He should use mineral oil in the loop instead of water, that way he could put a fan in the liquid and have it pushed through the fins much more efficiently.
i dont see aditional fans inside the colling liquid to make any difference, considering that fans are normally there to insure constant airflow... Liquid colling on a loop creates a continuus flow, you can clearly see that when he fills it as all the bits inside are floatin and swirlin around. so fans inside the installation would be completely pointless
i guess its not gonna be possible but try not to mix aluminium and copper ah and maybe if its possible try to match the waterflow with direction of the coolerfins? cooler fins?. any fins here?
@@TheUnjust117 hrm I'm not sure someone would have to model it in a thermodynamic simulation, perhaps the water from top to bottom causes it to pull enough water though the fins, otherwise, a fan forcing the cooler fluid through the fins on the heat sink would allow them to transfer heat faster. would be interesting to see a model of it though. but because the entry and output are on the "top" end I could see the fluid not passing through the fins so well, and so the fluids passive transfer of heat is being used more than a forced transfer of metal to the fluid.
I know this is a bit late, but if anyone plans to do this, or you are planning to do this again. Corrosion is caused by two issues here: 1, Water causing galvanic corrosion due to mixed metals. You need to use a liquid with a corrosion inhibitor, something like automotive coolant. The corrosion will mostly appear at the interfaces between the two metals, this means less heat transfer between the heat pipes and the fin stacks. 2, Silicone sealant. Most silicones are acetoxy cure, this means they release acetic acid as they cure. It smells like vinegar, and will cause corrosion on anything that it is contact with, or anything inside a sealed box with it. There are neutral cure silicones which should always be used anywhere where sensitive metals are involved like copper or aluminium. I've seen electronics sealed with silicone against water ingress, but because acetoxy cure silicone was used, they have corroded faster than if they weren't sealed at all.
Yes exactly what i was thinking. Make an evaporative cooler. Using radiator is just dumb. Why not use the proper watercooling kit? If you just make an artificial waterfall loop using a pump and waterproofed fan to cool wet fins, it would make more sense. Would be interesting to see how it performs against real water cooler
@@TierNone_LarperatoR I have a vague memory of someone cooling the entire computer by putting it in mineral oil, years and years ago. Tom's Hardware ca a decade ago, or something?
not 30 fps it add 300 fps your such a noob that has never tried to deal with liquid cooling and corrosion when done properly you get 300 fps minimum with corrosion when done right :p
Agreed! This would be the true test if this project is either to build something for the sake of building something unusual or it has some practical use. A lot of water cooling projects I see around are sadly the first thing, I mean its cool to create something weird and have fun by making it real but I would really like to see attempt of making something useless and available (either most of us have stock coolers or they are easy and cheap to get) into something functional either to save cash or as an emergency when you really need your pc to run and your part didn't arrive or got damaged.
The problem with an extruded cooler (both Intel and AMD) is that the fin structure is non-conducive to water flowing between the fins, which is where your better heat transfer comes from. If you never get flow through the fins, then the extra surface area doesn't mean crap and you'll wind up with mediocre performance. So you have to use a heat pipe cooler of some sort to get fins properly oriented in a way to be able to channel water between them to make full use of the surface area.
Now if you could build a cold plate to mount some custom heat pipes on the CPU, then pipe it up to the cold plate of the cooler, and orient the cooler on it's side to make the pipes only have 1 bend and to make tubing water into it more space efficient. The fabrication for that is fairly simple if you have a drill press and some background in jank machining, or a three axis vertical mill, but heat pipes are pricey, and aluminum or copper bar stock is expensive, so you are probably going to end up cheaper, cleaner, and more effective using a low profile heat piped cooler with horizontal fin heat pipes like an NH-L12S or an NT06.
You should use an automotive coolant , it has additives to protect aluminum engine blocks and heads from scaling and to carry heat efficiently. It will also protect cooper. Should work well with your project. It comes in green, red and white. Also it's good for 50,000 miles so you should never have to replace it.
