Can almost guarantee there were particles of it deposited on the CNC bed. How much of a problem this becomes over time after cleaning etc remains to be seen. I would have thrown the cooler in the trash and put something else on.
@@DanielKennedyaeos the Datron M8cube has a concrete polymer table with stainless steel conical inserts, so I'd expect the gallium to not have any great effect, maybe some cosmetic affect on lacquer but nothing structural. The T-slot table is anodised, so hopefully would escape
You went really above and beyond for this 2060. Owner is going to be really happy with this, unless he was hoping for a free waterblock or something lol.
@@JathraDH True, jokes beside softline starting kits are actully quite cheap. Got my alphacool kit for a discounted price of* 150ish €. GPU waterblock for 6700XT was almost the same price lol.
@@JathraDH I would have learned my lesson not to mess with a perfectly good card and never trying to copy Der8auer again by splashing liquid metal around in future when I don't really know what I'm doing. And also throw this card in the bin and buy something decent.
@@Rmx2011 Hm that's not a bad cost. What pump/rad size? If it actually came with one of their decent full copper rads and an OK pump id call that a pretty damn good price indeed.
@@JathraDH It was discounted back then, current price level seems to have actually gone up some :/ 280mm rad with a ddc310 pump. I think it was this kit: Alphacool Eissturm Hurricane
This was AWESOME!! I was expecting MUCH MORE Damage... so I think you're right about the nickle platig.. but there cvould have been the TINIEST of scratches through that nickle and that's why there was damage at all... GREAT JOB!!!
Does the nickel actually form a barrier to protect the aluminium from the gallium? Gallium will alloy with nickel too, its just that this doesn't damage the nickel, just potentially causes a little discolouration. But at the point where it has alloyed with the nickel? Well, it can basically just pass through the nickel surely? Seeing as the nickel is plated onto the aluminium.
The mill is now contaminated. That stuff that was milled off contains gallium in a form that can transfer to fresh aluminium. Also, the structural and thermal changes to the aluminium does not stop at the surface. Milling of visually affected area is not enough.
Superb video, Found it fascinating with the CNC drilling away the bad bits, I was wondering if any thin spacer plates would have to have been added but obviously not, Super nice to see this card put back to good working use.
The gallium probably absorbed into the aluminium a lot deeper than you milled off, so it will likely be very brittle still, even if it performs fine as cooler.
@@RealEpikCartfrenYT problem is it worth it for what looks to be a bottom rung tier AIB 2070. Aftermarket gpu coolers are expensive and often a PITA to put on right too.
Important to know. Also important discussion about material quality and purpose. Since it is just a cooler there's very little physical stress, and it should be fine.
Very good troubleshooting. I hope this makes it into some gamer news channel about liquid metal and what you found and gives you the proper credit for the findings.
What a perfect timing for this video. I was just in middle of delidding my 13900kf and applied liquid paste on the die and Rockitcool's full copper IHS. Temps between cores were so bad after the process that I had to remove the retainer system and pile up the cpu, ihs and water block layer by layer. If you ever attempt this, tape the IHS to motherboard before slapping the water block on. This way you can be sure it doesn't move when you screw the block on, as there will be nothing nothing holding the stuff together. As long as there is no tape between the ihs and water block, you can pull the tape off after you are done tightening down the cooler. No more throttling under Aida64 extreme.
During the materials science portion, grain boundaries were not mentioned for nickel. Zhou et al. 2021 “Effect of grain boundary engineering on corrosion behavior of nickel-based alloy 825 in sulfur environment” document reduction of grain boundaries’ extent with specific treatment. I think it is worth mentioning that the gallium diffused through the nickel grain boundaries to reach the aluminum grain boundaries where the damage took place, and molecular diffusion of gallium will eventually reach the aluminum even for a nickel barrier free of grain boundaries (tho molecular diffusion of sufficient mass for macro Ga/Al rxn may take many years).
if gallium gets on aluminum, such aluminum can be thrown away. Now he kind of shines. But in fact - corrosion with depth still continues - when penetrating, gallium itself DOES NOT BE CONSUMED and continues to destroy intercrystalline (intercellular) bonds. And you just delayed the moment. when the aluminum plate finally "breaks through" and ceases to perform the cooling function!
