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

Well it will depend how you take it to use, if you make such example: "give pistol to monkey and nobody's safe", same applies to all more serious stuff that works, just need be cautious and know how to use things to prevent bad happen. Few theoric options: train your monkey, dont give a gun to your monkey or just dont allow monkey to get to close to you as "bad can happen"

 

^^^I agree. And, I killed my monkey already. That sucker pissed me off one too many times.

It's totally worth it for the CPU, just don't make a mess and everything will be fine. There is no reason for anyone to get the Liquid Ultra on to any surface other than the die and copper plate unless they are not careful or don't follow instructions well. The GPUs have surface mounted components that are too close to the die for my comfort, but there is no need for enhance GPU cooling.

That video scared me off a long time ago, but I saw too many examples of good results. The video was good for creating proper awareness and respect for the product and was a good incentive to be careful.

 

I suspect that was the intended message. It is rather disconcerting to see though. Truth be told he did put a lot of it on the heatsink but wow, I'm sure one wouldnt want to use them on their desktop counterparts with Aluminum Heatspreaders......Or were they Copper?

 

it only reacts with exposed aluminium, so be careful and dont use it on aluminium heatsinks, and only apply a minimal necessary amount on a copper or nickel base and you will be safe. Whoever is finding info about this thermal compound would read the description of the product which clearly indicated not to be used with aluminium heatsinks. Although its ~60% of galium in this TIM its still dangerous to aluminium, so whoever tried most of the cheaper compounds on their desktops, they must have already ugraded from OEM aluminium heatsinks to something better with copper base heatsink and now upgrading to this TIM, would always find label saying where not to use it or people warning about, like we are doing now

 

I always wondered, the heatspreader on desktop CPUs. Are they Copper or Aluminium?

 

depends on the model of cpu and of the heatsink, google helps alot when you know the cpu model we're talking about and if its OEM heatspreader, otherwise refer to specific model, and again - google makes wonders when it comes to searching handy info such as heatsinks and such

 

Cool, thanks

 

You practically need to dip the aluminum into the liquid metal for it to be corroded. Aluminum has an oxide layer on it, if you don't scrub it off it forms a protective layer.

 

Finally applied some of this Liquid Metal. Wow oh wow!

4.5 ghz 1024M Bench is a walk in the park. Machine hovers on 80 degree's. I think I might be able to drop my flex down too now. Amazing stuff.

For those of you interested. I only applied the metal to the cpu. I didn't coat any of the heatsink.

UPDATE: 4.3 the bloomin' fans don't even bother spinning up lol

4.5Ghz as my new 24/7 OC lol . Need to work out how if possible to get 4.8Ghz for performance mode.

 

Sweet! Welcome to the Liquid Ultra club, Brother TBoneSan. Amazing doesn't begin to describe the results.

 

I am also no the liquid ultra. 4.2 24/7 Vid 0.. Loving it.

when i applied the liquid stuff on the cpu die as i was brusing it, alittle bit was brushed on the green part of the cpu the sorunding of the die, very very little i cleaned it right away with an alcohol whipe. Should i be worried? i dont think thats aluminum

 

Awesome, Brother Juang.

That should not cause a problem and I wouldn't be worried.

 

Try a wPrime 32M at 4.7GHz and see what your temps look like compared to my 3 week old application that has been run hard. Should be interesting to see how they compare with your fresh application. I just ran this one. No special cooling... just the CPU fan and Liquid Ultra. Flex was set at 75 and Pri Plane at 180A (would be 1440 in BIOS) using ThrottleStop.



http://www.rapidtables.com/calc/electric/Amp_to_Watt_Calculator.htm

 

Thanks mr Fox,



has anyone tried to clean this stuff out ? Like repasting? wonder how hard is it to clean it up ...

 

I can't seem to reach 4.7 with any kind of stability. I keep getting BSOD's.I've tried ranges of Flex 70 - 200 with no such luck. I've also had to set my Pri Plane ( core current limit ) stupidly high to deal with programs crashing.
I did manage a 32M run last night in TS but that was a once off.

