Repasting with CooLaboratory Liquid Ultra, any tips before I start?

can someone assist me Tbonesan you are in japan now right?

 

Yes I am, what's up?

 

Like Mr. Fox mentioned, most likely reason is being too stingy with paste. The first time I was way too paranoid and went too far with "as little as possible" and spread it out too thin, and temps were actually worse than with MX-4. Quickly cleaned off the stuff and applied more, and saw between 7-10C drop in temps.

Does not damage copper, but does bond to the heatsink. I now have a nice silver mirror finish thanks to Liquid Ultra lol. Tried gentle scuffing with the scuff pad but didn't seem to do anything. Since temps were still just as good, didn't bother sanding it away.

As for cleaning, soak a paper towel in alcohol, then very gently wipe it off the die in one smooth motion. I cannot stress the gentle part enough, too much force and some of it seems to harden and cling to the die. Luckily I was able to flake it off with fingernails. Also I think it slightly etched my die where I applied too much force with the towel, so be very careful.

 

Mr. Fox i think with CLU 5ghz barrier you can break and possibly game at 1200 or 1150 on the 880 m even??

 

Yes, I have already had the 4930MX at 50x4 with Liquid Ultra. Unfortunately, the crippled 18 motherboard doesn't allow significant GPU overclocking, so unless you disable SLI that kind of GPU overclock isn't possible. The laptop just shuts itself off because it doesn't have enough power to operate that way. I would not game with it overclocked like that anyway. Short bursts of number-chasing benchmarks is fine and it does not hurt anything, but gaming with the CPU and GPU overclocked that far for hours on end is probably not the best thing to do if you want the machine to last a long time. It's also not necessary. I get well over 100 FPS with 4.3GHz and stock GPU clock speeds with 780M SLI in almost all games.

http://valid.canardpc.com/nqhu1p

It is like a layer of solder. It's not actually etching the die. I thought the same thing. If you use a cotton swab in small circular wiping motion and repeated application of alcohol it will all come off the die, eventually, but if you try to smear it off it will look exactly as you described. On the copper heat plate, I just used the scuff pad and rubbed the snot out of it until it was all clean. Doing this actually polished the copper plate to a highly reflective finish. It was pretty awesome looking when it was done.

 

Ok, I reapplied liquid ultra and was more modest on my delivery and so far the results seemed to provide an improvement especially in stability so just wanted to say thanks for the vital feedback. One last question, Since it acts like solder when applying does the stuff harden up? Since its a laptop and mobile I don't want to have the fear of something being able to sling down into the system.

 

When I took mine apart for inspection (curiosity) it was a little crusty on the edges of the die where the surfaces are exposed to air, but the anaerobic surfaces (where the die and heat sink mate) were still a brushable liquid.

 

Personally do you feel that this stuff brings any danger to what I would call swag getting into the system from day to day movement. I personally wouldn't think so but you and several other knowledgeable and valuable opinions are probably the highest that can be offered and trusted so I would like to hear your point of view on this if possible.

 

I'd imagine it is perfectly safe - I haven't used it myself (yet) but as Mr. Fox mentioned, the edges appeared a bit crusty on inspection, so I doubt that any would leak out from between the heatsink and die with general day to day movement of the machine - the 'crust' probably helps to keep the TIM in, if anything. The only way that I would see it being a potential hazard would be if far too much were applied, potentially getting squeezed out all over the place......

 

Yup, Brother Stevie said it well.

I'm not worried about it at all. Just don't be sloppy and avoid applying it in excess and it should be fine. It's awesome stuff on the XM CPU. I don't remember exactly when I applied it now, but it has yet to show signs of degradation on the Alienware 18 or M18xR2... both holding strong with excellent temps. In addition to being benched at much higher overclocks, both machines idle at 4.3GHz with c-states disabled... so, that's pretty amazing.

 

Thanks for the reassurance I appreciate it.

 

Just an update since I last time I post on here.

I am still using the original paste that I put on. No need to repaste. The temps are still the same, I am running at 4.5 with a 3920xm running Boinc 24/7. I am running at 86C at 100% on all 4 cores/8 threads.

I just clean out the fans and dust on the heatsinks.

 

I am seeing similar temps on my 2920xm at 4.5 gigz. Cstates turned off and highest temp reached was 86c running 24/7. I did dial it back a bit to 4.2 though. It's very good stuff but it is a little nerve racking using it. It may have been just me but I used some right after I got the tube in mail on an Intel Nuc I purchased and put the end back on it and I stored it in a drawer at room temperature but when I re pasted my cpu and gpus 2 weeks later I found it very hard to squeeze any of the liquid out of the syringe like tube. I managed to get enough out for the cpu and when i was trying to squeeze out some for one of the gpus i was pushing soo hard that I snapped the plunger and it shot some liquid metal all over my mother board. I managed to get it all cleaned up but it was extremely nerve racking. It may just be an isolated incident but be mindful of it solidifying inside the tube after being stored. It looks the same but about 2/3rds of the tube was solid. Despite this I still recommend it just be careful.

