Sign in with
Sign up | Sign in
Your question

Possible temperature problem with E6600?

Last response: in CPUs
Share
May 23, 2007 5:34:17 AM

Hi,

I recently got E6600 with Asus Striker Extreme motherboard and Scythe Ninja heatsink. I've been reading the forums and I think my temperatures are rather high for idle values.

In Windows XP, IDLE the temperatures are: Tjunction 85C and Cores 52C/53C, measured with Core Temp v0.95. Intel's TAT does not work for some reason (Error Enumerating On Demand Clock Modulation support -error). Asus PC Probe II and Ai Booster show the temperature 47C for CPU.

I currently run the CPU at 1.400V core (from BIOS) and 350 MHz FSB, so 3150 MHz cpu. At any lower core voltage the CPU will not boot to Windows, it crashes at the spash screen or after that. I just actually noticed that 1.400 voltage is not enough for ORTHOS' Small FFTs - stress CPU to run for long without errors. With BIOS' automatic volatege setting for CPU core ORTHOS was stabile.

Do you think that the temps are high for normal IDLE use? Do you think that raising the voltage even higher or using automatic setting in BIOS is good solution for these FSB settings?

Thanks!
May 23, 2007 6:49:16 AM

I've got Antec Nine Hundred case. In the case, the big 20 cm fan above the CPU and the 12 cm fan behind the CPU are both on, both of these fans are taking the heat out of the case. The PSU is Nexus 500 W NX-5000, which also has 12 cm fan blowing the heat out.

The graphics card is Radeon X1300, which is highly inefficient :)  and not producing that much heat. Also got 2x Samsung Spinpoint HDDs installed.

Maybe the ambient temperature, i.e. the temperature difference inside the case, reduces the cooling efficiency so much that the CPU will not cool enough. I can try turning one one or two of the additional fans, which take some cool air inside the case and report how this affects the temperatures.

But is it normal to set such high voltage (more than 1.4V) to get the CPU to work stabily at 3150 MHz for example? CPU-Z shows that the core voltage is only 1.264 V. Does CPU-Z show the real voltage or is the BIOS setting of 1.400 V really the actual voltage, or just an upper limit?

Many thanks!
May 23, 2007 9:37:34 PM

I'm at 1.42 for 3750. If I remember correctly on stock settings my 6600 does almost 3300 (got a 47gig conversion running or I'd do a reset to check it again). When the air in the house is 60 my idle temp runs 2 degrees cooler than when it's 72, have no idea what the temps are in the case. Thermaltake matrix case.
Related resources
May 24, 2007 7:48:33 AM

If your idle temps are 52/53 what are the load temps? What are you hardware spec (case, HS, fans in the case, etc.) and what is room temp? If you're getting errors in orthos your o/c is not stable @ that voltage, either up the vcore or drop of o/c. Don't use that aisuite software for anything particularly monitoring temps. Coretemp and RMClock are much better with RMClock being best in my opinion as CT has been known to give inaccurate readings.

Bottomline is without seeing your load numbers, idling that warm is too high.

Be warned that all C2D and quad C2D chip are known to have irregularly shaped IHS (internal heat spreaders) meaning good contact between the HS and core can be difficult. You can consider lapping both the base of the HS and the IHS on your chip to give better temps. I knocked 7 °C off my load temps from the coolest core and over 10 °C off my load temps from the hottest core by lapping both the IHS on the Q6600 and the base of the ultra-120 extreme. See pics and temp graphs in those two threads.
May 24, 2007 12:41:25 PM

I just tested on ORTHOS Small FFTs stress CPU and the load temps were 72/72 Celsius when measured with Core Temp.

Some specs
Heatsink: Scythe Ninja
Case: Antec Nine Hundred (top, back (air out) and 1 front (air in) fans on)
Room temp: 25 C when measured in front of the case
MB: Asus Striker Extreme (newest bios)

I tested the setup with 8x core multiplier and 1600 MHz FSB (QDR) so 3200 MHz core, while memories are at 800 MHz 4-3-3-8. So this should be 1:1 memory FSB speed. Core voltage is now manually set to 1.450 V though CPU-Z says its still 1.264V. Idle temps are 53/53 and the load is the same as above.

Maybe the IHS is not flat at all... Do I need to protect the CPU somehow if I would like to lap it? And are the load temps dangerously high?
May 24, 2007 12:58:05 PM

I'm thinking perhaps that the Scythe Ninja is not fairing to well with the OC.

Still... your temps are still doable (okay range, though it could be better) from what I understand.

Hard to understand what the airflow is doing. That PC case does pack some fans though.

Can you show some pics on the case on how the Scythe Ninja is oriented?

Also, recently I found out speed step will run good (from what other guys say). I've even tried it with my dad's E4300, which is prime stable with speed step on. It still messes with my mind, seeing the speed jump. But since I'm over clocking at a lower speed, it really doesn't make that much of a difference to use it. But I left it on to see if my dad complains of anything. :lol: 

With speed step on (EIST & C1E), it will reduce your 3.2 to 1.8 ghz idle. In doing that, your idle temps should be lower (to help the Scythe Ninja), and when you put a load on the CPU, it should jump back to 3.2 ghz (though your going to be back at the lower 70's temps, perhaps). I think the E6600 would prolly benefit from it, even power consumption wise.

