I was going to lower my voltage for better performance but since it's a laptop I can't do that so I need some other way of getting better performance...
Actually Hardcore Overclocking I watched a video where TimmyJoe applied Liquid Metal to JUST the CPU surface, ignoring the IHS - I of coirse left a comment that he did it totally wrong, that due to the surface tension effects of Liquid Metal, ot MUST be applied to all both sirfaces that will contact each other. Not too surprisingly, his temp results were mediocre for Liquid Metal.
Actually Hardcore Overclocking *NEVER APPLY LIQUID METAL TO ANY SHUNT YOU PLAN ON USING FOR MORE THAN A YEAR!!* Louis Rossman hs a video where a Customer sent his Titan Xp to be repaired, and before then, Louis had no clue about Liquid metal, so at first he thought Nvidia just did a TERRIBLE solder job. Well it turned out that the Liquid Metal ATE AWAY AT THE SOLDER! So if you are going to be lazy, and Liquid Metal the shunts, you damned well better cover the solder pads of said shunt in Nail Polish! Or yanno, you COULD simply run a LITTLE jumper wire on top of the shunt and solder it, then tape it down.
Protip for application if you find the liquid metal is not flowing easily: Heat it up a bit. Gallium melts at 29c - and liquid metal has some additives that probably brings this a bit lower, but that's why it can be sort of syrupy at common room temperatures. Heating it up a little bit (and the CPU and the IHS too if you want) will make it more flowy and easily spreadable. You can heat the tube up in a glass of warm water, or just get a hairdryer on medium to heat up all the parts for a few minutes before you go to work.
“Because I keep seeing PEOPLE spraying liquid metal all over their stuff, like motherboards...” just say Linus, we have seen the video of him being a over excited teen on his first time with small tube in his hands.
in fairness, I think Steve at Gamers Nexus also did that to an x299 CPU some time ago. I don't remember if the Devil's Canyon also got the LM spray as well, or if that was someone else with LM. I have seen it a lot in vids around YT. They get LM, they get excited, then they squirt it all over there computer or laptop, lol.
@@Zarcondeegrissom they're both noobs. Only trust Roman aka Derbauer and Kingpin to do proper Liquid metal install videos. They're the real experts in extreme overclocking. Even the idiot doing this video is using a regular q-tip instead of the standard black cotton swabs by DAISO Japan that Thermal Grizzly includes in their Conductonaut.😏
I'm not gonna lie, this helped me immensely. I actually got a great paint job on my cpu and IHS. So my thermals are finally reasonable when attempting to run prime with my i5 4670k. So thank you buildzoid for making this video. Autofocus be damned.
Filler or not...good video. Thank you. Best liquid metal application video I've seen so far...and I've watched many dozen. How is your subscription count not over 1 million?!
Oh, and no matter how careful you are in applying liquid metal I'd totally recommend you insulate any contact points. It's a really solid failsafe for 5 minutes of work. The material cost (some conformal coating or similar) is trivial and really easily availiable - and if you just spray some in the cap and paint it on (preferably 2 layers 15 minutes between) it's also trivially easy to do. Even if it's just a 1% chance of the liquid metal moving that's a worthy insurance for an expensive CPU (and I've actually seen quite a few cases of it moving. Mostly probably because people apply too much of it. More user-error than the compounds fault - but then again your CPU doesn't risk frying if you make a sloppy mess with most thermal compunds. This absolutely does.
Thank you so much for this video, BZ! I just completed delidding my 8700k, and like an idiot, I would have tried to sqeeze the LM onto the die. Wouldn't have occured to me to squeeze it directly to the qtip. It worked great, and I'm sure saved me a mess to clean up! Thanks very much!
Thank God I found this video. Just bought a gaming laptop with a shitty liquid metal job and was going to redo it myself. Line you said, everyone just blobs it. This is surprisingly obvious after being shown but I never would have thought of it without your guidance. New fan here!
So it being this thin is not an issue in terms of it being able to fill all the imperfections? I guess under mounting pressure it should be enough even if super thin. I hear the miscoloration people seem to be so worried about is just a thing that happens from the gallium diffusing into the other metals - basically putting it's atoms in between in the gaps where it can fit in the other metal. Aside from being unsightly it should not have any negative impact on the thermal properties of the surface. Actually I hear it might even slightly slightly improve the thermal properties if anything - but mainly neutral. Diffusion in itself is not a problem in most metals - just the metals where it causes the other metal's bonds to weaken (which is what happens in alu.). The way it acts differently before and after the surface has been exposed to gallium before is no doubt because of this wetting effect the diffusion has. Before it's like water on really dry ground - it will tend to skip across, while after it's more like damp ground where it sticks to it more easily. That's obviously just an oversimplified analogy of the atomic forces involved here, but wetting and doping of materials (which is basically what happens here) is a well understood phenomenon you can read up on if interested. If anyone happens to have any hard data/testing to show results of how thermal conductivity of copper and nickel-plated surfaces change (of at all) after gallium doping then I'd be really interested to see it. I imagine because of this - if you wanted to be hyper-optimal you might be able to get a slight improvement if you bothered to do a re-application a month down the road from the initial application since at that point all the reactions will be done and the materials saturated. Probably unlikely to have statistically measurable effect (with the impresice measurements of CPUs we have to work with) since it's likely a trivial amount of the gallium that is diffusing in, but if for some reason you should happen to notice any drop in performance and you have eliminated all other factors (ambient ect.) then that's the first thing I would consider. After the second application I have no reason to suspect it shouldn't remain stable for a very long time.
1 very important question I would like to ask : You show how to spread it out perfectly, but what about air bubbles in between the IHS and CPU core ? Let's say the core is 15x15mm, it's 225 square mm. When air will be in between the IHS and core, the actual contact surface will be less then 225 square mm, and so the heat transfer is less good. Do you share my opinion ?