One option i guess is getting some copper sheet stock and milling down the whole thickness of that. It will make it more likely that you remove sufficient amount of gallium residue embedded in aluminium so it won't continue eating it. You can use paste both sides or die side LM, cooler side paste. Though i wouldn't say this card warrants LM, the themal impedance of the cooler is too large, and copper is already an improvement. How does one even nickel plate alu? It is a mystery to me.
Great video! My only thought was that after removing the corroded part with the CNC machine, you could have made the part of the cooler smoother. with high-grade sandpaper or the machine equivalent of it.
start CNCing it out and see if you can drop in/solder in a copper plate. That aly should be connecting to a copper plate or the heatpipes after all. Then solder and a torch should melt it and fuse it with good cover.
That was my thought as well, but it is a low tier GPU. Since it had the clearances and the liquid metal didn't go that deep this works and is much easier.
If you would go back to liquid metal, you could add a copper insert right? In the milled out area. If I remember correctly, it does not react with copper that agressively.
So informative! Great to learn a bit about materials used for cooling solutions and the potential risks. With the new generation parts putting out loads of heat, that kind of knowledge is quite important.
G'day Roman, Shiek, Makita & Michael, I think at 6:12 Makita was just saying "😾Why did they do that?" knowing how terrible it is to put Liquid Metal on an aluminium cooler, While I have heard many times NOT to use Liquid Metal on Aluminium it was very educational to see the actual result, 😲the cooler looked like it had just been turned to carbon, So happy to see you were able to do some mods to get it working well for Michael rather than it ending up as E-Waste 😁
I’m not a metallurgist…. But a silly question… when you bake aluminium foil on a metal tray the oven bakes the foil to the tray and solidifies to some degree …… would this work on the GPU cooler that was damaged?🙀🤔
Did you consider making a 0.5mm shim to replace the milled aluminum? Would the additional thermal interface (even if filled with solder) been worse than the missing mass of metal from the milling?
Does the nickel actually form a barrier to protect the aluminium from the gallium? Gallium will alloy with nickel too, its just that this doesn't damage the nickel, just potentially causes a little discolouration. But at the point where it has alloyed with the nickel? Well, it can basically just pass through the nickel surely? Seeing as the nickel is plated onto the aluminium.
It was very interesting. Hope you cleaned up the chips out of your milling machine after, it would have been bad to go to grizzly the next day and just find a pile of aluminium dust on the floor where your machine was 😂
That cooler is trashed. It should have been completely replaced. Its literally like mold in bread. The damage is wayyyyyyyyyyyyyyyyyyyyyyyyyyyyy deeper than what you can see.
The cooler is definitely ruined. Easy fix would be a cooler from a dead card. Gallium will eat most iron and aluminum alloys. In some cases, it can leave the metal in a condition where you can crumble it with your bare hands.
It's awesome if you know what you're doing. I don't think it's worth it using it on a air cooled GPU like this dude did, the die is large and should have no thermal issues with a decent compound and heatsink mounted properly. Only worth it on a overclocked liquid-cooled processor.
Great video, but I can't help but notice 1 small possible issue - the leveling during CNCing was done based on finstack and not on heatsink base. If the base was actually in parallel to the finstack, that should've been mentioned. And if it's not parallel, then, while you might've leveled standoffs with GPU hole, if the rest of heatsink is angled the contact may suffer somewhere due to different "GDDR height". But that also depends on what thermal pads were used. The problem could be present with good hard ones, and not present with poor-performing soft ones. I've seen a great deal of repair videos where GPUs suffered from bad temps due to users messing with cooling and assuming thermal pads can flatten when they mount the cooler.