And Purple coloured coolant too! But I agree with James. Running a Ethylene Glycol solution may yield better results and better longevity from corrosion.
Since you have different connected metals immersed in water, you'll need corrosion inhibitors to prevent the heatsink from getting destroyed by galvanic corrosion in a long-term run.
@@BiologyIsHot The problem with mineral oil for cooling applications is that it has less than 1/3 the specific heat as water, which means you need ~3X as much flow for the same performance.
@Teardown Dan its not gonna necessarily need 3x the flow, but 3x the liquid capacity to have the same thermal mass - thermal conductivity of mineral oil is pretty poor compared to water at about a 1/4, probably won't affect the CPU cooler side as much as there's ample surface area but the effectiveness of the rad could be an issue.
@@arlobubble3748 Increasing the total "thermal mass" by increasing volume only changes how long it takes for the loop to come up to temperature. It does nothing for actually moving heat from A to B. If you have a fluid with worse specific heat, you need more flow to achieve the same heat transfer rate because the fluid passing through the hot plate is warming up 3-4X as fast so you need 3-4X the flow to keep hot plate temperature the same, same goes for a radiator of a given size.
I would suggest to use a nickel-plated server heatsink like the Supermicro SNK-P0048PS. By using this you might be able to direct the waterflow well-balanced through all of the fins.
Version 2 definitely needs a closer fit to the cooler fins and a block on top so water can't flow over the top. It also needs a distribution tube on the inlet side so water gets evenly spread across the fins. Last thing, use a corrosion inhibitor. Get some 50/50 premix automotive antifreeze. Test that, then add some Redline WaterWetter and test again. Also test plain water with and without WaterWetter.
To be fair the air cooler has been sitting in a enclosed box with stagnate water keeping it humid and a tiny bit of air flow allowing for a heavy amount of corrosion unlike if it was just in water (Though the design needs a lot of work to get better flow)
It's worth revisiting. One change I'd make, is having the bigger sides against the fins so that the water is forced through the cooler, as apposed to having probably most of it running past it. You know, path of least resistance stuff.
You should try it wich a nickelplated or painted cooler, so it doesn't corrode. And maybe you could try to modify a fan to make it waterresistant and put in on the cooler ( just to look cool). I would also try to minimize the gaps on the side to foce the water though to fins.
@@heygek2769 well no but maybe you can force them to run by forcing power through them from a power supply? would be a lot easier than creating a water proof fan.
You could try a waterproof fan or mini fan, and maybe put the inlet on the bottom side of the cooler? Or split the tubes so you have more inlets at the bottom to separate the flow through the fins?
I think if he make the Water com out from the bottom part of the finstack it would force the water to travel from top to bottom and it would necessarily go through the fins. Also if he makes both inlet and outlet come from teh side of the enclosure then it probabyl would fit in the case with side panel on. BUT, it's not possible for a home-made fan to be waterproof. He would need to separate the motor and, well let's not get too much into this, you would really need a lot of engineering and tools to make an electric motor waterproof for submersion.
I think you need internal baffles, something to force the water to flow through the fins This was the project that made me sub \o/ I can't believe it was that long ago
Never expected to see a follow-up. What a treat this is. :) Imagine if you'd dried it out before storing. Also I'm really surprised the epoxy held since it's not only the seal, but also mechanically holding the entire block to the heatsink. Amazing.
It would be really cool to revisit the watercooled air cooler. A couple of suggestions: It could be fun to see how a big aircooler performs. I think something along the lines of Deepcool Assassin 3, or a Noctua NH D15, although fabrication probably will be difficult. Second suggestion would be to move the inlet and outlet, in a diagonel pattern. Inlet in the bottom and outlet at the top. That way the waterflow would move across the finstack, perhaps improving cooling.
