Did a successful repaste with Liquid Metal TIM (Thermal Grizzly conductonaut) on my P670SG. The results are amazing.
Before (IC Diamond, 2 year old) -
Prime95 small batch test (Max heat generation) max temp hit 99degC within a few seconds, had to shut it off.
GPU Stress test (Aida64) max temp 88degC
After liquid metal -
Prime95 small batch test (Max heat generation) max temp hit 81degC, stable 10 minute run.
GPU Stress test (Aida64) max temp 71degC.
Thanks @iunlock
-
-
-
I've got it all ready to be used; Conductonaut, CLUU (if needed), electrical tape and Fuji pads. The only thing that worries is me is that I travel A LOT with my laptop (pilot) and that I fear that the LM might be shaken off/dripping off somewhere when shaken too much while it's in a vertical orientation.
I'm most def using Conductonaut on die-IHS part, but how wise is it to use it on the heatsink... -
The grizzly kryonaut is just as good and for your worries I would just go with it instead of the LM just to ease your mind. I honestly believe LM was meant for stationary desktops that don't move all that much. -
Sent from my SM-G935F using Tapatalk -
-
Sent from my SM-G935F using Tapatalk -
Meanwhile we get a report of a user who says that his leaked out as a result of a laptop dropping and then when proposed the idea of placing the TP between 2 pieces of metal in a vice and dropping it ect to see what it would take to get it to leak it was stated by some people that it was expected that such a test would result in a leak.
-
I am inexperienced in all matters LM, but also far from stupid. I'm confident I can make the right choices and I'll practice on my old P670 first.
Now bear with me; if it is a fair assumption that leakage *could* happen in a worst case scenario, how do we mitigate that? I have an idea, tell me what you think.
LM, like any liquid, has a certain surface tension. I propose the following: I will find a sponge-like material that is very eh.. what's the scientific word, bouncy? and heat resistant to at least 105 degrees Celsius. I will test this in my own oven at that temperature for an hour or more and see if it melted at all.
This sponge material will then be cut into an open square that envelopes the die and gets clamped between the substrate and the heatsink, or IHS (both for the CPU and GPU). If cut flat enough it will not reduce the heatsink pressure on the GPU or the IHS pressure on the CPU die as those cores rise above the substrate.
This construction will allow air pressure to equalize (might be kinda important being a pilot and all and not wanting air to expand too much for example inside the IHS and therefor not wanting to glue it shut completely (is there a vent hole at all anymore in the IHS?)). Also the sponge membrane will catch any LM runoff where the LM surface tension will hold it within the membrane.
I don't want to use the silicon glue/gasket gunk as it hardens and I might have to repaste at some time and don't want it glued stuck too much.
Any somewhat 'breathing' material that resists 105 degrees Celsius and is easily compressible might do, as long as it is easily shaped and holds that shape while it is being applied.
Ideas? (Like on what kind of sponge might work) -
cut thin strips of it out or something?
Your main issue with that is going to be trying to cut it down to 1.0 - 0.5mm thick or smaller. -
OMG! Heat Insulation Sponge, its all there in those three words!
I'll look into that. It may of course be a little thicker than the die is high as it should be compressible, that's what creates the seal. As long as it is not compressed to be more of an impermeable solid again.
And one could even lay down a very thin strip of silicon glue and then glue thin shavings of that foam on it, making it a seal and a complete ring. -
Tell me this is a troll...
-
Sent from my iPhone using Tapatalk -
Well, I like to stop any LM leakage before it happens. Given the punishment a well traveled laptop gets (and I know) it is very reasonable, for me.
gOdFaThEr9 likes this. -
Yea I am interested so see you results as my laptop is placed within 3ft of 2 15" 1000w subs two times each day while in a vertical position. I worry that the low frequency vibrations will cause the liquid metal to leek out.
-
I don't know why all this madness about, the cure time is 30 min (at 75°) after this time is not liquid anymore
From 2012 at today 4 cpu repaste and various chipset with no floodingPete Light likes this. -
If it cures, why do people talk and/or worry about it dripping?