Just like he said it's not like traditional thermal paste. And the rules of the traditional paste bubble issue do not apply here. The rule is you want the liquid metal to be flatten enough that it sticks in the heatsink just like what paints do to your walls and by doing this the liquid metal will fill up the small gaps of the IHS (IHS if your using it in between heatsink cooler and cpu IHS). Then since it's not like thermal paste that will easily stick to the heatsink cooler cold plate. You also need to apply another one to the cold plate so that both liquid metal from top and bottom will touch each other and there you go you have connected a way for the heat to transfer.
I dont think thats much of an issue, what is in issue, and you can see it in video, is that maybe he applied too little. First time ever I did this, I applied too little and saw literally no difference in temp delta, and I applied around the same like he did here. Second time I applied more of it, so it was actually visible layer in there, not just like a reflective paint surface like in video, and I saw like 10 degree difference. I did however apply two layers of nail polish to protect the pins that were too close to the die, and since then I did that every time I delided any cpu afterward.
I've been looking for proper videos on how to apply liquid metal. All the top ones about liquid metal are people who don't know how to use it and blaming it on the liquid metal. It's good that you made a proper video, since I will be putting some on my gpu.
Thank you so much for this video. I want to apply liquid metal thermal paste on my PS4 Pro APU. The damn thing gets too hot and the fan is loud like a jet engine.
30 seconds of gentle cleaning with some very fine metal polish like autosol will 100% get rid of that stain on the IHS ... works perfectly on my nickel EK blocks
If the liquid metal is affecting the chip, in a fashion, that it is easier to apply a second time, could it be causing damage, that might cause problems down the road?
The stain only goes a few micrometers into copper, and doesn't affect structural integrity or performance in any way. The "drying up" thing people ovserve and performance degradation is just oxydation due to air getting in there, since fit between Heatsink and IHS is far worse than between die and IHS or in direct die cooling in laptops or GPUs.
Well the first time I used it, I put it in the freezer for 5 minutes or so and it was easy to apply. I think that heat is not a factor concerning the adhesiveness of liquid metal thermal paste
So what is your view on using a custom copper IHS with liquid metal? Would the LM diffusing into the copper cancel out any benefits from the larger IHS surface area? With 1.35v adaptive LLC 6 my temps are still not amazing even with a 280mm cooler.
How about applying on heatspreaders, between heatspreader and heatsink? Also, mine is a bit coppery, polished it some time ago - I know that it will form an alloy with copper, do you know if this process is ongoing (and I should re-apply it after some time as it "went" into the copper?)
how accurate those people saying the liquid metal applied might slip or come out and might touch the transistors ? of course a talking about liquid metal installed on a laptop with consistent move and shacking around and so on... could it happen ?! or is it really sticky ?
What about using a metal dish to as a delivery method? Squeeze into the dish then transfer to the surface and spread the with a cue tip on to the ihs? Then electrical tape right?
I used the brush included with Coollaboratory Liquid Pro, but this looks to be a superior method due to less spotty spread. I don't remember exactly what I did, but I recall squirting onto something and dabbing from there, like an easel. This way the brush didn't get gunked up, which I saw was an issue from reviews.
i used LM on both CPU and GPU, i used the tray tecnique, i squezzed a blob on an nickel plated "tray" (the ihs from a g3258) and then picked it up with a q tip
Old vid, I know but hopefully someone will see this and answer. Arent you worried about fibers from the q-tip from becoming stuck in the applied liquid metal? I know i sometimes get loose fibers from those when cleaning things with a q-tip and say, iso alcohol.
It’s not autofocus fault. No way it’s going to know you want to focus o this tiny object in the frame. Cup your hand behind it and have your hand and the smaller object taking up more of the frame and then it will autofocus for you.
I think me cleaning the surface with alcohol made it much harder to get it to stick to a flat surface. Wonder if slightly roughing up the surface using very fine sandpaper would allow it to spread easier.
how often should i reapply my liquid metal? just when i notice temps being out of whack other wise its good still? ive had it on my 8600k for about 4 months now and its still the same how long do you expect that to last?
when the temps go out of whack would be when I would check it. I've never run liquid metal long term. AFAIK Thermal Grizzly expect it to last like 2 years before needing reapplication.
Nogghan Had my delided 4670K runing 1.35V 4.4GHz for close to 5 yeara and the temps were the same since dat one. Used CLP on the die and Noctua NT H1 between the IHS and coolers. And it was not sealed, just held in place by the retention arm of the socket.
only if you apply way to much but if you do it like he does in the video it wont run at all its a thinly painted surface thats why he didnt even bother covering the surface contacts with a conformal coating, he knows the small amount cant possibly go anywhere and then itll be clamped by the IHS and then its really not going anywhere
Ok, thank you buildzoid, for putting things into nice perspective, including the potential damage. ill stick to my normal Pastes since im not interested in delidding anyway.
MrXanthios it shouldn’t affect performance??? I’ve used tissues and q tips to clean out thermal paste numerous times and it doesn’t affect performance whatsoever.
I want to delid an i5 4590 and an i7 6700k and use liquid metal on top of the i7's IHS. I also want to repaste a 1070 and a Vega 64 with liquid metal. Should I buy the 1g or 5g syringe?
Never even thought of putting it on the q tip first lol. I always used to put a pin head on the die. Will definitely use this from now on and when i get my 8086k next week. I always reseal mine though just makes it easier for resale and my ocd. Couldn't have the IHS in there all wonky :D. But i must say scariest moment ever in my life was delidding my 7900x with homemade custom delid tool. Worked fine though. Same tool to do my 8086k next week :p
What do you recommend using for removing liquid metal when cleaning? (in terms of solvents if any)I imagine gallium really doesn't react much to conventional solvents.