@@xxxSevenforxxx Well if the coolants are also not electrically conductive, then yes they should work as well, but using mineral oil to cool PC components is known to be safe and also not that expensive
Two ideas for future projects. #1: reverse flow to see if you get better or worse cooling. (used to work at a hydronic heating company and we used inverted flow in our water heaters and gained %20 more heat transfer.) #2: use a non conducive liquid for coolant and leave the fan mounted to the cooler for more even water flow through the cooler.
if you care about corrotion long term then use what cars use, that is destilled water with glycol and aditives. oil is not perfect and it also has lower thermal resistance than water so water will always win, so your best bet would be adding stuff to the water to stop the corrotion. If you are coming from the "oil cooler pc" kind of idea, thats only because its nearly imposible to keep destilled water destilled when you throw stuff in it and also when it stops being destilled, your whole pc is toast. Moneral oil may not be as good of a coolant but its waaaaaay safer for cooling live electronics. In a water loop, the coolant is isolated from the electronics so it doesnt really matter that its non conductive, so you are better off with a better coolant like water. Mercury would be a better coolant than water, but its super heavy, toxic and it eats copper and aluminum (it alloys on contact) so it wouldnt last long. Its all about which is the best and more convenient coolant for each situation.
I'd love to see a revisit of this. I have no suggestions for the cooler to use but the box can be made smaller. The top of the box could be right against the tops of the heat pipes. The sides of the box could be very close to the sides of the cooler. Inlet on one side and outlet on the other. You would then literally be forcing water through the fins. Maybe even have more than one inlet and outlet. Possibly 2 or 3 of each, spread over the height of the cooler. Water line distribution blocks are easy enough to find and aren't very expensive.
Water in the loop will all be the same temp. This is why it makes no difference what order parts are in the loop. Air in the line can cause air locking in the pump and inefficient heat transfer with in the rad
Aint that how its always supposed to be? Fill from the bottom and take from the top? At least thats how I know all of my coolers. Like, liquid heat exchanger thingies, not water coolers for yo PC, i mean the ones for cooling down say the vapor from a destillation
@@tatzecom water coolers rely on the physical contact of water. Air that gets trapped in the pump can also cause an imbalance and ruin the pump over time or worse cause air lock. When you take from the top the nossle needs to be fully submerged or else water gets into the loop this is why it is more pratical to take water from bottom
Yes a v2 would be great. After reading some of the other comments. Many have said what i was thinking too. But, here are some extra ideas. Mentioned already and agree with: 1. A coated/painted cooler 2. using the fans from the air cooler inside the water to act as a pump. some other ideas: 1. using a downdraft style cooler to fit in the case better 2. maybe use mineral oil instead of water 3. i like the idea of a smaller air cooler used with something that puts out more heat. ( we know this works. I don't find the d-15 with water to be that interesting. )
I don't understand how you only have 34.9k subscribers, the quality and frequency of your content is deserving of millions of subs. It boggles my mind how so many undeserving(in my opinion) channels can have millions upon millions of subs, yet a channel like yours has so few. I sincerely hope that someday soon your channel will start gaining subscribers in a big, nay YUGE way; I'm gunning for you brother, keep on truckin'.
I have a thought. Mineral oil?? It’s nonconducting and you could add a small fan for creating flow. The radiator should create enough cooling to keep the mineral oil cool theoretically. The problem would be the pump. It would have to be a very strong pump as the oil is much thicker and would need a stronger pump to circulate effectively.
@MajorHardware You should run a high ratio of whist vinegar and distilled water through the system with a filter in line somewhere. You could remove much or even nearly all of that scaling. Then, you should run distilled water through it a couple of times with the filter just to make sure everything is out as much as possible before finalizing on just distilled water. Then it should be back to nearly new functionality without building up new scaling. Also, and I'm sure someone else must have mentioned this in the comments, but your room temp was 2C higher than the original test and your CPU temp was 2C higher too. To me, that seems like it's working as efficiently as the first run through.
