Liquid metal is fine to use. It will stain the copper but the staining has no effect on cooling performance. Tested ad nauseum by multiple sources. I've been using LM since 2008, on bare copper, never had an issue with drying or staining as some people claim (erroneously).
@@lowkey423 I have an Intel 3570k running since years with LM between die and IHS and between IHS and cooler - for years. Never changed it, still performs like day 1 of application. I had a second system with that setup and removed the cooler after 3-4 years - copper had stains and the counterpart too. Nail cleaner WITH acetone removed all the liquid metal without anything left to be seen.
Question are going to test AM5 with lapping ? Debau8er said theres like almost 1.6mm of the IHS sticking out of the new LGA socket, I think getting ride of that extra & the nickle may just help Zen 4 quite a bit.
Very cool. Nicely done. I think it will no doubt give better results to use LM than to lap the zen2, and if you do that you probably just want to keep the nickel surface. It's not that you can't use LM with copper, but it will diffuse into the metal a lot more and produce very hard residue (actually a metal alloy that has been produced in reaction). Won't noticeably impact performance as long as the system stays as-is because that alloy also has great thermal conductivity, but it can make it a bit of a pain if you need to remove and reapply later at some point. It can be a tough cleanup if it's been sitting in there a while. With 2 nickel surfaces it is basically a non-issue as the reaction is very limited (basically only mild staining), Not worth 1-2c gain to have that extra work and worry - IMO. I think I would have not done this without protecting the pins though. It could be pretty easily done with a block of soft "plant-foam" (ask your mother lol) and sinking the pins carefully into that - or probably even easier - just tape down a piece of plexi or other surface (hell, even just some hard cardboard) on top of the pins. That will prevent you from accidentally push in any single pin too hard and the force will be spread out and make it much harder to have an expensive accident. I would also take the time to tape down the sides. You definitely don't want to leave any metal-dust in the pins, and I think it is a much better way to prevent dust getting there in the first place rather than just rinsing it very thoroughly after.
AMD IHS are notouriously concave - lapping them is certainly worth it, if you are not afraid of losing the warranty (or the warranty is gone). For even better results, though, the heatsink can be lapped as well - doing both saved me 5 degrees at full load.
It is best to stop no higher than 1500, if you intend to use compound - there should be some "key" to the surface/s so that the paste can actually live somewhere. If you're going further than 1500 you are into the territory where legit lapping plates are required for paste-less interface.
I undervolted mine using the offset in the BIOS. -.078 mV. It shows in the BIOS the voltage is at 1.008V at stock clocks. I am using a GIGABYTE B450M-DS3H motherboard. Using Cinebench r20 the core ratios are at 40.5x. My temps are not as good as yours sinch I do live in a hot country. CPU Tdie temp is at 71.9C at load.
No. It is a resault of using a heatsink. If you use a heatsink on CPU and you are not getting any scratches on IHS it means that heatsink never touch IHS. It means that thermal paste isoates the CPU. Thats why they invited IHS. To scratch. "Old fasion" CPU with uncovered CPU DIE was improved by IHS. IHS is protecting the CPU DIE from phisical damage too. So if you put too much thermal paste you will never scratch the IHS (and the thermal grease will be a full isolator). You have to know that thermal grease is a isolator compared to metals.
You can still use liquid metal on copper heatsink or copper (lapped) IHS. Need to make atleast 2 run-ins. First one will dry the liquid metal inside the copper and then few days later add a second time liquid metal. No problems for my copper heatsinks (CPU cooler plate or GPU cooler heatsink).
Viilutaja that’s funny because I’ve seen significant pitting on my last NH-D15 when I removed it when parting it out on eBay. But don’t take my word for it, check out Nexus Gamers response and de8ouwers response to these questions. Just go to their channel, and listen to them, as subject matter experts.
@@HydroKyl240COG der8auer even go as far as nickel plating a copper heatsink to protect it from reacting with liquid metal. yup, due time, you will damage the IHS.
The screenshot at 1:24 th-cam.com/video/XKig29P4FWo/w-d-xo.html what is it from? Like, these cpus, have different default voltage, some lower some higher?
