Is it me or this should have been done in one single uninterrupted pour as allowing for the metal to start to cool down and solidify before another pour will result in less than ideal properties (aka transition zones prone to cracking/easier to break)? *Just asking, definitely not my area of expertise
Very true, I was wondering why there wasn't a separation in the machined part, then he mentions it was the 3rd casting because the first one had separated.
I'm curious: Wouldn't it be easier to cast a simple cuboid of copper and then mill out the fine structure? Milling the fine structure in the graphite mold seems counterintuitive.
Yes you're right, that's what I wanted to do in the first place but I was fighting the problem of aeration, and the larger the volume of the casting the more chance of air in the final part, and since I wanted to end up with a very beautiful looking part I wanted to minimize air as much as possible.
I'd say this was the right way to do it. While casting a solid block would seem to be the easier option, it has some pretty big disadvantages. Firstly, the sheer amount of raw copper needed, which isn't the cheapest material (I know he's using scrap, but why use more than needed). Then there's the time to machine out the desired shape. Also, copper is notorious for not being easiest material to work with.
@@byrond8592 the air is the problem, the large you go the more air you will get, by staying small with the cross-sectional dimensions of the cast I reduce or completely eliminate the air contamination to get a solid part with no air or pitting. as you pour into a small cavity the liquid copper is protected from the air. that's one reason why they roll large bars of copper instead of casting.
I knew when you stopped pouring you were going to have to make a new pour lol. Thanks for clarifying that when you were done milling, most content providers would have just said " and here's the final product, turned out great" this is why I like your channel, plus I watch you and can see your wicked smart. Your very inspiring. Keep up the great work. I've missed your videos 👍🏻👍🏻👍🏻
Thanks! Yeah I kind of knew too when I was pouring it, I was like oh s***, but it was really heavy and I was losing my grip which is why I set it down and repositioned. But that gave it enough time to cool and become a solid. Thanks I appreciate it!
@@mmellow51 Probably mineral spirits or toluene to thin down the silicone so he could put on a really small even bead. The screws around the edge are doing all the clamping so the silicone is just there to seal.
You have no idea how happy it made me to see you properly heat that copper and graphite before pouring it. So many people can't get copper to that super high temp it needs to flow and never even heat the graphite then wind up with crap. Well done.
Watching how complicated the casting process was just blows my mind that they were doing it 6000 years ago. That must have been a lot of trial and error. Amazing!
He's literally making an artificial sun, with such a high CRI, colours won't look washed out, imagine office lights to actual sunlight, it's amazing. Also, I absolutely love molten copper.
You got it, this light in particular at 2 ft distance is six times brighter than the sun at the subject. I obviously have to diffuse it and treat it a little bit which takes away some brightness. But yeah the CRI of this is amazing and you said it perfectly, the colors won't look washed out they would look very natural, and an office with high CRI lights would probably be amazing to work in. Typically they use fluorescent lights which are very green in color and the CRI sucks probably like 80. If you ever see a fluorescent lamp with a spectrometer it's green is off the charts.
@@WarpedYT I've been wanting to build a custom light for my aquariums. To replicate the sun is my ultimate goal with it. This is so amazing! The applications for this kind of technology is limitless. I love lighting to say the least! It's crazy how just the spectrum alone can change the whole mood of a room! There's an artist named James Turrell who uses light as the medium for all his works. He literally warps your perception of everything from depth to color. I know that chip is experimental, but maybe you could point me in the direction of a similar chip?
Awesome to see good engineering process. Just my 2 cents, if you put the inlet for the coolant low on one side, and high on the other it may work even more efficiently. I think currently the coolant is pretty much just moving across the centre of the core. We had the same issue with our race car intercooler, when we moved the inlet and outlet, we saw a very big change in our intake temps. :-)
When I worked for a copper smelter (ASARCO) in the early 80s we had 3 types furnaces Reverb, Converter and Anode. When we did a pour of the copper out of the Anode furnace the moldes were preheated. Also we used green douglas fir inserted part way in and used a air piston to drive the inserted end of the log to the bottom below the copper surface. This helped to remove entrained O2 from the copper. Anodes were about 86% copper and went on to the refinery.
When an LED requires that level of liquid cooling, that alone speaks volumes about how powerful it is. This is obviously geared to the professional market, because those slow motion shots are amazing, and flicker free.
@@WarpedYT great video by the way. Thank you for sharing. I friend of mine had an amazing idea to machine graphite like that. I would like to share with you.
As an engineering student, this is the content I love to see. I love learning the process and you explaining mistakes that you've made, why they happened, and how you've corrected them is exactly what most engineering project channels on TH-cam lack. Thank you and congrats on the cast!
@@WarpedYT I am no expert, but I have watched a lot of people pour and cast copper (and other metals) into various shapes, and it is a RARE thing to see them end up looking this good. Third attempt or Thirteenth attempt. I think that your "through the flux" technique really helped keep the dissolved oxygen out of the copper. But man, those thin channels? INSANE.
Quick tip for calculating the amount of copper you need: Create the actual part in f360 alongside the mold, set the physical material, then check the body's properties for the volume & weight you need. No maths needed :)
Thanks, I should have done that, I didn't actually think of it. But at the end it was okay because I wanted to make it sort of visual where people could understand instead of just hiding behind the software calculations. But honestly if I would have thought of it I probably would have went that route since I did have an actual part drawing before even the mold.
Details on why the failed castings are definitely the most valuable to anyone wanting to begin casting. Also the transformation from scrap bin into the finished machined part was awesome!
Watching your videos is like watching Gandalf do his stuff - every single episode, without fail, is inspirational, useful and entertaining. Thanks for doing this!
I’m continuously blow away, awestruck & inspired by the amazing level of intelligence, ingenuity, know-how, creativity & extensive range of talents this dude has! An absolute badass of a human being!
At my job we design medical devices, and spend a lot of time working on heat transfer blocks. I've yet to see an optimized block that had a lot of straight parallel lines. The best ones always look organic. Sometimes a straight-lines block is selected because of manufacturing cost, but they never perform as well as the ones that make our prototype shelf look like an H. R. Giger alien autopsy. So, based only on the visual appearance of your block, I suspect that your design is not optimal, but I can't use our expensive machines at work to confirm it. Maybe contact one of the generative design software suppliers? They might be willing to design a part for you, just for the publicity of having you compare it to the straight-lines design.