 

Gently heating the syringe with a heatgun may help liquify it again. If you read the data sheet the melting point is 8C (~46F), so if you have a chilly room it will probably be thicker and more coagulated.

 

When I buy more I'll consider that. I just threw the tube away and re pasted my gpu's with arctic silver. Temps are at 40c to 45c at idle with AC5. The cpu was what really needed it.

 

Aaaand that's all! Came back to Spain with the summer starting here and I'm going to repaste this weekend. So not bad since I created the thread! So hot here to keep waiting.

Since the first page I've using 4.5ghz all the time no problem at all. Superb paste

 

Update: 1 molar hydrochloric acid is da bomb for cleaning off Liquid Ultra on the CPU, whether in partially oxidized form or fresh application. Long story short I had to take my heatsink apart, so I had to repaste Liquid Ultra. Turns out if you just dip a cotton swab in dilute HCl and rub it over the CPU die, it cleans off Liquid Ultra like it's some sort of magical magnet. I probably still wouldn't recommend people trying this on their own CPU, but I will most likely stick with dilute HCl for cleaning the CPU die from now on.

Since I made these bold claims it's only fair I test them... and so I did. Noticed there some spots left on the CPU die even after thoroughly wiping down with an alcohol soaked paper towel. The die felt perfectly smooth to the touch, but there were clearly spots of either oxidized stuff or smeared metal.

So guess what I did? I dabbed some cotton swaps in dilute (1 molar) hydrochloric acid, and applied to the CPU die. That's right folks, I applied gosh darn motherfrickin' ACID to the CPU die. Man I deserve +10 rep just for that. But it worked wonders, after some repeated rubbing all the spots completely dissolved away, and the die was as good as new again.

Had less luck with the heatsink. Whatever stuff was on there did not want to dissolve away completely, even though I could visually see it was thinning out and the cotton swabs were full of gray gunk. What I did find however was the acid softened and unbound whatever stuff was on the heatsink, so it could be scratched off with even a fingernail. I scraped it of with a box cutter just to get it done quicker, and a final polish with the scuff pad left a nice silver mirror finish on the heatsink without any of the patchy/blobby/spotty gunk. Felt smooth to the touch as well.

So in conclusion, acid works, especially effective on the CPU die. Now let's just hope my CPU doesn't magically disintegrate a few weeks down the road LOL. (furiously knocking on wood)

 

Hello,

Today I ordered the Liquid Ultra after reading this complete thread, I know I don't have a M18X, but this thread is big, I should report it here.

Sounds like awesome stuff, or you have the best temps ever or you short circuit your motherboard or so, really hardcore stuff xD!

Bought some for the CPU, I won't do my 680M with it because of the electronics so close to the GPU.
Should be here tomorrow, can't wait to use real metal xD.


Senshin

 

That should make a huge improvement for your CPU thermals. If your GPUs are running hotter than you would like them to, you can use a layer of electrical tape to cover the surface mounted components next to the GPU die for peace of mind when using Liquid Ultra on the GPUs. Covering them with a thin layer of high-temperature automotive RTV silicon might also work well, but I don't know that anyone has ever experimented with that. UltraGSM used that as a border or "fence" around the perimeter of his CPU die.

 

Does electrical tape have any heat impact? I was wondering about using that myself but on the CPU. Would certainly make me much more comfortable with the idea of using liquid ultra.

Sent from my HTC One_M8 using Tapatalk

 

Shouldn't have any impact as long as it doesn't touch the die. Although I'm not sure how hot the VRMs get so if the tape starts melting it'll be one big mess.

 

It should not cause any issues. Clevo GPUs don't need it because they have a plastic shield covering the components. Dell has nothing covering them up.

 

I removed the shield but the shield didn't do anything anyway as there was a huge puddle of paste underneath it and my VRMs are now impossible to clean.

Sent from my HTC One_M8 using Tapatalk

 

Well, that might have been from using too much. None has ever escaped from the place I applied it, so technically it should be fine with nothing. Using electrical tape or a thin layer of RTV silicon would simply be a bit of extra "comfort" for the apprehensive.

 

I use Liquid Pro on all my components now, its not really a big hassle to put on as long as you just have some patience

My first mistake was to put to much on the CPU which just ended with a BIG silverball rolling back and forward and eventually rolled of the CPU (LOL) and on the MOBO.