I had a 6900XT and used Alphacool Eisfrost Xtreme liquid metal on a nickel waterblock Alphacool Eisblock and the LM heavily pitted the block after 12 months. Ended up trashing the block, as great as the LM was for cooling I won't be using it again if it pits nickel. Temps were fantastic, but LM may not not ready for long-term use.
hi!i watched some of your videos.... i have a question! i put some conductonaut on an i5 6500 and i don't see an improvement! why? on stock tim in stress tests temps went like 67 degrees...with liquid metal......like 65degrees.....i use mx4 as thermal interface between ihs and cooler! thx
I read some time ago that one of the worst things to happen to a modern passenger plane is a gallium spill. You would have to replace huge parts of the airframe because you cannot really tell where it already went, and having your plane become brittle is not good.
My experience with liquid metal and GPU: Applied LM on 1080ti, tested, found only a minor difference (if any, because fan curve got adjusted and still kept about the same temperature, just running bit slower), scratched my head, reapplied paste back. There's a lot more downsides than benefits and I see no reason to use that for mid or entry level cards. Even 5 degrees C is nothing if its a difference between 65 and 60. People should stop be obsessed with temps. Undervolt if you can and stop
Actually the difference between 60c and 65c is astronomically MASSIVE in terms of GPU temperatures. 5 degrees is literally the difference between a top 5 and being out of the top 100 leaderboard.
@@NinjAsylum That's assuming you're an OC'er? I still can't see how it applies for most casual users. Applying LM will 100% stain your radiator (copper 100%, nickel plating the safest but prob not entirely secure either) and GPU die, and adds extra variable in terms of something that can brick your card. Heck there are reports of LM leaking PS5's even though they had padding to prevent spillage. I've seen a lot of vids with really expensive GPUs getting damage from LM. Just buy a good ventilated case and undervolt if possible. I'm not saying its straight bad, I used it myself but eventually grown up out of this
Step 1: Determine your maximum boost frequency & voltage Step 2: Use MSI Afterburner Curve Adjuster (Ctrl + F) and Click -100mv below your max voltage, hold shift, and raise the entire curve to your maximum boost frequency Step 3: Shift click the graph and drag your mouse to highlight all frequencies to the right of your desired target and hit Shift + Enter on keyboard to flatten the entire curve to the right, sometimes need to do it twice to completely flatten it. Step 4: Apply and enjoy Easy -10°C undervolt (There are videos on YT explaining the whole process if my directions don't help anyone)
LM sucks at gap filling, I've had applications where it was terrible and other where it's awesome (like one of my first applications led to a +50% core OC on a GTX 680M)
So it's an alloy of gallium and aluminum? And the aluminum is similar to work hardened aluminum? Interesting...just by looking id assume the aluminum had been oxidized with no previous knowledge...I seem to remember that after removal you should wait awhile to make sure you didn't scratch off the oxide layer and start the reaction again to be sure.
Great video. If you want another good video of what gallium does check out Lock Picking Lawyer video 593 where he uses some to totally destroy an aluminium padlock.
Nicely done! Though I'd dump the cooler inside electrolysis water with a phone charger, using a piece of copper/nickel/zirc to electroplate it back to normal.
de8auer you should take a look into the Gigabyte B650E Aorus Master that you had on your channel, it has a QC issue where it cant handle AVX intensive workloads or even very intensive CPU workloads sometimes. For the most part seems to be fine with 7600x CPU, feels very bad for people that have 7600x CPU with that motherboard and wont know if their motherboard is fault or not. It Even seems to work fine with higher core CPU if you lower the clock speed by allot.....
My Z790 Classified has a setting to automatically lower the multiplier on AVX workloads. I suspect it's how fine tuned the processors are these days, they're already overclocked the way we're getting them, and AVX is very intense. I know, that's Intel processors, but both Intel and AMD are doing the same as far as how their processors are tuned at the factory. It's recommended to run that setting I mentioned 2-3x lower than your set multiplier, so if you have the performance cores running 5.8, they'd only run at 5.5 or so under those AVX instruction sets, there's just too much heat generated and the stock clock already pushes it to the very cusp on water. I'm not saying there isn't issues with that board, just thought that might be a contributing factor.
Good work bud. Kaltverfestigung = is Work hardening in English and is the result of bending or forming Steel, Aluminium and Copper. I am an Englishman living in Germany and I struggle with the English pronunciations.