Whatever cooler or liquid you use next, you want to make sure the casing around the heatsink fits as tightly as possible. I'm guesstimating there's now a 5mm clearance around the heatsink, where water will pass though but little to no heat is transferred. This is essentially a pressure bypass. If you reduce this gap or ideally eliminate it, all the pressure from the pump can be used to push the liquid through the heatsink fins. It might be a good idea to turn to 3D printing instead of acrylic, to get that shape right. As for a heatsink, something like a Noctua NH-D9L could be an interesting route. It's a two-tower style cooler that you could turn into two chambers; one where you push the liquid up, overflowing into the next chamber with the outlet on the bottom. Much like the radiators are built.
Suggestion maybe - Rinse entire system with distilled water, perhaps a few times. Then fill with distilled water. The corrosion may be due to minerals/salts in the tap water. May have to create new cooler block due to old block corrosion leaching bad chemicals back into system. Just a thought.
Use PC cooling liquid. It has anti corrosive properties and more! Make a 2.0, with a better flow design over the fins, and use the big ole noctuah cooler or the be quiet dark pro! Would love to print this out and do it at home for an "extreme" water block! Keep up the good work, thank you for your time!
Use a Intel box cooler ;) They tiny, barely works stock. Will water make it usable? Also it's black so might take water better. Also they free and common in many shelves.
100% want a reboot of this. Adding a corrosive inhibitor instead of just DI water would help with your corrosion. Honestly not sure where that corrosion came from, cause if it was sitting on a shelf for most of that time it wasn't galvanic corrosion, which would be the big concern from using an aluminum finned heatsink with copper radiator tubes and pump internals. You could use a massive NH-D15 and put the water inlets on the sides instead of the top to get it to fit in a typical case. You could also use mineral oil in it, which would let you put a fan between your fin stacks on the D15 to minimize the pressure drop and get more even spread of coolant across the fins, but mineral oil also doesn't carry heat as well as water, so I'm not sure if that would help at all. Better option would be a low profile cooler like a NH-L12S or a NH-C14S or a NT06. It'll fit better in the case and give you a lower center of gravity so you'll have less strain on the motherboard when in a typical tower case. With the NH-12 or NT06 you can put the inlets on the side top or bottom, for the NH-C14S you'll want to put the inlet/outlet on the top or bottom because the sides are all kind of closed up and would be a bit restrictive. Put your top/bottom/sides all flush to the fins so you don't have much or any coolant that doesn't flow through the fins to maximize your coolant over the heat transfer surfaces. Leave about an inch of water box on each side that you can tap your inlet/outlet fittings into. That'll also give plenty of space for your flow to evenly distribute across all of your fins, since the fins will act as their own distribution plate. You can help that out by putting your inlet and outlet diagonally across from each other instead of straight across from each other, but I don't think that will make that much difference. And for testing accuracy, I'd like to see the deltaT from coolant temperature to CPU temperature, since that's what your waterblock is driving. You can lower your water temperature by changing pump/radiator/water mass/fans, but your waterblock will keep a relatively constant delta from CPU to water for a given power output. Also, testing back pressure across the cooler would yield some useful information, since higher flow rate does help somewhat, and all PC watercooling pumps are centrifugal, so lower back pressure makes more flow makes (to a limited extent) more cooling. Parallel pumps might be useful for this as well, since you want to maintain turbulent flow between your fins, and I'm not sure you'd get high enough velocity to cause that with only one pump. You'd need a flow meter and a D/P detector to measure that stuff. For anyone asking if building this sort of block is "worth it", I would argue that water cooling is literally never "worth it". Your return on investment for open loop water cooling is basically nothing. A decent AIO will get 95% of the thermal performance, which equates to 99% of the compute performance, at 1/8 the cost. The determining factor, to me, on whether this is "worth it" or not is it's performance against a similarly priced (sum of bill of materials for each setup) traditional water block. Even if it's equivalent, it's still "worth it" as a neat, unique mod that's just fun to play with. At no point does "practical" come into play when you are talking about open loop cooling. For practical, go with an AIO or the biggest air cooler your case can fit.