And you could have done a million repastes with LM, as long as none of those setups involve the system being subjected to extreme vibrations while vertical it is not representative for conditions on an airplane or near high vibration speakers.
These are valid concerns under very specific conditions. -
My laptop travel almost twice a year by airplane with me (Italy to Iraq ), anyway I'm talking of Coolaboratory Liquid Ultra, as you would see from my post in another thread. The CLU remain liquid for some time if not thermal stressed, but its cure time is very faster compared to other thermal compound -
1. I don't know you, never heard of you and haven't read that one post where you mention what LM you use so mentioning that isn't very useful, even less so when hinting at that mystery post and not pointing me to it. I like to learn though...
2.Would there be more talk there of curing? I do think that curing CLLU would beat non-curing TG-Conductonaut on drip-safety when not inside an IHS, right? Unfortunately you're the first one that mentions curing LM, so excuse me for wanting at least one or two others to chime in on that before I put my faith in it.
3. Almost twice a year you travel... I make roughly 3 landings a week (long haul) with the usual turbulence in between to boot. I dare say mine gets exposed to more crap than yours, not that it's a contest but I 'win'. I'd like to be 100% certain that leakage will not occur or has been prevented with my dam idea. -
yeah you must dont know me(and me neithr know you), but I have only 50posts so was easy to find but I'm not talking only of my personal post, if you try to search here and there on NBR (mostly on the alienware subforum) is easy to find "our"experience from past year till today, where is easy to find that CLU dried out in almost all system where was used.
Plus dont take my word too personal for you, I've an extreme system with heavy susteined load and temperatures -
Now that you mention it, I do recall discussions about dried out LM. It doesn't really sound like a good thing to be honest. Only CLLU dries out? How can LM evaporate I wonder... it actually sounds like even more reason to use Conductonaut then and employ a dam for safety. I like my LM like I like my women, wet B)
aaronne likes this. -
I think when they are talking about dry out they are talking about the tim going bad before they can use it.
Once applied LM of any kind that I have heard of will last forever. The issues with tim dry out once applied referrers to issues from last decade where grease based tims would dry out and lose performance over time and require repasting. I do not think anyone even makes tim that bad anymore but maybe it is possible to go out of your way to find some. However, it is those who have ptsd from those dry out days that still think they need to repaste their cpu block every year lol.
It is sort of like in the WC community you get those guys who think they need to drain their loop once or twice a year. Yea if your running your loop with the correct chemicals draining it is not going to do anything other than allow you to keep opening up the loop over and over to increase the odds of adding new outside contamination to it lol. (note that some people do drain their loop due to the use of dye and other colored coolant due to issues created from the use of such dyes and colored coolants ect) -
hey man, dry out is intended as solid and in need to scratch before a new application, but the solid CLU layer had all its thermal property in range (after 2 year or 4 year the CLU had still good temp°if didn't exceed 80°for too long, I think it works forever)
-
I've read someone mention using a LM at the center of a die and a regular paste at the edges as a way to safeguard from leakage.
What do you think about this?
Kryonaut + Liquid copper? -
But I think it might be easier to just use insulating electrical tape. personally did the 3m super 33 tape cover (method suggested by @iunlock). It was easy.... And it worked out beautifully....!
My temps are hovering at 65degC for both cpu(i7-4720hq) and gpu(gtx 980m), on mass effect andromeda, 1080 at ultra settings....i move my laptop (clevo p670sg) around a bit, haven't had any issues yet.
Sent from my iPhone using Tapatalkiunlock likes this. -
I think using a dam would be the best way and a liquid one like using other TIM might work real well. -
Sent from my iPhone using Tapatalk -
Sent from my iPhone using Tapatalk -
Edit: I'm not new to repasting. Try not to follow that mantra where anyone asking question and open to someone else's suggestions is automatically a noob or stupid.Last edited: Mar 27, 2017 -
iunlock likes this.
