AppHangB1 FarCry2/COH crash

[American197]

Excelent! Well, I just woke up, and this thing has been going for about 12 hours, and no crash/BSOD (obviously). I'll just hit stop, get some breakfast, and get back to you.

"Torture test ran for 11 hours, 50 minutes, 0 errors 0 warnings" :D

Is this proof positive that I do NOT have a dead stick?

PS, my temps are now into the high TEENS for my cores! Wow, from 68 to 18...quite a difference!

 

[smarteyeball]

Well that's a good sign Yeah, i don't think you have a dead stick.

The fact that it's run for that long and that it's passed memtest, seems to eliminate RAM from the equation for now.

I'd be looking at drivers now. Also, I'm guessing that the games themselves have been patched?

ahh, the joys of troubleshooting

 

[American197]

Hey Smarty, sorry, I keep editing this post quite a bit (3 times), and if you've read it already I may have put more stuff in it. From now on I'll just add new posts instead of trying to confuse you. If, in fact, I've even confused you. Or if, in fact, you've even read this yet....lol

Yeah, I currently have the latest patch of the game.

So get this...

Just loaded up XP32 on my 1TB...sp3, every update, etc (have NOT updated my mouse or keyboard driver tho...is this a problem?)...after everything is updated, I run COH and BOOM, crashes in like, 30 seconds. A brown SOD and an audio loop freeze...I havent had a restart yet, but I don't think that really matters.

EDIT: Just updated my monitor driver, went into COH and it once again crashed within a few seconds. I could'nt even build a unit, lol...

SECOND EDIT: I've installed the latest XP32 vid, sound, network, monitor, and chipset drivers, have the latest BIOS, etc etc etc and my system has crashed like, 5 or 6 times in COH, all within the first minute of running it.

SO. Correct me if I'm wrong, but it's NOT:

1) Vista/ DX10

2) Firefox

3) RAM (too many tests, tried different [but not all] sticks)

4) Nortion AV

5) Extra bloatware (on XP I just installed the bare minimum of EVERTHING)

....

Whats that leave? MOBO? A corrupted hard drive? RAM timings/BIOS settings? GPU? Power supply? Any thoughts?

I'm starting to think that it HAS to be my RAM or voltages...COH is crashing SUPER fast now, and XP can only use 2 sticks...faster crashes, because of less RAM (voltages/ AUTO settings show their flaws faster?).......

AHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH :sa:

 

[American197]

*sigh*...

well, I updated all the voltages and fooled around with them for a bit, did all the correct timings, etc etc, and I still crashed.

Strangely, when I would input my VCORE at 1.25 in BIOS it would read 1.24 on all of my sensor readings and apps in Vista. However, my VCORE reading would STAY at 1.24 (save for 2 dramatic examples)...as of right now, the thing bounces all around. I have NOT checked their readings in XP yet, but I'm sure they're the same. Is this normal?

Anyways, I was googling other options when my pc probe (mobo's temp/etc gauge) started going haywire (making crazy alarm sounds), saying that my VCORE went from 1.24 to 2.1, or around there. As soon as the alarms started, they stopped...so I quickly shut down and went back to AUTO. This was the SECOND voltage warning I got from pc probe TODAY...the other one said I went from 1.1 to 0.0 when I tried formatting my 1TB.

I've noticed the VCORE jumping around, sometimes from .9 to 1.24. It's supposed to be set on AUTO (I was told that was 1.2). Is this normal? Here's the warning log:

[01/14/2009 at 08:23 pm] Vcore Voltage Abnormal, 0.00 V
[01/14/2009 at 08:23 pm] Vcore Voltage become normal, 1.10 V
[01/15/2009 at 03:55 am] Vcore Voltage Abnormal, 2.04 V
[01/15/2009 at 03:55 am] Vcore Voltage become normal, 1.24 V

Could my crashes be due to spiking and dipping voltage?


Thanks again,

Sal

 

[smarteyeball]

Voltage fluctuations are normal when you have C1E / Speedstep /turbo mode enabled, or set to to auto (which is most likely enabled).

C1E and Speedstep basically lower your clock speed when the CPU is idle. C1E is more on/off (low clock speed/high clock speed) and Speedstep is (low/mid/high). They are 'power saving' devices on the chip that are designed to use less power when the CPU is IDLE or at low load.

