Sign in with
Sign up | Sign in
Your question

Q6600 + ASUS P5Q Pro = dead end

Last response: in Overclocking
Share
July 1, 2008 2:49:59 AM

Well, ive had my P5Q for about 3 weeks now, and had absolutely ZERO luck getting decent overclocks with decent temps. On my MSI P35 I was at 3.0ghz and around 60c full load with a crappy ASUS V-60 cooler. Now, with this ASUS P5Q and a Xigmatek S1283 w/OCZ Freeze, the stock idle/load temps are trash (39-44c idle, 55c load), and the OC temps are just pathetic, (68c with prime95 running set for max temp/usage, 55c idle). What gives here? Obviously this is way to high for a Q6600.

Ive reseated the cooler 4 times now, twice with both configurations (back and top). Ive made sure the paste is on perfectly, checked for contact marks, the cracks are filled in the Zigma, im not using too much paste, the cooler is seated with proper preasure. The chip is a G0 stepping and its VID is 1.3, dropping the voltage will cause the OC to fail, period.

Ive had this chip for almost a year now, its always ran hot eventhough its a G0. Ive tried 4 different coolers on this POS on 4 different motherboards and ive never been able to get the temps that some of these other people are getting, ive always been 5c-20c higher. Cores 0 and 1 have always been 4-6c higher than 2 and 3, ive checked the IHS and its not bent, deformed, concaved, or anything like that. Im about ready to sell this piece of crap and get a E8400.

I was getting 3.6ghz on my MSI P35 easily with lower temps using the ASUS V-60. I know the Zigmatek isnt overraited and its not the issue here, im putting the blame at 75% motherboard 25% CPU.
July 1, 2008 3:37:58 AM

lol. get a E4400 and either a P5Q Deluxe or (im gonna get in serious spit for this....) a evga 750i FTW which i seen reviews time and time again....and even one a 10 second Q6600 overclock to 3.6GHz with one voltage increase and 100% stable and cool. ill edit this with the URL of that particular site i was looking at last night.... seriously guys what's wrong with the 750i? i think its a perfectly good board once you update the bios and drivers... simple and powerful overclocks...
July 1, 2008 3:41:46 AM

Intel/AMD chipsets are the way to go. Nvidia chipsets are for clowns....or that is what you will become after juggling issue after issue.
Related resources
July 1, 2008 4:12:14 AM

Wow how about making sure you CPU is not concave or convex lapping mite be needed .
July 1, 2008 4:15:14 AM

Did you update the bios? That might help with temp errors although I seriously doubt it and what are you using to check temps? Coretemp, realtemp?
July 1, 2008 4:18:10 AM

spathotan said:
Intel/AMD chipsets are the way to go. Nvidia chipsets are for clowns....or that is what you will become after juggling issue after issue.



Sure Intel chip set are the best at OC the highest and are the most stable.
Wait ! this just in Intel 790i just ended world hunger and brought about world piece.


July 1, 2008 4:23:37 AM

shadowthor said:
Did you update the bios? That might help with temp errors although I seriously doubt it and what are you using to check temps? Coretemp, realtemp?


Im using Coretemp and HWMonitor, both are reporting the same exact temps across all 4 cores. Ive checked for BIOS updates already, and there is one but ASUS says it just fixes a compatability issue with a particular memory.

Also @ HoustonSerenity, I stated in my OP that the CPU is not concave or misformed in anyway so ive ruled that out.
July 1, 2008 4:24:02 AM

I had the same problems with that cooler. Its the Push pins, and it has to be seated in just the right manner. The middle heat pipe MUST be directed down the length of the cores.

You can say you have it seated right, tested junk, etc. But so did I.

After I trashed the push pins in favor of small bolts and screws, I get great temps. We made another thread that showed the direction the middle heat pipe has to go on a q6x00 series chip.

Another thing as well, just take your thermal paste and literally draw a line down each heat pipe only. Just a nice lil even line down the middle of the copper heat pipe. All three of them.

I mean, I had near the exact temps. 40c idle, and like 70c loaded. I was shocked that people said this rivals the TRUE.