Like Viilutaja said. Liquid metal is no problem with copper. But NEVER use it with aluminum. Delidding and using liquid metal should get about a 4C drop. Benefits, easier to lap just the IHS and get and extra 4C drop and better high stress performance. Downside, delidding a soldered IHS is more difficult. But if you want that extra temp drop ...
@@heksogen4788 I do not use "AUTO" for voltage setting. Automatic settings push CPU voltage too high (more that is necessary to get stable clocks). Use offset to get VID+OFFSET voltage. It gives lower temperatures with the same clocks. I use Ryzen 5 3600 @ 4,1GHz with ~1.18V. Every CPU got different standard voltage so you have to find your sweet spot.
It is just a option reviewed by me. Thats it. I done it on my CPU. I had no issues. It had not warranty (cpu from aliexpress). I have no problems :). It is 3-4 degrees lower temperature for free. Fun was for free too ;).
Niby tak ale efekt jest słaby bo chłodzenie jest słabe. Na AIO efekt powinien być lepszy. Mam HE1024 i cięższe chłodzenie z większą pojemnością cieplną (łatwiej chłodzące IHS) miało by lepszy efekt bo to krzem nie jest w stanie oddać ciepła na tak małej powierzchni więc by krzem był chłodniejszy przy tak małym styku z IHS trzeba jeszcze bardziej obniżyć temperaturę IHS.
Watch whole video. The dust was removed and this CPU still works perfect. Thats why people are not able to understand the content. They watch just small part of the video.
@@GameER Ye I saw using isopropyl, thats unnecesary risk imo, you first expose the whole surface to it and then try to clean it. I would preffer to prevent if from happening. Also IHS isnt hermeticaly sealed, theres always a small gap somewhere. So the dust can potentially get to the chip. But if it works it works :) Untill it doesnt.
@@GameER what sandpaper grits should i use? I want to sand a 7800x3d once i get it so around 1.2mm+polishing, what should i use for the first part (removing the material)?
![Lapping a CPU, the proper way [AMD 3950x]](http://i.ytimg.com/vi/sQw20rHvpRw/mqdefault.jpg)
You are one brave man with those pins! 😎
It is not so risky when you touch many pins in the same time.
I bought a cheap AM4 socket from eBay to do mine :)
Liquid metal is fine to use. It will stain the copper but the staining has no effect on cooling performance.
Tested ad nauseum by multiple sources.
I've been using LM since 2008, on bare copper, never had an issue with drying or staining as some people claim (erroneously).
How long did you leave the liquid metal before you replaced it?
@@lowkey423 I have an Intel 3570k running since years with LM between die and IHS and between IHS and cooler - for years. Never changed it, still performs like day 1 of application.
I had a second system with that setup and removed the cooler after 3-4 years - copper had stains and the counterpart too. Nail cleaner WITH acetone removed all the liquid metal without anything left to be seen.
Question are going to test AM5 with lapping ?
Debau8er said theres like almost 1.6mm of the IHS sticking out of the new LGA socket, I think getting ride of that extra & the nickle may just help Zen 4 quite a bit.
IHS is jut a piece of metal. Its easy to do lapping. Especially when there are no pins to damage (AM4 CPUs are easy to damage).
that UF thing embossed in the corner is actually raised a bit. Wonder if it hurts cooling at all
grate video
Fun to do for people who love tinkering but realistically it's not worth the trouble, it's a shame I expected better results
what you say @0:34 about high heat is absolutely correct because the highest failure rate amongst zen 2 chips has been the 3600s!
^^
Very cool. Nicely done. I think it will no doubt give better results to use LM than to lap the zen2, and if you do that you probably just want to keep the nickel surface. It's not that you can't use LM with copper, but it will diffuse into the metal a lot more and produce very hard residue (actually a metal alloy that has been produced in reaction). Won't noticeably impact performance as long as the system stays as-is because that alloy also has great thermal conductivity, but it can make it a bit of a pain if you need to remove and reapply later at some point. It can be a tough cleanup if it's been sitting in there a while. With 2 nickel surfaces it is basically a non-issue as the reaction is very limited (basically only mild staining), Not worth 1-2c gain to have that extra work and worry - IMO.
