No significant Temp drop after Lapping CPU's IHS...

Tiramisu

Weaksauce
Joined
Apr 28, 2007
Messages
125
I've already lost count how many hours I've spent lapping my E6700's IHS but the result was rather dissapointing. I'm basically getting the same load/idle temp as before or maybe 1 degree drop under full load Orthos small fft. Ambient temp is roughly the same at around 21 ~ 23 Celcius.

This is my first time doing lapping and I challenged myself to do the CPU's IHS for the first time. I think I did a fairly good job at the end. I managed to get the IHS to a semi-mirror finish with 2000grid and becouse I'm short in supply of 2500grid I couldn't finish it with a perfect mirror finish. At least it's all copper and nice and flat. I use both bar codes and graph paper to check for flatness and they shown perfectly OK. I also did the ruler and glass table test for flatness and they all seems preety flat to me.

My E6700 Batch# L633A557 is OC'ed to 3730mhz @ 1.48v (1.5125 BIOS) Idle temp is at 40 ~ 42c and full load stress Orthos goes up 66 ~ 66.5degrees (Was getting 67~68 degrees before)
I'm still using Zalman 9700NT cooler before and after the lapping.

I'm not sure what to blame then. Is it the Zalman cooler is having a hard time keeping temp down at such a high clock of 3730mhz and high voltage @ 1.5v that even a lapped IHS doesn't seems to help much. Now I'm thinking If I actually did a screwed up job, wouldn't the temp actually increase than before? but it doesn't and it dropped by 0.5 ~ 1 degrees only...

What do you guys think?
TIA
 
You should blame your cooler, what you should do is ditch that cooler and go for Tuniq Tower - 120 or THermalright Ultra 120 Extreme.
 
Yeah, before I'm going to get a Thermalright Ultra 120 Extreme in the next few days, I wanted to know why my Lapped E6700 doesn't show temp drop in IDLE state. This is certainly very werid as the Zalman 9700NT should be good enough to shown some temp drop in idle state. We can blame it for not giving further temp drop on a lapped IHS only under stress full load becouse of it has reached its max capability but this isn't the case...or perhaps I really did a screwed up job?!

Another thing, I use AC MX-1 thermal paste. It has being reviewed as a better TIM than AS5 and I put a vertical line on the centre of the IHS. This method also proves to work for all C2D so am I really so out of luck or what?? :confused:
 
Have you notice if the IHS was concave before lapping? if it is flat, then why should you expect any drop in temp.
 
maybe the zalman isn't flat either. I'm a little ignorant on this though since I haven't owned the 9700, although my vf900 was perfectly flat, so I'm guessing the same applies for you.
 
Thanks for the replies guys,

hmm...when I started with 400grid paper, I noticed that the centre comes off first and then slowly moves on to the 4 edges/corners....which is quite different than most of others that has a concave IHS. So I determined mine is a convex IHS??

Well the Zalman 9700NT is pre-lapped from the factory with a mirror finish like the Enzotech Ultra-X. Let's put zalman 9700NT out of the case for now. Doesn't really matter if the 9700NT is or isn't perfect flat becouse a lapped IHS doesn't actually drop my temp in IDLE/LOAD which is very weird.

Suppose if I also lapped my Zalman 9700NT, I'd expect even further temp drop. But lapping IHS alone should yield some temp drop instead of barely no temp drop watsoever....:(

Maybe I'll try lapping again but I doulbt I can do a better job than what I've now.
I basically got a mirror finished but compared to others with insanely true to life mirror finish, I'd state mine only as semi-mirror finish to be conservative. I only had supplies avaliable locally up to 2000grid. I hope I could get some 2500grid and even micro 25, 15 10...etc. but acheiving mirror finish is just pure cosmetics and yeild no actual perofrmance gain > 1000grid
 
hmm...when I started with 400grid paper, I noticed that the centre comes off first and then slowly moves on to the 4 edges/corners....which is quite different than most of others that has a concave IHS. So I determined mine is a convex IHS??

Sounds like you just answered your own question.
 
With a convex IHS, the heatsink is touching your core so temp dissipations is already optimal. Just change the heatsink to a beefier one and you will see better results.

With a concave IHS, the corners is preventing the heatsink from touching the core so the temps will rise higher and the thermal paste will interfere with the cooling.

 
What do you guys think?
TIA
I think "lapping" or SANDING is a huge waste of time, mostly due to the fact that people don't have the skill or equipment to really do it properly. (don't take that the wrong way). Last time I sanded my CPU was back in the BP6 days. My two Celeron 366's were sanded with great care and basically nothing really changed after the fact. They still ran at 550mhz with ease and the temps pretty much stayed the same.