 

I've pasted both my CPU and GPU with this stuff. It brought my temps down a lot on my CPU. My GPU, not so much. My AW shut down whiles playing Murder Soul Suspect. I opened up my AW and checked my GPU heatsink, much to my surprise, only a spec of Liquid Ultra was showing on the top left corner of the copper plate. From that, I gathered my heatsink copper plate wasn't flush against my GPU die. This's happening with the stock thermal pads, so I'm guessing I may need to change them for an exact measured thermal pad replacement. Any suggestions for the new aw 17 and 18 gpu heatsinks guys?

 

How realistic would it be to use fluxless solder to mate the heatsink to the CPU? I realize this is a pretty permanent solution, but assuming you can still access the screw on the socket (and thus release the CPU from the mobo), if you can pull it off this would mean that you would never have to repaste ever again, and the fluxless solder should be as good if not better than Liquid Ultra.

 

Hello,

So yesterday I installed the liquid ultra, all I can say is wow.
Application is easy as $#%^, just scared of it getting on my motherboard but didnt happen.

I first did a small grain of rice on the cpu, used the brush to do it good, then with the rest over i putted it on the heatsink, not perfect becuase i couldnt the old spot where the cpu where sitting. but roughly right haha, I think it was good enough anyway .

Put it back together, waited 1 min and started the beast, with very good temps also!!
Remember it was 27 degrees yesterday, max temp was around 80 - 82 after 10 min of stressing and it would idle back to 45- 50.
You guys should know I have 2pipe heatsink, so the temps youguys have I will never reach I geuss .

Before it could hit 90 degrees easy and went throtteling.
Also is there a cure time, becuase I installed it and went benching right away xD.

Very happy with it and hopeit doest degrade like my IC diamond, becuase it did in 1 year time, whatever other people are saying.... (anyway, my batch degraded so much after 1 year the hard diamond stuff was pushed outside the Bare DIE and on the bare die was like liquid/diamond mix thing, moreliuid, I could wipe it off easy what's not good).

I think I had a bad batch of IC diamond, but maybe not, don't know if there are people with same experience.



Senshin

edit: 1 small things, how does this stuff hard out? I mean the left over on the brush got hard in no time, like a few hours, whats'up with that?
Is that a reaction with air or so?
This also gave the good feeling it wouldnt go on my motherboard or so becuase of the harding out.

 

Hmmm... I'm really starting to get the itch to do this to my 880s and especially my 4940MX... So it didn't have any issues? I've heard of it rolling right off the die but I've also heard of IC Diamond scratching a GPU die so bad the logo isn't legible yet I've repasted my 880s a few times now with it and neither has any scratches that I've been able to discern.

Sent from my HTC One_M8 using Tapatalk

 

During a Valley benchmark, my second 880M card was reaching 93 degree while my first 880M card was around 85 degree.

I just repasted the GPU's with Liquid Pro (CPU is already done with Liquid Pro) and there is the result after a Valley benchmark.



The numbers speaks for itself..

 

Well none of MY repastes caused such a mess, that was the stock paste job... Pretty abysmal especially if you consider that my first time repasting in my life was repasting my Clevo. I tried cleaning it up but it tore my microfiber to shreds and its still there. Its an eye sore but it doesn't hurt the GPU so meh.

I'm starting to think I'll do this but I am a bit nervous it won't make a difference with the GPU that's overheating since the GPU heatsink doesn't seem to be flush with the die anyway. It would sure help the CPU though.

 

Taking a Dremel to the bottom cover and using 120mm fans on the U3 might be an alternative if you're really hesitant about Liquid Ultra. I saw a 5-8C drop across the CPU and GPUs even with just MX-4 on the 780M.

 

I took apart the 18 today and decided to have a peek at the Liquid Ultra. I'll have to dig up the date when I applied it (somewhere in this thread) to the 4930MX but I think it was something like 4 months ago. I have it running a minimum of 4.3GHz all the time and it never gets hots. To my surprise it was perfect still. It was still moist and none had moved off the core or anything. It looked about the same as the day I applied it. I was very pleased with what I found today.

Anyhow, I took some photos of that and I will post them later tonight or maybe tomorrow.

I decided to do a test again with IC Diamond since the test I did earlier in this thread was on the 3920XM and it was not with brand-new IC Diamond. To be fair, we will see how a brand new application of IC Diamond does on the 4930MX blast furnace. I'll do some serious overclocking and capture the temps, then I will do it over again with Liquid Ultra for comparison. I already know Liquid Ultra is going to win by a fairly wide margin, but some people still need more evidence that their Haswell needs it to avoid overheating.

