• 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.

Q9550 OC

M

Muze

n00b
Joined
Sep 21, 2009
Messages
20
So I was experimenting with OC my 2.8g Q9550. At 3.2, the temps seemed too high (while running a game CPU high was 79 according to Realtemp). I brought it back down to stock speeds for now. Here's the readings with PC at idle. I have aftermarket cooling.

SS-20090921185351.jpg


I feel like things should be running cooler. Am I right?
 
C1 Q9550 runs a little hotter than E0 Q9550.

Does seem hot though.

Probably your cooling.
 
Way too hot for stock settings. Re-seat your cooler. Is it a bolt through or a pushpin setup?
 
As the others stated, what are you using to cool the cpu?

Have you installed the h/s properly?

Have you applied the thermal paste / properly?

What voltages are you set at? Cpu-Z shows 1.36v. That's too much voltage for a Q9550 running stock.

Lots of unanswered questions. Let us know.
 
I'm no expert at OC. I didn't change the voltage so what you see is what it is, what should I lower it to? Thermalright heatsink with a nexus fan on it. It's held in by those long spiral screws. The case is an Antec with two outtake fans near the CPU. I just took it apart and reapplied thermal grease liberally. Right now the CPU is idling at around 51, stock settings.
 
Find out what your VID is supposed to be. My E0 is 1.28V I believe and I thought that was pretty high for this chip, so it shouldn't be above that.

And when you say you "reapplied thermal grease liberally", do you mean you put a bunch on it? Because you don't want to be doing that either.
 
Might be a hassle but you probably have to remount your heatsink. Use a pea size drop and make sure it is tightened firmly.
 
Exact model of motherboard?
Exact model of "Thermalright heatsink"?
Size/RPM of fan?

FWIW, my 6550 @ 3.3GHz.1,38v is at 61 degrees F under full load, idle <40
I think voltage spec for this CPU is 0.85V &#8211; 1.3625V, so I don't think your voltage is the problem
 
3943661292_fd94531ecb_o.jpg

My E0 idle temps.

Loaded is &#8804;60*C.

Of course, I be under water, arrrgghh!
 
I think there must be something wrong with my sensors.

First, I did do a better job with the grease as things are about 5 degrees cooler.

But the BIOS will report the most accurate temps? BIOS says about 31C, when I get into XP, Realtemp and Speedfan give a reading of 53C. Rebooting into the BIOS and it's still low 30s. What's going on here?

Also the voltage is set to auto. It puts it at 1.36. I did try lowering it but that didn't change anything.

Gigabyte X38T-DQ6

Edit: And I have the BIOS set to warn if the CPU goes over 60C, so my readings in XP must not be right. Plus I have yet to experience a stability issue.
 
Muze said:
I think there must be something wrong with my sensors.

First, I did do a better job with the grease as things are about 5 degrees cooler.

But the BIOS will report the most accurate temps? BIOS says about 31C, when I get into XP, Realtemp and Speedfan give a reading of 53C. Rebooting into the BIOS and it's still low 30s. What's going on here?

Also the voltage is set to auto. It puts it at 1.36. I did try lowering it but that didn't change anything.

Gigabyte X38T-DQ6

Edit: And I have the BIOS set to warn if the CPU goes over 60C, so my readings in XP must not be right. Plus I have yet to experience a stability issue.
Click to expand...

What heatsink are you using, as in model name/number, thermalright tells us very little, they have made ALLOT of heatsinks. BIOS can not be trusted, use realtemp, whats the ambient temp in the room where the PC is? Have you checked to see if the fan is running? And when it comes to applying the paste, less is better, you do NOT want to goop the stuff on. Keep your voltage low if you can, no reason not to, even if your temps don't change, you will save power. And you will not have stability trouble till way past those temps, TjMax is 100C for the Q9550.
 
Start by checking your ambient case temps .. if that is high then that explains a lot. I live in India where i see case temps approaching 40+'C and cooling becomes that much less effective. Improving case airflow is important.

Then follow up on what the rest have said..lower voltage etc.

Also, go by Real Temp or Core Temp temps..or even better the Delta to Tjunction that both provide which is the most accurate. As long as your delta value remains greater than 20 under load conditions you are good.
 