Hmm not sure if I'm maxing out my silicon or equally likely user error.

Can you have a peek at my XTU settings ( Flex is being set higher in TS my performance profile)

 

Those are way off, bro. Try setting these value in XTU first, then set ThrottleStop to match. Flex in the BIOS is XTU "Additional Turbo Voltage" divided by 3.90625, so 234.375mV is 60 Flex in the BIOS. Core Current Limit is Pri Plane. Your Core Current Limit is way high. I usually set my BIOS/ThrottleStop at the high overclock values and use ThrottleStop to lower them for profiles below my BIOS overclock. Going up in values often does not work as well for me, but going down for the lower ThrottleStop overclock profiles works well.

 

I'd like to see how it improves gpu temps at extreme high clocks for anyone brave enough to do it

Sent from my SM-N900T using Tapatalk

 

I believe ht_addict and [Nikos] already repasted their GPUs, but I don't know how they feel about extreme GPU overclocking. Actually, ht_addict has 7970M CF, so extreme GPU overclocking really isn't possible.

 

Thanks Bro Fox,
I'll try that when I get home this afternoon. It sure is fun and a little unbelievable being able to run these clocks without worrying about thermals. I can't wait to see what temps 4.7 hits - if it can keep under 90 that would leave some nice wiggle room for higher clocks again. I'm also wondering if we'll see any temp variance based on TIM application - you skim coating the HS and me not.

@Dandan

I didn't have the balls or wallet incase things go wrong. I might do it when I repaste this CPU and I can get an idea how or if it spreads.

 

At last

If time allows, tonight is the night for the new paste-job!



will go with LiquidUltra at first and when it runs out if it ever will, then will try the LiquidPro.

Something I kept on finding on random dark corners of the web about Pro curing and bonding between surfaces while Ultra version does not or if it does its very easy to remove and clean, as where Pro needs to be scraped and buffed off kinda way... sceptical myself now about Pro version, but will give Ultra a try that's for sure

Im gonna use it on unlapped unpolished standard 3pipe heatsink that can not obtain a sweet-spot with regular TIM's resulting in hotter temperatures during the load than any other people 3940XM's I have seen, so its time to find out if this stuff will patch the "donkey" heatsink I have in my alien

 

Ultra is a newer and product and supposedly better than Pro. If that's the case, I don't understand why they even manufacture the Pro any more. Seems kind of silly.

 

True... well there's a part of people that get stuck/psychologically attached to something they know that works, so I assume they continue to manufacture/sell the PRO due to such reasons, or they may have vast amounts of PRO version still available and are trying to get rid of what's left and then discontinue? nobody will tell you the real truth

 

Here is a 4.7 run. I might load up an earlier Windows image as I'm finding it hard to get stability.
Temps are friggin great

 

Ultra doesn't need to be sanded and buffed out like I've heard pro needs to right ?

Sent from my SM-N900T using Tapatalk

 

well they say it doesnt have to be and it doesnt stick as much and people describe LiquidUltra as easy cleaning vs LiquidPro.

I have just accomplished my repaste with LiquidUltra, and well at the moment had pretty hot room here ~25c + and laptop was on U3 pad with fans off, went in to bios and quick set Level3 factory OC and left alone while multitasking other things on laptop, CPU never hit more than 80c @ 4.0GHz x 8 core TS Bench hovered between 74 to 80c which I'm surprised, fans were setting on and off as laptop would just idle before quiet as sniper

Anyway, I think my donkey heatsink requires proper lapping BIG TIME, its not the first time I notice this crap pattern push out the TIM everywhere but not at this one specific bubble area. The paste you see is ~2 month old PhobyaHegrease Extreme (aka GC GELID Extreme)

The massive burn-in mark on CPU heatsink has been left since last time I applied freshy-fresh ICD portion and went playing with CPU OC tinkering spells for a few days, now as I think - could have a great-great thermal transfer have warped the heatsink by ICD? anyway ICD was failing me @ 4.4-4.5Ghz if recall correct as I was having thermal shutdowns at the peak temps past ~100c+ somewhere, and then ran out of play time with the options due to family commitments, work, household and such, eh, so I did repasted the CPU one more time since with P.H.X. and well, here you see great signs of ageing, where ICD was still looking fresh when I took CPU HS off ~4months later since application.