Build a D5 pump into the water box. Hear me out: Use a tower cooler like you used on this one. Do all the things I said in my other comment about tightening the box around the fins to get better flow through your fins. With the cooler oriented as it is at 1:28, put it on the bottom face, closest to the motherboard so it takes suction directly from the water box. Have it discharge out directly away from the motherboard. You could either print a pump top or get a pump/top that would be easy to mount. An XTOP would work, I think. Just tap four holes into the top of it, drill a couple holes on the mounting face, throw an o-ring between the pump top and the mounting face, maybe print the mounting face with a slight indent to receive said o-ring. It's settled. I'm buying a 3D printer after this deployment.
Oh an please make the opposite of this one too!!! Try to cool the water of an open Loop with aircoolers only, no radiators 😂 Could imagine that you use 2 CPU blocks for it. One on the actual CPU and the other one just in the line and then you can mount the aircooler on the coldplate of it 😂
I like what you did, but I will drop some science on you. You can pick whatever cooler you want to use, just make sure that the fins that run along the widest sides are even with each other. When you re-make the 2.0 version, make sure that the long outside plastic panels are touching the fins and that the bottom panel (closest to the cpu) is right to top of the bottom fin, no gap, same for the top panel as well. The water input and output sides make sure to leave at least 3/4” space, 1” would be better, this will help in the directional flow. On the water input fitting, use a 5/16” or 3/8” diameter tube that is caped or crimped at the end and the tube runs the full length into the box, cut a straight slice in the whole length of this tube, only on one side, use a dremel cutoff fiber mesh type disc (they are a whole lot safer) this will leave you about a 7/64" or so, wide slot. Aim the slot straight at the middle of the cooler fins, the output fitting that you used before is okay. If you want to make sure to get all of the air out of the system, make to output side panel taper to the fitting. By building the box this way, it will create a directional water flow that will majorly increase the cooling efficiency.
I'd be interested in seeing a second edition; i would recommend applying a clearcoat of paint or epoxy onto whatever radiator you use (prevents corrosion for longevity and won't affect efficacy very much) and attach the inlets and outlet the middle of the sides (the position they use now, but at the center of the skinny sides instead of both sharing the plane beyond the heat exchange plates).
I think the side that has the water cooled portion would be doing a majority of the work. The non water cooled side wouldn’t be completely useless though its effectiveness would depend on if a fan was used, placement and speed. It would be fun to play with though.
@@benjaminnyman8687 this is true but I'm sure there are a few options if it is a pull fan at the back exhaust out the back might work it would increase airflow
I seem to recall you didn’t use a radiator last time, which made the test kind of pointless. Glad to see you doing it right this time 👍🏼 I never saw the sequels, I’m gonna have to check those out, I never actually subbed until the tubing guides, that was awesome
Just a suggestion for the next cooler to add MoCool to the water. It is used by race cars to transfer more heat into the water, and assists with corrosion protection. I use it in both my Smart cars which have 700 cc intercooled turbocharged engines and struggle in summer to keep the temperature down. It could provide a better thermal conversion as well as giving your system longevity.
this is a year OLD WoW: it dosn't feel that long. ONLY a thought why didn't you just put one pie to you tape then you would be able to check it off the PC?????
I took a different twist on this about 10 years ago. I took a 1 rad water cooler and dipped the rad out the side of my case and into a large pot of water. Sometimes i put ice in too but usually left it without ice. It was the largest cooking pot i had in the house. As long as i topped it up every day (from evaporation) it stood steady no matter what voltage or multiplier i set it at. Truely the best solution if you don't mind having an open pot of water lying next to your open computer
Definitely revisit and revise this then stress test it. The flow through the fins could definitely be optimized with at least a tighter case around the cooling fins
I'd love to see a redesign with t he hoses connecting so that it fits in a case, using more acrylic to create channels to distribute the water through the heatsink evenly with a flow like you would have with air, and definitely with some anti-corrosion additives to the water. LOL Other improvements could be a temp sensor for a mobo header and, of course, RGB lighting.
