Why you should NOT buy Water Blocks made out of "Silver" - Thermal Conductivity measured & explained

hi welcome back to a new video a lot of things we cover on this channel are around the topic of cooling and this mostly involves also the topic of thermal conductivity but there are a lot of myths around this topic especially for example with upgrading Parts such as with this rocket cool upgrade Heats better which is made out of copper and then it's also commonly known that for example this Intel IHS or this AMD IHS they are also made out of copper so why would you need an upgrade heat sweater like how is this copper alloy maybe better than the alloy that Intel or AMD are using and then for example in the past I was also offering together with case King that was for the 8700k we were offering upgrade heat spatters with lab copper or also made out of very very clean silver so how could this improve your thermal conductivity and how do you even measure this that's what we're going to cover in today's video this video is sponsored by hetzner and their dedicated root service the ax line offers a wide range of servers depending on your required CPU power storage and memory it starts with the ax41 which features a ryzen 5 3600 with 64 gigabyte of memory and two 512 gigabyte and vme ssds for 34 euro per month the ax101 would come with 5950x 128 gigabyte of ECC memory and two 3.84 terabyte nvme ssds for 94 euro per month of course fully configurable with the OS you need additional storage and gigabit connection find out more in the link below what does a thermal conductivity even mean to explain this a little bit more in detail and also a little bit more simple we will imagine a wall made out of copper with a length of 1 meter and at the same time we imagine a height of about 100 meter and a width of about 100 meters now we want to transport heat through the length of one meter of our copper wall the thermal conductivity of copper is at about 400 watt per meter and Kelvin we could also say what per 1 meter and one Kelvin Kelvin is typically used to describe temperature differences this means for example you get up in the morning and your room has 20 degrees Celsius and in the evening it has 21 degrees Celsius then you would have a temperature difference of one Kelvin going back to our example with 400 watt per meter and Kelvin this means we are transferring 400 watt through our wall and at the given thermal conductivity this means that we will have a temperature difference of the surface of the left side on the wall and the right side of the wall of exactly one Kelvin for example if we would have our wall made out of aluminum which has roughly half the conductivity so about 200 watt per meter and Kelvin then we could only send 200 watt to maintain the max difference of one Kelvin between the Surface on the left side and the right side in summary the thermal conductivity means how much energy we can pump through one meter of material while maintaining a temperature difference of the surface of one Kelvin from input and output side before we get into the entire topic and measure some of these things you can see on the table I just just want to explain how we are measuring this and how we're going to calculate the thermal conductivity and that's why we have this tiny device which basically looks like a digital multimeter but it's far from that it's a device that can measure the electrical conductiveness on the surface of a material of a non-ferrid metal which means that you can basically measure how good is the electrical conductivity of a metal with this thing it sounds simple it's also very simple to use but I can tell you this calibrated device costs about 7000 Euro we're using this internally at thermody for example to measure conductivity and quality of different Alloys we're ordering for example if we want to check copper or if we want to just check our liquid metal for example with this device and also this sensor right here that contains an inductor and also an additional sensor we can perform an ECT and Eddy current testing it's basically the inductor in here if you hold it against a different material which is a non-ferrite material then this is causing a primary magnetic field this primary magnetic field is then causing the eddy current the eddy current again is causing a secondary magnetic field which you can detect with the sense so so basically one magnetic field current second magnetic field you can detect it and this way you can detect the electrical conductivity of the material going back to the beginning of the video you might ask yourself the question now why are we talking about electrical conductivity with this sensor while we are actually looking for the thermal conductivity and I can tell you it's pretty simple because both are directly connected in metals three electrons are responsible both for electrical and thermal conductivity for Metals there is also a second part for the thermal conductivity which is due to phonons inside the metal basically you have tiny vibrations inside the material that are transporting the heat but for metals that is only a very small portion and can be neglected so we can just purely go after the electrical conductivity the relation between electrical and thermal conductivity is written in the v demand franzlaw you can probably tell by the way how I pronounce it those are indeed or were indeed German scientists if we take this piece of copper for example and we will just look at normal Metals for a moment if you heat up this piece of copper both electrical and thermal conductivity will decrease the proportionality of this is written down in the vitamin France law this means if you know one so you know the electrical conductivity you can calculate the other for example the thermal conductivity that's also what you can see inside this child where you can see on top both copper and aluminum and the thermal conductivity over temperature it's very important that we always look at individual materials because they behave differently for example for copper and aluminum what I just said definitely counts for them both as you can see if the temperature increases to the right the thermal conductivity decreases but if you pay attention to Brass which is a copper sink alloy you will notice that it just behaves completely different and with increasing temperature the thermal conductivity is also increasing that is something you can observe for metal alloys whereas for the pure Metals it just behaves completely the opposite and that's something you can make use of for example for liquid metal alloys let's say the term aggressive conductor not because that's a metal alloy and that's beneficial because if you increase the temperature let's say for a mobile CPU then the thermal conductivity will be better if the CPU is running hotter so it's actually increasing in thermal conductivity