ASUS Z71 Thermal Pad Removal To Reduce Temperatures

rhino56 · Post by **rhino56** » Tue Oct 25, 2005 1:52 am

i sanded them off with 220 grit on a flat block that fit it. i used a sharpening stone as the block and just put the paper over it.
as soon as i started to hit the surface i went to a 600 grit then a 800.
i didnt go beyond that.

a lapping kit consists of autmotive sandpaper ranging from 600 to 1000 usually, if you wanted to go up to 2000 grit you could. i have a few times and then even buffed the surface with a leather buffing stone i have to get a good shine.
just get a piece of glass and lay it on a flat surface and put the paper on that. there is a real art to it and it takes practice to get a surface flat with hand lapping.
the small things like that copper chip are crazy to lap, i wore off my fingernails until they were pretty much bleeding on a couple fingers doing something that small.
if you look around the net youll see a few guides on how to lap, you'll also see many different views on what works best. you'll have to find what works best for you with it.
here are a few pics of things ive lapped
http://www.cryo-laboratory.com/upload/u ... %20122.jpg
http://www.cryo-laboratory.com/upload/u ... 760073.jpg
http://cryo-laboratory.com/upload/userf ... ockdd1.jpg

ive done alot more but you get the idea.
here is a decent thread about lapping
http://www.ocforums.com/showthread.php?t=220200

killswitch83 · Post by **killswitch83** » Tue Oct 25, 2005 6:20 am

That's some good info there rhino. Good thing about lapping, according to Arctic Silver's website, is that when you have a smoothly-lapped bottom on your HSF, it requires less thermal material, more than likely because it doesn't absorb into it, and it appears to be easier to clean. If I end up with a rough-bottomed HSF, I plan on lapping it. Good find man

thebugeyedcow · Post by **thebugeyedcow** » Thu Nov 03, 2005 4:12 pm

I found this heat reduction guide for the z71 v chassis and they bring up some interesting points about the sub-optimal construction of the CPU heatsink, and an easy fix. I think AS epoxy would work quite well.

http://www.overclockers.com/tips1210/

Post by **Apoptosis** » Thu Nov 03, 2005 8:32 pm

The article at overclockers is a joke. Shoving thermal compound around a heat pipe like that won't make a big difference. If anything solder should be used and would be the better way to do it. I also frown an articles that have conclusions without any data.

Solder it or leave it be.

killswitch83 · Post by **killswitch83** » Thu Nov 03, 2005 11:08 pm

not a big fan of their articles either eh? I thought the Wire Trick was a joke personally....there's not many people out there (save Yves...not knocking on him but he's the only one I know here who has tried it, though he has had success with it) who would risk doing such a thing in order to unlock their multiplier.....I just don't see them as much as a credible source. I used to, when I didn't know jack about OC'ing, but now it seems a little....well...you know.....I'll let you fill in that blank, rofl.

rhino56 · Post by **rhino56** » Fri Nov 04, 2005 11:12 pm

seems kinda funny that they would say

DO NOT use epoxy to mount the heatsink to the CPU core unless you want a permanent installation!

at the very end of the article. i mean why would you ever perma glue your heatsink to your cpu when it has a fastening system?

seems like they post anything on the front page anymore which is kinda sad.
still an awesome forum though, most people there are some of the nicest people also.

Kerii · Post by **Kerii** » Sat Nov 05, 2005 12:05 am

thebugeyedcow wrote:I found this heat reduction guide for the z71 v chassis and they bring up some interesting points about the sub-optimal construction of the CPU heatsink, and an easy fix. I think AS epoxy would work quite well.

http://www.overclockers.com/tips1210/

Good concept, poor execution.

Like Apoptosis said, soldering would've been better (and cleaner).

Though I admit to not being too comfortable with a soldering iron, maybe that ColdHeat thing I always see on TV would work.