Now that you have a 3D printer I would love to see a new experiment involving this it would be pretty awesome to come up with some sort of solution that actually works that you can just put onto a air cooler please continue this project
Perhaps a version people would want to replicate in real life? For example, how about using stock intel cooler (which almost everyone has) and some radiator coolant that doesn't cause corrosion?? Hobbyist 12V water pumps are for like 3$. These can be modified to be run by chassis fan outlets, which can change rpm by changing voltage. I can buy a second-hand ford fiesta radiator for like 4$ here. Add a few $s for hoses and if this 12$ design can perform as good as high-end custom water coolers, this would be really useful as an experiment. Odds are, a "waterblock" as large as an intel cooler and a radiator probably 10 times as large as a PC radiator would make a very silent and very cool system, maybe even without using any fans.
yessir v2.0! tighten up the sides and build a block off between inlet/outlet for force flow through the cooler! Heck with the same cooler even. I'm just as interested in seeing how much is on the table just in flow optimizing.
That be nice to see!Would be nice if there is extraction valve to remove the excessive liquid from it after use and a way to clence the cooler system after usage to prevent rusting ofc not attached to the rest the system while doing so, so u don't mess up the rest the system its awesome project build!Maybe re-attachable sides will do all that..
Yes, please revisit this. This time I think you should try to convert a factory air-cooler to water-cooling. Something round and low-ish profile. A factory Intel or AMD cooler is what I'm thinking of and this time add a little more focus on how the coolant moves in the chamber.
I hope you revisit the water-through-the-heatpipes idea in an aircooler with parallel water flow, plus a fan attached to the radiator. This would make it effectively an air cooled water block!
You should absolutely make a version 2 - however, you'll want to use a single metal type heatsink if you can find one - mixing metals is going to see accelerated corrosion unlike if you're using a single metal. You might be able to find one of the old all-copper thermalright heatsinks on ebay or the like, perhaps, would be the best bet, or you might get one of the oddball tower Zalmans that are all copper. You might also test to see whether alcohol might work better than water, given you have so much volumed mass to work with from the larger container, you don't need to worry about vaporization of the entirety of the fluid.
This is pretty interesting, not sure if it's actually better than just buying a waterblock. To prevent corrosion you should use a sacrificial anode - basically a blob of zinc you drop in the loop that takes all the corrosion so your aluminum doesn't. This is what salt water boaters do, you will find them in your home water heater as well. You could also use mineral oil, the trouble is that it does not conduct heat as well as water, and it's thicker than water so it's harder to pump. But it would be corrosion free and solve the big problem with oil immersion cooling which is deterioration of the rubber and plastic parts, since only the metal pieces touch oil. One problem with going to a bigger cooler is weight! The cooler itself is already pretty heavy and then you will add a couple of pounds of liquid to that. On a flat bench it wouldn't matter, but on a tower style setup it could be hard on the mounting apparatus.
Nice Project, but you should put a hot water tube at the middle of the box to prevent Air bubbles ... I think a project where we put cooper pipes passing through the fans, maybe Just 4 aside with some kind of joints on both sides, could be more light, secure, clean, cold, easy and cheaper than a entire box arround the cooler ...
In the next one, one side of the enclosure should be against the sides forcing water to have to flow through the fins instead of allowing it to flow around them. That would help with the volume of water and size of the cooler. Also I’d have the inlet at the top and outlet at the bottom so gravity helps out. Try to make it so that as much water flows across the fins as possible and make sure there’s no air between the fins before testing (saw the old video a bit ago and there was air in there)
I would like to see you do a build using something like a high proof alcohol and glycerin mixture instead of water. Something with a higher thermal load while being able to be chilled further without freezing or becoming more viscous. I imagine this will take some tweaking, but there should be some good example recipes on the net for "normal" liquid cooling setups. I bring this up because it might help solve your corrosion issues, while letting you dip that radiator into some ice water instead for some craziness.
