I've been having a problem with my CPU temps being way too high since I built this rig. My Q6600 with a mild overclock to 3.0GHz currently sits around 50-55 idle with a Zalman CNPS9700 (using the supplied STG-1 thermal grease and applicator)! I'm fairly certain it's not a case issue as temps only drop 3-4 degrees if I run it with the side of the case off. Everything is housed in a P180B with 4 120mm fans (1 in the PSU region, 1 up front, 1 in the rear, and one on the top of the case). Anyone have any ideas? Is it an issue of thermal grease? I've reapplied 3 times with no improvements, but each time I spread it on using the brush applicator that comes with the STG-1.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Q6600 temps way too high!
- Thread starter nite0
- Start date
Joe Average
Ad Blocker - Banned
- Joined
- Apr 6, 2008
- Messages
- 15,459
While I can't speak for being a cooling expert, I can tell you that "spreading" any thermal compound on a CPU is a bad idea, period. It's the biggest mistake that most people will make, especially those that are new to PC building, etc. Here's how you apply it on a quad:
Taken from this guide:
http://www.arcticsilver.com/pdf/appinstruct/as5/ins_as5_intel_quad_wcap.pdf
Seriously, there really isn't any better way to do it.
Spreading it out sometimes causes major issues, and it's really as simple as that explains: a single line of compound, as close to that particular size as you can get, and as thin - when the HSF is applied, it'll squeeze together and the paste will expand across the heatspreader and based of the HSF and fill in as required. Smoothing or spreading or heaven forbid brushing it over the surface isn't going to help nearly as much as letting simple physics - the compression on the compound caused by the two planes pulled together - can do.
Also, lapping that Q6600 is a relatively simple task and can lower your idle temps 5-10C, easily. Here's a thread with info from PCPer:
http://forums.pcper.com/showthread.php?t=441410
And just for the record, my Q6600 G0 with the stock Intel HSF (and AS5 applied exactly as those pics show) runs idle at ~48C if I push it to 3.2 GHz (8x400) but it's stable, without issues, I just choose not to run it at that speed. I actually just got some 400, 800, and 1000 grit sandpaper yesterday to lap it sometime this week, should prove useful. So if my stock Intel HSF keeps it at just under 50C at 8x400, I'm quite positive yours can do a helluvalot cooler since that HSF you have is way more efficient.
Good luck...
Taken from this guide:
http://www.arcticsilver.com/pdf/appinstruct/as5/ins_as5_intel_quad_wcap.pdf
Seriously, there really isn't any better way to do it.
Spreading it out sometimes causes major issues, and it's really as simple as that explains: a single line of compound, as close to that particular size as you can get, and as thin - when the HSF is applied, it'll squeeze together and the paste will expand across the heatspreader and based of the HSF and fill in as required. Smoothing or spreading or heaven forbid brushing it over the surface isn't going to help nearly as much as letting simple physics - the compression on the compound caused by the two planes pulled together - can do.
Also, lapping that Q6600 is a relatively simple task and can lower your idle temps 5-10C, easily. Here's a thread with info from PCPer:
http://forums.pcper.com/showthread.php?t=441410
And just for the record, my Q6600 G0 with the stock Intel HSF (and AS5 applied exactly as those pics show) runs idle at ~48C if I push it to 3.2 GHz (8x400) but it's stable, without issues, I just choose not to run it at that speed. I actually just got some 400, 800, and 1000 grit sandpaper yesterday to lap it sometime this week, should prove useful. So if my stock Intel HSF keeps it at just under 50C at 8x400, I'm quite positive yours can do a helluvalot cooler since that HSF you have is way more efficient.
Good luck...
- Joined
- Sep 14, 2004
- Messages
- 2,724
Your CPU cooler is part of the problem. My Zalman can BARELY handle the heat off my q6600 @ 3.2ghz. Temps peak as high as 75°C in Prime but I dont get any errors or crashes thank god.
I'd check out reviews for a new heatsink and decide whats best for you price/performance wise.
in Prime but I dont get any errors or crashes thank god.I'd check out reviews for a new heatsink and decide whats best for you price/performance wise.
- Joined
- Sep 14, 2004
- Messages
- 2,724
The 9700 is just a tad bigger. 110mm vs 92mm fan. Anywho they are decent performers on older chips but quads, specially overclocked, put out a little more heat than Zalmans can handle. Sure they work and are pretty, but are they the best solution? No.
Hey, I have a Q6600 G0 @ 3.0 Ghz too. Also have a Freezer 7 pro heatsink. My CPU idles at 30C. I haven't checked its temp under load but I know it only got up to 51C after 1 hour of Prime95. It may be a problem with your heatsink. What are your temps under load?
flak-spammer
Limp Gawd
- Joined
- Aug 24, 2007
- Messages
- 262
I'm going to ask why you would think 27C at idle is likely. First that would require you to have an ambient temperature lower than 20C more like 15-18C. Second the 9700 isn't even that good meaning that you'd have atleast a 12C delta between room temp and the CPU temp. It is more likely that you have your Tj Max set to the wrong values or you live in a freezer.
Maybe I'm wrong - I'd be happy to learn why! If it's any help, I'm using an Antec 900 with five fans. I'm reading the temps using CPU-Z and Speedfan.
My ambient temp is about 19C. We tend to keep the house in the upper 60s in the winter.
Guess I'll go google "tj max" to see if I can figure out what that is.