that is also the reason why we don't use it for liquid nitrogen overclocking because if you run this at like -100 or minus 200 the thermal conductivity is really bad and that's how you get to this equation basically you have to film a conductivity which is called the Lambda and it is a product out of the Lorenz number which is depending on the exact material the temperature in Kelvin and the electrical conductivity in Siemens per meter and this conductivity is what we measure with our device and to explain everything in a simple way again we are measuring the electrical conductivity with this device and it's directly related to the thermal conductivity which means if the electrical conductivity is increasing the thermal conductivity is increasing as well our surface conductivity meter is now set up and ready to go we have different parameters we can read on the display such as the conductivity that is always in relation to Copper that's called iacs so for example we have three different Metals right next to it this is pure copper this is rated with 100 iacs those three metal pieces came out from a lab and they are 100 precise like just a tiny piece of copper like this will cost you like 400 Euro so that's because they are calibration tools then we have brass next to it with 7.16 and then Titan which is actually a very bad conductor with 1.05 and you can also reach the frequency the tool is working with which is 60 kilohertz now if you just want to check it again we can check brass supposed to be 7.16 it's reading 7.14 so it's quite accurate going back to Copper it's reading 100.0 where it was reading 100.1 where it was reading 100.1 before and I was reading 99.9 back to 101. one aspect that's also important to know is the frequency you're measuring at with 60 kilohertz and this is determining how deep you're measuring inside the material this is a fairly low frequency you can also go up into like megahertz region if you go up into the megahertz region then you're only measuring like directly on the surface maybe like 0.01 millimeters on the surface with 60 kilohertz we're reaching a depth of about 0.4 to 0.5 millimeters the good thing about this is if you have tiny scratches in the surface then it's going underneath these scratches but it's going not too deep so we can also measure better thin materials and we can also measure materials that are like 20 centimeters thick what you also have to note is that the surface meter has a temperature compensation that means everything we're measuring is compensated to plus 20 degree celsius this only works for a certain temperature range but right now in plus 28 degrees Celsius this is still completely fine we also have a tolerance on this of zero point and five percent that also explains if we measure something twice it's maybe reading 99 second measurement it's reading 90 99.3 then 99.2 that is how you get those slight variations because of the 0.5 percent but for example if we change between the Intel and the AMD CPU and it's going from 99 to 95 that is a clear indication that there is a difference because that is far outside the tolerance one thing we cannot neglect is any kind of coating on the material or like the nickel plating which you have on heat spreaders right here we have a 12 900k that is lapped we have a 1200k that is not lapped so it's nickel plated we have a 5600x that is nickel plated and a lab 5600x now if we check just for reference again or 100 which is showing 101 if we go to the Intel CPU with nickel plating it's still very very close to 100 if we go to the left Intel CPU it's going close to 102. so actually the copper on the IHS of Intel is purer than our lab calibration one we have here which is already a great sign so that will definitely have a very good thermal conductivity whereas if we check the AMD heat spreader with nickel plating it's going down to 94 sometimes it's reading like close to yeah it's definitely going down and if we read the lapped AMD CPU it's going to 101.5 so there's almost no difference in the copper of both materials as you can see that is like measurement tolerance of the device even though it's you can see it's repeatable but if we check with nickel plating there is definitely a huge impact if we are now following the videoman France law and first starting with our copper reference that is listed with 100 iacs we measured 100.1 which then equals a thermal conductivity of 388. the 12900k was measured with 101.7 and that equals 394. the 5600x was slightly lower with 101.5 which equals 393 but I want to remind you again that the conductivity meter has a tolerance of 0.5 percent which then means if we take the tolerance into account both Intel and AMD CPUs have the same Heats better quality when it comes to the Copper itself however it seems like the nickel plating is quite a lot thicker on the AMD CPU which then results maybe in a little bit worse thermal conductivity overall and now going back to the beginning of the video we will take the rocket cool upgrade heat spreader and then just measure if this really is an upgrade or if it's in the end just the same copper it's actually the first time that I have one of these rocket cool IHS and now it kind of makes sense because there is a slight elevation in the center which is to compensate for the indium solder sheet that originally exists on the 1200k but if you want to replace it with liquid metal it makes sense that you're compensating the height difference with this small elevation because otherwise if you just compare the general shape of the rocket equal IHS versus a standard Intel IHS there's like no difference like the size difference is zero again taking the reference first so it's 100.1 checking rocket cooler it's 101.3 the rocket cool IHS in the end also has about 393 thermal conductivity which means that it will just perform the same as a stock Intel or AMD IHS but it has this additional elevation in the center which will cause your IHS to sit closer to the die if you want to replace your original IHS with this one and liquid metal then you will get better results with this than with the stock IHS because you will have a tinier gap between the die and the diehs but just purely conductivity wise there will be no difference if you are doing some sort of research about thermal conductivity maybe then ending up on Wikipedia checking values of metals you will definitely see on top silver which is listed as the highest conductivity for Metals that's why I got one of these tiny silver pieces which we are going to measure and check the thermal connectivity of course checking our reference again first it's 100.0 then checking to the silver 400 3.3 while 3.6 now calculated the thermal conductivity of silver you can see it's about 410 that's definitely a little bit lower than what you can find online sometimes you can find values around 420. the reason for that is that