Next time you retire it especially if you do a 2.0 run dry air through it until dries out to help preserve it. You can start by just setting it in the window and letting the sun heat up the inside and push out the moisture and then once you get most the moisture out put a piece of tubing across both ports to seal it.
If you do revisit this with another cooler i highly recommend finding that is all copper or all aluminum and make sure your radiator is the same material other wise your going to have that corrosion issue. If you have to use mixed metals please add something to the water in the loop to prevent the corrosion otherwise you might kill that lovely 360mm rad. :O
An easy way to get a little more efficiency out of it would be put input up top and out flow at bottom to make it a counterflow chiller, also shrinking the box will up efficiency as well forcing more flow over the fins.
If you do a version 2.0, make sure to electroplate the cooler with something corrosion resistant. As the copper aluminum are dissimilar metals, flowing water across it may cause electrolysis. May help prevent corrosion in future builds. I'd also try and place the fill and drain on opposite "Walls"/Corners to help draw flow across the entire surface area of the cooler. That's my ideas.
It'd be interesting to see a second version, but in it you should use a cooler with some corrosion-protective coating and minimize the free space at the sides of the heatsink, as flow around the heatsink doesn't help cooling. If I were to recommend a cooler, it would be the Noctua NH-U9S.
One thing I would suggest is using mineral oil to prevent the corrosion issue and also avoids the issue of leaks bricking the board. The only issue I see with what I suggest is that I dont know what pump could handle the viscosity of the oil.
![[MV] A leap of faith...ความเชื่อใจครั้งสุดท้าย (From "Petrichor the series")](http://i.ytimg.com/vi/U1piZH2CNXA/mqdefault.jpg)
We'd love a version 2.0 yes!
He should use mineral oil in the loop instead of water, that way he could put a fan in the liquid and have it pushed through the fins much more efficiently.
@@EVPointMaster such a good idea
i dont see aditional fans inside the colling liquid to make any difference,
considering that fans are normally there to insure constant airflow...
Liquid colling on a loop creates a continuus flow,
you can clearly see that when he fills it as all the bits inside are floatin and swirlin around.
so fans inside the installation would be completely pointless
i guess its not gonna be possible but try not to mix aluminium and copper
ah and maybe if its possible try to match the waterflow with direction of the coolerfins? cooler fins?. any fins here?
@@TheUnjust117 hrm I'm not sure someone would have to model it in a thermodynamic simulation, perhaps the water from top to bottom causes it to pull enough water though the fins, otherwise, a fan forcing the cooler fluid through the fins on the heat sink would allow them to transfer heat faster. would be interesting to see a model of it though. but because the entry and output are on the "top" end I could see the fluid not passing through the fins so well, and so the fluids passive transfer of heat is being used more than a forced transfer of metal to the fluid.
I know this is a bit late, but if anyone plans to do this, or you are planning to do this again.
Corrosion is caused by two issues here:
1, Water causing galvanic corrosion due to mixed metals. You need to use a liquid with a corrosion inhibitor, something like automotive coolant. The corrosion will mostly appear at the interfaces between the two metals, this means less heat transfer between the heat pipes and the fin stacks.
2, Silicone sealant. Most silicones are acetoxy cure, this means they release acetic acid as they cure. It smells like vinegar, and will cause corrosion on anything that it is contact with, or anything inside a sealed box with it. There are neutral cure silicones which should always be used anywhere where sensitive metals are involved like copper or aluminium. I've seen electronics sealed with silicone against water ingress, but because acetoxy cure silicone was used, they have corroded faster than if they weren't sealed at all.
