Q6600 running hot with Arctic Silver 5 and open case

S

Searinox

n00b
Joined
Jul 26, 2007
Messages
16
Got myself a Q6600 for my aging motherboard today. At 8GB RAM and quite decent specs I just don't feel like changing it.

My fan is running at maximum speed. It is almost winter. I used Arctic Silver 5 paste. I have an Arctic Cooling 7 heatsink on it. The CPU fan is set to max speed from BIOS and confirmed with SpeedFan in windows. There is no overclocking. The heatsink is fastened properly, unmoving. I even removed and re-attached it to check. The side of the case is currently removed.

The CPU idles at 69-70 degrees, with the cores around 53-57. What should I do?
 
If the side of the case is removed, then you more then likely are not passing through (front to back) the amount of air you potentially could with the case closed. Opening the case removes the desired vacuum effect. your fans will still move air of course.... what kind of case do you have? Did you remove the side panel because of the heat issue?
 
Yes, most definitely the mounting (assuming your house isn't set to 95f :p ). I had a Q6600 and I think my idle was around 50 with stock cooling. The Intel push pins can be a pain so double check your work.
 
Like others have said it sounds like mounting, the Q6600 is a very hot chip as well, I remember mine, very good CPU, cost $700 at the time.
 
Bought one of these. Applied AS5 instead of the stock paste and reassembled. Added another fan to blow on the new one. Closed up the case and let the room warm up. 54 degrees idle and after about 6 minutes in prime95 it hits 72 degrees. Let it warm up for a little and tried again. Hit 72 in just 3 1/2 minutes this time. Temperature under normal load is around 60-62. In other words I'll be on the hot side but slightly under manufacturer-specified limit(71 degrees) unless I'm explicitly stress testing. What do you think?

By the way should I care about overall CPU temperature or individual cores? The cores themselves seem well within the safe range.
 
Last edited:
It does sound like the heatsink is not mounted properly. More heat should be dissipating then that. Saying that, people have expressed how hot the quad i5's can get.
 
Searinox said:
Bought one of these. Applied AS5 instead of the stock paste and reassembled. Added another fan to blow on the new one. Closed up the case and let the room warm up. 54 degrees idle and after about 6 minutes in prime95 it hits 72 degrees. Let it warm up for a little and tried again. Hit 72 in just 3 1/2 minutes this time. Temperature under normal load is around 60-62. In other words I'll be on the hot side but slightly under manufacturer-specified limit(71 degrees) unless I'm explicitly stress testing. What do you think?

By the way should I care about overall CPU temperature or individual cores? The cores themselves seem well within the safe range.
Click to expand...

It will probably be fine, just try and stay under 90C, I ran my Q6600 at around 80-90C 24/7 for around 6 years and it was fine
 
Shinryu said:
It does sound like the heatsink is not mounted properly. More heat should be dissipating then that. Saying that, people have expressed how hot the quad i5's can get.
Click to expand...

I mounted the sink as well as I possibly could. There's a frame that only fits in one specific position and it was very firm. Then 2 screws which are always properly alligned with the socket and need to be screwed in as tightly as possible. Which I did. It's not the usual system where the sink clicks in at a determined height. It pushes against the cpu as hard as you can screw it in.

This isn't an i5. It's a much older model on socket 775 that's a decent cpu today and used to be a monster back in its day. The architecture is quite different and its max TDP on intel's site is 71 degrees.

So what does their temperature limit mean to me then? And what is more accurate to take into account? Core Temp or CPU Temp?
 
You have a poor contact between cooler and CPU. Or there is an issue with the cooler itself. Maybe a heat pipe is shot and not conducting heat away. I've had numerous Q6600's. Should not be getting those temps at stock clocks with that cooler. Hell, shouldn't get those temps with a stock cooler.

I'd be happy to sell you a good Q6600 that's lapped and will do 3.6Ghz at 1.4v with a Xigmatek S1283 cooler and Yate fan for a good price. :)
 
HydroBudz said:
You have a poor contact between cooler and CPU. Or there is an issue with the cooler itself. Maybe a heat pipe is shot and not conducting heat away. I've had numerous Q6600's. Should not be getting those temps at stock clocks with that cooler. Hell, shouldn't get those temps with a stock cooler.

I'd be happy to sell you a good Q6600 that's lapped and will do 3.6Ghz at 1.4v with a Xigmatek S1283 cooler and Yate fan for a good price. :)
Click to expand...

So what should I look for to see if I've done a proper job mounting it? And from your experience, how hot can they go before there's trouble?
 
Wait wait wait, lets roll back abit here. You keep saying that the core temps are different. What are they? Also, what program are you using to verify temps. If you're only using speedfan, try running core temp. Finally, what mobo are you running?
 
Jorona said:
Wait wait wait, lets roll back abit here. You keep saying that the core temps are different. What are they? Also, what program are you using to verify temps. If you're only using speedfan, try running core temp. Finally, what mobo are you running?
Click to expand...

I was indeed using speedfan. CoreTemp reports 50-52 core temps on all four cores while speedfan reports these to be around 10 degrees lower. Speedfan also reports CPU at 57 whereas CoreTemp does not appear to have an overall sensor. This is all happening while I have a skype call running so it's not exactly idle. The cores are each around 5-10% usage. The CPU revision is G0. Tjmax 100 what's that? Motherboard is ASUS P5B-VM SE with 1011 BIOS version I believe.
 