this is not as pure as what you can theoretically find this is 99.90 but there is also like 99.97 99.99 and the last few ditches definitely matter when it comes to Thermal conductivity that's why this is only about 410 if you want to go higher you have to check for higher Purity one more thing I want to test and highlight is gold because a lot of people have the false assumption that gold has a higher conductivity than silver you often see like HDMI cables or like audio cables that are gold plated also if you check on the back side of your Intel CPU for example all these pads they are made out of copper and they are gold plated the reason for the gold plating is not the better electrical conductivity it's just for protection against corrosion that's the only reason that's the same for all those connectors you have like the HDMI connectors it's simply because you always want to have the same conditions you don't want to get your connector to be worse because for copper for example like this one if you touch it it will corrode over time and your contact would be worse and that's the only reason why you would use gold but let's check the thermal conductivity of this and if we check the gold you will see it's almost 25 percent worse than copper you can see with the calculated value of below 300 it definitely would not make sense to build any kind of cooler out of pure gold it would make much more sense to actually use pure silver because of the high conductivity and we will also get to a water block that is made out of silver but before we get to that we will take another look at some copper Alloys because even though they have a very tiny amount of additional materials this will have a huge impact on conductivity again we're taking our reference first this is something that's called SF copper that basically means that this has a percentage of 0.015 to 0.04 of phosphorus still that means it's 99.9 copper so you could assume that this is pure copper but if we double check this you will see there is a huge impact on the conductivity compared to this normal plate right here that's something I just borrowed from Grizzly that's something we would use for making water blocks for example this will give you 100.4 so that is 99.99 and that is 99.9 but you see that it's a clear difference even though the phosphorus in here is only 0.015 percent now we are getting back to the kaplex kios which is this water block from Aqua computer and this is basically a silver special edition it's listing 0.925 Silver Edition on the back you might also notice Alloys sterling silver nine to five means that it's 92.5 percent of silver which also means that the rest is 7.5 copper usually and we already learned that most of the Alloys will not perform quite well so I can tell you this will be a surprise on the left we have the normal couple legs cryos which is the standard copper version if we measure this one you can see it's a very pure copper alloy the conductivity is great so this will definitely have a very good performance whereas if we measure the Silver version it's actually only reading 96.5 now going back to our chart we will notice that our silver cooler has a thermal conductivity of 390 and if you're now wondering why the silver block block with an aacs of 96.5 can have a higher conductivity than for example the SF copper with 98.9 that is due to the Lawrence number which is different for silver and it's significantly higher for silver than copper but still if we compare it to the Creos copper which is listing at 102.5 iacs it's resulting in 397. Chic joined us for the conclusion but the video has been taking so long that she already fell asleep taking a conclusion about our water blocks we have a clear winner and that is the aqua computer krios which has the purest copper of all the materials we tested it's better than all the ihs's also better than any of the other raw materials of copper which we tested at the same time the Biggest Loser we have is also from Aqua computer and that's the sterling silver edition with a price difference of 150 Euro because this is roughly 100 this is roughly 250. this product absolutely makes no sense sense at all I just don't get why it exists simply because you will pay much more and you will have worse performance at least on paper if you would measure it because this is 390 and this is 397 conductivity you probably won't be able to tell a difference in theory maybe it's like 0.1 degrees worse on your CPU but to be fair to Aqua computer if you go to their website they don't talk about the silver at all I mean their list that it's made of silver but they don't even say that it would be better they don't advertise this so I don't even know why the product exists I guess it's because people think it's made out of silver and then the conductivity must be higher I guess that's an assumption a lot of people have that's why they're probably wasting their money on this even though it's absolutely not worth it going back to the rocket cool IHS because that's something I often read in comments just get your 1200k and put your upgrade IHS on there made out of copper all the ihs's all the products we tested have about the same Purity for the copper the utilizers AMD IHS they have the same Purity as this one so just by buying this that's also made of copper you will not have a benefit on temperatures but what will give you a benefit on temperature is the elevation you have in the center which will make the IHS sit closer to your die and if you use Liquid Metal it will result in better temperatures and also this IHS should be more even than the stock Intel IHS the only thing I don't understand is why is this not nickel plated because definitely liquid metal with the gallium will eat your way into the pure copper if it's not nickel plated so that's something that's definitely missing on here and also if it's a chemical nickel layer that is very thin it will not have an impact on the overall performance of this thing so yeah don't really get it one more thing I don't understand is when I bought this online on Rocket cool the IHS looked much different to this it was also advertising that it has an increased surface area to the stock Intel IHS which this is not doing I don't have an issue with it because it's probably not going to help anyway but it's just weird that I'm getting a different product than what I actually paid for so that's pretty strange I'm not sure if they just forgot to update their website but yeah something they might want to do for the future all right I hope you enjoyed this video about thermal conductivity how to measure it how everything is explained and yeah it was a long video hope you enjoyed it see you next time bye foreign [Music] [Laughter] proportionality

Channel: der8auer EN

Views: 82,152

Keywords:

Id: zbWyZSuoh2w

Length: 21min 11sec (1271 seconds)

Published: Sun Aug 28 2022