I also tested the heat sink fit with pressure sensitive paper and the die pressure was amazingly uniform. Alienware did a fantastic job of fitting the CPU heat sink effectively. I will post a photo of that as well.

 

I highly recommend Liquid Ultra for the CPU. Not so much for the gpu though, as the copper doesn't seem to sit flush against the die.

 

Well, I lost 20-23 degree from the GPU It had Artic Silver 5 before I changed it.

 

I would repaste my GPU's again if I could get the copper plate to sit flush against my die.

 

Yeah that's one seriously sucky manufacturing defect...

Sent from my HTC One_M8 using Tapatalk

 

Might be just an anomaly with that machine, though. I don't have that problem on this Alienware 18. Cooling is great on CPU and GPU.

 

It's not a problem. I've managed to resolve the problem by switching out my thermal pads for the exact measurement thermal pads. Since doing that, I've seen significant improvements in my temps.

 

Which worked better, the 0.5mm or 1.0mm?

 

0.5mm. 1.00mm lifts the copper plate by 0.1mm on the top left corner facing the opposite site of the vram.

 

Where did you find that pressure sensitive paper, if I may ask? I looked for it some time ago, but didn't find it anywhere.

 

It was given to me. I will see if I can find out where he obtained it.

I think this is the source, but I'm not 100% certain. http://www.sensorprod.com/pressurex_micro.php

Here are other sources, but a different type (noted heat sink application) http://www.pressurex.com/ and http://www.sensorexpert.com/tactile-surface-pressure-indicating-film.php

 

Make sure the paper is rated for 28-80 PSI. The one with the highest sensititivty (2-20 PSI) will just light up all over the place and give you a false impression.

For referrence if I pinch the 28-80 paper really hard with my fingers I can just about get it to register, so 28 PSI of pressure really isn't all that much.

 

Oh man, I'm amazed at how much cooler your system runs now. It seems like a great choice to have made!

 

It was January 24th that I repasted the Alienware 18 with Liquid Ultra and it had not really shown any evidence of degradation. I almost hated to break the bond to start a new test with IC Diamond. At the rate it going it may have lasted a year (was already over 5 months of heavy overclocking). I can already see an increase in my loads temps with the fresh repaste without Liquid Ultra.

Here's the original post (also quoted in this thread on the same date):

Spoiler

Here is more information on the pressure-sensitive film: Fujifilm Prescale® Pressure Indicating Films

Here are some photos:

IC Diamond Application (taken yesterday) The third photo shows the pressure test result.

<iframe class='imgur-album' width='100%' height="750" frameborder='0' src="http://imgur.com/a/Rkmto/embed"></iframe>

Inspection of Liquid Ultra (taken yesterday) Look start at the 9th photo... amazing that 5+ months of heavy overclocking and it still had no evidence of degradation.

<iframe class='imgur-album' width='100%' height="750" frameborder='0' src="http://imgur.com/a/JUQvK/embed"></iframe>

 

This testing isn't working out due to 4930MX having such asinine temps. I'm getting some great benchmark scores on the CPU today, but having to use AC cooling to avoid thermal shutdown at higher overclocks. It cannot pass the Intel Processor Diagnostic at 4.5GHz due to the thermals being out of control. This was fully expected using something other than Liquid Ultra on this nuclear reactor turd that Intel haphazardly calls a CPU. I'm gonna have to get the Liquid Ultra back on this thing, but I'll do some gaming at 4.3GHz and report back on the thermals for that so the folks that are still too scared of using it can know what to expect with IC Diamond (also an excellent product, but Haswell just calls for drastic measures).

 

Good to know Liquid Ultra still holds strong even after nearly half a year of heavy usage. Though ever since I discovered that dilute acid cleans off Liquid Ultra like no tomorrow, I'm less concerned about it degrading into a puddle of white mess on my CPU and heatsink. I know I probably sound crazy but dilute acid really is effective at cleaning off any Liquid Ultra that has turned into white crud and just refuses to come off with anything else.

 

No more IC Diamond... back to Liquid Ultra.

Alienware 18 - 4930MX @ 4.3GHz with IC Diamond (98°C Core Max)	Alienware 18 - 4930MX @ 4.3GHz with Liquid Ultra (79°C Core Max)

I decided to go ahead and repaste the 780M SLI in the Alienware 18 with Liquid Ultra. Also decided to go ahead and test the idea of using an electrical tape shroud for the surface mounted components while I was at it. When the Liquid Ultra eventually fails (could be a long time since it shows no signs of degradation on the CPU at 5+ months in spite of some hardcore overclocking) I will post photos of what the electrical tape looks like when I take it apart.

Applied a small amount to the GPU die...	...took the residual from the brush and applied it to the heat sink

 

4.3GHz @ 1.16v *drool*