Grab copper and a drill, now!

Forge:
Where do you get the copper @.. And exactly what sizes did you get?
A3rOx:
Did the ones you buy work? Did you need the Sekisui thermal tape or did the stock tape work?

 

Nice to see a G51J owner turning to the darkside. That laptop has imo a lot of potential for copperblocks, since it has dual pipe from the cpu... double copperblock, hope you post some naked pics after the mod .

These are the ones that i bought, but wait for others to confirm before you place your order.

Copper Heat Sink For DDR DDR2 DDR3 RAM Memory RHS-03
Sekisui 5760 Dbl Sided Thermal Tape; 4ft, 48in X 1in

 

Abula speaks the truth.

and Soviet, I'm surprised you haven't done the "crossposting in my forum!?" thing yet.

 

The ones I've bought worked no problem. I strongly recommend though you do not use the stock tape (if u can even call it a thermal tape..).
Sekisui or 3M tape are your choices. I have both tapes and Sekisui is better than 3M in terms of how well it holds the block, but 3M works just fine.

 

MX-3 and the japanese tape arrived today. What's the curing time for the paste?

 

MX-3 is supposed to have 'no' curing time.

 

No thermal paste has zero curing time. There is always curing time regardless of whether the manufacturer says so or not. In regards for MX-3, the curing time is fairly short needing around 12 hours from my experience. Whether we can detect any further curing beyond a certain amount of time is unattainable as CPU and GPU thermal diodes cannot report non-integer temperatures to us.

 

Pls post if you get temp drops or higher temps with the mx3.

 

why would u get higher temps with better thermal paste (i guess most 3rd part therm. paste are better than stock OEM ones)?....unless u apply it wrong (too thick, too thin etc.)

 

Nice to see your back, though u were banned.

Well at first yea, i applied it way to thin, and the heatsink didnt fully seated, there was some pics of it also Forge had the same issue, then went a little thicker (not much, just slightly more on the farside), and temps were back to normal, still thinking on copper shim, but temps are fine atm, so i havent bothered.

 

Push the top part of the heatsink resting on the GPU die with your finger while you screw it down. That should help.

 

I just finished putting everything on about 10 minutes ago. So far the temps are exactly the same so I guess that's not a good thing. I also took those pads that covered the GPU ram off (they're sitting in a bag in case I need them) and replaced it with a thick enough layer of MX-3 to make contact. The CPU/GPU MX-3 layer was also thicker than I'd do with a desktop CPU. I tried making the thickness similar to the factory paste since it was surprisingly well placed on my system. Overall, I found MX-3 to be a huge pain in the to apply. I'd warm it up in hot water, apply it and it would immediately turn to this rubbery goo that was difficult to mold. I figured out a way around it by pre-heating the CPU/GPU core with a hair dryer so that it would keep the MX-3 in a goo state longer so I could apply it evenly. We'll see how it holds up after a few days of use, if I don't see temperature improvements then I'll likely try to reapply it again.

One thing that did suck for me is when I was spraying the insides of the laptop with compressed air, the piece of junk "void if removed" sticker tore in half with the force of air. So much for my warranty, thanks for making the sticker so flimsy Asus!


Temps for reference:
-----------------------
CPU: Intel Burn Test (Max Stress)
Before: 66C
Current: 66C

Furmark (1920x1080, post processing, 2x MSAA, xtreme burn mode)
Before: 98C
Current: 99C


More or less the same with the GPU running 1 C hotter. I'm hoping once the paste settles I will see some improvements. If not, I'll open it back up.

 

You can use pennies (or whatever the guatemalan equivalent is) to serve as copper shims.

 

I lol'ed. 10char

 

Thats a big gamble imo, if they aren't making contact for whatever reason, you could fry the memory. I think the heatsink isn't that even, at least was my case, so the pads do help to function as cushing for this unevenness. not saying what you did will not work but its risky.

I'm really laughing, not at your or anything disrespectful, just find it funny that air can voided the warranty. Mine broke even being super careful when i changed the CPU.

It should settle in with some cycles, but i don't think you gonna see more than a couple of degress difference.