Edit:

The lapping can help reduce the temps, though... I think I'm too timid to try that at this time. :lol: 

Edit:

Someone did start a thread here:

Lapping C2D

App403 did provide some links that might help you out.
May 24, 2007 1:55:06 PM

Quote:
Thermal Design 60.1°C @ 0.850V-1.3525V Intel® Core™2 Duo Desktop Processor E6600

I don't know if you wanted this info or not . . . :?: You need a Striker OC Guru . . .

My $.02: Can you drop the RAM ratio & voltage and up the FSB speed (allowing you to drop the cpu voltage) ???


I think that spec is for the Tcase (what I underlined below), not the Tjunction temp. Though I could be wrong.

Quote:
Thermal Specification: The thermal specification shown is the maximum case temperature at the maximum Thermal Design Power (TDP) value for that processor. It is measured at the geometric center on the topside of the processor integrated heat spreader. For processors without integrated heat spreaders such as mobile processors, the thermal specification is referred to as the junction temperature (Tj). The maximum junction temperature is defined by an activation of the processor Intel® Thermal Monitor. The Intel Thermal Monitor’s automatic mode is used to indicate that the maximum TJ has been reached.


The thermal trip is 85C or (mobile CPU 100C) for the Tjunction, if memory serves.

Edit:

Taken from the Temp guide:

Quote:
Scale

The temp scale shown below illustrates the normal ~ 25c Delta between Idle and TAT @ 100% Load, and the typical ~ 15c Delta between Tcase and Tjunction on an example system overclocked with 1.4 Vcore. 50c Tcase and 65c Tjunction are safe and sustainable temperatures.

Tcase/Tjunction

--60--/--75--75-- Hot
--55--/--70--70-- Warm <--he somewhere here
--50--/--65--65-- N
--45--/--60--60-- O
--40--/--55--55-- R
--35--/--50--50-- M
--30--/--45--45-- A
--25--/--40--40-- L

--20--/--35--35-- Cool


So he is sitting right at that line. :lol: 
May 24, 2007 7:59:01 PM

I dunno if I agree with the temp guide if you want your CPU to last longer. I'd say don't exceed the Intel number. If you don't care about the CPU, go w/ the guide. I'm guessing most people don't keep their PCs for longer than 3-4 years. It'll probably last that long running at 72C...?
May 24, 2007 9:06:29 PM

Just saying it's doable. The thermal shut down should help prevent the worse damage that heat can do to it. If 72C doesn't shut it down, it should be able to handle it, though it doesn't have to run that way with better cooling. Throttling should also kick in around.. 75C?

Even a Northwood core will shut down at 89-90C from what I remember, though it is a single core at 0.13 um.

If the guy could improve the cooling, I have no argument in getting it done.
May 25, 2007 7:46:35 AM

Quote:

Hard to understand what the airflow is doing. That PC case does pack some fans though.

Can you show some pics on the case on how the Scythe Ninja is oriented?


There is a pic inside the case below. I also put the directions of the air flows in that.

I was wondering if the HS could be turned 90 degrees clockwise, so that it would blow the hot air towards the big fan at the top... It just looks like it might not fit that way so I abandoned the project earlier. And the 12 cm fan behind the CPU should take the heat out anyways.


Picture of Antec Nine Hundred case with Scythe Ninja HS and Nexus NX-5000 power. MB Asus Striker Extreme, CPU E6600 @ 1.45 V 8x 400 MHz CPU
May 25, 2007 3:00:51 PM

Heh.. cable management could help, though I could understand that may be a pain, since the PSU is at the bottom.

Having 2 fans (well 3) so close kinda bothers me, pressures involved may hinder air flow.

That rear exhaust (120mm blue one), when you put your hand back there to feel air flow, does it deflect air back in?

The PC case I got for my dad, the rear 120mm exhaust was deflecting air back into the case because of the grill holes. I ended up cutting out the grill, which in turn stopped air being deflected back into the case. Thing is, I tested the air flow with nothing in the case, and use the PSU just to power up the case fans.

I have another example for air flow, in having too many fans can make things worse. I used to have a slot fan for my 6800 GS. With the fan below the video card, it did cause the RPMs on the card to go up. Which pretty much meant that they were too close together. And by removing the slot fan, the GPU ran cooler (around 3-4C cooler).

I would advise trying different configurations with your fans, like:

1.) turn or taking off the CPU fan. see what the 2 exhaust fans will do.

2.) turning off the rear exhaust - see what the cpu fan and top fan do alone.

3.) turning off the top fan - see what the cpu fan and rear exhaust does.

Just tinker around to see what does the best, even see if the HS can be turned to let the top fan pull air through it or what not.
May 25, 2007 7:06:49 PM

Quote:
Maybe the IHS is not flat at all... Do I need to protect the CPU somehow if I would like to lap it? And are the load temps dangerously high?


Since the C2D chips don't have any pins, you just want to take care to keep it dry while lapping. Also keep your hand dry. Wipe the IHS with a slightly damp cloth or paper towel such that it doesn't leave any drips of water.

Let us know what you end-up doing. Also, I think any core temp over the Intel spec is too high. 72 is over 61 so yeah, lap away and see how the temps improve.
!