Great point!, And you are absolutely correct about the parallel lines. However in this case I needed more flow, and I did in fact computer simulate all of my designs before making them. So one of the bigger issues I faced was the tooling I had to use to cut the slots with the max RPM I have available is not well suited to cut intricate shapes at that depth and level, I would have had to change my tooling very often. And with a very lengthy machine times I did not want to babysit this thing for 10 hours. But yeah the type of machine I had to use had a huge Factor in my final design, I also needed a higher flow rate and less pressure drop between the inlet and the outlet of the block. Your spot on though, that was a good comment.
@@WarpedYT If you want to squeak a little more performance out of your cooling loop, either add a little bit of isopropyl alcohol to it with a maybe a single drop of soap, or get a bottle of water wetter... It will help with the boundary layer flow by acting as wetting agent to allow even greater heat transfer. It doesn't take long for it to work either, put it in, and maybe 2-3 minutes depending on your flow rate to get everything distributed, you should be able to see it change. Might be an idea for a video if you can add it to a stable system while the pump is running, like add it to the reservoir once you hit your steady state 122F... see how much it changes.
@@kleetus92 are you into pc building by any chance, I have never heard someone mention this and I wonder if its viable in a PC water loop in terms of longevity.
@@TurkeyOW I am. It will work in any water based heat transfer system. My first time seeing it was in high performance car applications where the cooling system was being pushed to the limit, even with a bigger radiator. It may not solve all your problems, but it will give a few extra percent of heat transfer. You can buy a product literally called 'water wetter' in car parts stores... it's a clear redish pink liquid, apply per instructions.
@@TurkeyOW you may not see any significant improvement in PC application where we're deaing with very light loads (only a few undred watts), compared to car engine cooling. Also, acrylic parts hate alcohol :) it's generally advised not to use alcohol in custom loop coolants, or clean those parts with alcohol based cleaning solutions.
I wanted to say that this is a waste of talent that you don’t work for some big engineering firm (NASA, Boeing or whatever). But then I remember, if there are just a few kids who watch these videos and become an engineer, then you help advancing humanity even more. And not just as an entertainment.
Ha! Well thanks! I appreciate that, but I will have to mention I don't reveal where I work or what I do 😉. And that is exactly why I started my TH-cam channel, exactly what you said. The only reason we don't have more engineers mechanics and scientists with these kids growing up right now is because they don't know how much fun it is to know how to build things. I'm hoping to try to change that soon... I just got to post more videos.
Very nice work! I love how clean the finished product came out! The design of the block could be adjusted to get a more even flow across all of the channels, increasing the heat exchange. As it is now, the upper rows wont get as much flow through them as the coolant wont want to fight gravity to go through those channels. Placing the inlet in the top corner and the outlet on the opposing bottom can help this immensely. It could let you cut down on some material too if you end up making a bunch of these for this experimental LED. But hey, if it works it works!
I was actually thinking the opposite with the inlet at the bottom and outlet at the top. I feel like having the inlet at the top could cause the water in the other top corner and around the center to stagnate because the cold water would want to sink straight down, then across rather than trying to flow across the fins. Having the inlet at the top would at least evacuate the hottest water first. It might even be best to place the inlet right over the center so the center always gets the coolest water. That being said, I ain't no scientist so I could definitely be wrong haha
If he made the waterblock side that the led mounts on a mirror finish it would have helped alot in transferring the heat as well. I'm not sure what he used as a TIM but if it the block was as smooth as possible it would have needed less and cooled better.
I'll have to say this is really cool that you took the time to be precise. Every time you melt down the copper you burn off impurities to make it pure. It's really bright you can't see your hand. I can only imagine what happens at night. Be safe team
I know I had no choice I was so behind on this project, but finally it's behind me and the jet powered Tesla and the four months that the Tesla was broken down, time to move on to some free flying fun videos now
I am wondering why you went through the whole struggle with the graphite mold and the nasty dust everywhere, instead of machining it from solid copper or using a lost mold made of sand?
I assume he just didn't think about it at the time. You can't really say the mold reduced Machining time since he took like 12 hours to Machine the graphite anyway.
yes, thank you!. no, I need this light when I'm filming 1M FPS, Slow-mo Guys need more ideas and to unblock me from commenting on their channel, also they need to stop combing my comments section for ideas....LBVS
I would drill and tap a hole on top of the water block and use a bolt with a o ring to seal it, just to make priming the block easier and ensure its totally full
That's a great point, however since I want this to be a maintenance free device, I'm actually redesigning that back cover to allow it to automatically bleed and fill, this back cover was just for the video
I would also use a flexible sealant like Hylomar or instant gasket to seal the cover onto the block as the coefficient of thermal expansion of acrylic is about 4x that of copper.
On the Water cooled Copper heat sink, the water input connector should be placed further to the top and the output water should be placed lower. This would ensure all the water was being replaced, and not only some. This would help with the cooling.
I love watching this guy - his mannerisms and accents reminds me so much of my two college roommates from South Jersey. Plus he is brilliant and a heck of a machinist, mechanic, welder/fabricator, and really knows his electronics. Dang good filmmaker too. Glad I found this channel.
I just noticed his right hand is missing a finger, this tells me he has alot of full time experience in engineering. Love to be your assistant. Love all your videos 👍
Having that just be a solid block of copper is less efficient than it could be. You could find and order a vapor chamber sized up to a close fit, and use that as an interface between the diode and the waterblock. Or, failing that, you could press flatten a bunch of heat pipes and have them spread the heat across the water block. Of course, both options will need a lot of thermal compound or epoxy to actually get the job done, but the end result would be a more efficient use of the surface area within the waterblock, to get the heat away. Even as it is, you've got it running at just above spec for most silicon processors, which should be within spec for heatpipes and vapor chambers, so adding something to better spread the heat output could easily reduce the tempurature by several degrees.