After making the adjustments mentioned, and making sure that middle heat pipe runs right along the length of your cores, leaving the other two to get residual heat, I get (like now, since I am on that comp.) 32/32/ 34/34 idle on my q6600 at 3510 Mhz, and 58 c max.

Before said adjustments, very hot, like yours.

You have an excellent board, and cooler, though a High VID processor. But still. Your 3.6 loaded temps near 1.45 volts should be at about 63c.

Good luck!

--Lupi
July 1, 2008 4:26:45 AM

Here is the thing Lup, I have the Crossbow backplate, and I also applied the paste in a line across. I got it from the guide posted up here. Tomorrow ill redo the paste again, after that if nothing changes then I give up.
July 1, 2008 4:29:31 AM

The chip wont work. Just try it like I said? On all other heat sinks I had, I applied it to the processor, and all works well.

Not so, the Xiggy.

The spacers between the pipes contact the IHS first. Having the goo on the chip will make a good contact with the spacers, but poor contact with the heat pipes.

Clean them off. Draw a line down each heat pipe, and re apply with the proper direction for that middle pipe.

Lemme try and find the proper alignment thingie...

--Lupi
July 1, 2008 4:31:55 AM

Lupiron said:
The chip wont work. Just try it like I said? On all other heat sinks I had, I applied it to the processor, and all works well.

Not so, the Xiggy.

The spacers between the pipes contact the IHS first. Having the goo on the chip will make a good contact with the spacers, but poor contact with the heat pipes.

Clean them off. Draw a line down each heat pipe, and re apply with the proper direction for that middle pipe.

Lemme try and find the proper alignment thingie...

--Lupi


Hmmm.....ok. I applied the paste in a line across the chips on the CPU, NOT on the HS itself. Tired and old argument here, and im not calling you a liar by far, but im just having a hard time believing/understanding that the placement of the paste will make a 10-20c difference. PC is completly idle right now, stock vcore of 1.3, 1.5 CPU PLL (lowest, I took it off auto to see if it was over volting, nadda), and the ASUS EPU Six-Engine in maxium power savings mode and my temps are 43/42/38/40.
July 1, 2008 4:36:32 AM

I know. :)  No worries. It worked for me.

I bitched and bitched, then broke it down and re set the thing 10 times in a row, and wow. After paying attention to what was happening, it worked!

Can't hurt to try, right?

Still looking for that burried thread.

The cores go only one direction. if your three heat pipes are running over them, you'll have like 25% actual contact in the location that the CORES are really under.

The other way, one heat pipe goes DIRECTLY over the entire length of the cores. Leaving the other two to collect the left over heat.

The spacers are a fraction of a hairs breadth longer than the pipes. Placing the goo on the pipes themselves, will cause ALL the excess thermal paste to gather under the pipes, and NONE under the already to long spacers. Get it?

--Lupi
July 1, 2008 4:38:33 AM

Lupiron said:
I know. :)  No worries. It worked for me.

I bitched and bitched, then broke it down and re set the thing 10 times in a row, and wow. After paying attention to what was happening, it worked!

Can't hurt to try, right?

Still looking for that burried thread.

The cores go only one direction. if your three heat pipes are running over them, you'll have like 25% actual contact in the location that the CORES are really under.

The other way, one heat pipe goes DIRECTLY over the entire length of the cores. Leaving the other two to collect the left over heat.

The spacers are a fraction of a hairs breadth longer than the pipes. Placing the goo on the pipes themselves, will cause ALL the excess thermal paste to gather under the pipes, and NONE under the already to long spacers. Get it?

--Lupi


Hmm, alright. Right now I have the cooler on top/bottom, as in the fan blowing up towards the PSU. I had it mounted with the fan blowing towards the back and temps were exactly the same, so which way is the proper mounting to where the heatpipes can do their job?
July 1, 2008 4:39:07 AM

Also, if you have one of those cases with a CPU exhaust fan, making you always direct the heat sink left to right, as opposed to up and down, to align the CPU fan with the exhaust fan, you may have never even accidentally tried the correct direction!