My biggest concern is putting LM on the IHS. I have a 4770k, as long as I'm not being stupid, I shouldn't need to mask/cover the VRM's. I'm just a bit concerned since I won't have perfect coverage with the IHS like I would with the die.
you'd have to install the CPU with the system horizontal but once the socket hold down bracket is down that IHS isn't going anywhere. The main risk is applying soo much liquid metal that it spills everywhere when you put the IHS on.
Been running every delidded CPU vertically without resealing. It makes no difference if it's vertical or horizontal, it's not like it will drain downwards if that's what you're worried about. It should be as thin as a layer of paint, it won't go anywhere. Reasealing is pointless if you're just going to keep the CPU in the socket anyway. Reasealing might even hurt your thermals (check Gamers Nexus video, where even a very thin seal made it run a few degrees hotter than with no seal).
@Actually Hardcore Overclocking is it fine to cover whole zone around die with Kapton tape (to secure test points and other elements)? Substrate don't really heat from bottom, right?
It's not a silly question, but you don't need to worry, Tthe fibres are unlikely to make a measurable difference in temps, and wont impact any components of your system
I just used the small funnel in the package and then gently squeezed and applied a very small blob directly on the die then just spread it with a Qtip. I did it multiple times on a multiple CPUs, I never had an issue where the LM just shout out and made a mess everywhere.
Well there is still a human error there if you directly putting it in the die/IHS. Specially if your doing it inside the case and the case is not completely flat when titled (some case due to the stand design like mine) Then liquid metal will be happily sliding now in going down to the PCB. So Q-tip application before cpu application will have 100% no error. Unless your dumb and want to directly paint the PCB with silver coating ahaha
@@TrigrH Same, you can never rely on the syringe pressure 100% on these things, bubbles/pockets seem to be able to occur. Almost ruined an RTX 3070 with lm due to this, but managed to correct with a deep iso alcohol cleaning.
So you'd recommend against applying liquid metal on the ihs vs the bare die? Tempted to try out liquid metal but don't want to remove ihs to apply it bare die. My main curiosity is how it effects the longevity of the IHS
Why not use an *Acrylic Lacquer Conformal Coating* ? This is what it is made for. MG Chemicals 419C-55ML Acrylic Lacquer Conformal Coating, Clear ,55 ml Bottle - $12.95 at Amazon for a 55ml bottle. Which mL vs mL is cheaper than clear nail polish and it is a constant formula made for this type of thing.. This is what Professional's and Hobbyist alike use to protect their electronics.
I've never used liquid metal, but I think you may be referring to thixotropic flow properties. Like tomato ketchup, be careful when you squeeze the bottle!
Or you put the CPU without the IHS in the socket first then cover it with the IHS that way you can quickly close the lever and everything will be held in place without having any risks when moving the unsealed cpu, the IHS may slip out and bam, you got liquid metal everywhere.
Use a good quality q-tip. Or better yet, most liquid metal pastes come with high quality q-tips, use those. The fibers can cause voids, so quality matters.
Hey brotha! I've been trying to find info on uncore ratio and what I should be setting this at for an 8600k @ 5GHz, and what voltage is associated with uncore ratio to help with stability. Any pointers? I've read to keep it under the clock speed of cores not much info on the web.
with these testpoints under the ihs. what happens if you short them? is it like 12V to cpu or something? i took precautions and applied nailpolish on my 6700k
Jesus I've always thought of this, but you sound like Drift0r so MUCH. I don't think that's a bad thing and I love your videos man. Thanks for another great video. :)
I literally just tried this with Kryonaut on my freshly delidded and direct die cooled 14900KS hoping to avoid any long term dry out. The answer is an absolute no, with temperatures no better than the original Intel TIM plus Kryonaut. It was a very good thin coverage after dismantling, but replacing with Conductonaut literally dropped temps by 20C at 320W using a MoRa 420!
This felt oddly like watching Bob Ross painting videos, can you do an imitation video pls. “Now let’s paint some birds with the liquid metal, anywhere you like just have fun. *paints a liquid metal bird on all the critical board components *
Doesn't have to be a same video, but one that checks if it's possible to actually fit one under the IHS would be nice. If it's too thick, I don't think it will fit (I don't know though), but if it fits, it'd be interesting to see a performance comparison with liquid metal. Could to LM+Thermal Paste vs LM + LM vs LM + Graphene Thermal Pad vs GTP+TP vs GTP+GTP. Even if GTP + GTP loses out to LM+LM, if the loss is minimal, it could easily turn out to be the better choice simply because it doesn't have the inherent risk of LM.
Why do you need to apply to both and did you use that nail polish to protect? Also if you go bigger on the IHS the heat wont cause the liquid metal to drip down on the chip?
Pretty much. The reason is that liquid metal has a high surface tension (you can see it bead on the surface of the die if you apply it directly to the die) and when you put the ihs over the pcb, this surface tension will prevent the LM from "wetting" the underside of the IHS. Imagine trying to wash an oily plate with water. The water can't get on the plate because of the oil. While this isn't really a "correct" analogy, the effect is the same.
Ha! Ha! You covered the installation; but, how do you remove the stuff and cleanup later? The closest that I've come to Liquid-M is IndigoX and it's fairly easy once you melt it onto your CPU. Just a real "B" to get off.
So about the resealing.. Like if its for testing then I can see that doing it like this is no issue, but for a full time system is the tension of the hold down system enough to keep the IHS in place? For the rest, wel great vid!
Some say sand the die with grit 1500 gently... I should make the LM stick easy. I only tried in the past on a polished cpu, damn! I had to abbandon it and give up! Was impossible to stretch out... was just a "ball" silver that could not stick
You’re funny bro. Awesome informational video. I appreciate the tutorial very much. Also, I thought you couldn’t use Q-tips due to the cotton fibers. Glad to know that they’re just mini paint brushes, Very helpful
@@czdaniel1 well I did the delid added liqi metal and I used the grizzly thermal on top of the IHS in relpacement of the old thermal..you don't put liquid metal on top of IHS I'm using a Hyper Evo 12 it has copper heatsink..