Thanks for that info, you're absolutely correct, however I wanted to keep the manufacturing as simple and cheap as possible. I will get the temperature down further once I had a large radiator outside the unit. Or I might redesign it later on all together to be even smaller. In one of my earlier iterations I did in fact try the heat pipe idea, it did not work. If I had a better mill I could have totally went smaller on the fins. I didn't fact try to go as soon as possible on the back plate, but what I ended up with at a certain point was a hot spot right where the chip mounts on the liquid side of the block, I had to change up the geometry of the inside of the water block to allow a more even heat spread. I may end up redesigning it later on but for now I just need the light to work for a project... Thanks I always love the intellectual comments cuz it gets my gears turning!
Wouldn’t you get better heat transfer from copper to coolant if the intake and exhaust ports for the coolant were offset? (Intake at the top left and exhaust at the bottom right for example)
Intake at bottom, exhaust at top for better effect. ( Heat rises, self venting, counter current circulation to a larger surface area of the copper block)
i was thinking that as well, plus microfinning (probably wouldnt even need to be that "micro") all the horizontal channels to dramatically increase the actual surface area and help with heat transfer.
2 relative easy sugestions to make it even colder: 1. Install input (or incoming stream) fitting on lower part and output fitting on upper part of plexiglass.(BC thermodynamics say that cold gas/liquid is allways at bottom and warmer allways at top, with my suggested setup You should be able to remove more hot coolant from cooler) Or even easier, just rotate block 90° relative to ground, so input fitting on bottom, output on top. 2. Polish surface where LED chip is mounted on block (or whole block plane it whould be easier), You will get better contact between 2 planes and better heat transfer from chip to cooler. Otherwise amazing cooler 👌🏻 I like this project 👍
That's a great idea, as for the cover this was actually only temporary because I need to use it in any orientation, so I actually redesigned that entire backplate to allow it to Auto bleed as well as deliver coolant in any position. As for polishing the back part that I should have done and I did not think about but that makes absolute perfect sense. I'm sure if you magnify down on that back plate there are valleys and peaks for sure... I totally have to do that. Thanks!
That's a Nichia turbo array led prototype?!? I think yes! What an absolute unit here! New level! To be used in IMAX projectors etc! It's amazing - how can I get my paws on one of these? Love it!
Wow Matt! You keep raising the bar on showcasing a mix between technology, wrenching, creating, and imagination. I wish people grasped how impressive what you do is considering you aren’t working at a multi billion dollar R&D company.
Thanks! Yeah the last few episodes I've been trying to make them hybrids. Doing exactly what you said , next thing I'm going to be working on is just making more videos and making it more organic so I can be more myself and more of a character. A lot of times I'm very rigid because I need to make sure I hit all of my points throughout the video to keep the story solid, especially with complicated subjects. This is necessary for me to show the videos I'm going to be showing in the near future where I'm going to let people into some of my actual projects.. thanks for the kind words!
My question, why not just pour a solid block of copper and then machine the piece from that? Instead of making a graphite mold, and then pouring the copper into it? Just to clean it up with the mill in the end anyway.
For a couple of reasons but mostly I'm aiming to create some mind-blowing high-speed videos... Been working with high-speed stuff for a long time now and the biggest problem is lighting.
Thank you! It's been tough to get things moving over here! I have a way I want to do things and a certain vision but now I just canceled out everything else in my life to focus on this and my projects.
Way to increase efficacy and design on this would be to: 1. machine a o-ring channel into the copper to better seal and easy removal to clean the water block. 2. No sure about mill or limitations but cutting smaller, denser fins into the copper will give more surface area to exchange heat. Also cross cutting the channels will give more surface area as well. Overall the finished product worked and looked great!
All your points are right on the money, I didn't machine or no ring because I didn't make the wall thick enough because I was trying to save weight. As for the fins, that's exactly why I didn't cut them thinner, that was as small as I could go on that Mill with that RPM all that would have taken a week to mill that thing. Same thing with the cross cutting, my original design idea was cross cut and leave tapered spikes for the heat transfer but with the current Mill that I'm using, very low RPM, that would have taken an eternity...lol. thank you
A simple crucible fitting loop on the end of a rod as a pouring shank would have helped you a lot. Scissor type pouring shanks are both uselessly awkward and dangerous. You must do a pour like this in one go, stopping and starting will always give you discontinuities......Martin
Nice job... Its probably already been suggested but you should try the following: 1: current design turn the heatsink 90 degrees 2: modify the current design with your inlet hose top left corner and outlet hose bottom right 3: modify number 2 and have the fins on the block at 45 degrees The idea is for all the above to increase the flow efficiency of the fluid in the block thereby reducing the temp. (you could also try radiator coolant ???)
What some overclockers like to do is throw the whole radiator in an ice bath. If you use something like acetone or alcohol to cool with, you should be able to get it colder than that with no problems. You could also get some extra fans and do a push pull configuration to get a bit better airflow over the fins of the radiator.
Just a tip, you could just check the full volume of the part in fusion, compared to the volume without the extruded channels by right clicking on the body to inspect it.
hate to be late but worth a shot. What about adding an O-ring and making a checkered pattern, it would add a bit of extra machining time but I wonder how this would affect performance. If adding these helps make the block hold more pressure, you could use a stronger pump and go longer and dial the pump back and extend the life of the pump.
Just quit a job where i was machining all sorts of heatsinks just like this but out of blocks we extruded, so seeing it done basically backwards from nothing is pretty neat
Thank you for sharing your learning process with us! Saves me time when i go to do this kind of thing myself. I learned; use flux BEFORE hand, when it fully melts, then add your metal
when you make your housing for this thing, i would add a second radiator in the loop and a small fan with a duct to blow a little air over the front of the LED to push away that superheated pocket of air, convection will move some but forcing cool air across it will help.
usted vive el sueño de todo ingeniero, vivir para crear, construir, aprender y compartir con orgullo todo lo adquirido por el bien de la humanidad, mis respetos señor, usted es fenomenal
That block is too beautiful to enclose in sheet metal. If your plans can accommodate the idea of it being open, you should (in my opinion) consider a way to do so.