If you have speedstep and power saving enables, the BIOS may be reporting "throttled-down" CPU temp.
What is exact model of Thermalright CPU cooler and fan details.
There are a lot of helpful people here, but they need info/data to be more helpful.
 
Even 1.36V seems kind of high. My Q9550 runs at 1.10V when set at auto. Actually 1.36 is the max voltage recommended by Intel. I'd definitely dial that back. The range for this processor is 0.85V to 1.3625V.

http://processorfinder.intel.com/details.aspx?sSpec=SLAWQ

I think if you get the voltage down it will help. It would still help to know exactly what hsf you are using. Hope you get it all worked out.
 
I would load optimized settings in your bios, save the settings & reboot to windows. When your desktop is loaded, open those same programs and show us the screen shot. That should give us some answers.
 
Thanks for your replies, I had to dig through my email to find the receipt.

# THERMALRIGHT, Ultra-120 eXtreme CPU Heatsink

# ARCTIC COOLING, MX-2 High-Performance Thermal Compound

Ambient temps are Temp1, Temp2, etc? in the 30s. Current idle:

http://www.mashedbuddha.com/images/SS-20090922120004.jpg

The PC case is currently without one side door, which is facing the inside of the cabinet the PC is in. There is only one door on the cabinet. There is some heavy duty sound proofing material in there. There's also an intake and outtake fan that I built into the cabinet but I rarely run them to reduce noise. This rig is running pretty quiet compared to the jet engines I've had years ago but still I would save only about another 5 degrees, and it still runs too hot outside of the cabinet. Some pics of the set up in a minute.
 
Here are some pics. I know you guys are going to yell at me to take it and put in the open room. It's cool to the touch, only feels hot when gaming.

0922091220b.jpg


0922091220a.jpg


0922091221.jpg


House has central air, usually between 71-73

I took it out, put it on the floor and will post temps again.
 
Last edited:
Mini P180's do not have sufficient air flow to begin with, and you have it in a casket!
 
Readings with the case out on the floor and wide open (PC off, all the fans run when on)

SS-20090922123344.jpg


DSC01964.jpg


DSC01965.jpg


DSC01966.jpg


Case is ANTEC, Performance One P182 Black Mid-Tower Case
 
Thanks for posting the pics. The only missing pic is the pic I most wanted to see. Can you show us a CPU-Z screen at stock settings. I wanted to get a read on your cpu vid. I agree you shouldn't have your pc enclosed. Keep your pc on the floor, outside the desk, with the side panel on the case. Then report your temps.
 
lodingi said:
Thanks for posting the pics. The only missing pic is the pic I most wanted to see. Can you show us a CPU-Z screen at stock settings. I wanted to get a read on your cpu vid. I agree you shouldn't have your pc enclosed. Keep your pc on the floor, outside the desk, with the side panel on the case. Then report your temps.
Click to expand...

Here it all is:

http://www.mashedbuddha.com/images/SS-20090922173318.jpg

After that I lowered the voltage to 1.30 and it reports at 1.28 with no change in temp.
 
Last edited:
It's probably normal temps for the Q9550 C1.

If you want to go for silence you should look into going to liquid cooling. It would probably lower your system temps.
 
hotbrass said:
Even 1.36V seems kind of high. My Q9550 runs at 1.10V when set at auto. Actually 1.36 is the max voltage recommended by Intel. I'd definitely dial that back. The range for this processor is 0.85V to 1.3625V.

http://processorfinder.intel.com/details.aspx?sSpec=SLAWQ

I think if you get the voltage down it will help. It would still help to know exactly what hsf you are using. Hope you get it all worked out.
Click to expand...

That's not the max recommended voltage, that's the VID range. I believe max voltage is 1.45V.
 
Frobot said:
That's not the max recommended voltage, that's the VID range. I believe max voltage is 1.45V.
Click to expand...

Correct, allot of people get that mixed up.

And to the OP could you now please post LOAD temps with the case out of the desk? As idle temps mean nothing. Also, have you tried doing the penny/pressure mod on the TRUE?
 
Last edited:
Muze said:
Finding some stuff on this pressure mod. What are they referring to when something is.. lapped?

after 20 minutes of Stalker:

http://www.mashedbuddha.com/images/SS-20090922194048.jpg
Click to expand...

Please DL and run at least 5 passes with Intel Burn Test (IBT). And post temps from that.