@@Simmymind360 I see. We are using different terminology. The "I" in IHS when I speak the term stands for "internal" and is referring to the internal component between the transistors/die and the protective lid that you removed upon which your Hyper-Evo goes on.
Saif Sameer I'd keep it unsealed just to make life easier if you do end up replacing the LM. Gamers Nexus has also found you get slightly better thermals with the cpu unsealed due to the smaller gap between the IHS and CPU die.
This is the exact way to apply! Its unbelievable how many videos show them applying it to the die or ihs directly. 😂 no way you should risk doing it directly. Your way is the best and how i always do it! Another great video!
no, hell it's not worth in most cases even when pushing high oc on ambient, on the 7(8)700k it's worth only if you are trying to push it hard (1.4-1.45v) edit: as the user tommihommi1 pointed, you can use a lower fan speed if you go with LM
ehh if you are running an intel cpu 75c is way too low to be concerned,below 90 everything is fine,since tjmax is 100 and intel laptop cpus that are basically the same as desktop parts run at tjmax for years with no perfromance decrease
You could, i've done that before. Keep in mind, however, that liquid metal is electrically conductive and thus a spill onto the motherboard or any SMDs near the GPU die could cause damage.
You should definitely seal the capacitors right next to the die, if there are some. I've also used nail polish before, works rather well in my opinion. Just make sure to go over it a few times as it's hard to get full coverage with just one pass.
3 coats is what i normally go for, but heavily dependent on how well the capacitors are coated. I usually check that visually. I'd definitely wait more than 10 minutes between each pass, 30 minutes is how long i normally wait.
you can get better performance even from a solderd cpu, but not that much (yes you can delid a soldered chip but it's stupid and you have a considerable risk of ripping the die right off the substrate)
Maybe a bit late but i'll be glad if someone explains to me why no apply liquid between the top of the cpu and fan. Not native english speaker, thanks!
Please make a review for q-tips. There are too many brands and I can't decide between them X D Good and informative video. Even if im not trying this with my cpu (it is an older one). ^^
The lesson here is: destroy it with over voltage, not with liquid metal.
well, it's funnier that way :D
I was going to lower my voltage for better performance but since it's a laptop I can't do that so I need some other way of getting better performance...
@@akali6636 delete ur self
@@lonewolf3021 lmao why
This is kinda a filler video because I'm doing a ton of 2400G testing.
I heard in your last stream that you were partially wanting to put it under LN2, is there still a chance for that? Because I'd love to see that.
yeah there's a chance of that. My next LN2 delivery is still some time away.
This is pretty much most helpful video from you I've ever watched. So, not really a filler.
Actually Hardcore Overclocking I watched a video where TimmyJoe applied Liquid Metal to JUST the CPU surface, ignoring the IHS - I of coirse left a comment that he did it totally wrong, that due to the surface tension effects of Liquid Metal, ot MUST be applied to all both sirfaces that will contact each other.
Not too surprisingly, his temp results were mediocre for Liquid Metal.
Actually Hardcore Overclocking *NEVER APPLY LIQUID METAL TO ANY SHUNT YOU PLAN ON USING FOR MORE THAN A YEAR!!* Louis Rossman hs a video where a Customer sent his Titan Xp to be repaired, and before then, Louis had no clue about Liquid metal, so at first he thought Nvidia just did a TERRIBLE solder job. Well it turned out that the Liquid Metal ATE AWAY AT THE SOLDER!
So if you are going to be lazy, and Liquid Metal the shunts, you damned well better cover the solder pads of said shunt in Nail Polish! Or yanno, you COULD simply run a LITTLE jumper wire on top of the shunt and solder it, then tape it down.
Fun fact, Thanos asked Buildzoid how to overclock the infinity gauntlet after watching this video.
Protip for application if you find the liquid metal is not flowing easily: Heat it up a bit.
Gallium melts at 29c - and liquid metal has some additives that probably brings this a bit lower, but that's why it can be sort of syrupy at common room temperatures. Heating it up a little bit (and the CPU and the IHS too if you want) will make it more flowy and easily spreadable. You can heat the tube up in a glass of warm water, or just get a hairdryer on medium to heat up all the parts for a few minutes before you go to work.
“Because I keep seeing PEOPLE spraying liquid metal all over their stuff, like motherboards...” just say Linus, we have seen the video of him being a over excited teen on his first time with small tube in his hands.
in fairness, I think Steve at Gamers Nexus also did that to an x299 CPU some time ago. I don't remember if the Devil's Canyon also got the LM spray as well, or if that was someone else with LM. I have seen it a lot in vids around YT. They get LM, they get excited, then they squirt it all over there computer or laptop, lol.
Linus has done this several times
No we haven't. Weasel boy is unwatchable.
A A I feel that a priming is due, same can happen with other thermal compounds, the safest is the thermal tapes/strips. Depends on application too.
@@Zarcondeegrissom they're both noobs. Only trust Roman aka Derbauer and Kingpin to do proper Liquid metal install videos. They're the real experts in extreme overclocking. Even the idiot doing this video is using a regular q-tip instead of the standard black cotton swabs by DAISO Japan that Thermal Grizzly includes in their Conductonaut.😏
I'm not gonna lie, this helped me immensely. I actually got a great paint job on my cpu and IHS. So my thermals are finally reasonable when attempting to run prime with my i5 4670k.
So thank you buildzoid for making this video. Autofocus be damned.
Instructions unclear. Applied liquid metal to TH-camrs. They became very unstable...
I dabbed it on the q-tip but it dabbed back, what do i do?
stop hating on it
Hit a dab off your rig.
That will show that dab who really dabs and who not to fuck with.
@@nexxusty I hit a dab of liquid metal now I clock higher
I’m literally so pissed off right now trying to apply this fucking garbage bullshit onto my shit
Filler or not...good video. Thank you. Best liquid metal application video I've seen so far...and I've watched many dozen. How is your subscription count not over 1 million?!
