Need some help, getting BSODs and hangups

[Sparky]

I've been battling strange hangups for awhile now that happened when I touched the case. Yeah, really weird stuff. I replaced the PSU when my Antec died and that didn't help. In the process I changed cases. I RMA'ed my mobo thinking that was the problem, and also the gfx card (that I know did have problems).

Now I have all the pieces and parts put back together, and it worked great for a few days. Then BAM, got hit by a BSOD. Later this week the system froze when I touched the case. Now today I've had two BSODs. I saved the info from the two today, maybe you can help me track down the problem.

First one:
IRQL_NOT_LESS_OR_EQUAL
STOP 0x0000000A (0x3ACD3E84, 0x0000001C, 0x00000000, 0x80502544)

Second one:
DRIVER_IRQL_NOT_LESS_OR_EQUAL
STOP 0x000000D1 (0x8B832DB8, 0x00000002, 0x00000000, 0xBA13D7A1)
portcls.sys adress BA13D7A1 base at BA13C000, datastamp 41107F13


I am running an Asus A8R32-MVP, naked opteron 165 @2.8GHz, G.Skill 2GB PC3200 2.5-3-3-6 1T @ stock (crap OC ram), X1950XT, SB Audigy 2 ZS. I know I am overclocked but the CPU was ran under orthos for 16 hours with no errors so I don't think that is it.

If anyone can help me out I would greatly appreciate it!

[Sparky]

Well I was running the windows based memtest because I was told to try it out. My RAM had already passed 14 passes of memtest86+ so I doubted it could be the RAM but at this point I was willing to try almost anything. This is what I got just now:



I had to have two instances of memtest running since one couldn't handle the 2GB of RAM I have.

Wonder why memtest86+ didn't report an error but this did? Could it be maybe my RAM isn't bad, but my OC is screwing with it? I am running a 133 divider, with stock timings except I adjusted the max async latency to 10 (otherwise it wouldn't POST). Ideas? I am selling a 2GB kit of Corsair XMS, would it be worth maybe sticking that in there for now to see if it changes anything?

[Jupiler]

Maybe this can help :

http://support.microsoft.com/kb/921401/en-us


// edit :

you might wanna give that Corsair kit a go, and see if the problem goes away.

[Sparky]

I just got yet another BSOD. This time it was a STOP 0x0000008E (0xC0000005, 0xBFA4A653, 0xAA6737EC, 0x000000) with ati2cqag.dll listed as being the offending file.

When I get a chance I'll put the Corsair RAM in. With the frequency I've been having these BSODs lately it shouldn't take too long to figure out whether it helps or not

[[XC] 4X4N]

It's got to be ram related, either a setting or a bad stick. Try one stick at a time. Run memtest, you should be able to get over 500%. If it does, then go with the other stick. I'm betting that you have one that is bad.

[Sparky]

windows memtest or memtest86+?

[tiro_uspsss]

I just got yet another BSOD. This time it was a STOP 0x0000008E (0xC0000005, 0xBFA4A653, 0xAA6737EC, 0x000000) with ati2cqag.dll listed as being the offending file.

When I get a chance I'll put the Corsair RAM in. With the frequency I've been having these BSODs lately it shouldn't take too long to figure out whether it helps or not

the offending file: ATI2cqag.dll - try re-install / change ur drivers for ur video card

[[XC] 4X4N]

windows memtest or memtest86+?

Memtest for windows. I'm not a believer in memtest86 anymore. Half the time I can't even boot into windows where I can pass for hours with memtest86.

[Sparky]

the offending file: ATI2cqag.dll - try re-install / change ur drivers for ur video card

It has done this with the other drivers though, and others are using these without problems (Catalyst 7.3).

Memtest for windows. I'm not a believer in memtest86 anymore. Half the time I can't even boot into windows where I can pass for hours with memtest86.

Maybe not because the RAM was unstable but the CPU was? Anyway I'll run it. Currently I put that Corsair kit in my PC, we'll see if that changes anything.

This instability caused both F@H clients to EUE, the GPU client and the CPU client. Bummer... But I want to run them through the night to keep everything but the RAM the same on the system to see what happens... argh this is so frustrating.

[Spawne32]

you know i was having this same problem when i just got finished building this new rig, seems that my OCZ ram didnt have enough volts supplying it and i had to bump it up manually to 2.0v

[Sparky]

you know i was having this same problem when i just got finished building this new rig, seems that my OCZ ram didnt have enough volts supplying it and i had to bump it up manually to 2.0v

Hmm that is a thought, I've only ever used the default voltage setting (read: auto ). I could try that I suppose, see what voltage these modules are rated at and set it manually.

It happens when you touch the case ?? Maybe your case isnt grounded or something I would take the system apart and put it back together on a cardboard box to see

It has done it between two cases, two PSUs, and two mobos. Also did it when it was sitting running on my desk in between cases. These BSOD's I've been having aren't from touching the case, I've left it alone.


I stuck the Corsair XMS in my system last night around 12 then went to bed. When I got back to my PC about 11 hours later (hey, it's the weekend, I didn't have to get up ) it was still running. I've been using it since and so far no BSOD. Too early to tell anything yet, but at least this time it has made it almost 13 hours now without a BSOD while last night I was lucky to go for an hour or two before it would do it.