Milling out the deterred aluminium thins the heat plate i get it. But will this not have effect on the heat distribution throughout the plate with long time usage ? Interesting topic Roman
I mean the volume removed was extremely low. I'm more surprised about the fact that liquid metal TIM actually contains Gallium and that people are "ok" with using any amount of Gallium in their PC's. There's a reason why that shit is flat out banned and is a criminal offense to carry on aircrafts.
Nice repair!! Glad to see a simple milling was all it really needed. As for the voltage sensor, im keeping my eyes out for the official release! I know its probably late notice and already thought of, but you sgould also have an alternate version that has the power cable plug in from the front side of the computer 90° rotated on the x plane from what you have there, it will give more functionality to people with smaller cases in the future due to cable management, it'd alao make for neater cable management for cards with the pcie power at the end of the card, like evga stuff. Keep up the great work :)
People really don't know what gallium does til they experiment with it. Its actually very surprising it didn't completely destroy that cooler. I would certainly be careful around a ford pickup with it lmao.
Wow perfect timing. Migrating to a new case and repasting cpu after 3 years. Looked at paste vs liquid metal. Paste it is! Yikes. Your cats name is Makita? I love her power tools....
Awesome work. I hope your CNC mill didn't get hit with any gallium debris during the milling.
Ya. Creepy problem. 😑
Can almost guarantee there were particles of it deposited on the CNC bed. How much of a problem this becomes over time after cleaning etc remains to be seen. I would have thrown the cooler in the trash and put something else on.
@@DanielKennedyaeos - I'm sure Derbauer knows what he's doing ;)
He's a problem solver, not avoider.
@@DanielKennedyaeos the Datron M8cube has a concrete polymer table with stainless steel conical inserts, so I'd expect the gallium to not have any great effect, maybe some cosmetic affect on lacquer but nothing structural. The T-slot table is anodised, so hopefully would escape
Now the CNC mill starts to chip away a month later from this video.
The explanation of the chemical reaction was interesting, I appreciated that part especially.
It's not a chemical reaction, but I concur.
This is absolutely fascinating. Thank you for the explanation of grain boundary diffusion.
You went really above and beyond for this 2060. Owner is going to be really happy with this, unless he was hoping for a free waterblock or something lol.
I think if the owner had a custom loop they wouldn't be bothering with a 2060 lol.
@@JathraDH True, jokes beside softline starting kits are actully quite cheap. Got my alphacool kit for a discounted price of* 150ish €. GPU waterblock for 6700XT was almost the same price lol.
@@JathraDH I would have learned my lesson not to mess with a perfectly good card and never trying to copy Der8auer again by splashing liquid metal around in future when I don't really know what I'm doing. And also throw this card in the bin and buy something decent.
@@Rmx2011 Hm that's not a bad cost. What pump/rad size? If it actually came with one of their decent full copper rads and an OK pump id call that a pretty damn good price indeed.
@@JathraDH It was discounted back then, current price level seems to have actually gone up some :/
280mm rad with a ddc310 pump.
I think it was this kit: Alphacool Eissturm Hurricane
Very interesting the reaction between the metals, thanks for taking the time to explain it.
As a machinist I appreciate the steps you take in troubleshooting. Well done and another nice video.
Im a ventriloquist and I greatly appreciated this boost to my knowledge as it helps alot with ventriloquism
This was AWESOME!! I was expecting MUCH MORE Damage... so I think you're right about the nickle platig.. but there cvould have been the TINIEST of scratches through that nickle and that's why there was damage at all...
GREAT JOB!!!
Does the nickel actually form a barrier to protect the aluminium from the gallium? Gallium will alloy with nickel too, its just that this doesn't damage the nickel, just potentially causes a little discolouration. But at the point where it has alloyed with the nickel? Well, it can basically just pass through the nickel surely? Seeing as the nickel is plated onto the aluminium.
Good video as always, I love how technical you get with the hardware, helps us learn more :)
Great video! Allways wanted to see the damage from the liquid metal on aluminium. Also nice that you managed to fix the viewers card 👍
That was very nice of you! Nice Mill work!