Oh, and no matter how careful you are in applying liquid metal I'd totally recommend you insulate any contact points.
It's a really solid failsafe for 5 minutes of work. The material cost (some conformal coating or similar) is trivial and really easily availiable - and if you just spray some in the cap and paint it on (preferably 2 layers 15 minutes between) it's also trivially easy to do.
Even if it's just a 1% chance of the liquid metal moving that's a worthy insurance for an expensive CPU (and I've actually seen quite a few cases of it moving. Mostly probably because people apply too much of it. More user-error than the compounds fault - but then again your CPU doesn't risk frying if you make a sloppy mess with most thermal compunds. This absolutely does.
Great guide! 2 weeks ago I delidded and did exactly this to my 8086k and went from 5.1GHz @ 1.275V to 5.3GHz @ 1.425V with custom water cooling.
Thank you so much for this video, BZ! I just completed delidding my 8700k, and like an idiot, I would have tried to sqeeze the LM onto the die. Wouldn't have occured to me to squeeze it directly to the qtip. It worked great, and I'm sure saved me a mess to clean up! Thanks very much!
Thank God I found this video. Just bought a gaming laptop with a shitty liquid metal job and was going to redo it myself. Line you said, everyone just blobs it. This is surprisingly obvious after being shown but I never would have thought of it without your guidance.
New fan here!
Man the timing is perfect since my delid tool and liquid metal are being shipped now. Thanks for the info!
So it being this thin is not an issue in terms of it being able to fill all the imperfections? I guess under mounting pressure it should be enough even if super thin.
I hear the miscoloration people seem to be so worried about is just a thing that happens from the gallium diffusing into the other metals - basically putting it's atoms in between in the gaps where it can fit in the other metal. Aside from being unsightly it should not have any negative impact on the thermal properties of the surface. Actually I hear it might even slightly slightly improve the thermal properties if anything - but mainly neutral. Diffusion in itself is not a problem in most metals - just the metals where it causes the other metal's bonds to weaken (which is what happens in alu.).
The way it acts differently before and after the surface has been exposed to gallium before is no doubt because of this wetting effect the diffusion has. Before it's like water on really dry ground - it will tend to skip across, while after it's more like damp ground where it sticks to it more easily. That's obviously just an oversimplified analogy of the atomic forces involved here, but wetting and doping of materials (which is basically what happens here) is a well understood phenomenon you can read up on if interested.
If anyone happens to have any hard data/testing to show results of how thermal conductivity of copper and nickel-plated surfaces change (of at all) after gallium doping then I'd be really interested to see it.
I imagine because of this - if you wanted to be hyper-optimal you might be able to get a slight improvement if you bothered to do a re-application a month down the road from the initial application since at that point all the reactions will be done and the materials saturated. Probably unlikely to have statistically measurable effect (with the impresice measurements of CPUs we have to work with) since it's likely a trivial amount of the gallium that is diffusing in, but if for some reason you should happen to notice any drop in performance and you have eliminated all other factors (ambient ect.) then that's the first thing I would consider. After the second application I have no reason to suspect it shouldn't remain stable for a very long time.
1 very important question I would like to ask : You show how to spread it out perfectly, but what about air bubbles in between the IHS and CPU core ? Let's say the core is 15x15mm, it's 225 square mm. When air will be in between the IHS and core, the actual contact surface will be less then 225 square mm, and so the heat transfer is less good. Do you share my opinion ?
Just like he said it's not like traditional thermal paste.
And the rules of the traditional paste bubble issue do not apply here.
The rule is you want the liquid metal to be flatten enough that it sticks in the heatsink just like what paints do to your walls and by doing this the liquid metal will fill up the small gaps of the IHS (IHS if your using it in between heatsink cooler and cpu IHS).
Then since it's not like thermal paste that will easily stick to the heatsink cooler cold plate.
You also need to apply another one to the cold plate so that both liquid metal from top and bottom will touch each other and there you go you have connected a way for the heat to transfer.
I dont think thats much of an issue, what is in issue, and you can see it in video, is that maybe he applied too little. First time ever I did this, I applied too little and saw literally no difference in temp delta, and I applied around the same like he did here. Second time I applied more of it, so it was actually visible layer in there, not just like a reflective paint surface like in video, and I saw like 10 degree difference. I did however apply two layers of nail polish to protect the pins that were too close to the die, and since then I did that every time I delided any cpu afterward.
I've been looking for proper videos on how to apply liquid metal. All the top ones about liquid metal are people who don't know how to use it and blaming it on the liquid metal. It's good that you made a proper video, since I will be putting some on my gpu.
Thank you so much for this video. I want to apply liquid metal thermal paste on my PS4 Pro APU. The damn thing gets too hot and the fan is loud like a jet engine.
Did you apply it to your ps4 pro? Did it make a difference
I'm here for the same reason. I'm going to apply it to my ps4 too!
30 seconds of gentle cleaning with some very fine metal polish like autosol will 100% get rid of that stain on the IHS ... works perfectly on my nickel EK blocks
If the liquid metal is affecting the chip, in a fashion, that it is easier to apply a second time, could it be causing damage, that might cause problems down the road?
The sole tip "Liquid metal diffuses into the copper" was very helpful information to know. Both my GPU and CPU coolers are pure copper HS.
The stain only goes a few micrometers into copper, and doesn't affect structural integrity or performance in any way.
The "drying up" thing people ovserve and performance degradation is just oxydation due to air getting in there, since fit between Heatsink and IHS is far worse than between die and IHS or in direct die cooling in laptops or GPUs.
I heat the tube up with a hairdryer till it is really hot. It is MUCH easier to control.