Turbomode is new to the i7's. From what I understand of it, Turbomode disables unused cores and 'overclocks' cores in use. For example, if an app is only utilizing 2 cores, turbomode disables the un-utilized cores and 'seeds up' the other 2 cores. If only 1 core is being used, then turbomode disables 3 cores etc.

Reported voltages in Windows are often lower than whats actually set in the BIOS. CPU-Z is good for watching your voltages. I would probably trust that over the PC probe.

However those spikes to 2.04 (if accurately reported) are a cause for concern and could very well be part of the.

In BIOS, I would temporarily disable C1E, speedstep and turbomode. That way your chip will be running at it's rated speed and set voltage, without fluctuations, apart from Vdroop (voltage droop which is normal.)

If that helps stabilize things, then try turning the features back on one at a time. Disabling these features won't hurt your chip

Since XP doesn't have DX10, you can rule that out.

It's unlikely that it's the hard drive.

Disable Norton when running games. It's possible that Norton is causing mischief. (personally I would ditch norton anyway, but that's just my opinion. Avast is a good free A/V )

 

[American197]

Excellent advice. I'm gonna start checking my PS right now, and I'll disable those features. I'm not sure, but I believe I disabled one or more of them earlier when toying with voltages, but I ended up getting a one-time-only 'Overclocking Failed' error. When I put the feature back on, the error went away, if I remember correctly. I'll look at my journal...I've started keeping a journal of voltages tried, methods tried, etc etc! lol

To be clear, my voltages have only spiked/dropped off (that I know of) 3 times, all in the last 24 hours. What would be an example of ''stable'' voltages? Lets say if I set my VCORE to 1.25...any idea of correct fluctuation?

Never been a big fan of Norton, that's for sure. I think WHEN (thinking positive, lol) i figure this whole thing out I'm gonna reformat everything, and get rid of Norton. I'll certainly look into Avast.

I know I've been a pain, but seriously, I appreciate all of your help. I would be in Google nightmare-ville if it wasnt for you! Thanks again,

Sal

 

[American197]

Once again, you're dead on the money. I turned those programs off in the BIOS (had to googlehunt for the C1E) and my voltage is STABLE. As in, hasnt moved or changed in 5 minutes (1.14v). I used to watch the thing dance around from moment to moment.

My 3.3v is at 3.1v, my 5v is at 4.97v, and my 12v is at 12.42v. Now, from what I've been reading, you don't want a voltage swing of more than 4% in either direction, or you risk instability. However, my 12v is about 4% off already, just idling. This seems to be a problem.

I have 3 7200 hard drives, an optical, a usb/sd card/floppy all-in-one, and my 4 channel HAF case fans hooked up (presumably) to this 12v...is it too much for it to handle? (I say presumably because I read that the 12v powers anything with a motor, including fans, HDs, and opticals. I'm guessing about my floppy).

Any info you have on this would obviously be most appreciated. I realize this isnt (seemingly) Vista related, but it sure helps ...any personal thoughts on the i7s hyperthreading? I've been toying with the idea of turning it off.

Stranger still, my PS was making a super weird noise before I turned off those BIOS modes you told me about, and now the noise is gone. This is crazy stuff man. What a first build, hmm?

........spoke too soon. Although the voltages are rock solid, I just had that alarm go off again, and the VCORE voltages rose to 2.04. I'm wondering if this is just a glitch, or if my PS is going out. Any thoughts? Should I shut down?

 

[smarteyeball]

Hyper threading only seems to advantage apps that makes use ouf multi core setups or are aware of HT. One advantage of disabling HT is that it may make your temps a little lower. Either way it can't to turn it off and test.

Again, monitor your CPU voltage within CPU-Z and not the PC Probe. Run P95 or IBT for a few minutes and watch what the voltage goes up to. If it really is spiking to to over 2 volts, then that it is a serious cause for concern.

Are you using PC probe to monitor your 12v, 5v and 3.3v or BIOS?

Try Everest Ultimate and check the sensors. (It's a good app to have regardless)
Lavalys - Comprehensive IT Security and Management

The problem with software monitoring apps is working out which one is 'accurate'.


What sort of strange noise is the PSU making? Those volt variations are a little sus. especially the 3.1v. The 12.54v reading should be within spec, but is still a little high. It may be worth returning / swapping the PSU, then again the PSU may be perfectly fine :huh: I guess it depends how much fluctuation is going on under load. ie is the 12v spiking higher than 12.54 or dipping below 12v.

A smooth first build is no way to learn! It's all part of the 'fun'


(Sorry for the poor responses, I've got a splitting headache and can't think straight ATM )

 

[American197]

No problem man, take care of yourself first!!!

