I should have mentioned this, as well. Sorry about that. It is also possible that there was some ESD damage to a module in one of the sets. If the damage was not too severe, it may have taken some time to show symptoms.

You may also find Avoid Static Damage to Your PC | PCWorld interesting further reading as to why RAM problems can occur. It basically says that ~400 Volts of electrostatic discharge (ESD) can ruin RAM, but you will not feel the discharge until it reaches ~1400-3500 Volts (that link says 3500-4000 Volts, but it varies based on the climate and how dry it is). When RAM receives a shock that we cannot feel, it can work for a while, and then slowly fail over time until symptoms become noticeable. That is one of the many reasons why so many users end up on these forums wondering why their systems are suddenly not working.

So for future reference, make sure you practice good electrostatic discharge (ESD) habbits and keep yourself grounded when touching PC components.

For ESD Safety, follow these steps:

1. Shut down and turn off your computer.
2. Unplug all power supplies to the computer (AC Power then battery for laptops, AC power for desktops)
3. Hold down the power button for 30 seconds to close the circuit and ensure all power drains from components.
4. Make sure you are grounded by using proper grounding techniques, i.e. work on an anti-static workbench, anti-static desk, or an anti-static pad. Hold something metallic while touching it to the anti-static surface, or use an anti-static wristband to attach to the anti-static material while working.

Once these steps have been followed, it is safe to remove and replace components within your computer.

 

[Bsbsod]

I had heard of mixing and matching being problematic, i didn't realize that chips of the exact same type, by the same manufacturer, purchased at the same time could be so different!.

i'm hoping that this fixes the remainder of my issues. I'm going to continue my normal routine in the hopes of exposing any issues. If nothing comes up in a few days I will get back to you to close this thread for good.

Wish me luck.

 

[writhziden]

I did not know you purchased them at the same time and they were supposed to be the same type. That is very odd. You must have gotten a batch that was mixed with some older version still sitting on the shelf. I think if you explain that to one of their representatives, they will probably replace the whole set. If you only send back one pair and get a new pair, you will be faced with the same problem of mixing and matching since the new pair will be newer than the pair you keep. You would be without your computer for a time with that option to send back all RAM, though. Just something to think about.

Best of luck!

 

[Bsbsod]

Ok, well here's a massive update:

In addition to everything I've tried before I have now

1. Bought new RAM to match my 5-5-5-18 corsair (its the exact same stuff, same voltage now)...and still it BSOD'd

2. Found my Motherboard was running hot, so I took the NorthBridge heat dissipator off, cleaned up the crap the manufacture put on there and put arctic silver on that baby (dropped temps by over 20 degrees!)...and still it BSOD'd

3. Run Memtest, Furmark, Prime 95 and all that stuff for extended periods of time 6+ hours each and no issues were found....and still it BSOD'd

4. Completely re-wired my computer to ensure maximum airflow (I'm still not over the temps in my case) and still...you got it...it BSOD'd.

Okay...


So we're right back where we started, except my case is damn near turning into a fridge at this point (injecting humor to diffuse tension). Latest BSOD message was once again, related to NVIDIA drivers. I've since COMPLETELY and i do mean COMPLETELY ripped every shadowy presence of the NVIDIA drivers and then rolled back to a previous driver to test stability. I'm also running a profile on my fan to raise their speed and ensure the card is not burning up the way it was before (70 degrees under load)

I'm hoping i dont crash again. If I do I will post the next BSOD that comes up. And likely start thinking of reversing back to playing my games on an abacus.

 

[writhziden]

That sounds like a lot of work you put into your problem. What happens when you run with the old RAM only? Do you still BSOD?