Well the first time I used it, I put it in the freezer for 5 minutes or so and it was easy to apply. I think that heat is not a factor concerning the adhesiveness of liquid metal thermal paste
@@alihusein4299 LM melting point 29C, warming up helpful
So what is your view on using a custom copper IHS with liquid metal? Would the LM diffusing into the copper cancel out any benefits from the larger IHS surface area?
With 1.35v adaptive LLC 6 my temps are still not amazing even with a 280mm cooler.
How about applying on heatspreaders, between heatspreader and heatsink? Also, mine is a bit coppery, polished it some time ago - I know that it will form an alloy with copper, do you know if this process is ongoing (and I should re-apply it after some time as it "went" into the copper?)
how accurate those people saying the liquid metal applied might slip or come out and might touch the transistors ? of course a talking about liquid metal installed on a laptop with consistent move and shacking around and so on... could it happen ?! or is it really sticky ?
What about using a metal dish to as a delivery method? Squeeze into the dish then transfer to the surface and spread the with a cue tip on to the ihs? Then electrical tape right?
*No T1000's were harmed in the making of this video.*
what kind of thermal tape should i use with liquid metal on laptop to protect it from spilling or to protect components nearby
I used the brush included with Coollaboratory Liquid Pro, but this looks to be a superior method due to less spotty spread. I don't remember exactly what I did, but I recall squirting onto something and dabbing from there, like an easel. This way the brush didn't get gunked up, which I saw was an issue from reviews.
0:22 you should of edited in the ltt clip where he does that on a laptop and get LM all over the board.
i used LM on both CPU and GPU, i used the tray tecnique, i squezzed a blob on an nickel plated "tray" (the ihs from a g3258) and then picked it up with a q tip
Old vid, I know but hopefully someone will see this and answer. Arent you worried about fibers from the q-tip from becoming stuck in the applied liquid metal? I know i sometimes get loose fibers from those when cleaning things with a q-tip and say, iso alcohol.
It’s not autofocus fault. No way it’s going to know you want to focus o this tiny object in the frame. Cup your hand behind it and have your hand and the smaller object taking up more of the frame and then it will autofocus for you.
I think me cleaning the surface with alcohol made it much harder to get it to stick to a flat surface. Wonder if slightly roughing up the surface using very fine sandpaper would allow it to spread easier.
Today in "How to paint with liquid metal" presented by our favorite host BuildZoid...
... or is he LiquidZoid? :)
how often should i reapply my liquid metal? just when i notice temps being out of whack other wise its good still?
ive had it on my 8600k for about 4 months now and its still the same how long do you expect that to last?
when the temps go out of whack would be when I would check it. I've never run liquid metal long term. AFAIK Thermal Grizzly expect it to last like 2 years before needing reapplication.
It can last indefinitely, since the main killer is oxidation, and that's not a problem if everything fits snugly and your IHS is re-sealed.
Nogghan Had my delided 4670K runing 1.35V 4.4GHz for close to 5 yeara and the temps were the same since dat one. Used CLP on the die and Noctua NT H1 between the IHS and coolers.
And it was not sealed, just held in place by the retention arm of the socket.
Actually Hardcore Overclocking does the lm run down over time, when having the mobo vertically?
only if you apply way to much but if you do it like he does in the video it wont run at all its a thinly painted surface thats why he didnt even bother covering the surface contacts with a conformal coating, he knows the small amount cant possibly go anywhere and then itll be clamped by the IHS and then its really not going anywhere
Gallium bonds pretty much itself to all metals at the surface atomic level.
So be careful don't spill it lol!
What about filling a felt tip pen with the stuff? Then paint it on.
Hello, I tried this today but the liquid metal doesn't stick to the surface, it remains a bubble
Ok, thank you buildzoid, for putting things into nice perspective, including the potential damage. ill stick to my normal Pastes since im not interested in delidding anyway.
what about the fibers from the q tips, aren't they gonna end up together with the liquid metal?
MrXanthios it shouldn’t affect performance??? I’ve used tissues and q tips to clean out thermal paste numerous times and it doesn’t affect performance whatsoever.
I want to delid an i5 4590 and an i7 6700k and use liquid metal on top of the i7's IHS. I also want to repaste a 1070 and a Vega 64 with liquid metal. Should I buy the 1g or 5g syringe?
1 g should be enough but i would get 2g for the future
@@Robyne_u61 Thanks for the advice.
Never even thought of putting it on the q tip first lol. I always used to put a pin head on the die. Will definitely use this from now on and when i get my 8086k next week.
I always reseal mine though just makes it easier for resale and my ocd. Couldn't have the IHS in there all wonky :D.
But i must say scariest moment ever in my life was delidding my 7900x with homemade custom delid tool. Worked fine though. Same tool to do my 8086k next week :p
Aren't the white qtips kind of bad because of the cotton fibers, that could easily break off ?
What do you recommend using for removing liquid metal when cleaning? (in terms of solvents if any)I imagine gallium really doesn't react much to conventional solvents.
TheStigma I've seen people remove liquid metal with towels and isopropyl (i think that's how you spell it) alcohol.
My biggest concern is putting LM on the IHS. I have a 4770k, as long as I'm not being stupid, I shouldn't need to mask/cover the VRM's.
I'm just a bit concerned since I won't have perfect coverage with the IHS like I would with the die.
Man, man, man...thank you for the post. Made applying the LM way easier. I used the technique you showed. Worked amazingly. Down 12+C
My processor i7 4790k. gaming Temps are in the 85-90°C rang. reduced my temperatures best solution liquid metal or thermal paste?
can i use conductonaut on cooler master hyper 212 led turbo?
Can you safely run your CPU vertically without resealing? I never saw anybody use it that way, mostly on a test bench.
you'd have to install the CPU with the system horizontal but once the socket hold down bracket is down that IHS isn't going anywhere. The main risk is applying soo much liquid metal that it spills everywhere when you put the IHS on.
My 6700k has been running vertical on LM for about 8 months without a problem, so can at least make it that far.
