Overclocking a Q6600 on Asus P5KC

Howard Goldstein wrote:
> On Mon, 12 Nov 2007 02:22:55 +0100, Herzog <her.DeleteThis@zog.it> wrote:
>> Hi,
>> as subject... it runned at 3 GHZ nicely for a while, but suddently
>> it started rebooting as soon as Windows loaded with a blue screen
>> and c000021a error. I suspect it has something to do with voltage
>> leaks since it started
>
> "Voltage leaks" as in "ground loop"?
>
>> giving error when i removed and reinstalled a KVM switch.. as soon
>> as i plugged it this thing started. Another time it restarted when
>> i plugged LAN cable in, and some times it started without
>> reckognizing mouse. I reinstalled XP twice but it kept doing the
>> same... my vendor solved the problem re-installed XP and lowered
>> overclock a bit, now it seems it runs fine at FSB 330 (instead of
>> 333), he put also some more fans so it runs nicely at 55° at full
>> CPU usage. Any advice to get it even more safe ? i just change FSB,
>> should i also change some other value ?
>
> 55C at full load sounds terrific.

Indeed.

> I don't have the answer but I have experienced similar instability at
> 333x9 after weeks of great stability, but on a striker extreme, not a
> P5K.

I find this disconcerting (although I don't have a Qxxxx). Is there
something that cases Core 2 CPUs to become less overclockable with time?

> I'm back down to 300x9 for the time being but in the meantime
> have completely reworked the chipset cooling -- the 680i combination
> and default cooling left the sinks on the two chipset controllers far
> too hot to touch

I can't recommend the iCute S901 case enough for cooling of all your
ancilliary devices. I don't know if it's available in continental America
(if indeed that's where you are) but it's cheap and does an excellent job of
supplying airflow to the NB, SB and MOSFETs. You'll see it in this short
review of (3) cases:

http://pcworld.co.nz/pcworld/pcw.nsf/reviews/56FC71DAD9E5C67BCC2571ED000C70D2

It's cheap and cheerful and does an excellent job. No need for extractor
fans as the positive pressure in the case is high. The back of the case is
well-perforated to allow the air out freely.

> - and replaced a 430W power supply with a 750W
> supply that supposedly has improved ripple.
> sure.
>
> (I see around 42-50C across the 4 cores while idling and 1.325v vcore,
> peaks at 62C on the hottest core under full load)

Sorry Howard, I forget, are you water cooling or air? I've seen such high
idle temps mentioned before, perhaps it's just a feature of the Qxxxx
series? Or water cooling? (Less quick to respond to die temp changes due to
the resevoir of heat). My E4500 @ 3.3GHz, air cooled with a good cooler,
might hit ambient + 34°C under TAT, but drops to ambient + 5°C within 30
seconds of idling.
--
Cheers,
Shaun.
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC 

This message is not archived
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC 

Howard Goldstein wrote:
> On Thu, 15 Nov 2007 13:24:19 +1300, ~misfit~ <misfitnz DeleteThis @yahoot.com.au>
> wrote:

<snipped>

>>> I'm back down to 300x9 for the time being but in the meantime
>>> have completely reworked the chipset cooling -- the 680i combination
>>> and default cooling left the sinks on the two chipset controllers
>>> far too hot to touch
>>
>> I can't recommend the iCute S901 case enough for cooling of all your
>> ancilliary devices. I don't know if it's available in continental
>> America (if indeed that's where you are) but it's cheap and does an
>> excellent job of supplying airflow to the NB, SB and MOSFETs.
>> You'll see it in this short review of (3) cases:
>>
>>
>> http://pcworld.co.nz/pcworld/pcw.nsf/reviews/56FC71DAD9E5C67BCC2571ED000C70D2
>>
>> It's cheap and cheerful and does an excellent job. No need for
>> extractor fans as the positive pressure in the case is high. The
>> back of the case is well-perforated to allow the air out freely.
>
> Wow look at the size of that fan. And here I thought the 18cm or so
> fan on the Antec 900 was pretty big. Is it as noisy as it looks like
> it'd be? Yes I am in NA, I don't see it at the usual suspects over
> here though (yet).

It's not noisy at all. It runs at about 450rpm and is all but inaudible.
However, it sure moves some air!

>>> - and replaced a 430W power supply with a 750W
>>> supply that supposedly has improved ripple.
>>> sure.
>>>
>>> (I see around 42-50C across the 4 cores while idling and 1.325v
>>> vcore, peaks at 62C on the hottest core under full load)
>>
>> Sorry Howard, I forget, are you water cooling or air? I've seen
>> such high idle temps mentioned before, perhaps it's just a feature
>> of the Qxxxx series? Or water cooling? (Less quick to respond to
>> die temp changes due to the resevoir of heat).
>
> I'm air cooling -- poorly at that, Phil mentioned it a couple of weeks
> ago and I'm pretty sure he's right (*). Here's where I am with a
> right now after replacing one of the 12cm Antecs in the P182B with a
> quieter but more powerful Scythe (2 degrees better with that fan
> swapout)
>
> dev.cpu.0.temperature: 48
> dev.cpu.1.temperature: 48
> dev.cpu.2.temperature: 42
> dev.cpu.3.temperature: 45
>
> (*)Ambient is 29, it's 38 inside the case, poor ventilation in this.