Always a joy to watch real pros doing real pro work.
The mill is now contaminated. That stuff that was milled off contains gallium in a form that can transfer to fresh aluminium.
Also, the structural and thermal changes to the aluminium does not stop at the surface. Milling of visually affected area is not enough.
Superb video, Found it fascinating with the CNC drilling away the bad bits, I was wondering if any thin spacer plates would have to have been added but obviously not, Super nice to see this card put back to good working use.
Awesome job the owner sure will be happy, and also not to complicated of a repair straight forward and practical.
The gallium probably absorbed into the aluminium a lot deeper than you milled off, so it will likely be very brittle still, even if it performs fine as cooler.
yeah i would just grab another cooler from another graphics card and slap it on
@@RealEpikCartfrenYT Raijintek sells VGA heatsinks too although for some you need to forego the gpu backplate.
@@RealEpikCartfrenYT problem is it worth it for what looks to be a bottom rung tier AIB 2070. Aftermarket gpu coolers are expensive and often a PITA to put on right too.
Important to know. Also important discussion about material quality and purpose. Since it is just a cooler there's very little physical stress, and it should be fine.
It would actually be interesting to see it opened up 6 months from now to see if there's any visible change to the area under pressure.
It's so cool you have your own engineering playground to just make stuff every nerds dream ugh lol what a video glad you did this for the guy
You can add thin sheet of cooper to that area and melting it together with some soldering paste (then sanding it a little bit)
Copper and aluminium also corrode when put together
Another great video, always informative. Thanks.
Very good troubleshooting. I hope this makes it into some gamer news channel about liquid metal and what you found and gives you the proper credit for the findings.
What a perfect timing for this video. I was just in middle of delidding my 13900kf and applied liquid paste on the die and Rockitcool's full copper IHS.
Temps between cores were so bad after the process that I had to remove the retainer system and pile up the cpu, ihs and water block layer by layer. If you ever attempt this, tape the IHS to motherboard before slapping the water block on. This way you can be sure it doesn't move when you screw the block on, as there will be nothing nothing holding the stuff together. As long as there is no tape between the ihs and water block, you can pull the tape off after you are done tightening down the cooler.
No more throttling under Aida64 extreme.
Excellent work saving that coldplate.
During the materials science portion, grain boundaries were not mentioned for nickel. Zhou et al. 2021 “Effect of grain boundary engineering on corrosion behavior of nickel-based alloy 825 in sulfur environment” document reduction of grain boundaries’ extent with specific treatment. I think it is worth mentioning that the gallium diffused through the nickel grain boundaries to reach the aluminum grain boundaries where the damage took place, and molecular diffusion of gallium will eventually reach the aluminum even for a nickel barrier free of grain boundaries (tho molecular diffusion of sufficient mass for macro Ga/Al rxn may take many years).
if gallium gets on aluminum, such aluminum can be thrown away. Now he kind of shines. But in fact - corrosion with depth still continues - when penetrating, gallium itself DOES NOT BE CONSUMED and continues to destroy intercrystalline (intercellular) bonds. And you just delayed the moment. when the aluminum plate finally "breaks through" and ceases to perform the cooling function!
One option i guess is getting some copper sheet stock and milling down the whole thickness of that. It will make it more likely that you remove sufficient amount of gallium residue embedded in aluminium so it won't continue eating it. You can use paste both sides or die side LM, cooler side paste. Though i wouldn't say this card warrants LM, the themal impedance of the cooler is too large, and copper is already an improvement.
How does one even nickel plate alu? It is a mystery to me.
I always appreciate a cat jumpscare in a tech video!
Wow what a great video! Thanks! I learned a lot!
Just amazing...keep up the good work
Great video! My only thought was that after removing the corroded part with the CNC machine, you could have made the part of the cooler smoother. with high-grade sandpaper or the machine equivalent of it.
Excellent work Roman
start CNCing it out and see if you can drop in/solder in a copper plate. That aly should be connecting to a copper plate or the heatpipes after all. Then solder and a torch should melt it and fuse it with good cover.