Anyways, smearing it all over was a bit dumb, using a syringe and needle to inject the compound in between the gaps would've been better. They're plentiful and cheap.

killswitch83 · Post by **killswitch83** » Sat Nov 05, 2005 12:19 am

oh yeah, ColdHeat, the soldering iron that cools within like a second or something after you use it right? that's an awesome creation, especially if you don't want to melt the laminate off a PCB while soldering a connection, or elsewise....of course, that could be the Newcastle talking (bought 2 six-packs and feelin good!), but at any rate, yeah, solder is better than smearing thermal epoxy anyday

infinitevalence · Post by **infinitevalence** » Sat Nov 05, 2005 2:31 am

Dont get to happy about the cold heat it cant be used on electronics as it will fry them. for soldering in the heatsink... well thats a different story. Also i never even thought to do this. maybe this weekend i will pull out my heatsink off my asus laptop and give it a go... or maybe i should do some school work... god i need a vacation.

killswitch83 · Post by **killswitch83** » Sat Nov 05, 2005 2:38 am

you tellin me? lol, Newcastle night (which is still ongoing even as we speak) is my only reprieve from school and a full-time job to support my having a house. school work is just natural, lol. speaking of which, going for another, lol

OgreMustCrush · Post by **OgreMustCrush** » Sat Nov 05, 2005 11:26 pm

Heh, the coldheat. I got one of those at costco, was on sale, and came with a fancy case, and the wire cutter/stripper thing. Works fine for small wire joins, but as for soldering that heatpipe to the base, no way. You need some serious heat for that, remember, it is an assembly designed to pull away heat. With the coldheat, all you will do is drain your batteries.

As for frying electronics, in my experience it only does it if you bridge contacts. Fried a couple cheap LEDs with the thing. Of course, the tip isn't small enough to do real delicate work anyway, so your not going to want to use it to do a vmod on your motherboard.

As to get the heat to solder the base the the heatpipe better, butane torches work well. Or if you're on the cheap, go to somewhere with nothing flammable around, cut the base off a pop can, flip it over so the concave part is up, fill it with some alcohol, clamp whatever your soldering above it, light, and solder.

rhino56 · Post by **rhino56** » Sat Nov 05, 2005 11:58 pm

i should try it and see what my results are.

Bun-Bun · Post by **Bun-Bun** » Thu Nov 10, 2005 9:12 pm

Hey guys, got a bit of a problem. I made a alumium shim on the CNC mill at school. Got it to exactly 1 mm x 14mm square and I then lapped it and cleaned it and cleaned the heat sink and GPU and I also sanded down the ridges on the heat sink. I thne applied thermal compound and put it all back together.

Turned the laptop on and everything was fine. So I continued to test my temperatures. After half an hour everything was running fine and stable and I was running 15 degrees cooler then before. I thought it was great. Then all of a sudden the laptop lock up and woudlnt respond. So I restarted it and tried to boot again. It booted but was very sluggish and then restarted with a BSOD. Now when I turn it on there are graphical glitches on boot and I get BSOD unless I start with the basevideo option. Then when I try to activate the drivers and change resolution I get a BSOD.

So I am conviced it is the video card but what the hell caused it to do this?

anyway, im sure asus isn't going to warranty it and I don't have time to wait for a RMA (last time I did it with my current supplier and it took 3 weeks) because I need it for school. So I am wondering what everyone's opinion is on what causes this so I can not do it again and where can I get replacement video cards for the A71v and am I limited to getting another 6600 Go or are there more options for me?

Thanks...

killswitch83 · Post by **killswitch83** » Thu Nov 10, 2005 10:46 pm

I think you're pretty well limited to the 6600 Go, due to the fact that if you look at laptop chassis that have removable video cards and compare the sizes of other mobile GPU's, they all vary in PCB size. I think it was in the notebook forums here I saw a pic that someone posted that had 6600 6800 and 7800 all in different sizes. I would ask Apop about that, as he's the resident lappie guru

Bun-Bun · Post by **Bun-Bun** » Fri Nov 11, 2005 12:33 am

Thats what I figured. I e-mailed Asus to see if they have any other cards released for this laptop. I got it back up and running with out artifacts. But at stock clocks the card is unstable. I have to underclock, run it with powermizer set to battery mode and not run anything 3d intesive games.
The second I do it locks up. But for school it's fine for now. I have it cleaned up really nice with the orignal thermal pads so as soon as my other Asus laptop comes back from RMA (cracked LCD) I will transfer the guts back to it and RMA my new one... sad.

Bun-Bun · Post by **Bun-Bun** » Fri Nov 11, 2005 9:52 am

Ok what the hell... when I went to bed my laptop would lock up after even the slightest 3d application running. This morning I wake up overclock ti for the hell of it and go play quake 4. I played for AN HOUR!! perfectly not even the slightest artifact. Thats a awesome game btw.