I work with alot of cooling components for very powerful computers and hands down the best way to do a liquid cooled block is to have a completely hollow cavity and fill it with fin stock the will allow coolant through and then solder or braze a cover over it. the fin stock maximizes the area that the coolant is touching and it can wick away heat insanely good
I agree with you, that was the way I wanted to go, however I needed something very strong in robust to use on set for film and my TH-cam videos. I can almost guarantee you that if I just soldered a cover over some thin stock, during the first shoot I would have coolant all over the place and a burned out cob. When filming even TH-cam videos and you're in a hurry or you're caught in a situation stuff gets knocked down hit abused, this light is not going to just sit on the desk I can assure you. That being said I completely agree and in the future when I get my next CNC machine in a few months, I may do some really high RPM intricate milling on another piece just like this to make an improved version. Unfortunately the automatic manual machine I'm using now is based on a dinosaur machine.
here's an idea for cooling. liquid cooling. but not just any liquid cooling, but one that both makes use of a very thermally efficient fluid, and basically high pressure pulse width modulated high capacity liquid to air intercooler. the idea is to have the fluid have a periodically interrupted flow to help the intercooler which itself is liquid cooled by a highly chilled liquid, to help with more efficient transfer of heat energy. think 2 cooling circuits running parallel to eachother through the intercooler block in such a way that it can constantly be kept cool from one resivoir of a cooled liquid while absorbing heat and more quickly cooling the liquid to the main circuit. no idea if it would work, but it seems like it could be a more efficient way to do it than a single intercooler, and not need a second one to compensate.
One design flaw is that you should've machined thinner fins closer to each other and then cross-cutting them, making many "pins" instead of channels. This way you would've created way more heat-dissipating surface in contact with water, increasing the efficiency exponentially; this technique is used in PC watercooling and it works great
It might have been mentioned already (don't want to read 1K comments to find out!) - but a suggestion for your port positions in the block cover (as an automotive guy!) - Put the inlet near the top, and the outlet near the bottom vs side to side. Makes for better flow through your core.
Brilliant work! It looks like you might get short-cycling of the coolant by having the inlet and outlet in the center, but if the chip is small enough and centered on the water block, that's probably not a bad thing, as the greatest flow rate of the coolant will be at the center, where the most heat is being generated. Are you using a thermal paste between the chip and the block to maximize the heat transfer from chip to copper?
Maybe you should try a led->peltier->waterblock sandwich. You will cool the hot side of peltier with the waterblock while the cold side cools the led with sub ambient temps.
Wow this channel never ceases to amaze me. They always make me want to get off ass and do something. Thank you again for bringing this content to the people
What do I think, I think that you have more talent in your little finger than most people will ever have in their whole lives. You do amazing stuff. Love it!
![ละลาย (LALALYE) - DAOU PITTAYA ต้าห์อู๋ พิทยา [OFFICIAL MV]](http://i.ytimg.com/vi/wo6r9TgausI/mqdefault.jpg)
Is it me or this should have been done in one single uninterrupted pour as allowing for the metal to start to cool down and solidify before another pour will result in less than ideal properties (aka transition zones prone to cracking/easier to break)? *Just asking, definitely not my area of expertise
He was saying that it was too heavy and he was losing his grip.
He recast it 2 more times to get it right... says in the video
Very true, I was wondering why there wasn't a separation in the machined part, then he mentions it was the 3rd casting because the first one had separated.
@@kingcosworth2643 Yey to basic physics based logic :)
It looks like he worked around that by superheating the final pour over the furnace to allow the whole item to re-melt into one item
I'm curious: Wouldn't it be easier to cast a simple cuboid of copper and then mill out the fine structure? Milling the fine structure in the graphite mold seems counterintuitive.
Maybe playing with molten copper and building new skills is the goal. 🤷♂️
Yeah probably counter intuitive but got the same result lol
Yes you're right, that's what I wanted to do in the first place but I was fighting the problem of aeration, and the larger the volume of the casting the more chance of air in the final part, and since I wanted to end up with a very beautiful looking part I wanted to minimize air as much as possible.
I'd say this was the right way to do it. While casting a solid block would seem to be the easier option, it has some pretty big disadvantages. Firstly, the sheer amount of raw copper needed, which isn't the cheapest material (I know he's using scrap, but why use more than needed). Then there's the time to machine out the desired shape. Also, copper is notorious for not being easiest material to work with.
@@byrond8592 the air is the problem, the large you go the more air you will get, by staying small with the cross-sectional dimensions of the cast I reduce or completely eliminate the air contamination to get a solid part with no air or pitting. as you pour into a small cavity the liquid copper is protected from the air. that's one reason why they roll large bars of copper instead of casting.
I knew when you stopped pouring you were going to have to make a new pour lol. Thanks for clarifying that when you were done milling, most content providers would have just said " and here's the final product, turned out great" this is why I like your channel, plus I watch you and can see your wicked smart. Your very inspiring. Keep up the great work. I've missed your videos 👍🏻👍🏻👍🏻
Thanks! Yeah I kind of knew too when I was pouring it, I was like oh s***, but it was really heavy and I was losing my grip which is why I set it down and repositioned. But that gave it enough time to cool and become a solid. Thanks I appreciate it!
@@WarpedYT you have a great channel I really liked the diesel powered go cart and then same cart jet powered great stuff man I love all your videos
@@byker4lyfe1 bring the diesel kart back!
Great thing! Did the mould survive all the casting tests or were there some more?
You're*
Every man‘s dream: Have all the machinery and tools and you can build almost anything from scratch.
Very inspirational! 👍
Great engineering accomplishment while also very educational and entertaining!
Thanks this one was really tough.
@@WarpedYT you are great engineer 👍
@@WarpedYT What was that mixture you made 7:28 and why use it? It looked like Silicone/RTV and something else.
@@mmellow51 Probably mineral spirits or toluene to thin down the silicone so he could put on a really small even bead. The screws around the edge are doing all the clamping so the silicone is just there to seal.
You have no idea how happy it made me to see you properly heat that copper and graphite before pouring it. So many people can't get copper to that super high temp it needs to flow and never even heat the graphite then wind up with crap.