BababooeyHTJ
Supreme [H]ardness
- Joined
- Jan 21, 2009
- Messages
- 6,951
While I can't speak for being a cooling expert, I can tell you that "spreading" any thermal compound on a CPU is a bad idea, period. It's the biggest mistake that most people will make, especially those that are new to PC building, etc. Here's how you apply it on a quad:
Taken from this guide:
http://www.arcticsilver.com/pdf/appinstruct/as5/ins_as5_intel_quad_wcap.pdf
Seriously, there really isn't any better way to do it.
Spreading it out sometimes causes major issues, and it's really as simple as that explains: a single line of compound, as close to that particular size as you can get, and as thin - when the HSF is applied, it'll squeeze together and the paste will expand across the heatspreader and based of the HSF and fill in as required. Smoothing or spreading or heaven forbid brushing it over the surface isn't going to help nearly as much as letting simple physics - the compression on the compound caused by the two planes pulled together - can do.
Also, lapping that Q6600 is a relatively simple task and can lower your idle temps 5-10C, easily. Here's a thread with info from PCPer:
http://forums.pcper.com/showthread.php?t=441410
And just for the record, my Q6600 G0 with the stock Intel HSF (and AS5 applied exactly as those pics show) runs idle at ~48C if I push it to 3.2 GHz (8x400) but it's stable, without issues, I just choose not to run it at that speed. I actually just got some 400, 800, and 1000 grit sandpaper yesterday to lap it sometime this week, should prove useful. So if my stock Intel HSF keeps it at just under 50C at 8x400, I'm quite positive yours can do a helluvalot cooler since that HSF you have is way more efficient.
Good luck...
I don't know, Ive tried that and it seems to take way too much TIM to get an even spread on the cpu. When I pull the heatsink after doing that I can take a credit card to both the heatsink and cpu and remove excess thermal paste. I just spread a thin even layer onto my cpu with a credit card and I allways get good consistant temps. Thats just my experience, it seems like Im in the minority with that.
I also highly doubt that lapping your cpu will net you 10c drop in temps, it helps but not by that much.
The DTSes used in 45nm Core 2 CPUs are not accurate below about 50C. You should ignore any readings below that number since they can be off by more than 15C in either direction.
Whoops, missed that. But yes, the 65nm chips are inaccurate as well, although not as much.
BababooeyHTJ
Supreme [H]ardness
- Joined
- Jan 21, 2009
- Messages
- 6,951
Im sorry but I have not seen any hard evidence of this unless your chip has a stuck sensor like my Yorkfield and woldale.
It's from Intel themselves. The sensors are designed to be very accurate close to Tjmax since that's what they're actually for, but they have a steep derating curve the farther away they get since they aren't meant to read lower temperatures. The stuck sensor issue is a different thing entirely.
BababooeyHTJ
Supreme [H]ardness
- Joined
- Jan 21, 2009
- Messages
- 6,951
Makes sense, Ive never trusted idle temps anyways. Explaines alot too.
Get rid of the thermal compound that came with the cooler. It's junk. Get MX-2 or OCZ Freeze and apply the thin line method as described previously.
Even with Zalman 9700, you should not be seeing 50-55 idle temps. I had my Q6600 at 3.4 with Zalman 9700 and it barely broke 40c during idle and rarely went over 70c at load (depending on room temps - my room temp reached 85f at times due to no A/C).
I'm surprised nobody has asked this yet- you mentioned that you are on a mild overclock, but what is your voltage?
My Q6600 idles around 50-53c if i leave the voltage set to AUTO at 3.15ghz, but it idles at 43-45c if i manually set the voltage to 1.24v. I find that especially with mild overclocks, the AUTO setting puts in too much voltage.
Try working your voltage down on your oc. Mine can probably go lower too but I dont have the time to change it, stress test it, change it, stress test it, etc. Too many games to play!
My Q6600 idles around 50-53c if i leave the voltage set to AUTO at 3.15ghz, but it idles at 43-45c if i manually set the voltage to 1.24v. I find that especially with mild overclocks, the AUTO setting puts in too much voltage.
Try working your voltage down on your oc. Mine can probably go lower too but I dont have the time to change it, stress test it, change it, stress test it, etc. Too many games to play!
uhh... the Q6000 series C2Q Kentsfield CPUs are 65nm guys...
anywayz i m using the Q6600 @ 3.0GHz with a Xigamtek S-1283 and Antec P182...
my temps are 31C idle and as for load 45C in prime and 50C in IBT
+1 on auto voltage, especially if you're oc'ing even just a little. Mines on water, but still idles high 30's and loads low 60's at 1.52v / 3.55Ghz. I would tend to agree that for the most part your idle temps really don't matter much, but at the same time if they're already in the 50's either something's wrong or you live in a sauna.
I've played with the voltage to get it as low as possible, so it's not on auto. Don't remember the exact value off the top of my head.
I ordered a tube of MX-2, so I'll apply it using the directions posted earlier and keep the thread updated.
I think low-mid 30s idle temps are a reasonable expectation for a Q6600 with a decent fan and a mild overclock.
Thanks for all the input guys.
I ordered a tube of MX-2, so I'll apply it using the directions posted earlier and keep the thread updated.
I think low-mid 30s idle temps are a reasonable expectation for a Q6600 with a decent fan and a mild overclock.
Thanks for all the input guys.
Hence the entire conversation you quoted...
What's the voltage and the ambient temp of your computer? Can you add a side fan to your computer case?