bluescreen - twice

[briligg]

ok, i'm confused... cpu-z shows all kinds of memory data, but neither the number 1066 or 1600 appear anywhere.

so, you're saying i can't just put 'memory up' to 1600mhz and forget about it...

 

[streetwolf]

534 is really twice that; 1068 (1066). you multiply this value by 2.

here's mine:

 

[streetwolf]

show me the CPU screen.

 

[briligg]

ok...

 

[streetwolf]

Notice that you are running less then 2.67Ghz? That's because you probably have EIST (Speed-Step) enabled in the BIOS. That lowers your multiplier thus lowering CPU speed when the CPU is doing little work. Most overclockers disable this as well as a few other similar options.

You really need to read the manual and check out web sites to understand how to o/c the i7920. It's more complicated then previous Intel processors.

Use the canned speeds, 920, 940 etc. This will adjust your other parameters to the proper values. Once you understand what's going on you can set things to MANUAL and change all the settings to your liking.

 

[briligg]

hm.
i've got some reading to do then... i had hoped there would be a quick fix.

i'm going to look around for a simple set of changes i can make to get my system stable.

is there some way i can check system stability? i mean, maybe the mb was set at all default before. maybe it was never oc'ed.

 

[streetwolf]

get Real Temp 3.0. It uses Prime95 to stress your machine. Both are free.

Aren't you stable at the default values? You should be.

Here's a few links to digest:

http://hothardware.com/Articles/Overclocking-Intels-Core-i7-920-Processor/

http://www.xbitlabs.com/articles/cpu/display/core-i7-920-overclocking.html

http://www.tweaktown.com/articles/1672/stock_to_clocked_core_i7_920_at_3_8ghz/index.html

Once you get an understanding of the 920 PM me if you have any questions.

 

[briligg]

thanks for all your help so far. i installed pc probe ii from the rampage cd, and it's been giving me temp warnings for the mb and nb. and i tried using 'memory up', briefly, setting it at 1600 mhz, since they say other values will be adjusted accordingly, and then installed pc probe, and it gave me the temp warnings and 2 or 3 voltage warnings too before i shut down and restored the default settings.

my case is a minitower, it has two noisy 80mm fans that run all the time. my comp guy is getting a regulator made for them, they don't have the 3 leads necessary to plug into the mb's monitoring program. and i am getting a proper 120 case fan to put in the side of the case, but there isn't much room. the cpu really hasn't done any demanding work so far, i really didn't think there would be an issue with these things.

 

[briligg]

alright, i downloaded real temp and prime95. i started running the small fft sensor test and it caused a bluescreen within seconds, with the same clock interrupt error. i booted up and ran a small fft test from within prime95 and let it run for 16 min. i stopped it because core 0 had reached 96 deg., which made me uneasy. but there hadn't been any errors or warnings.

i'm running memtest right now, 3 copies testing 1500 mb each, and there is about 700 mb unused. they are at 220% coverage so far, no errors.

i googled 'clock interrupt error', and on justanswer.com someone with this problem was advised to *clear cmos, *check bios has the right voltage set for the ram, and *check clock speeds for components are set correctly in bios. it mentioned other things i've already done, too, like updating bios and setting it to default values. so in a bit i'm going to try those things.

there is clearly an overheating issue under load, and i'm going to get that looked after, but really the bsod's i've had up til now all occurred when the machine wasn't doing anything demanding. (except for the prime95 bsod).

and the thing that strikes me as really the strangest of all - i can't find the dump files. they don't seem to exist.

​

 

[briligg]

okay, after reading a bit about clearing cmos, i guess that's not a good idea... the guy on justanswers with the same bsod errors has the same memory as me - ocz pc3 12800, maybe it's just running hot??? it's located under the dvd drive, bad airflow there, case fans are at the other end of the box. memtest doesn't show any errors so far, after 300% coverage (3 copies @ 1500mb, now running 3 @ 1800 mb). i took the side off the case, for now. this ram has a spec of 1.65V, but is that the recommended setting or an upper limit? my bios has dram bus at 1.5V.

still concerned about the lack of minidumps... what if this keeps happening, and i never have minidumps...

cpu cores at 70 to 75 degrees with memtest running...