I've been trolling around the asus forums quite a bit, picking up info, etc etc

The voltage errors (nearly 60 times in the last 2 days) have been going off like bombs over here...I would imagine that if they were real, my computer would be Yahoo Serioused. (Like that reference? LMAO I loved him when I was a kid!). However, that doesnt change the fact that theyre driving me crazy. No other sensors have detected a voltage increase, however, and many people have complained about this sort of thing with PC PROBE II...so, I think I may just cautiously ignore it.

I DID check my power supply today using a multimeter and a molex...I was only able to check the 12v (steady for 20 min at 12.5) and the 5v (5.02). Both were tested while idle, and during prime95, with NO voltage variation. I'm gonna go buy a PS tester tomorrow, just to make sure. I'll get back to you with those results.

 

[smarteyeball]

It was bloody hot here today and I think it fried my brain a little. I used to love "Young Einstein" as a kid. Poor old Yahoo is a "joke that Australia try's to forget" these days

Have you tried COH etc with the RAM and Voltages manually set yet? (check your journal. It's a good idea. Unfortunately I start by scribbling notes and then get lost amongst the mess )

Yeah, PC Probe doesn't have a good rep for reliabilty. I'd still keep an eye on the vcore (CPU volts) just in case. It's feasible that something that is still set to AUTO in the BIOS may be overvolting the chip. Possibly turbo mode if you still have it enabled.

With the ASUS boards, when you get the 'failed overclock' message, you can usually just carry on with the current settings without reverting settings you've changed. If you were to get that everytime you restarted etc, then you'd think twice about adjusting something.

Amongst your combinations, have you tried:

RAM = manual settings (manufacturer specified timings / volts)
Vcore =manual settings
C1E / Speedstep: disabled
Turbomode: disabled


Just a thought, are you powering the mobo with the 4pin ATX or the 8 pin ATX ? if it's only the 4 pin, try the 8 pin. If it's already the 8 pin, then nevermind

Have you tried double checking that both the video cards 6pin power cables are secure?

Double checked ALL connections are secure?

(sorry to state the obvious, but it can't hurt to be sure)

On the plus side, at least you know more about BIOS settings, RAM and CPU than you did a week ago. Fingers crossed you get it working sooner than later

 

[Klaudius]

First off I want to say, "I'm sorry"
ASUS is both one of the most highly reccomended boards because of features, and one of the most amazingly frustrating boards on the market.

Here is my quote for you "the bleeding edge is no fun when its your blood"

I still have one machine that has an ASUS motherboard, and it has PLAGUED me for about 4 years with random crashes, etc
Driver updates are sketchy, and if you don't speak mandarin Chinese its hard to get support.
Additionally about 50% of the ASUS video cards I have used crash horribly or need to be sent back to the manufacturer.
At one point I just switched to Intel boards just to get away from the pain. But i seriously missed all the features.

Random crashes in my long experience are in 3 major categories, CPU, Video card and sound. Sometimes its memory, but with todays quality control its insanely rare. Also I wouldent even really say windows anymore, as windows is pretty stable these days, if you notice even on these forums it usually comes down to a driver/program from a manufacturer or a piece of hardware failing.

One: CPU
-the CPU being seated wrong
-the cooling unit not being attached correctly(bad seal etc).
-Also there being fur/dust in the cpu fan blocking airflow.
-Cpu fan pin connectors snapped
-CPU fan connectors touching case and welding to it
-CPU fan connectors touching power supply and shorting
-CPU mis detecting type and CPU voltage.
-wrong core voltage setting on CPU
-Bad bios battery out of the box making machine loose settings
-Bad NVRAM chip loosing settings on every reboot
-bad Memory slot, not the chip, but the slot itself.


Two: Video card
-The video card having a heat sink not attached right, fan acutally falling off (ASUS of course),
-Memory frying on the card, (that was a bfg 6800 the back side of the card was smoking and the memory looked like someone shot it). This was out of the box, no overclock.
-Video card not being seated correctly
-Not enough power getting to the video card, (too many drives etc)
-Video card being too close to another card, and not getting enough ventillation.
-Wires from some other device touching the video card and shorting it or getting cought in the GPU fan.
-drivers1, NVIDIA in general is good, but once about every 10 generations of driver, they put out a spectacularly bad driver. So bad in fact it becomes stuff of legends.
-drivers2,ATI drivers are in general bad and in particular horrible, but the hardware is better than NVIDIA for features. If you can get one that runs, don't change the driver if you can help it, you have about a 60% chance your new driver will cause crashes and lockups with ATI.