Well done.
Thank you, yep that's for sure the graphite conducts way too much heat plus you don't want any moisture. But you know what's up
Is this because copper is so thermally conductive and loses heat fast?
I am really IMPRESSED with the quality of that 3rd casting
Thank you
Same. It looks like a production item with how clean it came out
Watching how complicated the casting process was just blows my mind that they were doing it 6000 years ago. That must have been a lot of trial and error. Amazing!
6000 years ago the process is like euv today
He's literally making an artificial sun, with such a high CRI, colours won't look washed out, imagine office lights to actual sunlight, it's amazing. Also, I absolutely love molten copper.
You got it, this light in particular at 2 ft distance is six times brighter than the sun at the subject. I obviously have to diffuse it and treat it a little bit which takes away some brightness. But yeah the CRI of this is amazing and you said it perfectly, the colors won't look washed out they would look very natural, and an office with high CRI lights would probably be amazing to work in. Typically they use fluorescent lights which are very green in color and the CRI sucks probably like 80. If you ever see a fluorescent lamp with a spectrometer it's green is off the charts.
@@WarpedYT I've been wanting to build a custom light for my aquariums. To replicate the sun is my ultimate goal with it. This is so amazing! The applications for this kind of technology is limitless.
I love lighting to say the least! It's crazy how just the spectrum alone can change the whole mood of a room! There's an artist named James Turrell who uses light as the medium for all his works. He literally warps your perception of everything from depth to color. I know that chip is experimental, but maybe you could point me in the direction of a similar chip?
Awesome to see good engineering process.
Just my 2 cents, if you put the inlet for the coolant low on one side, and high on the other it may work even more efficiently. I think currently the coolant is pretty much just moving across the centre of the core. We had the same issue with our race car intercooler, when we moved the inlet and outlet, we saw a very big change in our intake temps. :-)
Makes sense. Better liquid distribution over the heat sink.
It warms my heart to see someone who cares about QUALITY as much as quantity. Great work!
When I worked for a copper smelter (ASARCO) in the early 80s we had 3 types furnaces Reverb, Converter and Anode. When we did a pour of the copper out of the Anode furnace the moldes were preheated. Also we used green douglas fir inserted part way in and used a air piston to drive the inserted end of the log to the bottom below the copper surface. This helped to remove entrained O2 from the copper. Anodes were about 86% copper and went on to the refinery.
When an LED requires that level of liquid cooling, that alone speaks volumes about how powerful it is. This is obviously geared to the professional market, because those slow motion shots are amazing, and flicker free.
Oh man. This would be great for my large tesla coils' bridge.
Awesome work!
I'm gonna check that out is that on your channel?
@@WarpedYT yes, it i!!! His channel is amazing. Im a big fan
@@WarpedYT great video by the way. Thank you for sharing. I friend of mine had an amazing idea to machine graphite like that. I would like to share with you.
Ok feel free to share, you can DM me on Instagram or email "Carbon12TV at Gmail Dot Com"
Thank you glad you liked it .. it was so much work
As an engineering student, this is the content I love to see. I love learning the process and you explaining mistakes that you've made, why they happened, and how you've corrected them is exactly what most engineering project channels on TH-cam lack. Thank you and congrats on the cast!
With all the machining involved, would it be more cost efficient and easier to cast a slab of copper and then mill it all out?
Yes absolutely
Have you milled copper? If he broke drills and taps how well would milling have gone?
Showing failures is good. Working past them is better. Showing how you did it, the best.
Thank you, considering that I almost gave up on this project, I appreciate that
@@WarpedYT I am no expert, but I have watched a lot of people pour and cast copper (and other metals) into various shapes, and it is a RARE thing to see them end up looking this good. Third attempt or Thirteenth attempt. I think that your "through the flux" technique really helped keep the dissolved oxygen out of the copper. But man, those thin channels? INSANE.
Quick tip for calculating the amount of copper you need: Create the actual part in f360 alongside the mold, set the physical material, then check the body's properties for the volume & weight you need. No maths needed :)
Thanks, I should have done that, I didn't actually think of it. But at the end it was okay because I wanted to make it sort of visual where people could understand instead of just hiding behind the software calculations. But honestly if I would have thought of it I probably would have went that route since I did have an actual part drawing before even the mold.
Details on why the failed castings are definitely the most valuable to anyone wanting to begin casting. Also the transformation from scrap bin into the finished machined part was awesome!
Watching your videos is like watching Gandalf do his stuff - every single episode, without fail, is inspirational, useful and entertaining. Thanks for doing this!
Thanks!
I’m continuously blow away, awestruck & inspired by the amazing level of intelligence, ingenuity, know-how, creativity & extensive range of talents this dude has! An absolute badass of a human being!
At my job we design medical devices, and spend a lot of time working on heat transfer blocks. I've yet to see an optimized block that had a lot of straight parallel lines. The best ones always look organic. Sometimes a straight-lines block is selected because of manufacturing cost, but they never perform as well as the ones that make our prototype shelf look like an H. R. Giger alien autopsy.
So, based only on the visual appearance of your block, I suspect that your design is not optimal, but I can't use our expensive machines at work to confirm it. Maybe contact one of the generative design software suppliers? They might be willing to design a part for you, just for the publicity of having you compare it to the straight-lines design.
Great point!, And you are absolutely correct about the parallel lines. However in this case I needed more flow, and I did in fact computer simulate all of my designs before making them. So one of the bigger issues I faced was the tooling I had to use to cut the slots with the max RPM I have available is not well suited to cut intricate shapes at that depth and level, I would have had to change my tooling very often. And with a very lengthy machine times I did not want to babysit this thing for 10 hours. But yeah the type of machine I had to use had a huge Factor in my final design, I also needed a higher flow rate and less pressure drop between the inlet and the outlet of the block. Your spot on though, that was a good comment.
@@WarpedYT If you want to squeak a little more performance out of your cooling loop, either add a little bit of isopropyl alcohol to it with a maybe a single drop of soap, or get a bottle of water wetter... It will help with the boundary layer flow by acting as wetting agent to allow even greater heat transfer. It doesn't take long for it to work either, put it in, and maybe 2-3 minutes depending on your flow rate to get everything distributed, you should be able to see it change.