That was my thought as well, but it is a low tier GPU. Since it had the clearances and the liquid metal didn't go that deep this works and is much easier.
amazing work....just amazing!!
Looking at the thickness of those thermal pads, it's a pretty low end 2060. Putting liquid metal was a really weird decision.
If you would go back to liquid metal, you could add a copper insert right? In the milled out area. If I remember correctly, it does not react with copper that agressively.
think that would be an option too.. place a 0.5mm copper shim on the die. Good thinking!
@@groenevinger3893 It would just alloy with the copper and "seep through" it, like it did to the nickel plating.
Now that is good content.
Great Video
So informative! Great to learn a bit about materials used for cooling solutions and the potential risks. With the new generation parts putting out loads of heat, that kind of knowledge is quite important.
This is interesting and also a lesson. Read before doing anything, do your research, and ask questions if you feel like you need to.
G'day Roman, Shiek, Makita & Michael,
I think at 6:12 Makita was just saying "😾Why did they do that?" knowing how terrible it is to put Liquid Metal on an aluminium cooler,
While I have heard many times NOT to use Liquid Metal on Aluminium it was very educational to see the actual result, 😲the cooler looked like it had just been turned to carbon,
So happy to see you were able to do some mods to get it working well for Michael rather than it ending up as E-Waste 😁
I’m not a metallurgist…. But a silly question… when you bake aluminium foil on a metal tray the oven bakes the foil to the tray and solidifies to some degree …… would this work on the GPU cooler that was damaged?🙀🤔
Great job. I didn't expect that this could be fixed
Would gallium also have grain boundary diffusion into copper? What about a plated copper?
i love this kind of content
the wireview might solve alot of problems with space between case glass and gpu power connector aswell :) so looking forward for it to be released.
Did you consider making a 0.5mm shim to replace the milled aluminum? Would the additional thermal interface (even if filled with solder) been worse than the missing mass of metal from the milling?
11:22 where can one get this? I meant the 8pin power reader thingy.
Does the nickel actually form a barrier to protect the aluminium from the gallium? Gallium will alloy with nickel too, its just that this doesn't damage the nickel, just potentially causes a little discolouration. But at the point where it has alloyed with the nickel? Well, it can basically just pass through the nickel surely? Seeing as the nickel is plated onto the aluminium.
It was very interesting. Hope you cleaned up the chips out of your milling machine after, it would have been bad to go to grizzly the next day and just find a pile of aluminium dust on the floor where your machine was 😂
im not sure but i think a pro like derbaurer, would be smart enough to clean the card before putting it back together..
@@groenevinger3893 He's talking about the CNC machine.
@@FR4M3Sharma indeed!
steel chassis, concrete polymer table, stainless inserts, only the t-slot table would disappear :D
That cooler is trashed. It should have been completely replaced. Its literally like mold in bread. The damage is wayyyyyyyyyyyyyyyyyyyyyyyyyyyyy deeper than what you can see.
Another great video
Very interesting indeed, cheers from Argentina 🇦🇷 🍻
I need that machine lol!!! Very precise job!! Well done!!!
Dude, I have a Palit RTX 2070 Blower X with a waterblock, How can I get a nice air cooler to fit it?
2:38 the black stuff is conducive dust.
And it can go literally everywhere!
The cooler is definitely ruined. Easy fix would be a cooler from a dead card. Gallium will eat most iron and aluminum alloys. In some cases, it can leave the metal in a condition where you can crumble it with your bare hands.
So many liquid metal horror stories for me to even consider bringing that stuff close to my hardware
It's awesome if you know what you're doing. I don't think it's worth it using it on a air cooled GPU like this dude did, the die is large and should have no thermal issues with a decent compound and heatsink mounted properly. Only worth it on a overclocked liquid-cooled processor.
Really excited for your wattage display gor the gpu to release
Well done!
Thanks for more great content keep it up.
Great fix! Thanks!
rofl... Fix for something that NEVER should be on a low end GPU.
"Good handling" thermalpaste? How fast i can take corners before slipping?