-
Sent from my iPhone using Tapatalk -
Just mask off the surrounding area of the die(s) as suggested with 33+ and don't overdue the amount of LM used.
Stick with what works and is suggested here.As you can imagine, a lot of trial and error has gone into this. There are reasons for what is recommended and with good purpose and... Reason.
It's fun to be creative, but don't mess with mixing, lining and going experimental here.
The moment your LM comes in contact with another paste it will separate and cause issues. Tried and tested...
Also, use just the right amount of LM and you'll be fine. I've taken apart numerous systems and the LM stayed put very well.
The 33+ tape is just a free insurance policy. That's all. It's not to be heavily relied on or used as a crutch to undermine the importance of the actual amount of LM applied.
Happy repasting...
Cheers
::iunlock::gOdFaThEr9 likes this. -
Condescention is not my favorite dish and I hate sounding like an old fart by bringing up being into hardware modding since the C64.
It's quite clear already to most folk that it's not for the faint of heart.iunlock likes this. -
I'm stilll very much into containing possible LM runoff and will be open to any suggestion (including foil? as I read somewhere here). My system gets quite a bit of travel abuse (my avatar is well chosen, except the skin color, unless it's a tan) so if any run off could exist, it will, it'll shake loose I guarantee. Murphy's Law most def applies to me here.gOdFaThEr9 likes this. -
My main daily driver (MBPr) has been LM'ed for years and that thing has gone through some major turbulence. I've opened it up a few months back to check and everything stayed put very well without any leakage. (Using the right amount is the most important thing here.)
Also,... Well I have a crazy, but logical theory in why blocking off the perimeter can do more harm than good. Pm me when you can and we can chat.
Cheers.
::iunlock::tijgert likes this. -
iunlock likes this.
-
^please update thread after with what you learned.
-
Best advice too is to apply a tiny bit to the heatsink first well away from the mobo just to ensure there are no air bubbles in the tube causing it to shoot out unexpectedly.
I've done four or 5 repastes now mostly on the same 2 units and every time I've had very very little run off the die. I suspect the only reason it did was when I took the heatsink assembly off or put too much on like iunlock said. The LM has an excellent surface tension
Sent from my SM-G935F using Tapatalk -
Sent from my iPhone using Tapatalk -
This is my AORUS X5v6 with stock thermal paste:
This is after a repaste with CLLU:
Test done with the CPU @ 4Ghz and the GPU with +100Mhz on core.
I'm not happy with the lack of improvement in GPU temps.
I want to repaste the memory modules;
What's recommended?
Also any advice in regards to the lack of improvement in temps on the GPU?
Edit: Got to this thread via google, just realised it's in the alienware thread. My apologies, still would like some advice. Thanks.Last edited: Apr 1, 2017 -
@iunlock Quick question regarding LM, is it ok to use LM if im going to be moving and traveling with my laptop constantly or would a paste be better? Thanks
-
If it makes you feel more comfortable, I'd also recommend going with Kyronaut. It's pretty much fail proof with no worries at all.hmscott likes this. -
Thanks for the quick reply. Will give the Kryonaut a try first and see how that goes. Thanks for the help greatly appreciate it.
iunlock likes this. -
Just follow the repaste guide and you should be golden.Vasudev likes this. -
jeanjackstyle Notebook Evangelist
Just to share a bit of my experience with CLU: I have been on a p650sg for two years and a half now, repasted from day 1 with cooling liquid ultra, and I move around a lot with the laptop in my backpack (home/work on bike, car and plane trips four times a month).
first: I can confirm that CLU does dry out, at least in my laptop: after 8-9 months I would say, thermal performances are not as good as before (65-->80 °c on gpu for instance), and when you open the thing, LM has turned into a hard cake and was blackened especially in one corner, both on gpu and Cpu. Part of it was still liquid though. And cleaning it was not that hard, though polishing the heatsinks was indeed tiresome.