You can have a look at some snaps I took during this evening:

















I have cleaned heatsink and CPU die surfaces pretty well with 99.9% pure Isopropyl alcohol, and used very little amount of LiquidUltra - size of half rice grain if not less, then carefully spreaded it out trough the die with little brush provided(kept CPU on paper towel outside the laptop in case one small random dizasterous drip woul'd have decided to do some wild run trough the nearby components ) .

Before I put in CPU back to its socket I have used thin line of automotive RTV silicone gasket maker to make little border just around the CPU die so this which-craft TIM has no chance of escaping his workplace even if it wanted to (used it before on some cars, so know how useful it would be in this application, even after its cured its easy to take apart and its still soft)

Now its time for me to crash to the pillow, tomorrow's gonna be a hard day and well, again, if time allows, I may do some proper OCing in this.

Well, considering the amount required for repaste - this tube should last at least 5-10 repaste cycles which is alot, but only thanks to viscosity of the TIM, otherwise it'd be other case

Is it just me, but do I still think my temps are high(higher vs others) compared to folks who tried it in this thread with same/similar cpu? I blame my donkey Heatsink, look at it and have your judge

nite all!

 

i dont know if the burn marks from IC diamond would affect temps. accounting to mr fox and others it shouldnt.

 

Damn thats a lot of TIM on that. The burn marks shouldnt be a problem as they are not in the region of contact.

 

That simply looks like stained copper to me, not burn marks. Mine had some dark color around the edges of the paste, but not nearly that much. As long as the heat sink is not warped I cannot imagine why the copper having a different color around the perimeter of the die would affect the cooling. I suspect it is merely a chemical reaction that caused the bare surface to tarnish.

Many that PhobyaHegrease Extreme (aka GC GELID Extreme) really dried out fast for 2 months. I turned to what looks almost like cracked dirt from the bottom of a dry pond. I haven't seen paste do that in a long time.

 

well there you go you've seen everything now, and take in account I dont game at all, just desktop use more than 12h/day everyday, and most times more, but hey what do I know... maybe just bad batch of a paste? IDK, now will see how well LiquidUltra will last. What do you think of my temps?

@ standard clocks TS bench 1024



@ all standard bios settings but TS setting only raised multi's to all cores @ 44

(also both tests are done on U3 pad with no fans powered on)


I will install XTU sometime today and will try to use one of the 4.5+ GHz profiles if ,Im allowed by you's, to test my system(pm me known satable XTU profile please anyone..?)

BR

 

Hard to tell what your temps are really like at only 3.7 Ghz. I've found my idle temps actually up a tiny bit but under load is where it's at. Can you try 4.3 or something? You shouldn't need XTU for that. If you want I can PM you an XTU profile for higher clocks.

Side issue I was having issue with TS locking up my computer when launching. Even after deleting it making a new launch directory to create new settings. I don't believe the crash is due to incorrect settings. It all happened around the time I was trying to hit 4.7ghz.
I may have fixed it before I went to bed last night by deleting/disabling iRST cache acceleration and starting it again. Anyone had this issue before ?

 

how are you at 1.5v at 4.7ghz? what flex are you using and are you at 47x multiplier? I tend not to go towards 1.4v on my cpu it'll just kill it eventually due to heat, I have mine running at 4.7 at 1.388v but thinking to drop to 4.6 to lower that voltage even more.

edit: nvm apparently mr.fox is also at 1.5v, maybe cause I don't OC my GPU.