Great Job!!
Good job man!
a 0.5mm copper sheat on the space milled would not be better?
Borderline impossible to bond in regards to thermodynamics and flatness.
This was my thought too
Its very hard to see that gainward is a part of palit, my palit 2060 dual is identical
Great video, but I can't help but notice 1 small possible issue - the leveling during CNCing was done based on finstack and not on heatsink base.
If the base was actually in parallel to the finstack, that should've been mentioned.
And if it's not parallel, then, while you might've leveled standoffs with GPU hole, if the rest of heatsink is angled the contact may suffer somewhere due to different "GDDR height".
But that also depends on what thermal pads were used. The problem could be present with good hard ones, and not present with poor-performing soft ones.
I've seen a great deal of repair videos where GPUs suffered from bad temps due to users messing with cooling and assuming thermal pads can flatten when they mount the cooler.
I would assume he probed the GPU side surface of the cooler off camera and adjusted it to level.
I learned in 3 minutes from der8auer more than what I learned in a full year of Science of Materials during my Mechanical Engineering college!
Me either... collage was great
Great job
I had a 6900XT and used Alphacool Eisfrost Xtreme liquid metal on a nickel waterblock Alphacool Eisblock and the LM heavily pitted the block after 12 months. Ended up trashing the block, as great as the LM was for cooling I won't be using it again if it pits nickel. Temps were fantastic, but LM may not not ready for long-term use.
hi!i watched some of your videos.... i have a question! i put some conductonaut on an i5 6500 and i don't see an improvement! why? on stock tim in stress tests temps went like 67 degrees...with liquid metal......like 65degrees.....i use mx4 as thermal interface between ihs and cooler! thx
Usually the whole part is thrown out because gallium is hard to get rid of...
I read some time ago that one of the worst things to happen to a modern passenger plane is a gallium spill. You would have to replace huge parts of the airframe because you cannot really tell where it already went, and having your plane become brittle is not good.
Why not just copper shim between the milled cooler and GPU? would've been less work
And maybe nickel-plate the shim.
My experience with liquid metal and GPU:
Applied LM on 1080ti, tested, found only a minor difference (if any, because fan curve got adjusted and still kept about the same temperature, just running bit slower), scratched my head, reapplied paste back.
There's a lot more downsides than benefits and I see no reason to use that for mid or entry level cards. Even 5 degrees C is nothing if its a difference between 65 and 60. People should stop be obsessed with temps. Undervolt if you can and stop
Actually the difference between 60c and 65c is astronomically MASSIVE in terms of GPU temperatures. 5 degrees is literally the difference between a top 5 and being out of the top 100 leaderboard.
I went from like high 80c to high 73/74c on my 3060ti swapping out the factory thermal paste and ramping up the fan curve
@@NinjAsylum That's assuming you're an OC'er? I still can't see how it applies for most casual users. Applying LM will 100% stain your radiator (copper 100%, nickel plating the safest but prob not entirely secure either) and GPU die, and adds extra variable in terms of something that can brick your card. Heck there are reports of LM leaking PS5's even though they had padding to prevent spillage. I've seen a lot of vids with really expensive GPUs getting damage from LM. Just buy a good ventilated case and undervolt if possible.
I'm not saying its straight bad, I used it myself but eventually grown up out of this
Step 1: Determine your maximum boost frequency & voltage
Step 2: Use MSI Afterburner Curve Adjuster (Ctrl + F) and Click -100mv below your max voltage, hold shift, and raise the entire curve to your maximum boost frequency
Step 3: Shift click the graph and drag your mouse to highlight all frequencies to the right of your desired target and hit Shift + Enter on keyboard to flatten the entire curve to the right, sometimes need to do it twice to completely flatten it.
Step 4: Apply and enjoy
Easy -10°C undervolt
(There are videos on YT explaining the whole process if my directions don't help anyone)
LM sucks at gap filling, I've had applications where it was terrible and other where it's awesome (like one of my first applications led to a +50% core OC on a GTX 680M)
What did he think such a cheap card would have a nickel plated headsink? He clearly put some thought beforehand as he used nail polish on the SMDs.
