Random hard locks, but prime stable

T

thrawn86

Gawd
Joined
Nov 10, 2005
Messages
697
I'm no noob, but I'm stumped on this one. specs in sig, brand new components.

Problem:
random hard locks (screen and image freeze, have to force power off) at random intervals.

attempted fixes:
rails are good, oc is stable in both prime, memtest and games (though it happens at stock/below stock), hard drives passed SeaTools tests, new video drivers installed (it happens sometimes when I'm watching a movie), and I ever ran a virus scan for good measure.

I'm completely stumped here. I remember having a similar problem a LONG time ago, and it turned out to be the power supply, but my voltage rails are just fine, and my Kill-a-watt is reporting load power draw thats well below its rated 420w (and besides it freezes at idle/low load just as often as full load)
 
In my case, it was a USB device. Totally random lockups @ idle, just reading through the forum, and or @ full load playing COD4, etc.
 
thrawn86 said:
attempted fixes:
rails are good, oc is stable in both prime, memtest and games (though it happens at stock/below stock), hard drives passed SeaTools tests, new video drivers installed (it happens sometimes when I'm watching a movie), and I ever ran a virus scan for good measure.

I'm completely stumped here. I remember having a similar problem a LONG time ago, and it turned out to be the power supply, but my voltage rails are just fine, and my Kill-a-watt is reporting load power draw thats well below its rated 420w (and besides it freezes at idle/low load just as often as full load)
Click to expand...

How are you determining that your voltages are fine? With a DMM or some software? And since you mentioned the Kill-A-Watt, I have to mention this:

Paul_Johnson said:
Yes and a quick search would turn up this topic a million times over. Here is the recap:

1) APFC can fool Kill-A-Watts into giving you abnormally low readings (some times giving better than 100% efficiency)

2) Power supplies derate with temperature anywhere from 2w/c above a nominal rated at value to 10w/c.

3) Kill-A-Watt's and most power meters sample too slowly to catch transient loads (the Transient load from our tests is 117w and is COMPLETELY missed by Kill-A-Watts).

4) Power supplies last longer if you stay in the 40% to 60% range of their output.

5) power supplies are quieter if you stay in the 40% to 60% range of their output.

6) Power supplies are cooler if you stay in the 40% to 60% range of their output.

The power meters in UPS software are just as bad. You have to spend some change before you get anywhere near an accurate power meter when your PSU has APFC.
Click to expand...

Oh and try this:
[Testing the GPU]
Use Furmark or ATI Tool to stress test the GPU
 
Danny Bui said:
How are you determining that your voltages are fine? With a DMM or some software?
Click to expand...
both. speedfan is actually dead on with the voltages, aside from detecting my 12v line as "2.18" (multimeter reads 12.18 lol) also the powerstream has adjustable rails and under/over volt indicator lights, so I can tell pretty quick if its dropped below 11.5 or so.
Danny Bui said:
Oh and try this:
[Testing the GPU]
Use Furmark or ATI Tool to stress test the GPU
Click to expand...

I'll load up ati tool, haven't used it in awhile. I hope thats the problem, theres a 4890 with my name on it :0
 
Last edited:
locked up after 3 hours of prime and atitool artifact test. the tests didn't fail, it just completely locked up.

event viewer isn't telling me much either, aside from the fact that the previous shutdown was unexpected :p

going to try with all my drives disconnected tomorrow and wiping all my drivers, and I suppose after that its on to stress testing in linux or reformatting :(
 
just an update in case anyone else has suggestions.

atitool and prime are still stable for 12+ hours, rails are still spot on and OC and stock speeds, and memtest passes but i'm still getting hard locks ( screen freezes with looping sound. ) I know its also not software related, it locks up almost immediately when I boot to the windows 7 cd.

even though the psu seems fine, it seems to be about the only thing I can't directly test or haven't tried swapping out at the moment, so i'm looking at ordering a bfg ls-550.
 
I have almost exactly the same problem.
I am getting hardlocks when playing CPU-intensive games. In Suprememe Commander for example I will have a hard lock 95% in a skirmish game with a few AIs. Most of the time at the end of the game. Screen will freeze completely and the sound will loop and go around me in a circle (5.1 sound).

Prime95 however is rock solid, as is furmark & memtest and the temps and voltages all are fine.

I already tried to remove one memory stick and swap their position, but that didnt help at all. However when I run my computer at stock speeds I dont get any hard locks.
Also tried billions of different BIOS settings. I also tried only overclocking a bit, but it wont go away. Only at stock speed its gone.