--Lupi
July 1, 2008 4:39:31 AM

I am looking!!!

--lupi
July 1, 2008 4:48:48 AM

Lupiron said:
Here!!

Note the indent and lil arrow location locater!

http://i83.photobucket.com/albums/j298/ericeod/sidebysi...

--Lupi


Lol, that is the same pic I used to determine how I should put the paste on my chip. So now I need to remove the cooler and turn it back around facing towards the back so that middle pipe can run down that line, WITH paste on the pipes instead of the cpu this time.

I noticed when I took the cooler off there was.....no paste on the cooler or the CPU at all, except the tiny little spot where I had put the one line on the chip. So basically im just getting a little bit of cooling from 1 pipe, as opposed to 3 pipes and all the alluminum fins that go with it.
July 1, 2008 4:56:31 AM

That could hurt. Its the spacers!

Place it on the pipes, a nice lil line down all 3 almost the whole length, and re seat it like it needs to be!!

Hopefully it helps.

I knew the spacer part happens.

I am just lazy and dont like to type so many things out, since I am one of the On Duty guys that is here alot.

Hehe, hear alot of the same things, and have to type the same junk.. And man, I am fing lazy!

Let me know how it works!

--Lupi
July 1, 2008 10:16:24 PM

Well I just finished reseating and redoing the paste Lup. No change. 38/38/34/35 with stock voltage and clock. Its about 70f in this room, maybe 75f.
July 1, 2008 10:32:15 PM

Just did a quick test with Prime95 for max heat, I let the 1024k's finish all 10 test before stopping it, temps peaked at 54/54/49/49. Ive seated the cooler perfectly, exactly as described and ive read. All 4 screws are tightened down all the way, just everything. If I was into OCing when I first bought this proc I would have RMA'd it for heat. Too late now. These temps are exactly the same as they were with my ASUS V-60.

For reference....here a is a ROUGH picture of the bottom of the cooler, the grey lines are obviously the OCZ Freeze that I applied in a line, the length of the lines and the width are however very accurate to what I put on there.

July 1, 2008 10:42:44 PM

Wow. I have never seen one that is simply hot to be hot, hehe.

You've tried the tricks I know, so I am out of info!

Perhaps your spacers are a bit taller than the ones on my cooler?

I have planned to lap my xiggy since I noticed that the spacers were taller, so maybe I will give that a go, and post the info, so you can try it as well.

I don't know how much thermal paste you have, but you may wanna test to see if that is the case, with the spacers.

You mentioned that you removed it, only to find that the paste didnt spread at all, meaning very little contact. Perhaps you should yank it off again, take a piece of glass, place a small square of smoothed out heat transfer goo on the glass with yer finger, making sure its an even but thin later.

Then take your cleaned xiggy and press it firmly to the glass, using the pasted area for contact with the heat pipes.

You can then look under the glass, and see exactly what is making contact! Who knows, maybe you will notice something else preventing it from contacting the processor IHS.

Dunno, man! Sorry.

My xiggy works like a champ. (And I havent removed it since I got it working good! IE, I tried like 10 times, and this last time it worked, so I didnt take it back off, heheeh! Scared that I will have to re seat it 20 times before it works again!!)

--Lupi
July 1, 2008 10:50:06 PM

I checked for that as it had crossed my mind, but to the eye it didnt appear that bad. Sadly, believe it or not the only cooler ive ever used that gave me great temps at 3.0 was a Scythe Ninja Mini in a microATX build on a Intel G33. Im gonna switch to AMD later this year/next year anyways, so....ill just deal with this for now, ill just put it at 2.8 or something and leave it. Thanks for all the help Lup.
July 1, 2008 10:57:46 PM

Bah! That sucks. You can still over clock it to where the Prime small ffts test reaches 70c, because you will Never see those temps in real life! Unless testing for them!

But use Small ffts to determine temps. Large doesnt help at all. No idea why it says that its for max heat and power!

AMD?? Man, talk about no over clocking power! Sad!