When I pulled the G.Skill from my PC I noticed the modules were really warm. How warm is too warm for RAM? They have heatspreaders, THIS is the kit I have.

[apexracing]

G.skill seems to have a problem with their heatspreaders, they don't attach them very well with thermal tape. Instead of cooling the RAM, it actually heats them up...

Try running the RAM without the heatspreaders and place a fan in front of them.

[Sparky]

G.skill seems to have a problem with their heatspreaders, they don't attach them very well with thermal tape. Instead of cooling the RAM, it actually heats them up...

Try running the RAM without the heatspreaders and place a fan in front of them.

Taking them off would void the warranty though...

I'm thinking that if my system ends up having no more problems with the Corsair, I'll RMA the G.skill kit and sell the replacements, and buy the corsairs from my friend that I am selling them for.

[[XC] 4X4N]

I have rma'd with Gskill before. No problems, and sent me the replacements within a couple of days.

Did you ever try using one stick at a time? That would tell you for sure if its bad.

[justin_c]

goldmemory, not memtest, nor memtest86. goldmemory will fail without the first 5 minutes compared to a dozen passes with memtest.

[Sparky]

goldtest costs money for anything more than a quickie test

4X4N, I haven't tested just one stick at a time yet. I want to see what happens with running this Corsair kit first. If everything runs A-OK, I'll get about finding out whether the G.skill is bad or maybe just doesn't like the timings set at the CPU overclock I have, and if it is bad which stick. Trying one thing at a time here.

Well it has been running for about 21 hours now with no BSOD.... still too early to tell, but it is a little bit comforting. RAM is easily RMA'ed, my CPU, well... it isn't, since I popped the lid off it. So if it was the CPU I'd be SOL.

[Sparky]

Well the system had been running find for about 22 hours or so, and I was playing a game and the system hardlocked. Hit the reset, no POST Powered off, back on, and I was up and running again... for now.

Argh this is getting annoying enough I'm almost tempted to sell almost everything I know works fine and buy all new parts

Is it possible that the traces of arctic silver on the top of my CPU could be causing problems? It is a naked opteron 165, and there are stains of arctic silver on top from a mishap and it is around several of the little chips around the core. Or maybe I damaged the chip when I delidded it, although there is nothing physically damaged looking on the top. Or maybe my overclock isn't totally stable, even though it passed a 18+ hour orthos run.

Come to think of it, all my problems started when I sold my single core 3700+ and got this opteron. Up until then it was extremely stable. I hate my PC right now

[[XC] 4X4N]

I doubt you damaged your cpu. I have taken off the ihs on many amds, and haven't hurt one. I had a venice 3500 that had all 4 corners chipped, plus a little crack in the middle of it, and it still ran at 2.7 24/7.

Is there a fry's near you? They have the e4300 and a junk ecs board for $119 right now. I'm tempted to get one myself, but I have enough for now and will wait until the q6600 goes below $300. I know it costs to change, but in f@h you will about double your output from the opty. When I had my 170 running at 2.9ghz, it would be about 17-18 hours per wu. My e6600 at 3.4ghz, would do them in about 7hours.

[Sparky]

I doubt you damaged your cpu. I have taken off the ihs on many amds, and haven't hurt one. I had a venice 3500 that had all 4 corners chipped, plus a little crack in the middle of it, and it still ran at 2.7 24/7.

Is there a fry's near you? They have the e4300 and a junk ecs board for $119 right now. I'm tempted to get one myself, but I have enough for now and will wait until the q6600 goes below $300. I know it costs to change, but in f@h you will about double your output from the opty. When I had my 170 running at 2.9ghz, it would be about 17-18 hours per wu. My e6600 at 3.4ghz, would do them in about 7hours.

I'm not aware of any Fry's around here. Even if there was, I wouldn't want to get a junk mobo, and if I did I'd still need to buy new memory, those C2D chips call for DDR2 right?

I have my doubts about a budget C2D outdoing my opteron, at least at stock on a crap mobo. I run the GPU client on one core anyway so it isn't that big a deal.

[sc00p]

Sparky... this is a wild guess... but do you still have all these crunchers at your dorm room (listed at your sig)?

Another guess... it just seems a bit random errors... so could it be you are now at point when general circuitry from the wall is affected and causing some ripple for the rig (->dirty current ).

I had once random errors with one rig (reason: folding-addiction) until I found nice UPS (remote line, 3500VA, ~60kg).

You would think at this scenario all your rigs should behave alike, but not necessarily...

...btw: this is not devious plan to get you to shutdown any folders

[Sparky]

Yes all the crunchers are in my room. Half are on the same circuit as my PC, the others are on the bathroom circuit. My PC is behind an APC 725VA UPS. I've had trouble at friends' houses though too.

[Sparky]

Interesting....

Just started an orthos run - didn't make it 2 minutes before it errored out on core 0. I was running the blend test. Will try a small FFTs test next to see if it is CPU or RAM related.

[Sparky]

Small FFTs failed after 19 minutes. Downclocked the CPU to 2.7 from 2.8GHz, running a blend test again. 21 minutes so far, this is a good thing. Previously wasn't making it 5 minutes before a failure on blend.

*fingers crossed*