I also found MX3 a huge pain to apply, i waisted like half the tube just trying it to make it stick always end up in my finger/bag. I just went with small rice application but tested/wasted like 4 application just to know the exact size that would cover the gpu and the placement that i had to since the sink doesn't install parallel but with some angle, so pressure pushes more to one side, hard to explain typing, but just did like 4 tests and one worked really well, then just clean it and repeat the same procedure.

 

How much ram did you use for your "maximum" ? it changes depending on what you have available.

 

Well as i said before copper in guatemala was stolen from cables and even water counters, but not so much anymore since copper price went down some. But we dont have any copper coins .

But Forge pointed me to some really nice copper samples that play great for doing shims, Sample Copper Sheet Pack $7.50 come with 5 different thickness, so its a good pack to do shims with.

 

Yeah I kept the pads just in case I notice the ram isn't making contact with the MX-3. I'll open it up in a few days to check. I did apply a fairly thick amount to ensure that it does though.

Oh I found it amusing too though I'm crying inside. Soon as it ripped I thought, "there goes about $100 worth of resale value".

I was hoping I'd see a difference right off the bat but that was wishful thinking. If I do see a 2C drop in the long run I'll be happy.

I should have done that but I got lazy towards the end and decided to just go with what I'd applied. It was a PITA doing the ram, NB, GPU and CPU with that difficult paste.

 

I'm in XP so it tops off at 2 gb when running the test.

 

hm...k I think my ICD7 lowered my CPU temps a bit then. I was getting a max of 61C in orthos before, and Intel Burn test is supposedly more intensive. I'm maxing @60C w. 2,200mb.


*edit* currently maxing @51C in orthos at stock voltages. Makes me believe my CPU's original paste was very poorly applied.

 

Stop.... stop.... your making me open it and put the ICD7 .

 

I facepalmed at all of page 52.

 

after about 48 hours of settling I tried furmark again. I maxed at the same 88C, but this time it took 732 seconds (12.2 minutes) to max. 2 of those minutes were a steady 87C.

 

Sounds about right since the P8700 has half the cache of the T9600 and runs with lower voltage. Well here's hoping MX-3 helps with temps. Was ICD7 easy or difficult to apply?

 

BTW, did you see any drops on NB with the MX3?

 

ICD 7 is a PITA to apply. its thicker than MX-3 iirc, and you really need to press just to get it out. I applied it without warming it tho, using a credit card to spread a thin-ish layer.

 

What are your full furmark settings? Can you try 1080p, 2xAA, post processing and xtreme burn? I want to compare temps. My stock clocks are 550/950/1375.

 

I don't do the stock clocks, as I'm undervolted. I can only do asus stock underclocks atm. and how long do you want me to run the test? (the full furmark settings are what forge has been using, so we could do direct comparisons. 1366x768 @8xMSAA+ extreme burning mode.

 

That must've been fun.

 

I hear its even harder than MX3, although my tube is sealed atm so i cant say for sure. But if you go for another paste after in case you arent satisfied, then go Shin-Etsu X23-7783D.

 

X23 is harder to apply than MX-3.

 

Just did a 10 minute furmark run @ 1920x1080 + 2x AA+ xtreme burn+ post processing:




hit 88C max, drops to 81C when fans kick in, then fans kick off and back to 88C..gotta love that rollercoaster.

 

At least it's not 90+.

And, not offense to Joker and Abula, but is MX-3 really that hard to apply? Although it's definitely harder to apply than AS5, I've had no real issues with it. Then again, I've cleaned up and reapplied it at least a fifteen times, for the GPU and CPU. Maybe it's the practice.

Also, I don't think the MX-3 layer in place of the thermal pads was a good idea, Joker.


And I noticed better temperature right off the bat, by the way.

 

I found it to be tedious to apply. It dries very fast and it's hard to get a good even spread with it. AS 5 in contrast is much easier since it doesn't harden when applying. We'll see if the MX3 for the ram was a bad idea or not in a few days. I'll be opening it up to see if they were all making contact. If they were, then I'll stick to MX3 instead of those pads.