--Lupi
July 1, 2008 11:01:53 PM

Well I completely give up now. Set it to 2.7 and I get a BSOD 20 seconds into Prime95. It has to be this motherboard, I didnt even have these problems on my G33 or P35 with inferior cooling. This board HAS to be putting too much voltage to the CPU and its just not being picked up, its my only explination.
July 1, 2008 11:07:29 PM

Okay, lets start at the beginning! What was the VID of the chip in Core Temp?

Whats the Core Voltage in CPUz.

What is the Core Voltage set to in the Bios?

July 1, 2008 11:07:55 PM

Lets look for that power!
July 1, 2008 11:09:13 PM

Lupiron said:
Okay, lets start at the beginning! What was the VID of the chip in Core Temp?

Whats the Core Voltage in CPUz.

What is the Core Voltage set to in the Bios?


VID is 1.3

Core Voltage in CPUz is 1.120 (wow........)

Core Voltage is set to 1.3 in the BIOS
July 1, 2008 11:10:46 PM

Uhhh, yeah! Enable Load Line Calibration!

That will bomb out about any processor!

Turn it on and re test!

Then we will get to tracking things down!

--Lupi
July 1, 2008 11:11:14 PM

Disable speed step! Iest!
July 1, 2008 11:12:04 PM

LLC is enabled and speed step and the other are disabled as well, it was the first things I did when I got the board. CPUz never reported voltage that low with my MSI P35, it was always around 1.28.
July 1, 2008 11:14:31 PM

Under load, what does it do? Crank prime 95 small ffts and see what happens to the Voltage in CPUz.

We need to see if it goes up anyways.

--Lupi
July 1, 2008 11:16:37 PM

It goes up to 1.208. Also.....you said IEST, what is that exactly? IS that the same as C1E state, because I think that is set to enabled in my BIOS. This is all at the stock speeds btw.
July 1, 2008 11:37:50 PM

Double checked BIOS. Load line is Enabled, speedstep is disabled, C1E is disabled, spread spectrum is disabled, something called Intel Virtulization Tech is Enabled.
July 1, 2008 11:54:14 PM

Hmmm.. Man. Your Core Voltage shouldn't be near that low!! What Bios you runnin???

And your Temps at 0 load tell me that the cooler is on wrong, hehehe.

Should go in pairs of temps. 39/38/34/37 is just wrong.

My xiggy has 35/35/ 31/31

That means improper contact with the surfaces. :( 

1.120, that is amazing.

Thats NVidia chipset worthy.

Try 1.350 in the Bios, and see if its just a range thing.

The CPUz voltage should be near 1.3250 of that 1.3500!!

For it to drop and droop so low, that easily explains your stability problem.

Do you have HWMonitor? That shows Temps and voltages as well. Make sure they match up!

And back to the Bios.. updated it???

--Lupi

July 1, 2008 11:56:33 PM

I have the release BIOS. There is a BIOS update for the board on the ASUS site but it say it just addresses a problem with a certain memory. Also my temps have been offset like that with every cooler ive used, cores 2 and 3 have always been around 4c cooler.

I should mention I have the ASUS Six-Engine running, and its set on auto. It controls the power sent to the CPU, Video card, fans, memory, chipset and hard drives. Could this be an issue?
July 2, 2008 12:02:18 AM

Yes, it could, if it is in power save mode, or some such!

Disable it and see if you get closer to the voltage we want!!

And I know, but usually the pairs are the same. Like 50/50, then 45/45 when Idle. never 50/50 then 45, then 49.

The core pairs are twins! IE, each quad has two core pairs. Basically take two core two duos and combine them into one. So the cores should be equal to eachother. Most of the time.

We need real power to your Processor!

--Lupi
July 2, 2008 12:04:55 AM

Ok uninstalled the program, CPUz now reads 1.256. But I dont think this is the issue, because the program dosent activate when overclocking.