Three: Sound card
-Soundblaster,(and to be fair this is pre X-fi cards) ......it causes more pain to more people than i can believe. The battlefield sieries alone which was supposed to be soundblaster compatable only supported Hardware acceleration in one out of the 20 chip sets they had available at the time. Add to this generally buggyness and slow updates and you have a contender for pain of the year award.
And the soundblaster laptop cards ? You can forget that. 75% of them that I have installed cause random crashes, and need to be returned.
Notice that these days most people have onboard sound? Thats because it works and its easy to auto configure.
-Not installing the lastest sound card driver, sometimes you can get away with windows version of the driver, but newer drivers generally do not crash the "hardware" sound options on most games.

I cant explain to you how twitchy the ASUS boards seem to be, its like they are tiger in a cage just waiting to shred you at any opportunity. Look up "problems with ASUS motherboard" and just look at the massive amount of hits.

Good luck, I dont even know what to tell you.
I dont even know which type of religion to point you too for the correct prayer.

 

[smarteyeball]

First off I want to say, "I'm sorry"
ASUS is both one of the most highly reccomended boards because of features, and one of the most amazingly frustrating boards on the market.

ATI drivers are in general bad and in particular horrible, but the hardware is better than NVIDIA for features.

lol, yeah. They're a PITA to tame

As a first time ATI user, I can now understand the vitriol people have aginst ATI drivers. I used to sit back and laugh. Now I'm one of them

 

[American197]

Smartey,

Dude, I dunno how anybody could not like Yahoo, he was great!

So many strange things are happening man, I don't even know. This asus mobo is looking more and more like the problem...quite a few people over on the asus forums are having the problems as well. Apparently my buddy found out that you have to unplug your computer from the wall for the voltages to correctly be programed in to BIOS. I mean, really?

I ran INTEL BURN TEST, and it seems to have worked. It just sat there for 30 min, and my temps only went up to the 50s while my CPU load never went past 60%, so (thinking that it wasnt doing its job) I exited out of it...Now when I try it again, it says my computer fails the test within 1 second. My buddy over at asus forums said he had never seen anything like it.

I had tried COH (on XP), yes, and it crashed. However, I left it on all night last night (on pause) and it seemed to work, but I'm not really sure. It somehow restarted my game that I had been playing. Odd. I'm gonna play it throughout the day, and see if it crashes.

LOL, no problem about the obvious. It is the 8 pin, yes. And yes, I've checked my connections over and again.

You know, you're right. I do know a lot more (and am much more comfortable) about BIOS, and computers in general (thanks to your help!)...I suppose that was the point. I guess me having to deal with a bogus asus mobo for a bit is worth it...lol

Klaudius,

Hey, I appreciate all those tips. I mean, this has been ridiculous, but I guess that's what I get for getting ''bleeding edge''. However, if it turns out that my mobo is the culprit (90% sure) I am NEEEVVVEER buying an asus product again. I mean, their American (corporate?) headquarters is like, 7 miles away from my house! I should be able to walk over there and get support...ugh, and their web site...I feel like I'm a fly in a web.

Oh well, I'll keep plugging away. Thanks for your time!

 

[smarteyeball]

Apparently my buddy found out that you have to unplug your computer from the wall for the voltages to correctly be programed in to BIOS. I mean, really?

I'm sorry, What?! If the PC has no power, then how are you supposed to even access BIOS, let alnoe enter settings..... Unless you unplug the PC, leave it sit for about 5 mins and then re-plug it in.

With IBT, you have to select 1) maximum stress and n (no) for error detection. That way each pass will give a resiudal value and residual (norm) value. If any of the residual (norm) values are different, then something is unstable. For best results, you should realy be running it in a x64 OS.

If COH was paused, then chances are the card wasn't at 'full speed'. It may be crashing when it ramps up to it's full clocked speed :huh:

Try not to let one board sour you on a brand My first P5Q PRO board had to be RMA'ed because of a faulty DIMM slot. The second board no problems. Although by now, your frustration/dissapointment levels would be getting pretty high

 

[American197]

Yeah, apparently the BIOS voltages would only take hold AFTER he unplugged his machine and plugged it back in. I dunno, but thats what he said.

I just played COH for a couple of hours, and I thought I had fixed the problem, but then my machine randomly restarted.