O'clocking is mostly based on the luck of the draw. You either get a great CPU, a shitty cpu, or one in between. You'd be better off removing the IHS although I think that's really a bad idea unless you have money to burn or someone is giving you CPU's. I put the idea of lapping/sanding up there with "burning in my cpu" - it's all in your mind. I know someone is going to reply and give me an example of an extreme decrease in temperature after lapping but those cases are rare.

Unless you are in a temperature controlled room - you really can't rely on any minute temp changes
 
With a convex IHS, the heatsink is touching your core so temp dissipations is already optimal. Just change the heatsink to a beefier one and you will see better results.

With a concave IHS, the corners is preventing the heatsink from touching the core so the temps will rise higher and the thermal paste will interfere with the cooling.


DAMN IT...so maybe I destroyed a nicely convex IHS base that's suppose better to work without lapping. Well, if that's the case, at least my temp didn't rise but instead, it remain the same or maybe just a degree drop under the same ambient temp before/after.

I lapped my IHS becouse I got nothing better to do for the past few days and I've being waiting for the Thermalright Ultra-120 Extreme to arrive and it just arrived right after I finished all the work...LOL so sooner or later I'm going to take everything apart again...

I think "lapping" or SANDING is a huge waste of time, mostly due to the fact that people don't have the skill or equipment to really do it properly. (don't take that the wrong way). Last time I sanded my CPU was back in the BP6 days. My two Celeron 366's were sanded with great care and basically nothing really changed after the fact. They still ran at 550mhz with ease and the temps pretty much stayed the same.

O'clocking is mostly based on the luck of the draw. You either get a great CPU, a shitty cpu, or one in between. You'd be better off removing the IHS although I think that's really a bad idea unless you have money to burn or someone is giving you CPU's. I put the idea of lapping/sanding up there with "burning in my cpu" - it's all in your mind. I know someone is going to reply and give me an example of an extreme decrease in temperature after lapping but those cases are rare.

Unless you are in a temperature controlled room - you really can't rely on any minute temp changes

I totally agree, on the luck of draw part as I've bought 3 C2D already and each performs quite differently and each fits in your three describtions literally. extremely poor vs regular vs better than average.

You're refering to the ambient temp will effect the temp of your CPU, it's true, but suppose a lapped CPU will give you more headroom in OC as your ambient temp gets higher. Let's say if the clock speed remains the same but when hot summer days come, you're forced to lower your clock speed on profile down to suit the ambient temp at full load, but a lapped IHS will help you in this case as it will usually drop roughtly 5c in average so this allows you to keep your OC setting without forcing to down clock your speed if ambient temp rises.
 
You should blame your cooler, what you should do is ditch that cooler and go for Tuniq Tower - 120 or THermalright Ultra 120 Extreme.

my expert analysis says you have reached your maximum efficiency with your current cooler.
 
my expert analysis says you have reached your maximum efficiency with your current cooler.

Your expert analysis...:D ;)

Thank you! advice well taken;)

but why wouldn't my "IDLE" temp drop ANY at all?! I doubt the 9700NT cooler is this bad that it simply couldn't do any better during idle state..??

Another quck question for you though, how much AC MX-1 TIM would you put in btween a lapped IHS and HS? :)
 
Your expert analysis...:D ;)

Thank you! advice well taken;)

but why wouldn't my "IDLE" temp drop ANY at all?! I doubt the 9700NT cooler is this bad that it simply couldn't do any better during idle state..??

Another quck question for you though, how much AC MX-1 TIM would you put in btween a lapped IHS and HS? :)

Not too much and not too little? lol j/k. just put a small thin line like applying AS5.
 
just because the Zalman looks shinny doesn't mean it is NOT concave. I lapped my Zalman 9500 before I went with water cooling and it was REALLY concave. After lapping both my CPU and Zalman 9500 I got about 10 deg drop in temps.... In fact there is a link floating around that shows how bad the surface on Zalman can be. Im too lazy to dig it up though.

Recommendation: LAP your Zalman, in fact thats what you should have done before even lapping the CPU.
 
just because the Zalman looks shinny doesn't mean it is NOT concave. I lapped my Zalman 9500 before I went with water cooling and it was REALLY concave. After lapping both my CPU and Zalman 9500 I got about 10 deg drop in temps.... In fact there is a link floating around that shows how bad the surface on Zalman can be. Im too lazy to dig it up though.

Recommendation: LAP your Zalman, in fact thats what you should have done before even lapping the CPU.

Yeah, that's true, I wasn't feeling lapping the HS but only the IHS to see how far my mileage will go. Looks like it didn't move one bit...lol

I'm not going to bother with my zalman 9700nt as I'll soon ditch it and get a Thermalright Ultra 120 extreme or alike.

I'm just a bit dissapointed for the time I've spent lapping the IHS and get basically nothing. To think on the bright side, at least I didn't do a screwed up job for a first timer and everything is working perfectly OK just like b4
 
Back
Top