Might be an idea for a video if you can add it to a stable system while the pump is running, like add it to the reservoir once you hit your steady state 122F... see how much it changes.
@@kleetus92 are you into pc building by any chance, I have never heard someone mention this and I wonder if its viable in a PC water loop in terms of longevity.
@@TurkeyOW I am. It will work in any water based heat transfer system. My first time seeing it was in high performance car applications where the cooling system was being pushed to the limit, even with a bigger radiator. It may not solve all your problems, but it will give a few extra percent of heat transfer. You can buy a product literally called 'water wetter' in car parts stores... it's a clear redish pink liquid, apply per instructions.
@@TurkeyOW you may not see any significant improvement in PC application where we're deaing with very light loads (only a few undred watts), compared to car engine cooling.
Also, acrylic parts hate alcohol :) it's generally advised not to use alcohol in custom loop coolants, or clean those parts with alcohol based cleaning solutions.
I have never seen anyone hold a pen like you do. Good video. 10/10
I wanted to say that this is a waste of talent that you don’t work for some big engineering firm (NASA, Boeing or whatever). But then I remember, if there are just a few kids who watch these videos and become an engineer, then you help advancing humanity even more. And not just as an entertainment.
Ha! Well thanks! I appreciate that, but I will have to mention I don't reveal where I work or what I do 😉. And that is exactly why I started my TH-cam channel, exactly what you said. The only reason we don't have more engineers mechanics and scientists with these kids growing up right now is because they don't know how much fun it is to know how to build things. I'm hoping to try to change that soon... I just got to post more videos.
Skill, knowledge, patience and having the right tools. The recipe for success.
Very nice work! I love how clean the finished product came out! The design of the block could be adjusted to get a more even flow across all of the channels, increasing the heat exchange. As it is now, the upper rows wont get as much flow through them as the coolant wont want to fight gravity to go through those channels. Placing the inlet in the top corner and the outlet on the opposing bottom can help this immensely. It could let you cut down on some material too if you end up making a bunch of these for this experimental LED. But hey, if it works it works!
I was actually thinking the opposite with the inlet at the bottom and outlet at the top.
I feel like having the inlet at the top could cause the water in the other top corner and around the center to stagnate because the cold water would want to sink straight down, then across rather than trying to flow across the fins. Having the inlet at the top would at least evacuate the hottest water first.
It might even be best to place the inlet right over the center so the center always gets the coolest water.
That being said, I ain't no scientist so I could definitely be wrong haha
If he made the waterblock side that the led mounts on a mirror finish it would have helped alot in transferring the heat as well. I'm not sure what he used as a TIM but if it the block was as smooth as possible it would have needed less and cooled better.
Anything made out of copper looks great in my opinion.
Outstanding. Every project you do is cool AF. Looking forward to you describing the significance of the type of light you are using.
I'll have to say this is really cool that you took the time to be precise.
Every time you melt down the copper you burn off impurities to make it pure.
It's really bright you can't see your hand. I can only imagine what happens at night.
Be safe team
Thank you !
You've been teasing your subscribers about for so long, it's cool to see the finished product and the full design process behind it.
I know I had no choice I was so behind on this project, but finally it's behind me and the jet powered Tesla and the four months that the Tesla was broken down, time to move on to some free flying fun videos now
@@WarpedYT Can't wait!
Mount the cooler so the fins are vertical so you get even flow over all the fins.
Son of a biscuit I am simply enamored with your process, production, and product. Literally my favorite content creator on the planet 🔥
I am wondering why you went through the whole struggle with the graphite mold and the nasty dust everywhere, instead of machining it from solid copper or using a lost mold made of sand?
My question too! Whats the answer?
I assume he just didn't think about it at the time. You can't really say the mold reduced Machining time since he took like 12 hours to Machine the graphite anyway.
Followed the posts you were posting, and you blew my expectations away. This is what the Slo-Mo guys need when filming in 1M fps.
yes, thank you!. no, I need this light when I'm filming 1M FPS, Slow-mo Guys need more ideas and to unblock me from commenting on their channel, also they need to stop combing my comments section for ideas....LBVS
I would drill and tap a hole on top of the water block and use a bolt with a o ring to seal it, just to make priming the block easier and ensure its totally full
That's a great point, however since I want this to be a maintenance free device, I'm actually redesigning that back cover to allow it to automatically bleed and fill, this back cover was just for the video
@@WarpedYT Awesome! I can't wait to see the completed version.
I would also use a flexible sealant like Hylomar or instant gasket to seal the cover onto the block as the coefficient of thermal expansion of acrylic is about 4x that of copper.
On the Water cooled Copper heat sink, the water input connector should be placed further to the top and the output water should be placed lower. This would ensure all the water was being replaced, and not only some. This would help with the cooling.
notification squad, Have a nice weekend guys!🔥🔥🔥
Thanks you as well !
I love watching this guy - his mannerisms and accents reminds me so much of my two college roommates from South Jersey. Plus he is brilliant and a heck of a machinist, mechanic, welder/fabricator, and really knows his electronics. Dang good filmmaker too. Glad I found this channel.
Dude! Your channel is absolutely insane, loved every second of this video!
Thanks!!
I just noticed his right hand is missing a finger, this tells me he has alot of full time experience in engineering. Love to be your assistant. Love all your videos 👍
*this man is the definition of an Engineer going God Mode.* 💪🏻🔥
Thank you !
Having that just be a solid block of copper is less efficient than it could be. You could find and order a vapor chamber sized up to a close fit, and use that as an interface between the diode and the waterblock. Or, failing that, you could press flatten a bunch of heat pipes and have them spread the heat across the water block. Of course, both options will need a lot of thermal compound or epoxy to actually get the job done, but the end result would be a more efficient use of the surface area within the waterblock, to get the heat away. Even as it is, you've got it running at just above spec for most silicon processors, which should be within spec for heatpipes and vapor chambers, so adding something to better spread the heat output could easily reduce the tempurature by several degrees.
