Q6600 Heat Issues - Need Help

[leftheaded]

ask me questions and i'll get the details you need. any help is appreciated. thanks people

Here are the kind of temps I'm seeing:
clock/idle/load/peak
2.4GHz 45C/69C/74C
3.2GHz 46C/71C/84C

Pretty lame my temps are the same whether I OC or not.

i've tried all sorts of stuff, but i'm still at a loss. tried changing the airflow in the case, tried blowing an 18" floor-fan through my case with the panels off, reseated the HS... my HS seems to be flat, so I don't see any reason to lap it. the cpu might be slightly concave ... I tested it's flatness when i tested the HS, but I didn't take photos

 

[Nanan]

What voltage are you using? What fan is on the heatsink? Case airflow good?

If you have your voltage up high then you are going to see higher chip temps.

 

[leftheaded]

What voltage are you using?

It was AUTO in the BIOS. It was under 1.3, but I can't remember exactly. Will check tonight

What fan is on the heatsink?

CPU Fan: Scythe S-FLEX S-FDB 120mm Fan SFF21E (49 CFM)

Case airflow good?

I think so? I've got 5 fans in my P180B just as they recommend. two exhaust up top (one SFF21E, 49cfm & one SFF21D, 33cfm), CPU fan (SFF21E, 49cfm), one intake in the front middle (SFF21D, 33cfm), and the bottom fan (antec stock) between PSU and HDDs (blowing towards the PSU).

I thought putting an 18" floor fan through it would be plenty of airflow. It's certainly nothing I can do all the time, but it did knock off 2-5C if I recall?

 

[leftheaded]

first, thanks for all the help here.

ok, i got home and took a lot of pictures. hopefully, this sheds some light on the problem and can point me to a solution...

now to check flatness (liquid and razor this time)... the photos of the razor test are pretty hard to take, but they did reveal something I wasn't expecting. My last flatness test with ketchup led me to believe my TRUE was flat and my CPU might be slightly concave.

The razor on the cpu did reveal subtle amounts of light, but it seemed more or less flat... it didn't seem to be able to rock at any angles and there weren't any valleys that stood out as problems. I'd say it's "flat" enough.

The razor on the TRUE was fine when checking perpendicular angles. The diagonal check revealed the unexpected - the edges had enough space to rock... maybe only a millimeter or two, but that's plenty more than enough to be a problem. one corner was worse than the rest.

meanwhile, I mounted the stock HSF from Intel and have had it running prime95 for 2 hours (stock cpu, stock BIOS). It's at 71C under load, 45C idle. It's just barely within spec (71C is the max per Intel warranty). the thing that stands out as different compared with the TRUE is that all four cores are with 2C now. core 2 & 3 were about 8-10C lower with the TRUE under load. Is this a huge clue? does it imply that the stock HSF is flater than my TRUE??


here's the TRUE mounted in my case:

pulled it off and here's the base and cpu AS5 leftovers:

here you can see the washer i used with the TRUE:

here's the cpu all cleaned up (L726A819):

the glass is level & I have no reason to think it isn't flat:

the liquid (my son's prescription for his sinus infection )

Here's the cpu liquid tests (seems flat here):

here's the TRUE under liquid (i've juiced up the saturation and overexposed it to emphasize the problem area... it's obvious):

so, now the question is... is the TRUE the problem? And if I lap the TRUE how many degrees should I see the load temps drop?