Second, I did experience a leaking of LM in my laptop. After one of these repaste, everything was looking fine, but after a few days I did notice that my 980m disappeared from device manager, and that every field in gpuz relating to this gpu was empty. So I did open the laptop, removed the heatsinks, and while Cpu was ok a big blob of CLU was making its way around the die, effectively shorting several transistors. After cleaning it, and repainting the die, everything runs fine, never had an issue since, performance is the same.
I don't know if I have been lucky, but several days with this blob out did not damage the laptop, while effectively shorting the gpu.
My 2 cents -
It is true that with CLU, if the contact is not good between the die and HS (gap), due to the heat and air (rapid temp spikes from one extreme to the other), the CLU will oxiliquidize? (Just made that up, but hope yall will get the drift.)
I have yet to experience this with Grizzly Conductonaut (TGC), but it's not to say that TGC is exempt from the possibilities.
The main thing here is to make sure that when using LM, that the die is making an even contact with the HS. If not, LM is not recommended. Instead, use traditional paste like Kyronaut or Gelid Extreme.
LM surprisingly cleans up pretty well off of the motherboard, but you didn't hear that from me. I've done a lot of controlled 'what if's' on an experimental mobo with leakage scenarios. If LM gets up under any of the chips, bye bye mobo...
For anyone using LM for the first time, you own it to yourself and motherboard to;
1. Ensure that your die is making an even contact with the HS. (Use pressure paper if you can...it's worth it to know the details.)
2. Apply LM, screw down the HS, let it sit for a few minutes and take the HS back off to gauge the amount that you have used. Too much? Too little?
3. Keep in mind that when you lift the HS off, lift it up as straight up as possible to avoid LM getting onto the mobo. You can also rearrange the LM again and reuse it. Add more if needed.
-
As for my experience, I have never had any issues with TGC and it's been perfectly fine in and out of airports. I've opened up numerous systems that have been LM'ed for one reason or another and they have all stayed put very well. CLU is a bit thicker (higher viscosity), but that doesn't mean anything, because TGC is a lower viscosity (more liquidy) but it also stays put just as well.
LM has a smaller tolerance level and is less forgiving than traditional paste. True. Therefore, take caution and your time with it...if done correctly, it's fantastic. If not, then it's a nightmare.
In fact, I was connected with @nerdstaz all day as he did his first LM on his AW and did a great job. The temps are fantastic and the core differentials are amazing. Darn near even. Success...
Look at this FS run at 44x, untuned and staying in the upper 60C's and scratching 70C....pretty darn good.
Here is his run at stock clocks:
Last edited: Apr 7, 2017Pete Light, pressing, MiSJAH and 2 others like this. -
Can you point me toward that 'pressure paper'? Where can I get it and does it have an official name/brand?
-
I used CLU on my cpu and after few months when I cleaned the heatsink it had solid bumps formed on it. I tried to sand it off with included abracive pad but CLU bumps are really stuck to copper. I guess I have to do a heatsink lapping now.
Anyone have an idea why it solidifies like that? -
jeanjackstyle Notebook Evangelist
You can look at several topics here, the globally accepted idea is that with small gaps/ low heatsinks pressure (especially on laptops) part of CLU is exposed to air, and when it heats up it liquioxidize (thanks for the new vocable iunlock !). Meaning it bakes into this solid gump.
You can either lap your heatsinks (but I found out that CLU did diffuse a bit in copper, so do not sand them too much waiting for a change in color), for me soaking for some time the bumps in acetone or isopropyl alcohol usually allows me to remove them with a piece of cloth. -
Will try acetone next time.
After lapping I saw that it made little holes (like rust) into my heatsink. Just around the die where excess clu were pushed out. Most important is that temperatures are still same. Even if it gets mostly solid, it still performs.
[Liquid Metal Showdown] Thermal Grizzly Conductonaut vs Cool Laboratory Liquid Ultra / Pro
Discussion in '2015+ Alienware 13 / 15 / 17' started by iunlock, May 11, 2016.