Thanks for that info, you're absolutely correct, however I wanted to keep the manufacturing as simple and cheap as possible. I will get the temperature down further once I had a large radiator outside the unit. Or I might redesign it later on all together to be even smaller. In one of my earlier iterations I did in fact try the heat pipe idea, it did not work. If I had a better mill I could have totally went smaller on the fins. I didn't fact try to go as soon as possible on the back plate, but what I ended up with at a certain point was a hot spot right where the chip mounts on the liquid side of the block, I had to change up the geometry of the inside of the water block to allow a more even heat spread. I may end up redesigning it later on but for now I just need the light to work for a project... Thanks I always love the intellectual comments cuz it gets my gears turning!
Wouldn’t you get better heat transfer from copper to coolant if the intake and exhaust ports for the coolant were offset? (Intake at the top left and exhaust at the bottom right for example)
Intake at bottom, exhaust at top for better effect. ( Heat rises, self venting, counter current circulation to a larger surface area of the copper block)
i was thinking that as well, plus microfinning (probably wouldnt even need to be that "micro") all the horizontal channels to dramatically increase the actual surface area and help with heat transfer.
@@za7v9ier yup and you will get more even flow when the pressure of the water is forced up.
2 relative easy sugestions to make it even colder: 1. Install input (or incoming stream) fitting on lower part and output fitting on upper part of plexiglass.(BC thermodynamics say that cold gas/liquid is allways at bottom and warmer allways at top, with my suggested setup You should be able to remove more hot coolant from cooler)
Or even easier, just rotate block 90° relative to ground, so input fitting on bottom, output on top.
2. Polish surface where LED chip is mounted on block (or whole block plane it whould be easier), You will get better contact between 2 planes and better heat transfer from chip to cooler. Otherwise amazing cooler 👌🏻 I like this project 👍
That's a great idea, as for the cover this was actually only temporary because I need to use it in any orientation, so I actually redesigned that entire backplate to allow it to Auto bleed as well as deliver coolant in any position. As for polishing the back part that I should have done and I did not think about but that makes absolute perfect sense. I'm sure if you magnify down on that back plate there are valleys and peaks for sure... I totally have to do that. Thanks!
Beautiful work!
What was the solvent used to dissolve the silicone?
Mineral spirits... Nice trick not many know. Also goop and toluene work well but that solvent is nasty stuff.
Had the same question. Super thanks
Thank you for the copper casting tip about adding extra flux to the crucible first, heating the flux till its melted then adding the copper //
That's a Nichia turbo array led prototype?!? I think yes! What an absolute unit here! New level! To be used in IMAX projectors etc! It's amazing - how can I get my paws on one of these? Love it!
awesome....Glad to have you back making videos.
Thanks! I'm glad to have finally pulled everything together now
Wow Matt! You keep raising the bar on showcasing a mix between technology, wrenching, creating, and imagination. I wish people grasped how impressive what you do is considering you aren’t working at a multi billion dollar R&D company.
Thanks! Yeah the last few episodes I've been trying to make them hybrids. Doing exactly what you said , next thing I'm going to be working on is just making more videos and making it more organic so I can be more myself and more of a character. A lot of times I'm very rigid because I need to make sure I hit all of my points throughout the video to keep the story solid, especially with complicated subjects. This is necessary for me to show the videos I'm going to be showing in the near future where I'm going to let people into some of my actual projects.. thanks for the kind words!
glad to see you used copper wire, it is always the cleanest copper 'scrap' that you can get.
You are a very clever man. I am in awe of your technical versatility.
Well this was sick.
Thank you !
My question, why not just pour a solid block of copper and then machine the piece from that? Instead of making a graphite mold, and then pouring the copper into it? Just to clean it up with the mill in the end anyway.
Love me some high efficiency led cobs
I wondered why you needed such a bright light. Very cool video!
For a couple of reasons but mostly I'm aiming to create some mind-blowing high-speed videos... Been working with high-speed stuff for a long time now and the biggest problem is lighting.
Getting better and better content. Keep it up bud.
Thank you! It's been tough to get things moving over here! I have a way I want to do things and a certain vision but now I just canceled out everything else in my life to focus on this and my projects.
Dudes got one man army but for machining and making things. Guy could do anything
Thank you for the tip about the flux before melting the copper to keep out gasses. Awesome channel.
Your welcome! And thank you
Way to increase efficacy and design on this would be to:
1. machine a o-ring channel into the copper to better seal and easy removal to clean the water block.
2. No sure about mill or limitations but cutting smaller, denser fins into the copper will give more surface area to exchange heat. Also cross cutting the channels will give more surface area as well.
Overall the finished product worked and looked great!
All your points are right on the money, I didn't machine or no ring because I didn't make the wall thick enough because I was trying to save weight. As for the fins, that's exactly why I didn't cut them thinner, that was as small as I could go on that Mill with that RPM all that would have taken a week to mill that thing. Same thing with the cross cutting, my original design idea was cross cut and leave tapered spikes for the heat transfer but with the current Mill that I'm using, very low RPM, that would have taken an eternity...lol. thank you
@@WarpedYT Like i said before, the part turned out awesome! Really enjoyed the whole process.
A simple crucible fitting loop on the end of a rod as a pouring shank would have helped you a lot. Scissor type pouring shanks are both uselessly awkward and dangerous. You must do a pour like this in one go, stopping and starting will always give you discontinuities......Martin
Crucible fitting loop, I will look that up, yes I don't like this scissor type at all, it always makes me nervous
Nice job...
Its probably already been suggested but you should try the following:
1: current design turn the heatsink 90 degrees
2: modify the current design with your inlet hose top left corner and outlet hose bottom right
3: modify number 2 and have the fins on the block at 45 degrees
The idea is for all the above to increase the flow efficiency of the fluid in the block thereby reducing the temp.
(you could also try radiator coolant ???)
This was both inspiring and entertaining. I am now going to capitalize on this and stop procrastinating.
What some overclockers like to do is throw the whole radiator in an ice bath. If you use something like acetone or alcohol to cool with, you should be able to get it colder than that with no problems. You could also get some extra fans and do a push pull configuration to get a bit better airflow over the fins of the radiator.