Been running every delidded CPU vertically without resealing. It makes no difference if it's vertical or horizontal, it's not like it will drain downwards if that's what you're worried about. It should be as thin as a layer of paint, it won't go anywhere. Reasealing is pointless if you're just going to keep the CPU in the socket anyway. Reasealing might even hurt your thermals (check Gamers Nexus video, where even a very thin seal made it run a few degrees hotter than with no seal).
@Actually Hardcore Overclocking is it fine to cover whole zone around die with Kapton tape (to secure test points and other elements)? Substrate don't really heat from bottom, right?
Nail polish? For conductors. (5775c annoying to apply liquid metal)
am i silly to be paranoid about fibres from the q-tip mixing with the liquid metal and staying on the cpu?
It's not a silly question, but you don't need to worry, Tthe fibres are unlikely to make a measurable difference in temps, and wont impact any components of your system
after a quick google i saw that conductonaut comes with its own qtips for paranoid individuals such as myself. carry on.
I just used the small funnel in the package and then gently squeezed and applied a very small blob directly on the die then just spread it with a Qtip.
I did it multiple times on a multiple CPUs, I never had an issue where the LM just shout out and made a mess everywhere.
It happened to me, I used the funnel, slowly applied pressure then it shot out half the tube on the cpu and 10cm along the table.
Well there is still a human error there if you directly putting it in the die/IHS.
Specially if your doing it inside the case and the case is not completely flat when titled (some case due to the stand design like mine)
Then liquid metal will be happily sliding now in going down to the PCB.
So Q-tip application before cpu application will have 100% no error.
Unless your dumb and want to directly paint the PCB with silver coating ahaha
@@TrigrH Same, you can never rely on the syringe pressure 100% on these things, bubbles/pockets seem to be able to occur. Almost ruined an RTX 3070 with lm due to this, but managed to correct with a deep iso alcohol cleaning.
So you'd recommend against applying liquid metal on the ihs vs the bare die? Tempted to try out liquid metal but don't want to remove ihs to apply it bare die. My main curiosity is how it effects the longevity of the IHS
so what would I expect if I liquid metalled the ihs to heatsink? (specifically for a noctua l9i) ~1-3C?
Wouldn't using a cleanroom swab be better, since you won't get the (albeit minor) cotton lint on the die.
yes it would but I doubt it would make a measurable difference in temperatures.
3:14 is it ok to cover the golden thingys around the die with something like a clear nail polish "just in case" :D
yes it is.
Use kapton tape. Very easy and safe. Link to Amazon www.amazon.com/gp/product/B006ROR6JQ/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1
Why not use an *Acrylic Lacquer Conformal Coating* ? This is what it is made for.
MG Chemicals 419C-55ML Acrylic Lacquer Conformal Coating, Clear ,55 ml Bottle - $12.95 at Amazon for a 55ml bottle. Which mL vs mL is cheaper than clear nail polish and it is a constant formula made for this type of thing..
This is what Professional's and Hobbyist alike use to protect their electronics.
I've never used liquid metal, but I think you may be referring to thixotropic flow properties. Like tomato ketchup, be careful when you squeeze the bottle!
Or you put the CPU without the IHS in the socket first then cover it with the IHS that way you can quickly close the lever and everything will be held in place without having any risks when moving the unsealed cpu, the IHS may slip out and bam, you got liquid metal everywhere.
So this is why I got the "Buildzoid is now Live!" notification at 1 AM EST....
Applying the paste on the q-tip instead of the cpu die and making a mess. Genius
I was going to say that die was not a virgin but BZ already mentioned that.
Is it possible to apply too much liquid metal?
Question, serious question here, I'm hoping someone can help me.
Will the q-tip fibers be a problem if they get mixed in?
Use a good quality q-tip.
Or better yet, most liquid metal pastes come with high quality q-tips, use those.
The fibers can cause voids, so quality matters.
Hey brotha! I've been trying to find info on uncore ratio and what I should be setting this at for an 8600k @ 5GHz, and what voltage is associated with uncore ratio to help with stability. Any pointers? I've read to keep it under the clock speed of cores not much info on the web.
with these testpoints under the ihs. what happens if you short them? is it like 12V to cpu or something? i took precautions and applied nailpolish on my 6700k
Jesus I've always thought of this, but you sound like Drift0r so MUCH. I don't think that's a bad thing and I love your videos man. Thanks for another great video. :)
Kevin Gee I think he looks like him too
Stupid question: instead of liquid metal can I use regular thermal compound like the thermal grizzly Kryonaut between the die and the ihs?
you can but the thermal performance on high heat loads sucks very badly.
Actually Hardcore Overclocking at the high heat range it performs worse than the stock Intel paste?
i see someone is now a pcmasterrace man
performs better but not as good as LM
I literally just tried this with Kryonaut on my freshly delidded and direct die cooled 14900KS hoping to avoid any long term dry out. The answer is an absolute no, with temperatures no better than the original Intel TIM plus Kryonaut. It was a very good thin coverage after dismantling, but replacing with Conductonaut literally dropped temps by 20C at 320W using a MoRa 420!
This felt oddly like watching Bob Ross painting videos, can you do an imitation video pls. “Now let’s paint some birds with the liquid metal, anywhere you like just have fun.
*paints a liquid metal bird on all the critical board components *
Paint happy clouds on the silicone ⛅️
Now do a same video with that Graphene thermal pad.
Doesn't have to be a same video, but one that checks if it's possible to actually fit one under the IHS would be nice. If it's too thick, I don't think it will fit (I don't know though), but if it fits, it'd be interesting to see a performance comparison with liquid metal. Could to LM+Thermal Paste vs LM + LM vs LM + Graphene Thermal Pad vs GTP+TP vs GTP+GTP.
Even if GTP + GTP loses out to LM+LM, if the loss is minimal, it could easily turn out to be the better choice simply because it doesn't have the inherent risk of LM.
The point is most channels are avoiding it for some weird reason.