 

Well, we're approaching week # 5 and it's still doing a superior job. Muy bueno!

 

Im myself pleasantly surprised how cool my laptop idles most times I see it jumping between 29-45c and if heavily loaded on some multitasking never hits beyond 70c, I have also went in to bios and set fan tables to kick in at higher temps as its just not necessary for fans to spin up earlier where temperatures never cross the limits or even if they do it would be for a split of a second.

Very happy with this TIM, and considering Im using this "donkey" heatsink Im finding temps pretty good! Next time repasting will do serious lapping on this heatsink to see if we can improve even more

 

I emailed the company and they said there is no need to repaste unless you change out the components. It does not loose its effectiveness over time.

 

well that will be surely tested under some of our heavy loaded beasts over time, but other than that this is even better news

 

Never need to repaste..?! Never looses effectiveness...?!
That sounds too good to be true. No complaints here

 

you guys talking about the pro or ultra? iirc they are two similar but also very different TIM.

 

Were OBVIOUSLY talking about ULTRA, what are you referring to???

 

I was thinking, I know many say don't re paste gpu with this, but what if we put masking take around the border of the die to cover those metal parts while we paint and spread this out, like painting. Then once done, remove the tape and put down the heat sink. Think that will solve the danger problem of application to the gpu?

Sent from my SM-N900T using Tapatalk

 

read my post 5 my posts back

 

You can use a silicon dam like Brother UltraGSM did or masking off everything during application would be safe, too. If your GPUs are not running too hot it is best to leave them alone. They are more sensitive to ESD and sometimes they can be finicky making good contact with the PCI-e slot and start beeping, etc. It's not worth the risk of hassles if everything is working right, but if you need to repaste then I have no doubt this stuff will do a great job on the video cards, too.

 

from what I've read pro is easier to clean up than ultra as they are two different TIM. as per tomhardware pro is better than ultra in most cases in cooling.

 

technical specs of the compounds disagree with your statement. Only the time will show how easy it is to clean (if we actually need to replace this TIM ever haha!)

and many tests I've seen performed with both TIM's Ultra was 1-3 degrees ahead in almost all the tests so thats why I cant get the point where have you found so much info outweighting the LiquidUltra vs former TIM?

 

no idea, but I put tomshardware a lot ahead than many other reviews, I mean its tom we're talking about lol

do you have both ultra and pro and your ultra runs cooler?

 

Ive both on hand but only Ultra on application and I'm not feeling too keen to try Pro just for a "crack" to find out. Enough to read the label, if it says ~"after any amount of time removing this application will require GRINDING TOOLS AND METAL POLISHING AS USED IN AUTOMOTIVE INDUSTRY", so dont need go taste the Mexican chilli if you know it will burn your throat inside-out I think. Knowing Pro dries and creates solid bond between surfaces puts me off even thinking about it putting it on most expensive laptop CPU DIE I am referring to manufacturers manual details enclosed to the packaging of each. Where Ultra TIM doesnt mention a thing about bonding, just easy cleaning.

Hey why dont you give Pro a try, you know , we're Tom talking about here, right?

 

I'm not Tom, my name is Yoshi, nice to meet you.

 

i was actually thinking of that. with it being solid imo the TIM last quite a long time, one time application sort of thing.

 

Would be keen to see someone try that and report later, months down the line how it performs etc etc, and if removal needed experience and such

 

Hello all!

So i order liquid pro and liquid ultra maybe will some test my M18xR2 CPU,but little fear...
Anyone see heatsink result when again take apart?
I have seen some terrible results... CoolLaboratory's Liquid Ultra on a Macbook Pro - Jurriaandevos.nl

 

It's awesome to see how this TIM is holding up!

Do you use Throttlestop for monitoring the temps of the cpu rather than HWINFO? I see TS displaying temps around 8 degrees lower than HWINFO does usually.