 

I think he's trying to kill his GPU RAM so he can have an excuse to buy a new Clevo.

 

I've aplied ICD7 according to manufacturers instructions. I actually found it easier to apply than any other paste (AS5 etc) because there is no need to spread it on the core.

This is the guide which I used: http://www.innovationcooling.com/applicationinstructions.htm.
As to the size of the blob I was guessing, I opted for something smaller than what was mentioned in the guide, because they were referring to desktop CPUs'. Results? 1-2C drop on CPU and 1C drop on CPU [under max load]

PS:

I voided my warranty in a similar way. I was cleaning the inside with compressed air, when suddenly I pushed the sticker with the small pipe from the can and it tore. The best part was when my GPU died 2 weeks after I voided the warranty.

 

As Soviet has already explained, the way ICD 7 is applied according to the manufacturers only really applies to Desktops. In desktops there is alot more pressure put on the die, which flattens the paste allowing it to sit evenly on the chip, there is not enough pressure from the heatsink in most laptops to do this, and especially in a laptop with a (apparently commonly) uneven heatsink you will get an uneven spread of the paste.

 

Starting with post #159 to #167. http://forum.notebookreview.com/showthread.php?t=423878&page=16

 

While i dont disagree that the soviet way is the best way, back in the day when i applied it, it was really hard, i put a small size to be spread with the platic bag wraped to my finger to find it wont stick to gpu, and started sticking to the bag, then decided to put more so the excess will stay on the die, but it still went to the bag/finger, so i tried the creditcard way, same deal it spread semi fine the first stroke, then the second it lifts etc. leaving a kinda shadow residue on the gpu die, but the paste went to the plastic bag or the cc.

So i went back to the desktop style, thinking that with my uneven heat sink i would need to do some tests for me to see how it spread after installing (fearing that the farthest side might not get enough paste as it happen with AS5), so i did 4 different tests in different positions and qtys, everytime screwing it the gpu heatsink, till the 4th i saw full conctact with all die and the heatsink, i clean it again with artic clean and just tried my best to put the same amount into the same place hoping it would spread the same as it did in the 4th try.

I might try with heated paste, maybe if it becomes as manageable as AS5 then ill be fine, but i did had a hard time on MX3.

 

I haven't had issues with it hardening on me, at least not prematurely. For the GPU, I just squirt a blob on the die, spread it out with a putty knife, put the copper shim on top, and repeat on the shim. I've applied it when it was 80F and when it was 70F, and there wasn't any appreciable difference either time.

Maybe I got a really exquisite tube of MX-3.

 

Maybe you got MX-H2O.

 

I thought that was KGB top secret.

Oh snap.

 

Or repackage MX2, jk. Maybe ill go get a Putty knife, maybe thats the key. I can barely imagine how u do the shim spreading twice the mx3.

 

Doesn't have to be a putty knife. Just something hard and thin. I tried a razor one time and it was just as easy.

 

I had a little bit of trouble myself when sandwiching MX-3 on my copper shim on my northbridge. That was when I just switched over from MX-2 and had gotten used to the ease of MX-2. You will get used to working with "harder" pastes like MX-3 and ICD7. It just takes practice. Look at Forge, he remounted over 9,000 times and he has mastered the art of spreading MX-3 now.

Off topic: This is me ranting.

http://forum.notebookreview.com/showthread.php?t=428334
http://forum.notebookreview.com/showthread.php?t=428173
http://forum.notebookreview.com/showthread.php?t=428052

 

Offtopic, any of you have tried Unigine Heaven benchmark

 

It would make a great screensaver.

 

Too bad I can't run DX11.

 

I'd make a 6600 joke, but I am aware of karmic backlashes.

I wonder if I should try it, for giggles. Eh, maybe tomorrow. Today. Whatever.

 

The benchmark can be run DX9, DX10 or DX11.

Took my Desktop GTX285 for a spin on it, 81C on the benchmark i feel G51 and maybe clevos melting..... well probably clevo wont, but i ran it dx10.