Ahhh, yeah, porr case ventilation is one thing I *won't* abide. What's the
point of buying an awesome CPU cooler and sticking it in a hot-box?

> The cooler is an Arctic Cooling Freezer 7 pro (only US$19, or 15 euros
> in real money) that has a deeply grooved base manufactured in a way I
> just can't figure out how to lap the silly thing because the 4 ears
> that secure the cooler to the PCB are affixed to the base and extend
> below it blocking and they're a lot narrower than my piece of glass

Get a narrower piece of glass?

> Check out the difference between pairs of CPUs! Loaded or unloaded
> (above is unloaded) cores #0 and 1 are consistently cooler by 7-10
> degrees. Do you think I might have a bum quad core with a funky
> internal connection to the heatspreader? The cooler is definitely
> flat against the cpu. The o/s should be tossing processes at whatever
> CPU is ready and loading them all equally over time.

There is usually a 3 - 4 degree difference between the temps of my cores.
One is always hotter than the other (except with TAT running, they're both
the same or within a degree then). I put it down to variations in the
thermistor or, less likely due to their verey close proximity (and the fact
that my two cores are both on the same bit of silicon), the contact with the
heat-spreader thing. Thermistor methinks.

>> My E4500 @ 3.3GHz, air cooled with a good cooler,
>> might hit ambient + 34°C under TAT, but drops to ambient + 5°C
>> within 30 seconds of idling.
>
> That sounds great. Is that in the iCute case?

Hehee! It *is* great. Yes, it's in the iCute. I doubt I could keep the temps
that low in a different case unless I modded the case to duct room air
directly to the cooler as I did with previous builds/cases. Where the iCute
wins out over the ducting method is that it cools the whole mobo. Very well
too I might add. Also, not only is this the most powerful machine and
highest OC I've had in a while, it's also the quietest by far.
--
TTFN,

Shaun.
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC 

This message is not archived
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC 

Howard Goldstein wrote:
> On Thu, 15 Nov 2007 16:08:16 +1300, ~misfit~ <misfitnz.TakeThisOut@yahoot.com.au>
> wrote:
>> Howard Goldstein wrote:
>>> On Thu, 15 Nov 2007 13:24:19 +1300, ~misfit~
>>> <misfitnz.TakeThisOut@yahoot.com.au> wrote:
>>
>> <snipped>
>>
>>>>> I'm back down to 300x9 for the time being but in the meantime
>>>>> have completely reworked the chipset cooling -- the 680i
>>>>> combination and default cooling left the sinks on the two chipset
>>>>> controllers far too hot to touch
>>>>
>>>> I can't recommend the iCute S901 case enough for cooling of all
>>>> your ancilliary devices. I don't know if it's available in
>>>> continental America (if indeed that's where you are) but it's
>>>> cheap and does an excellent job of supplying airflow to the NB,
>>>> SB and MOSFETs. You'll see it in this short review of (3) cases:
>>>>
>>>>
>>>> http://pcworld.co.nz/pcworld/pcw.nsf/reviews/56FC71DAD9E5C67BCC2571ED000C70D2
>>>>
>>>> It's cheap and cheerful and does an excellent job. No need for
>>>> extractor fans as the positive pressure in the case is high. The
>>>> back of the case is well-perforated to allow the air out freely.
>>>
>>> Wow look at the size of that fan. And here I thought the 18cm or so
>>> fan on the Antec 900 was pretty big. Is it as noisy as it looks
>>> like it'd be? Yes I am in NA, I don't see it at the usual suspects
>>> over here though (yet).
>>
>> It's not noisy at all. It runs at about 450rpm and is all but
>> inaudible. However, it sure moves some air!
>
> I bet. I'll look for it out this way.

It's worth looking for, believe me.

