do the latest true's still need to be lapped for optimal lga775 (Q6600) cooling?

[LOUISSSSS]

i've heard that they purposely curved the base to match the convex (concave?) IHS of the Q6600 (and others?) but that usually led to higher temps in general, do the latest TRUE120s that i can buy now still need to be lapped for optimal cooling?

does the cpu need to be lapped for the true120 to at least compete with a Tuniq?

 

[aigomorla]

bull dookie.

louis lap that sucker.

You'll be happy with the end result more then there bull dookie.

 

[LOUISSSSS]

wtheck is bull dookie lol

 

[aigomorla]

Originally posted by: LOUISSSSS
wtheck is bull dookie lol

dookie, as in crap...

 

[LOUISSSSS]

what exactly is crap?? the hypothesis that the curve of the TRUE's base is curved on purpose to match the curve of some intel CPUs?

 

[Quiksilver]

He means it's a load of bullshit.
It was just an excuse Thermalwrong started to use because of their poor Q/C.
There a small convex curve to Intel's IHS's; yes, but how bad that curve is varies from processor to processor; so Thermalwrongs bases will never match up perfectly; in some cases they can way off.
The way I see it, you have two choices; lap both the TRUE12 and the CPU IHS; or leave them both alone.

 

[LOUISSSSS]

Originally posted by: Quiksilver
He means it's a load of bullshit.
It was just an excuse Thermalwrong started to use because of their poor Q/C.
There a small convex curve to Intel's IHS's; yes, but how bad that curve is varies from processor to processor; so Thermalwrongs bases will never match up perfectly; in some cases they can way off.
The way I see it, you have two choices; lap both the TRUE12 and the CPU IHS; or leave them both alone.

so i'd be EVEN WORSE OFF with a lapped true and an unlapped IHS?

 

[katank]

Yes, potentially. I got great results from lapping both though.

 

[Quiksilver]

Originally posted by: LOUISSSSS

Originally posted by: Quiksilver
He means it's a load of bullshit.
It was just an excuse Thermalwrong started to use because of their poor Q/C.
There a small convex curve to Intel's IHS's; yes, but how bad that curve is varies from processor to processor; so Thermalwrongs bases will never match up perfectly; in some cases they can way off.
The way I see it, you have two choices; lap both the TRUE12 and the CPU IHS; or leave them both alone.

so i'd be EVEN WORSE OFF with a lapped true and an unlapped IHS?

Pretty much, there is a greater chance of having worse temps if you lap one but not the other than if you just left them alone (it is a given that it would be worse than both lapped.)

 

[chizow]

Its a coin-flip either way, just look at the possibilities:

Unlapped:
()
)(
((
))
||

1 Lapped:
(|
)|
|(
|)
||

2 Lapped:
||

Basically, even with 1 lapped, you'd end up with equal probability of obtaining better results than worst-case unlapped, but those results might be worst than best-case with both unlapped and everything was spooning perfectly. Of course lapping both means you can reasonably put all that nonsense to rest.

 

[BonzaiDuck]

LOUIISSS --

AigoMorla is correct. Quiksilver is misinformed, but no since getting into a p***ing contest about it.

I'd spoken directly with ThermalRight's tech support several times over the last few years, and our last discussion was around mid-2007 concerning the TRUE.

The convex base is deliberate. It is not a QC problem of sloppy manufacture; it may be better described as a QC problem of misguided design. TR gave me definite reasons for the choices they made.

I cannot be sure that the convex base-design is still implemented on other models like the IFX-14. I recall that they made the Ultima-90's the same way, though.

Even so. ThermalRight coats their heatsink-base on all models I know of with nickel-plating. Similarly, the Intel IHS is also coated with nickel. Nickel has a significantly higher thermal resistance (or lower thermal conductivity -- same thing) than copper.

TR chose to do this -- or they say -- to prevent corrosion. They say that lapping voids the warranty because it may damage the welds or solder-joints between the heatpipes and the heatsink-base. Meanwhile, SVC, who had marketed a "custom-lapped" TRUE not long ago, insists that the lapping doesn't risk any real damage to the heatsink. [They confirmed that they use a jig to do the lapping. You won't damage the heatsink if you avoid holding it by the fins or heatpipes when you lap it.]

The only trouble with lapping a convex base: you want to get it started so that it isn't cock-eyed. I used a pair of wooden blocks and a small clamp. After there is a flat surface that replaces the convex ridge evenly on either side, you can removed the blocks and clamp. But you should be able to do it without using that approach.

I say -- LAP THAT SUCKER! Lap the TRUE and lap the IHS of the CPU. Temperature reductions from TRUE-stock-performance between 4 and 10C are almost guaranteed, depending on the processor's effective thermal wattage.