I think I would have not done this without protecting the pins though. It could be pretty easily done with a block of soft "plant-foam" (ask your mother lol) and sinking the pins carefully into that - or probably even easier - just tape down a piece of plexi or other surface (hell, even just some hard cardboard) on top of the pins. That will prevent you from accidentally push in any single pin too hard and the force will be spread out and make it much harder to have an expensive accident. I would also take the time to tape down the sides. You definitely don't want to leave any metal-dust in the pins, and I think it is a much better way to prevent dust getting there in the first place rather than just rinsing it very thoroughly after.
Witam walazzka :)
Dobry filmik mimo niezbyt dużych zysków z lappingu
Pozdro!
How do I not make it lap unevenly
AMD IHS are notouriously concave - lapping them is certainly worth it, if you are not afraid of losing the warranty (or the warranty is gone).
For even better results, though, the heatsink can be lapped as well - doing both saved me 5 degrees at full load.
It was a CPU fro Aliexpress so the warranty was "Just an illusion" th-cam.com/video/uY4cVhXxW64/w-d-xo.html
I’d say 3ish degrees is what you might expect. Not a bad drop at all, imo.
did you lap the heatsink as well though?
Yes. More flat surface = better heat transfer (and less isolator - thermal grease).
@@GameER sweet, always pays to ask :D
gratitude for your reply,
Thank you.
Did u apply thermal paste?@@GameER
@@dispelledword7818 Of course it is necessary
I believe that you should only have to go as high as 1500 with the same results, at least if you still plan to actually use thermal compound
It is best to stop no higher than 1500, if you intend to use compound - there should be some "key" to the surface/s so that the paste can actually live somewhere.
If you're going further than 1500 you are into the territory where legit lapping plates are required for paste-less interface.
1:37: Are you sure your knife is perfectly flat as well?
I undervolted mine using the offset in the BIOS. -.078 mV. It shows in the BIOS the voltage is at 1.008V at stock clocks. I am using a GIGABYTE B450M-DS3H motherboard. Using Cinebench r20 the core ratios are at 40.5x. My temps are not as good as yours sinch I do live in a hot country. CPU Tdie temp is at 71.9C at load.
Get better cooler ;)
Nice video, @2:42 did your CPU come with those scratches/scuffs?
No. It is a resault of using a heatsink. If you use a heatsink on CPU and you are not getting any scratches on IHS it means that heatsink never touch IHS. It means that thermal paste isoates the CPU. Thats why they invited IHS. To scratch. "Old fasion" CPU with uncovered CPU DIE was improved by IHS. IHS is protecting the CPU DIE from phisical damage too. So if you put too much thermal paste you will never scratch the IHS (and the thermal grease will be a full isolator). You have to know that thermal grease is a isolator compared to metals.
@@GameER thank you for your response
You can also directly lap the die inside
You can still use liquid metal on copper heatsink or copper (lapped) IHS.
Need to make atleast 2 run-ins. First one will dry the liquid metal inside the copper and then few days later add a second time liquid metal.
No problems for my copper heatsinks (CPU cooler plate or GPU cooler heatsink).
Viilutaja that’s funny because I’ve seen significant pitting on my last NH-D15 when I removed it when parting it out on eBay. But don’t take my word for it, check out Nexus Gamers response and de8ouwers response to these questions. Just go to their channel, and listen to them, as subject matter experts.
@@HydroKyl240COG
der8auer
even go as far as nickel plating a copper heatsink to protect it from reacting with liquid metal. yup, due time, you will damage the IHS.
The screenshot at 1:24 th-cam.com/video/XKig29P4FWo/w-d-xo.html what is it from? Like, these cpus, have different default voltage, some lower some higher?
That's a lot of work for 1-3C cooler.
the difference between golden simple and bronze???
silicon lottery.. some samples are better performers than others, its like baking a batch of cookies.. their not all quite the same.