E8400 E0 @ 4 GHz
P5Q @ 445 MHz
2 x 2 GB Corsair XMS
Enermax 625W PSU
EVGA GTX 260 SC (also happened with a BFG 260 GTX)

Right now I think its either my CPU, the mainboard or maybe the PSU, memory or maybe the sound card (creative x-fi gamer).
But I cant check any of the components because I dont have any spare.

Please keep us updated if you find a solution.
 
Last edited:
OP said it happened at stock and below - which has me thinking it could be related to the NB + 4 full DIMMs.

Delissa, I think you just need to bump up your NB voltage. When I moved from 1x1 to 2x2, my OCs were no longer stable.
 
Mav451 said:
OP said it happened at stock and below - which has me thinking it could be related to the NB + 4 full DIMMs.

Delissa, I think you just need to bump up your NB voltage. When I moved from 1x1 to 2x2, my OCs were no longer stable.
Click to expand...

Ah my bad didn't see that. And yea I would agree with you. If it's happening at stock and below, looks like your NB voltage would probably be the culprit. I had huge problems being stable on my Q9550, but after a small bump in NB voltage, from like 1.25 to 1.29, I'm fully stable.
 
Tried upping the NB voltage already. Right now I am at 1.36V and it still freezes.
 
Yeah I dunno.

1) You said it goes away at stock (so hardware problems, CPU/Mobo/GPU/X-Fi are not the issue). Neither are drivers then.
2) Have you tried bumping the VTT two steps? I remember that I was still crashing in L4D, even after a healthy boost to NB and vcore, and in the end a small bump in VTT got me stable.
2b) USB devices - go through device manager untick "save power" option. Also disconnect anything you don't need to be using.
 
VTT? I dont find that option anywhere in my BIOS.
I have my PLL at 1.54 if thats what you mean.

Ill try the USB devices, but I only have an XBox360 controller receiver connected besides keyboard and mouse.
 
thrawn86 said:
just an update in case anyone else has suggestions.

atitool and prime are still stable for 12+ hours, rails are still spot on and OC and stock speeds, and memtest passes but i'm still getting hard locks ( screen freezes with looping sound. ) I know its also not software related, it locks up almost immediately when I boot to the windows 7 cd.

even though the psu seems fine, it seems to be about the only thing I can't directly test or haven't tried swapping out at the moment, so i'm looking at ordering a bfg ls-550.
Click to expand...

Any update for you? Did you try raising your other voltages?
 
It's probably called FSB Termination then. Yeah, I doubt the controller receiver is an issue then. I found that it's certain thumb drives or 2.5" USB HD's that can cause BSODs. I guess it's less relevant here, since your issue is hard locks and not BSODs.
 
Some diagnostic advice: Software diagnostics are never 100% accurate. In fact my experiences tell me that they can't be relied on most of the time by themselves. Prime 95 and other stress tests have some diagnostic relevance as they are fairly good for stability testing, but they are not diagnostic tests by design. When you have a problem nailing down an issue between hardware or software the easiest solution is to start swapping hardware if that option is available to you. (Most people do not have duplicates or similar hardware found in their main machines laying around for diagnostic use.) I'd start with RAM, video card, drives, etc first before trying to do a CPU or motherboard swap. Put a volt-meter to your PSU and monitor the rails in hardware. Hard locks can be a symptom of PSU problems. Though I'm normally not one to say things about PSU's diagnostically speaking as they are often blamed for things that they can't be responsible for. (Such as drops in FPS etc.) Still you should check it given that hard locks can and do occur with faulty PSUs. Though random reboots are more often likely than a hard lock when that's the issue.

More often than not, if an issue like this is truly not a software issue and is a hardware problem, it will be memory, hard drive or motherboard related in that order of likelyhood. Memtest 86 is not really a good diagnostic test when the results are considered alone. A problem with Memtest 86 results may indicate a problem but those problems can be created as a result of bad BIOS settings. There may also be a problem with the motherboard or system where other hardware relates to the memory subsystem. In such cases the memory channels on the motherboard, chipset, or memory controller problems may be the culprit. In cases which an Intel Core i7 or AMD Athlon/Phenom/Phenom II processor is being installed, Memtest 86 faults can also indicate a problem with the CPU's integrated memory controller. Memtest 86 is a good tool but diagnostically it only narrows down a potential issue in one subsystem that spans several parts. Additionally problems reported by Memtest 86 do not conclusively prove that there are no additional problems with the system outside the memory subsystem. In some cases more than one piece of hardware will be responsible for a myriad of problems.
 
You must log in or register to reply here.
Back
Top