>>>>> - and replaced a 430W power supply with a 750W
>>>>> supply that supposedly has improved ripple.
>>>>> sure.
>>>>>
>>>>> (I see around 42-50C across the 4 cores while idling and 1.325v
>>>>> vcore, peaks at 62C on the hottest core under full load)
>>>>
>>>> Sorry Howard, I forget, are you water cooling or air? I've seen
>>>> such high idle temps mentioned before, perhaps it's just a feature
>>>> of the Qxxxx series? Or water cooling? (Less quick to respond to
>>>> die temp changes due to the resevoir of heat).
>>>
>>> I'm air cooling -- poorly at that, Phil mentioned it a couple of
>>> weeks ago and I'm pretty sure he's right (*). Here's where I am
>>> with a right now after replacing one of the 12cm Antecs in the
>>> P182B with a quieter but more powerful Scythe (2 degrees better
>>> with that fan swapout)
>>>
>>> dev.cpu.0.temperature: 48
>>> dev.cpu.1.temperature: 48
>>> dev.cpu.2.temperature: 42
>>> dev.cpu.3.temperature: 45
>>>
>>> (*)Ambient is 29, it's 38 inside the case, poor ventilation in this.
>>
>> Ahhh, yeah, porr case ventilation is one thing I *won't* abide.
>> What's the point of buying an awesome CPU cooler and sticking it in
>> a hot-box?
>
> Point taken. Here's the before and after with the side off and a 12cm
> fan blowing up a typhoon into the case:
>
> case closed
> dev.cpu.0.temperature: 45
> dev.cpu.1.temperature: 46
> dev.cpu.2.temperature: 38
> dev.cpu.3.temperature: 42
> Thu Nov 15 10:12:39 EST 2007
>
> open w/ blower
> dev.cpu.0.temperature: 43
> dev.cpu.1.temperature: 43
> dev.cpu.2.temperature: 37
> dev.cpu.3.temperature: 39
>
> Every degree or two counts. I have another quiet Scythe with a lot of
> ooomph to add to the box. Hoping that'll do it

It could.

>>> The cooler is an Arctic Cooling Freezer 7 pro (only US$19, or 15
>>> euros in real money) that has a deeply grooved base manufactured in
>>> a way I just can't figure out how to lap the silly thing because
>>> the 4 ears that secure the cooler to the PCB are affixed to the
>>> base and extend below it blocking and they're a lot narrower than
>>> my piece of glass
>>
>> Get a narrower piece of glass?
>
> I thought I could cut the one I have now but can't find the glass
> scoring doohickey anywhere.

Don't make the mistake I have. "Old" glass is hard to cut, if you're
intending on cutting glass it's best if you're using new stuff. It changes
on a molecular level over time.

> I confess that I think I'll be
> disappointed if I do lap the thing. The cooler's fins are stacked
> together very tightly -- there's no more than 1/16" between them.
> What little air doesn't bounce off the fins takes a shortcut out the
> side of the open-frame fan. And the base, well the many grooves are
> quite pronounced. Arctic Cooling claims its to improve compatibility
> with their MX-2 goop. Because of all this it's probably not worth the
> effort.

Hmmm, that cooler got reasonable reviews (from the quick look I just had). I
do a great deal of modding of coolers/cases etc. Perhaps you could replace
the fan with a more 'directed' one, even shroud it a little? Hot glue is the
best thing since sliced bread.

> In my quest for the unachievable (performance but silence) some wee
> hours browsing turned up a couple of candidates, the Noctua UH-12F and
> the Scythe Mine rev B. I continue to seek advice and recommendations
> from you and others here in the froup

I'm afraid I have little experience with coolers other than my own, the
Thermaltake Mini Typhoon. It seems quiet to me and is bloody efficient.

>>> Check out the difference between pairs of CPUs! Loaded or unloaded
>>> (above is unloaded) cores #0 and 1 are consistently cooler by 7-10
>>> degrees. Do you think I might have a bum quad core with a funky
>>> internal connection to the heatspreader? The cooler is definitely
>>> flat against the cpu. The o/s should be tossing processes at
>>> whatever CPU is ready and loading them all equally over time.
>>
>> There is usually a 3 - 4 degree difference between the temps of my
>> cores. One is always hotter than the other (except with TAT
>> running, they're both the same or within a degree then). I put it
>> down to variations in the thermistor or, less likely due to their
>> verey close proximity (and the fact that my two cores are both on
>> the same bit of silicon), the contact with the heat-spreader thing.
>> Thermistor methinks.
>
> Hmmm my E6750 is always within 1C and the E6600 (may it rest in
> pieces) was too, unloaded. Another difference between the dual core
> E4XXX and E6XXX?

Could well be.
--
TTFN,

Shaun.
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC 

"Howard Goldstein" <hgoldste.DeleteThis@mpcs.com> ha scritto nel messaggio
news:1195071989.32495@news.queue.to...
> 55C at full load sounds terrific.

Air cooling but very effective, fully taking advantage of all Colermaster's
case air passages.
original intel CPU fan plus 3 case fans.

> I don't have the answer but I have experienced similar instability at
> 333x9 after weeks of great stability, but on a striker extreme, not a
> P5K. I'm back down to 300x9 for the time being but in the meantime
> have completely reworked the chipset cooling -- the 680i combination
> and default cooling left the sinks on the two chipset controllers far
> too hot to touch - and replaced a 430W power supply with a 750W
> supply that supposedly has improved ripple.
> sure.

power is 750 W here too.
Anyway techicians didnt fund any problem here, so it's same PC and now it
runs.
It just seems that some KVM or lan cable connection did a temporary voltage
peak, and probably OC'd systems are quite more sensible to that.
Anyway, from now on i'll use all my nodes with a remote management software,
KVM times are past here.

H
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC 

This message is not archived
 >> Stay informed about: Overclocking a Q6600 on Asus P5KC