**EDIT** Nevermind, it just bottomed back to 1.120

This might sound incredibly stupid, but is it possible that the VID is wrong, and the chip's stock is NOT 1.3, but much lower? But when I experimented with this, anything under 1.3 would not boot when I had it at 3.0, and when I had it at 3.0 well the temps were too high......ugh....
July 2, 2008 12:17:24 AM

No, the VID is what it says it is, more than likely. It's prolly the Bios. Just update it anyways, and hope they fixed what they didn't know was wrong.

It's worth a try! Happens all the time. I never go off of the update advice from the Bios guys. (Unless its a major issue.)

It is slightly possible that your bios is corrupted. But Anything is better than an initial release Bios!

Takes 5 mins. (usually less, DL the bios, un zip it, toss it on a flash drive, press alt+f to execute ez flash, it will find the drive, and the file, and you just hit enter. ) 1 min later you load from saved profile for your old bios back, but since yer at stock, just go set the basics up and save.

Can't hurt.

--Lupi
July 2, 2008 12:18:32 AM

Ok ill try the BIOS update now. Deathly afraid of this lol.
July 2, 2008 12:18:38 AM

But with load line calibration enabled, you should get very close the voltage you set in the Bios. Like easily half what you have. Ine is .0250 on my newer board. x38. So your's should work, baring NVidia!

--Lupi
July 2, 2008 12:21:57 AM

Lupiron said:
But with load line calibration enabled, you should get very close the voltage you set in the Bios. Like easily half what you have. Ine is .0250 on my newer board. x38. So your's should work, baring NVidia!

--Lupi


My board is a P5Q, Intel not Nvidia :pt1cable: 
July 2, 2008 12:28:01 AM

No kidding? Coulda fooled me with that VDrop!

:) 

Seriously, it's an error there.

Mine is .0250 on an older board than yours, my newest board! That's excellent, and if ASUS went from getting it right to the bottom like NVidia, they seriously blow.

But if it were'nt for the other people with those boards having nothing like your amount of VDrop, i'd possibly consider that if I was exceptionally bored.

Hehehe!


--Lupi
July 2, 2008 12:29:51 AM

Ok about to flash the BIOS. If this dosent help anything then im gonna go ahead and RMA the board for a replacement. Ill probably get an MSI P45 or a X38/X48.
July 2, 2008 12:42:19 AM

Hmmm...
July 2, 2008 12:45:19 AM

Ok BIOS is flashed to the new version. CPUz says 1.280 now. But....dammit spread spectrum and all that is enabled, its at 1.6ghz, forgot to turn it off. But as far as temps go, no difference. I set the cpu voltage manually to 1.3
July 2, 2008 12:49:33 AM

At least you get the voltage you are paying for with heat now! Your Overclock will be much higher, even at the same voltage.

And much better when you decide to add more.

Just test with prime small ffts to 70c max, as you'll not see those temps real world.

Damn cooler!! Kick it once in the fin for me.

--Lupi
July 2, 2008 12:52:21 AM

Ok. Keeping a close eye on CPUz and HWMonitor for any spikes/drops in it, gonna watch this for awhile before I do any OCing.

This board dosent have a heatsink for the top row of Mosfets, so I ordered a Thermalright one today for it. Im also going to replace the NB and left mosfet heatsinks with Thermalright also if I can get a stable OC at 3.0. Im assuming better/actual cooling on the mosfets will help lower CPU temp?
July 2, 2008 1:04:12 AM

Uhh. prolly not! Damn it. It is your cooler! Since the CPU temps are measured from Diodes INSIDE the chip, one under the IHS, and the others embedded in the cores themselves, outside heat wouldn't be reflected much at all.

Must smash!

Err, I mean, take good care of that heat sink.

--Lupi
July 2, 2008 1:08:07 AM

Ok its at test 6 and temps have hit 59/59/55/55. No way in hell this is gonna get 3.0.

Also, I havnt exactly mentioned how ive been secuting the heat sink. Im using the Crossbow backplate screw kit of course. What im doing is putting the HS up on the CPU, and screwing in the screws about halfway in diagnials. So top right first, then bottom left, top left, bottom right, and I go in that motion until they are fully tightened one after another. And im screwing them in midly, I go until the screw stops with the preasure im applying, I havnt used anymore force.
!