To your knowledge, what would cause a random restart?

Oh, and what does it mean if I fail BURN TEST? What does that point to? CPU? MOBO? RAM?

Thanks again,

Sal

 

[smarteyeball]

it sounds like the 'taking hold' was similar to resetting. That sounds more feasible.

I know that feeling that you think you've finally nailed it only to discover you didn't. My commiserations.

As to what caused it, probably the same thing that's been plaguing you all along... In other words, I don't know.

When you say you fail IBT, do you mean that your residual(norm) values aren't matching? I'd lean towards a CPU setting.

What changes did you make that let you play COH for longer? That could be a path to follow.

With IBT running and CPU-Z montoring, what is your voltage maxing to?

For example, this is my chip at idle with C1E enabled:



At speedstep load it rises to 1.256v and when my chip clocks up to 3.5ghz at full load, that voltage rises to 1.264v.

If you know what it's maxing at, you can work out if it's other settings beside the vcore at play. Things like GTL ref settings etc. Although normally, you would only need to adjust them when overclocking.

 

[smarteyeball]

Also, download CORETEMP:
Core Temp

and find out what your chips VID is. VID is the voltage setting that intel sets/recommends for each individual chip to ensure stability at that chips stock speeds. ie Even though 3 people have a 17 920, they may also have different VID's.

For example, mine is unfortunately 1.2500. (The lower the VID, the higher the voltage/overclock headroom)



Once you know your VID, enter that value into BIOS in your Vcore. That way, it's set to run stable at stock speed according to intel manufacturing.

 

[American197]

LOL, thanks for your co-misery...I'm right about to revert to an older BIOS, because I think mine is reporting voltages incorrectly...(its been happening to people on the asus forums)...

I've just downloaded a great program by CPUID, makers of CPU-Z, called CPUID Hardware Monitor. However, when revewing its specs, it says my manually set DRAM voltage is 1.66 when I set it to 1.5 in BIOS, and we all have heard about the 1.65 'limit' for the x58 chipset... Everything else matches up perfectly, so I trust the program over my BIOS...thoughts?

I have COREtemp, and have never noticed that VID before...now I know why...it's (for some reason) greyed out on my COREtemp. Yep, VID and 'Revision' are greyed out. Thoughts?

Oh, I deleted BURN TEST, reinstalled, and fired it up again, and passed 5 tests at the hardest level. My temps never went over 60, which is super odd to me, but good, right? I'm gonna run it for maybe 15 passes, and then see what happens...

My CPU is set to 1.23v, and it reads at a constant 1.2v. It doesnt change when under load. I'll put it to auto and test it out, and get back to you. *EDIT*...just tested it, and it was around or just under 1v at idle and 1.2v solid under full load...I changed my DRAM voltages to auto as well and they stayed at 1.66v...that worries me...any particular program you use to check your DRAM voltages? The only one I've found is Hardware Monitor...

Ps, thanks for taking the time to copy and paste those highlighted windows, I appreciate it!!

 

[smarteyeball]

Forgot to link HWMONITOR originally. It was one the same d/l page as CPU-Z anyway It's a good app.

The greyed out VID in CORETEMP may be because the app isn't updated to read i7's yet.

Everest *should* report your current RAM voltages if I remember correctly (I'm in W7 atm and have it installed)

60c in IBT is pretty good temp wise. Remember, you aren't overclocked and don't have a stock cooler What temp loads are you getting with p95?

That 1.66v sounds like overvolting to get those lower 8-8-8-20 timings. It's odd that it won't stick at 1.5v.

Maybe you need to start from scratch. ie Clear the CMOS and re-enter all your settings manually. Check your MOBO manual. There should be a jumper on the board that you change for about 10secs which clears all the settings.

(Switch off, unplug power -this is where that unplugging trick makes sense- change the jumper and enter values again.)

 

[American197]

""(I'm in W7 atm and have it installed)"...I'm confused, whats W7 and atm? LOL, Noobishness strikes again!

You're probably right. I may just format everything clear CMOS and reset my BIOS... or I'll be lazy (can I call myself lazy after all this? lol) and just clear CMOS...what do you think? Any benefit in wiping the OS?

I downloaded EVEREST, but it was only a trial version, and did not (from what I saw) have DRAM voltages in there...they had 3.3v, 5v, and 12v, with 12v being greyed out.

Anyways, I'll go ahead and clear the CMOS, and revert to a different BIOS, test it, and get back atcha

Peace,

Sal