Those are great ideas !!
and slap some Noctua iPPC 3000RPM 120mm on the rad! instead of those cheap DC fans from 1995 :)
Freaking Amazing. The slow motion was as sharp as that knife.
Thank you
Thats bright as hell. What a cool project
May your hardwork and ideas bear fruit soon
Just a tip, you could just check the full volume of the part in fusion, compared to the volume without the extruded channels by right clicking on the body to inspect it.
I know but I wanted to make it more visual for those who don't use CAD
Future reference, with scrap copper you need to melt and recast a few times. Maybe make some blocks next time then when it's super clean cast it
hate to be late but worth a shot.
What about adding an O-ring and making a checkered pattern, it would add a bit of extra machining time but I wonder how this would affect performance. If adding these helps make the block hold more pressure, you could use a stronger pump and go longer and dial the pump back and extend the life of the pump.
Solid copper water tubing and a larger radiator/pump will knock those temps down (this only will help if it's more a static install).
Just quit a job where i was machining all sorts of heatsinks just like this but out of blocks we extruded, so seeing it done basically backwards from nothing is pretty neat
Thank you for sharing your learning process with us! Saves me time when i go to do this kind of thing myself. I learned; use flux BEFORE hand, when it fully melts, then add your metal
Yes that method worked amazing for me.
Your cinematography skills are on point!
Thank you
Thanks for the great videos. I can appreciate the hard work and patience you put into these videos.
Absolutely gorgeous piece! It's a work of art and science
when you make your housing for this thing, i would add a second radiator in the loop and a small fan with a duct to blow a little air over the front of the LED to push away that superheated pocket of air, convection will move some but forcing cool air across it will help.
Love that idea, that makes perfect sense I believe it is exactly that, a superheated pocket of air.
usted vive el sueño de todo ingeniero, vivir para crear, construir, aprender y compartir con orgullo todo lo adquirido por el bien de la humanidad, mis respetos señor, usted es fenomenal
That block is too beautiful to enclose in sheet metal. If your plans can accommodate the idea of it being open, you should (in my opinion) consider a way to do so.
I may actually make a exoskeleton and leave it open, it will look even crazier because the final design has 6 small tubes. Thank you!
@@WarpedYT I look forward to whatever you think up! I just found your channel, so I am definitely subscribed now.
I work with alot of cooling components for very powerful computers and hands down the best way to do a liquid cooled block is to have a completely hollow cavity and fill it with fin stock the will allow coolant through and then solder or braze a cover over it. the fin stock maximizes the area that the coolant is touching and it can wick away heat insanely good
I agree with you, that was the way I wanted to go, however I needed something very strong in robust to use on set for film and my TH-cam videos. I can almost guarantee you that if I just soldered a cover over some thin stock, during the first shoot I would have coolant all over the place and a burned out cob. When filming even TH-cam videos and you're in a hurry or you're caught in a situation stuff gets knocked down hit abused, this light is not going to just sit on the desk I can assure you. That being said I completely agree and in the future when I get my next CNC machine in a few months, I may do some really high RPM intricate milling on another piece just like this to make an improved version. Unfortunately the automatic manual machine I'm using now is based on a dinosaur machine.
here's an idea for cooling. liquid cooling. but not just any liquid cooling, but one that both makes use of a very thermally efficient fluid, and basically high pressure pulse width modulated high capacity liquid to air intercooler. the idea is to have the fluid have a periodically interrupted flow to help the intercooler which itself is liquid cooled by a highly chilled liquid, to help with more efficient transfer of heat energy. think 2 cooling circuits running parallel to eachother through the intercooler block in such a way that it can constantly be kept cool from one resivoir of a cooled liquid while absorbing heat and more quickly cooling the liquid to the main circuit. no idea if it would work, but it seems like it could be a more efficient way to do it than a single intercooler, and not need a second one to compensate.
One design flaw is that you should've machined thinner fins closer to each other and then cross-cutting them, making many "pins" instead of channels. This way you would've created way more heat-dissipating surface in contact with water, increasing the efficiency exponentially; this technique is used in PC watercooling and it works great
"I'm not Tony Stark!" Great Job!!! Love your channel
Time to put that LED into a nice reflector bowl with a projector lens and a sharp cut off. 👀
That's exactly what I did, I'll be using it for some upcoming videos/projects
Rock That Silly String! ftw
It might have been mentioned already (don't want to read 1K comments to find out!) - but a suggestion for your port positions in the block cover (as an automotive guy!) - Put the inlet near the top, and the outlet near the bottom vs side to side. Makes for better flow through your core.
I've got no words. Just great! very impressive!I i enjoyed seeing it very much
Awsome video 💞 love from India 🇮🇳🇮🇳
Thank you
Love your project videos.. Keeps the mind curious.. Thanks.
Finally a new video
I know I have to stop picking these really difficult subjects, all of these are very challenging recently.
Brilliant work! It looks like you might get short-cycling of the coolant by having the inlet and outlet in the center, but if the chip is small enough and centered on the water block, that's probably not a bad thing, as the greatest flow rate of the coolant will be at the center, where the most heat is being generated. Are you using a thermal paste between the chip and the block to maximize the heat transfer from chip to copper?
Maybe you should try a led->peltier->waterblock sandwich. You will cool the hot side of peltier with the waterblock while the cold side cools the led with sub ambient temps.
Wow this channel never ceases to amaze me. They always make me want to get off ass and do something. Thank you again for bringing this content to the people
Appreciate that you mention the metric units also✌✌✌✌
Insane, why not?! Lol love the attention to detail
That's a friggin great idea! Great light!
Dude you would be the perfect neighbor to live next to
🇺🇸🇺🇸🇺🇸🇺🇸
Perfect... Hello Brazil...👏😁
After watching this. I feel casting a solid copper block and machine that down would have been faster and better option.
What do I think, I think that you have more talent in your little finger than most people will ever have in their whole lives. You do amazing stuff. Love it!
Thank you, I appreciate that but the one thing I've had the most trouble with is just making more videos. These latest projects have been really tough
I definitely would love to see more metal casting on this channel.
I am going to do some more