Why do you need to apply to both and did you use that nail polish to protect? Also if you go bigger on the IHS the heat wont cause the liquid metal to drip down on the chip?
The IHS on the Threadripper 1950x is solder, is that better?
I'm fairly certain the surface tension of the liquid metal will keep it together up to temperatures beyond CPU temps
So any old Q-tip will work? Also, is it 100% necessary to apply to both the die and IHS?
Pretty much. The reason is that liquid metal has a high surface tension (you can see it bead on the surface of the die if you apply it directly to the die) and when you put the ihs over the pcb, this surface tension will prevent the LM from "wetting" the underside of the IHS. Imagine trying to wash an oily plate with water. The water can't get on the plate because of the oil. While this isn't really a "correct" analogy, the effect is the same.
Ha! Ha! You covered the installation; but, how do you remove the stuff and cleanup later? The closest that I've come to Liquid-M is IndigoX and it's fairly easy once you melt it onto your CPU. Just a real "B" to get off.
So about the resealing.. Like if its for testing then I can see that doing it like this is no issue, but for a full time system is the tension of the hold down system enough to keep the IHS in place?
For the rest, wel great vid!
Yes, it's more than enough. Resealing is pointless if you're just going to keep the CPU installed for a long time anyway.
alright, thats something to mess with once the liquid metal arrives ^^
So what is liquid metal? Mercury?
Wapn Perfo gallium
Awesome video. Thanks for posting.
Some say sand the die with grit 1500 gently... I should make the LM stick easy. I only tried in the past on a polished cpu, damn! I had to abbandon it and give up! Was impossible to stretch out... was just a "ball" silver that could not stick
"On todays episode Actually Hardcore Overclocking teaches us how to stir the paint "
has anyone ever used conductonaut and then put thermal paste ?
where is linus when you need him?!
You’re funny bro. Awesome informational video. I appreciate the tutorial very much. Also, I thought you couldn’t use Q-tips due to the cotton fibers. Glad to know that they’re just mini paint brushes, Very helpful
Thermal grizzly liquid metal comes with q-tips
Can I use Lm in-between the Die and IHS .and on top of IHS normal thermal paste?
Why on Earth would you go through all the delid work only to skip the simple step of removing the original cheap thermal paste?
@@czdaniel1 well I did the delid added liqi metal and I used the grizzly thermal on top of the IHS in relpacement of the old thermal..you don't put liquid metal on top of IHS I'm using a Hyper Evo 12 it has copper heatsink..
@@Simmymind360 I see. We are using different terminology. The "I" in IHS when I speak the term stands for "internal" and is referring to the internal component between the transistors/die and the protective lid that you removed upon which your Hyper-Evo goes on.
Why cotton not safe
This is so informative and scary that I put back my Liquid Metal back and requesting refund through Amazon.
That capacitor on your gpu seemed quite interesting. What was that for?
Thanks for this video.. Do you advice to unseal the cpu to check things up after some time?
Saif Sameer I'd keep it unsealed just to make life easier if you do end up replacing the LM. Gamers Nexus has also found you get slightly better thermals with the cpu unsealed due to the smaller gap between the IHS and CPU die.
Is it alright if i try to use it in a laptop?
This is the exact way to apply! Its unbelievable how many videos show them applying it to the die or ihs directly. 😂 no way you should risk doing it directly. Your way is the best and how i always do it! Another great video!
Thanks dude. I mean that.
Is it really worth it for a mild overclock where temps never go above 75?
no, hell it's not worth in most cases even when pushing high oc on ambient, on the 7(8)700k it's worth only if you are trying to push it hard (1.4-1.45v)
edit: as the user tommihommi1 pointed, you can use a lower fan speed if you go with LM
if you want lower fan speeds, go for it.
i think exactly the same, on the cpu i did it just because i already had the lm lying around (used on the gpu), i tought: "why not?".
ehh if you are running an intel cpu 75c is way too low to be concerned,below 90 everything is fine,since tjmax is 100 and intel laptop cpus that are basically the same as desktop parts run at tjmax for years with no perfromance decrease
i'm currently using my rog strix g laptop, could i use liquid metal also for my gpu chip?
You could, i've done that before. Keep in mind, however, that liquid metal is electrically conductive and thus a spill onto the motherboard or any SMDs near the GPU die could cause damage.
@@MrLostGravity a friend of mine told me to apply nail polish around it, also found some youtubers do it.
You should definitely seal the capacitors right next to the die, if there are some. I've also used nail polish before, works rather well in my opinion. Just make sure to go over it a few times as it's hard to get full coverage with just one pass.
@@MrLostGravity would 3 coats of nail polish do? at least 10 min interval of application?
3 coats is what i normally go for, but heavily dependent on how well the capacitors are coated. I usually check that visually. I'd definitely wait more than 10 minutes between each pass, 30 minutes is how long i normally wait.
I'll just buy an AMD CPU and don't bother with delidding.
Matteo Kropej and just live with an oven for a computer
Gavin Jones *triggered
Matteo Kropej reeeeeeeeeee
you can get better performance even from a solderd cpu, but not that much (yes you can delid a soldered chip but it's stupid and you have a considerable risk of ripping the die right off the substrate)
unless you buy a ryzen 3 2200g or a ryzen 5 2400g those are sadly TIM not indium solder
WHAT? I CAN'T HEAR YOU - I HAVE GALLIUM IN MY EARS!
I always used the syringe, works fine form me. I also used a bit less on the cpu in fear of overdoing it. Regardless it all works fine.
Maybe a bit late but i'll be glad if someone explains to me why no apply liquid between the top of the cpu and fan. Not native english speaker, thanks!
Great video, are u a artist on the side?
Please make a review for q-tips. There are too many brands and I can't decide between them X D
Good and informative video. Even if im not trying this with my cpu (it is an older one). ^^
Thank you for this vid. It was a great help.