So what the heck... im gonna keep stress testing it for awhile and then I will try the shim again... computers make no sense...

kenc51 · Post by **kenc51** » Fri Nov 11, 2005 10:11 am

Try and remove the HSF off the eGFX card and apply some arctic silver or ceramic (ceramic is prob better for a laptop)... this sound like the GFX card is overheating... also make sure your custom shim is not loose and causing a short or something...you could have misplaced a screw and is loose inside the case....then if it happens to cause a short, it could cause too much voltage to the GFX card, overheating it (or even kill it)

If it comes down to it, remove the shim and use the standard cooling to be sure its not faulty equipement

Post by **Apoptosis** » Fri Nov 11, 2005 10:25 am

remove the shim and clean the core to see if you cracked it (too thick of shim) or scratched it!

Bun-Bun · Post by **Bun-Bun** » Fri Nov 11, 2005 2:10 pm

First off I designed the shim according to the same specs as Rhino. When it came off the lathe it was exactly 1 mm thick. After lapping it was .91 mm thick.

I already did those things and it was not cracked or scratched. even with the stock thermal pads it was still locking up and not being stable. When it first locked up with the shim in place it was only at 83 degrees C. With the thermal pads at the same clock it was 106 degrees C. Both reading are after 30 minutes of running Real-time HDR IBL.

so after it locked up the first time I opened it up and cleaned it really good to make it look brand new so I could RMA it. Put it back together with teh stock thermal pads and turned it on and it still artifacted on boot and windows would not recognize the card. Couple hours later windows reconginzed the card and all artifacts went away but the second I opened Real-time HDR IBL it locked up and the Nvidia driver test detected stock speeds as unstable.

So I left it alone and talked to my gf over google talk and fell asleep. Woke up this morning and for some odd reason decided to ramp the clocks up to 275/600 and played quake 4 and it worked perfectly. After 2 more hours of stress testing I decided to try the shim again. Installed the shim and made sure I didnt move the shim at all when I installed the heatsink. Ran Real - Time HDR IBL for an hour stable at 83 degrees C. Decided to try overclocking. Took the clocks to 330/680 and ran it again. Ran for an ohour at 92 degrees C. Last time I tried these clocks with the thermal pads it was 116 degrees C before it locked up.

So now it is working and I can't figure out why... 3dMark05 score is now 2462 and games run awesome and its perfectly stable.

I have one theory. I used Artic clean the clean the crappy old AC5 that I had and I think maybe the 1st solution might of gotten underneath the memory chips and underneath the whole chip. So when I used the 2nd solution I wouldn't have been able to get rid of it all. So maybe the 1st solution left over was shorting out the core or memory and cause my problems and then after the laptop being on all night the left over solution evaporated making it work now.

So to sum up

Thermal Pads
Chipset 53°C idle 57° max load
CPU 61°C Idle 75° max load
GPU 67°C Idle 102° max load

Shim and AC5
Chipset 45° Idle 53°C max load
CPU 56°C Idle 63° max load
GPU 64°C Idle 82° max load

*GPU stock clock
*CPU is Dothan M725 @ 533 FSB 2.16 GHZ

what are your guys thoughts on my theory of the Artic Clean?

killswitch83 · Post by **killswitch83** » Fri Nov 11, 2005 10:46 pm

well, arcticsilver.com states that AS 5 isn't conductive, but it IS capacitive, which means it can it can discharge a little electricity if you get it in the way of close-proximity circuits, which is quite possibly what happened. I had the same time happen the first time I used AS 5 on a Socket A 3200+. Accidentally got a little bit near the L4 and L5 bridges. I cleaned it off with 91% rubbing alcohol, but at first it didn't work out too well, it kept wanting to give BSOD's and act funny. I cleaned the core off and used a clean Q-Tip to try and get the rest of the AS 5 off the bridges, once again using the 91% rubbing alcohol. After that, it worked fine, no problems except when my buddy tried to install his old ass HP All In One and really old picture software from HP...gave him hell on that one, lol

. But yeah, you want to make sure you keep the AS 5 away from the PCB on the processor and mobile video card, and surrounding circuits, and of course not use too much because it can run a little. Good to hear you got it stable and running nicely