Lapping when when you take a HS or CPU and lap it, "Lapping is a machining operation, in which two surfaces are rubbed together with an abrasive between them, by hand movement or by way of a machine." In the case of most CPU's and HS's you use something of known flatness, such as plate glass and some wet/dry automotive sand paper ranging from 320 to 3000 grit to make sure the two mating surfaces (the base if the HS and the top of the IHS) are flat and have the most contact area for heat transfer as possible.
 
I ran it on Standard:

Time (s) Speed (GFlops) Result
25.205 35.4645 3.469841e-002
24.998 35.7583 3.469841e-002
25.040 35.6975 3.469841e-002
25.030 35.7125 3.469841e-002
25.034 35.7064 3.469841e-002

Don't know what any of that means. But it said it was successful and fully stable at the end.
 
Muze said:
I ran it on Standard:

Time (s) Speed (GFlops) Result
25.205 35.4645 3.469841e-002
24.998 35.7583 3.469841e-002
25.040 35.6975 3.469841e-002
25.030 35.7125 3.469841e-002
25.034 35.7064 3.469841e-002

Don't know what any of that means. But it said it was successful and fully stable at the end.
Click to expand...

Don't need the results from it, you are not having stability issues, I need readings from realtemp wile it is running.
 
Muze said:
I found a washer and sanded it down, put thermal paste on the bottom and put it under the TRUE bracket, retightened it. A 4 to 7 degree reduction.

Before washer

http://www.mashedbuddha.com/images/SS-20090922173318.jpg

After washer

http://www.mashedbuddha.com/images/SS-20090922223428.jpg
Click to expand...


Maby your TR is deffective? :confused:

I do think that your heat sink is not seated evenly because core 1 is about 10 degrees hotter than core 4.

You don't have to put pressure on your heat sink to make it sit right. Just enough pressure to maintain a good contact to the CPU. Good thin even layer of thermal paste is important.
 
BlueFireIce said:
Don't need the results from it, you are not having stability issues, I need readings from realtemp wile it is running.
Click to expand...

ok I grabbed screenshots as it peaked for each pass.

Pass 1

http://www.mashedbuddha.com/images/t/SS-20090922224752 copy.jpg

Pass 2

http://www.mashedbuddha.com/images/t/SS-20090922224826 copy.jpg

Pass 3

http://www.mashedbuddha.com/images/t/SS-20090922224855 copy.jpg

Pass 4

http://www.mashedbuddha.com/images/t/SS-20090922224940 copy.jpg

Pass 5

http://www.mashedbuddha.com/images/t/SS-20090922225007 copy.jpg

This is with the washer installed.
 
Outside. I just put it back in without all the soundproofing. The cabinet fans are plugged in and will switch on if the PC case gets warm. So far it's about 1-2 degrees hotter in there than out.
 
Alright. They look a bit high, but certainly better than what you were getting before.
 
Yeah it's about a 10C reduction overall since I started the thread. Concerned about it being uneven across the cores, but I just don't know how to seat it more evenly. The biggest improvement came from the washer. I may purchase a new heatsink and lap it.
 
Differing temperatures between cores often has nothing to do with contact between the heatsink and the CPU, but rather is caused by differences between the cores themselves. Nothing to worry about. As long as your temps are far enough away from Tjmax (100C), you're golden.
 
Muze said:
I found a washer and sanded it down, put thermal paste on the bottom and put it under the TRUE bracket, retightened it. A 4 to 7 degree reduction.

Before washer

http://www.mashedbuddha.com/images/SS-20090922173318.jpg

After washer

http://www.mashedbuddha.com/images/SS-20090922223428.jpg
Click to expand...

What do you mean by "sanded it down"? Did you sand the washer? Did you put the paste on the washer? That's not needed, the only reason for the washer is to add pressure, also, is this a normal washer with a hole in the middle? If so it might not work the best, as the TRUE bracket has a screw or "bump" in the middle and may reduce the pressure that might have other wise been added. Wile your temps are better, they still seem high for stock with a TRUE.
 
Yes I sanded each side of the washer. I did put paste on it because it seemed as though that's what they were doing in other threads (in other forums). It's a regular washer, and I thought the point was that the bump goes through the hole? I'm sure it could be much better, or done differently, I did it in 5 minutes.
 
You must log in or register to reply here.
Back
Top