 

[zigzag3143]

even if no minidumps there should be some mention of an app failing in event viewer

 

[briligg]

0 events hardware, ie, key management service, media center, win powershell. 3 events antivirus, information level. nothing shown on the different diagnostic pages...

 

[SquonkSC]

If BSOD is invoked by ram there is nothing to dump.

Did the compguy think of increasing the voltage of the ram when he overclocked it?

greetz

 

[briligg]

i haven't had a chance to talk to him, i don't know what he did. i'm going to see him today.

but it isn't overclocked now, and i did get a bluescreen as soon as real temp started it's stress test...

 

[SquonkSC]

i haven't had a chance to talk to him, i don't know what he did. i'm going to see him today.

but it isn't overclocked now, and i did get a bluescreen as soon as real temp started it's stress test...

Sounds like your ram is damaged by the overclocking. Hope I'm wrong.

Good luck.

 

[briligg]

eek!
we removed the dvd drive that was right over it, now it has plenty of head room. especially since the side of the case isn't on right now. and now there is room to put a really oversize fan into to side (after which i'd really like to take out the two noisy, bitty fans at the back). the good news is, initial testing with memtest doesn't show any errors so far... now i've done everything i can to keep the ram cool, i'll put memtest back on and let it run a good while...

so, 1.5 V for the ram is good? how can i monitor it's temp?

 

[SquonkSC]

eek!
we removed the dvd drive that was right over it, now it has plenty of head room. especially since the side of the case isn't on right now. and now there is room to put a really oversize fan into to side (after which i'd really like to take out the two noisy, bitty fans at the back). the good news is, initial testing with memtest doesn't show any errors so far... now i've done everything i can to keep the ram cool, i'll put memtest back on and let it run a good while...

so, 1.5 V for the ram is good? how can i monitor it's temp?

Everest Ultimate Edition.

Type everest in google.

greetz

 

[briligg]

ok, everest is pretty complete... but i don't see physical data like temperature...
i had pc probe on earlier, it gave me temps for nb and sb, don't remember if it had ram... i'm going to reinstall it...

i'm sort of swimming in data here, but i'm going with the idea that the bsods were caused by overheated ram - that seems to fit everything. i'm not even going to stress test the system until i get better cooling installed. memtest was going fine, but it was causing the cpu temp to slowly creep up. i shut it down with core 0 at 76 deg c, after about a half hour running.

solving that mystery about why there are no minidumps made everything fall into place. thanks squonksc, for that. streetwolf has also been a big help. i'll check back once we've resolved the heat issue and i've run a proper stress test. cheers, all.

 

[SquonkSC]

ok, everest is pretty complete... but i don't see physical data like temperature...
i had pc probe on earlier, it gave me temps for nb and sb, don't remember if it had ram... i'm going to reinstall it...

i'm sort of swimming in data here, but i'm going with the idea that the bsods were caused by overheated ram - that seems to fit everything. i'm not even going to stress test the system until i get better cooling installed. memtest was going fine, but it was causing the cpu temp to slowly creep up. i shut it down with core 0 at 76 deg c, after about a half hour running.

solving that mystery about why there are no minidumps made everything fall into place. thanks squonksc, for that. streetwolf has also been a big help. i'll check back once we've resolved the heat issue and i've run a proper stress test. cheers, all.

Everest does have temps sensors for everything, including ram.

Read my PM, I can show you how.

 

[briligg]

update - i got the everest sidebar running (didn't notice that option earlier), and the memory timing was 7-7-7-20. the bios must set it that way on default. ocz recommends 7-7-7-24, so i reset it to that. i didn't change anything else.

my comp guy says he never oc'ed. he checked asus's website and my memory is rated as compatible with this mb, so he left it at that. he advised checking each memory module by putting in one at a time and testing it. sounds reasonable, but if i don't get another bsod i'm going to leave that - kind of a bother. if i get problems after i put in proper cooling and do a stress test, i'll do it. he also felt that although a ram problem could certainly explain the missing minidumps, a software issue is also possible. he asked for the error code on the blue screen - but there isn't one, the only numbers are the technical info #s at the bottom, which i assume is not what he meant. i guess that points even more to the ram.

oh, and still no temps for the ram. maybe my mb doesn't have a sensor for that. nb, sb, and gpu are always a little warm, 70 - 75 deg. c.