Last edited:
Searinox said:
So what should I look for to see if I've done a proper job mounting it? And from your experience, how hot can they go before there's trouble?
Click to expand...

Pull the cooler and look for proper contact in the thermal paste. You'd notice if there were air pockets or anything strange. The paste should be uniformly spread from edge to edge on the CPU's heat spreader.

From the technical spec's the Q6600 will throttle at 100C, but I've seen them become unstable in the 80-90C range. I'd try to keep it under 80C at load. With your cooler you should be in the upper 40's low 50's at load. The one I have left is in the mid 60's at load running at 3.6Ghz with 1.4v.
 
HydroBudz said:
Pull the cooler and look for proper contact in the thermal paste. You'd notice if there were air pockets or anything strange. The paste should be uniformly spread from edge to edge on the CPU's heat spreader.

From the technical spec's the Q6600 will throttle at 100C, but I've seen them become unstable in the 80-90C range. I'd try to keep it under 80C at load. With your cooler you should be in the upper 40's low 50's at load. The one I have left is in the mid 60's at load running at 3.6Ghz with 1.4v.
Click to expand...
From what you say I should be within the safe range, even though the temps are hotter than they should be. I've so far removed the old and new fans several times and each time the paste covers the socket uniformly and a little of it has pooled over the edges where it's been squeezed out. There are no uncovered portions, it's just that the paste appears "smudgey" and shows traces, like butter over bread.
 
Sounds like too much HSF compound. Can you post a pic of how much you're using?

If you apply too much it acts like an insulator instead of a conductor of heat...
 
I can't check BIOS at this moment but both CPU-Z and Speedfan are reporting Vcore around 1.2-1.24v. I took the heatsink off. AGAIN. And I wiped the CPU clean. I then smudged what was left of it into the thinnest possible layer on the sink's contact surface, reducing its volume to less than half of what it was. The temperatures appear unchanged. Maybe lower by a degree, but I'll let the computer run for longer before I decide if there was indeed a drop.
 
I can't stand the smear applications of grease. It rarely worked for me. One drop, about the size of a grain or two of rice in the center of the chip. Then use the heat sink to spread it out.

I'm out of ideas though. You really shouldn't be that hot.
 
My previous cooler, which was able to keep an E5400 within 45-58 degrees idle to load, was letting my Q6600 idle at 70 degrees...... Could I be dealing with a broken sensor?

Btw, regarding the difference between CORE temps and TCASE temp.... shouldn't the cores be HOTTER than tcase because they're the actual heat source?
 
Last edited:
I cannot complete a prime95 test for realtemp calibration because after 1 minute of running the temperature hits 72 degrees and I refuse to let it go further. I wish there was a way to change the duration of the test realtemp wants to attempt because I'm afraid I won't survive into prime95 small FFT test even 3 minutes.
 
You're not going to kill the processor. It'll throttle itself and shut off long before it dies from heat.
 
I ran the program. And I have absolutely no idea what to do with it. I looked up google for threads on other forums and none mention a specific way to use it nor is there much on how to interpret the results. What do I do with the results of this test?
 
If I'm understanding correctly, all you're looking for while running Prime95 is load temps and stability. If you're aren't getting BSOD's or Prime 95 isn't erroring out after running it for a few hours, you have a stable system. All you really need to be concerned with here is how hot your CPU is getting while Prime95 is running. 72 degrees won't kill it, so let it run for a while and see what the max CPU temp gets to. Worst case scenario, the Q6600 gets too hot and shuts itself down (it starts to throttle down at 91C), which will keep it from getting damaged.
 
I swapped the cooler with this. I made it blow air away from the back and then also placed a case fan at the back, so fresh air is drawn from outside and blown onto the heatsink, and then the fan at the middle picks it up and pushes it further to the front.

After 2:30 hours of warmup, Idle temps went down by 2 degrees - to 52. Prime95 now takes 15 minutes - up from 4 - to reach 71 degrees and doesn't go over that. There is also about 5 degrees less variation between idle and normal usage. So yes there's a performance improvement. But I'm still idling above 50. Another weird thing I discovered later: if I do the small FFTs test - which isn't claimed to be max heat - I actually get higher temps. In about 8 minutes I can hit 72. Yay for more weird crap.

This brings back memories of the very first cpu I had on this motherboard. It was an Intel Pentium 524 at 3.06Ghz with hyper-threading with this very same setup. The thing idled at around 50-ish and went up to 76 during very intense load(not stress test). But that was with a stock cooler and during the summer heat and I had fewer coolers and more hanging PSU wires. This one has no excuse.

Since I started idling at 70 and brought it down to 52, I'll consider the 18 degree drop an improvement justified by the coolers' claimed performances. If this were my old processor, which idled at 48, it would now be at 30, an excellent temperature for any cpu. Sadly, I'm stuck with this 52C hothead. And really I guess it is what it is and can't be helped. At this point I'm just going to give up and leave it at this. Thanks for all the help.
 
Last edited:
As an Amazon Associate, HardForum may earn from qualifying purchases.
You must log in or register to reply here.
Back
Top