More of this!
awesome fix!
So it's an alloy of gallium and aluminum? And the aluminum is similar to work hardened aluminum? Interesting...just by looking id assume the aluminum had been oxidized with no previous knowledge...I seem to remember that after removal you should wait awhile to make sure you didn't scratch off the oxide layer and start the reaction again to be sure.
Great video. If you want another good video of what gallium does check out Lock Picking Lawyer video 593 where he uses some to totally destroy an aluminium padlock.
The reason it was damaged was just user error, right? Using LM on an aluminium heatsink
Well done, thou I would not even put that card in the same room as the CNC lol.
Nicely done! Though I'd dump the cooler inside electrolysis water with a phone charger, using a piece of copper/nickel/zirc to electroplate it back to normal.
There are more then 4 people interested in the digital power reader. Great Video.
You do nice work.
Not yet available?? Give us the Wireview nao! 😁
Will there be a Wire View for vertical mount cards?
its called triple junction points between grains and represents least and most stable energy state. this condition is highly wanted.
6:08 Makita is a beautiful cat 🐈🙀🥰
de8auer you should take a look into the Gigabyte B650E Aorus Master that you had on your channel, it has a QC issue where it cant handle AVX intensive workloads or even very intensive CPU workloads sometimes.
For the most part seems to be fine with 7600x CPU, feels very bad for people that have 7600x CPU with that motherboard and wont know if their motherboard is fault or not.
It Even seems to work fine with higher core CPU if you lower the clock speed by allot.....
Thanks for mentioning this I had planned to upgrade using that board specifically for heavy workloads on a x7950 3d. Not anymore...
My Z790 Classified has a setting to automatically lower the multiplier on AVX workloads. I suspect it's how fine tuned the processors are these days, they're already overclocked the way we're getting them, and AVX is very intense. I know, that's Intel processors, but both Intel and AMD are doing the same as far as how their processors are tuned at the factory. It's recommended to run that setting I mentioned 2-3x lower than your set multiplier, so if you have the performance cores running 5.8, they'd only run at 5.5 or so under those AVX instruction sets, there's just too much heat generated and the stock clock already pushes it to the very cusp on water. I'm not saying there isn't issues with that board, just thought that might be a contributing factor.
Good work bud. Kaltverfestigung = is Work hardening in English and is the result of bending or forming Steel, Aluminium and Copper. I am an Englishman living in Germany and I struggle with the English pronunciations.
Why did the owner actually bothered using LM to a 2060.
Milling out the deterred aluminium thins the heat plate i get it. But will this not have effect on the heat distribution throughout the plate with long time usage ? Interesting topic Roman
I mean the volume removed was extremely low.
I'm more surprised about the fact that liquid metal TIM actually contains Gallium and that people are "ok" with using any amount of Gallium in their PC's.
There's a reason why that shit is flat out banned and is a criminal offense to carry on aircrafts.
what's the name this tester for molex tg-pmd-p18r?
Love that Datron mill.
Marvelous 🤓🍷
This was a cool video.
Doesn’t the aluminium once affected by the Liquid Metal alter its thermal conductivity negatively?
Does heating it help?
Nice repair!!
Glad to see a simple milling was all it really needed.
As for the voltage sensor, im keeping my eyes out for the official release! I know its probably late notice and already thought of, but you sgould also have an alternate version that has the power cable plug in from the front side of the computer 90° rotated on the x plane from what you have there, it will give more functionality to people with smaller cases in the future due to cable management, it'd alao make for neater cable management for cards with the pcie power at the end of the card, like evga stuff.
Keep up the great work :)
People really don't know what gallium does til they experiment with it. Its actually very surprising it didn't completely destroy that cooler. I would certainly be careful around a ford pickup with it lmao.
Nice repair but nobody has a whole CNC 3d printer milling machine to perform that kind of work 😊
Very interesting video
Wow perfect timing. Migrating to a new case and repasting cpu after 3 years. Looked at paste vs liquid metal. Paste it is! Yikes. Your cats name is Makita? I love her power tools....