What is that lid made of?
The IHS is made of copper but coated with nickel.
Great trick, i realy learned something
Like Viilutaja said. Liquid metal is no problem with copper. But NEVER use it with aluminum. Delidding and using liquid metal should get about a 4C drop. Benefits, easier to lap just the IHS and get and extra 4C drop and better high stress performance. Downside, delidding a soldered IHS is more difficult. But if you want that extra temp drop ...
hey you easily can use liquid metal on a bare copper IHS, but nice vid
Parabens pra quem traduziu para PT-BR !!! :)
Great video. At which grit did you have to use water?
800 usually.
Should i do this with my 9900KF?
You can try. ^^
Which sandpaper u used?
Music?
a zrobisz materiał o lapingowaniu amd
Niet ;)
WOW you decreased 2.5 Celsius. Thats a World RECORD.
As it is for free - its a RECORD ;).
very gentle?? you are pushing the fuking pins my friend
I have done it more gantle than your words. Cpu survived. Your gantleness not.
What the hell is that 1.1V ryzen. Are you undervolting?
No. My Ryzen 5 3600 has a very low VID. 1.12V for stock voltage.
@@GameER what frequency you get with this voltage? Is that voltage value under stress? What's your vcore?
@@heksogen4788 I do not use "AUTO" for voltage setting. Automatic settings push CPU voltage too high (more that is necessary to get stable clocks). Use offset to get VID+OFFSET voltage. It gives lower temperatures with the same clocks. I use Ryzen 5 3600 @ 4,1GHz with ~1.18V. Every CPU got different standard voltage so you have to find your sweet spot.
You can use liquid metal on copper with no problems...
IMO is a poor gain, I own a 3700X and wouldnt risk or invest time/money for 3-4degrees, honestly
It is just a option reviewed by me. Thats it. I done it on my CPU. I had no issues. It had not warranty (cpu from aliexpress). I have no problems :). It is 3-4 degrees lower temperature for free. Fun was for free too ;).
Napracowałeś się dla mizernego efektu :P btw. dobry materiał.
Niby tak ale efekt jest słaby bo chłodzenie jest słabe. Na AIO efekt powinien być lepszy. Mam HE1024 i cięższe chłodzenie z większą pojemnością cieplną (łatwiej chłodzące IHS) miało by lepszy efekt bo to krzem nie jest w stanie oddać ciepła na tak małej powierzchni więc by krzem był chłodniejszy przy tak małym styku z IHS trzeba jeszcze bardziej obniżyć temperaturę IHS.
dude the ihs is soldered ?
yes
use a piece of flat glass
It's not a bad idea, but you'd need grooves to collect the material. Otherwise it will collect and foul the work.
Lol just removed some material. Lap had no effect. 1 degree ...gfto use non conductive paste. Safer and well...safer.
Liquid bs is a disaster.
This is not lapping. This is sanding.
If interested you can see what real lapping is here:
th-cam.com/video/sQw20rHvpRw/w-d-xo.html
water
Exlosing pins to metal dust lol. This is how to not do it.
Watch whole video. The dust was removed and this CPU still works perfect. Thats why people are not able to understand the content. They watch just small part of the video.
@@GameER Ye I saw using isopropyl, thats unnecesary risk imo, you first expose the whole surface to it and then try to clean it. I would preffer to prevent if from happening. Also IHS isnt hermeticaly sealed, theres always a small gap somewhere. So the dust can potentially get to the chip. But if it works it works :) Untill it doesnt.
That's not lapping, that's sanding. Two different processes.
whather plzzzz¡¡¡
İşlemciye rtx ekledi eleman
Liquid Metal is far better than this i think
So basically it's a huge waste of time.
Not if you enjoy the fun of trying something new!
@@GameER what sandpaper grits should i use? I want to sand a 7800x3d once i get it so around 1.2mm+polishing, what should i use for the first part (removing the material)?
@@inqizzo Use lower grade sandpaper. Start from 80/100/120 and go up.
Absolutely a waste of effort
This isn't lapping