Responding to this in the hope it gets seen because this should be taken I to account.
th-cam.com/video/u6zsHqdNRQ8/w-d-xo.html
I’d like to see a more efficient design being tested, perhaps with the water flow going through the fins better... go for it!
Seconded.
Use one of the lower profile colleges like a noctua l12s or the c14s
Another vote for an improved efficiency design.
Yes exactly what i was thinking. Make an evaporative cooler. Using radiator is just dumb. Why not use the proper watercooling kit? If you just make an artificial waterfall loop using a pump and waterproofed fan to cool wet fins, it would make more sense. Would be interesting to see how it performs against real water cooler
also USE a painted heatsink (black preferrably)
Have a normal liquid cooling setup.
*Then liquid cool the radiator.*
Didn't LTT do this with one of their projects?
lmao
@@TierNone_LarperatoR I have a vague memory of someone cooling the entire computer by putting it in mineral oil, years and years ago. Tom's Hardware ca a decade ago, or something?
Don't forget to add the water chiller
Juan Pretorius yooooo that’s big brain
Everyone knows corrosion adds at least 30fps to your liquid cooler
I was wondering where the extra performance i got came from.
not 30 fps it add 300 fps your such a noob that has never tried to deal with liquid cooling and corrosion when done properly you get 300 fps minimum with corrosion when done right :p
@@raven4k998 can confirm. It also increased RAM for me
@@raven4k998 Yes, but this method allows an extra 7hp and 40 Exp
Yes increase coolers surface area.
do it with a stock cooler .. thats a way to make use of the useless intel coolers
Sealing around a stock cooler would be a lot more work.
@@drackar yea but still do-able
Agreed! This would be the true test if this project is either to build something for the sake of building something unusual or it has some practical use. A lot of water cooling projects I see around are sadly the first thing, I mean its cool to create something weird and have fun by making it real but I would really like to see attempt of making something useless and available (either most of us have stock coolers or they are easy and cheap to get) into something functional either to save cash or as an emergency when you really need your pc to run and your part didn't arrive or got damaged.
The problem with an extruded cooler (both Intel and AMD) is that the fin structure is non-conducive to water flowing between the fins, which is where your better heat transfer comes from. If you never get flow through the fins, then the extra surface area doesn't mean crap and you'll wind up with mediocre performance. So you have to use a heat pipe cooler of some sort to get fins properly oriented in a way to be able to channel water between them to make full use of the surface area.
Now if you could build a cold plate to mount some custom heat pipes on the CPU, then pipe it up to the cold plate of the cooler, and orient the cooler on it's side to make the pipes only have 1 bend and to make tubing water into it more space efficient. The fabrication for that is fairly simple if you have a drill press and some background in jank machining, or a three axis vertical mill, but heat pipes are pricey, and aluminum or copper bar stock is expensive, so you are probably going to end up cheaper, cleaner, and more effective using a low profile heat piped cooler with horizontal fin heat pipes like an NH-L12S or an NT06.
If it's so good with an old i5 I would love to see it on any modern 8 core cpu!
3900X that bish
It's weird seeing a maker in the comments. Like seeing your priest at the bar. Or a bear walking on its hind legs.
I have a 3950x but I'm to big of a sis to try it on it
@@MajorHardware do it, do it, do it, do it, do it.
@@MajorHardware do it, do it, do it, do it, do it, do it
You should use an automotive coolant , it has additives to protect aluminum engine blocks and heads from scaling and to carry heat efficiently. It will also protect cooper. Should work well with your project. It comes in green, red and white. Also it's good for 50,000 miles so you should never have to replace it.
I don't think the pc is moving 50k miles so i think you'll be fine
And Purple coloured coolant too! But I agree with James. Running a Ethylene Glycol solution may yield better results and better longevity from corrosion.
Since you have different connected metals immersed in water, you'll need corrosion inhibitors to prevent the heatsink from getting destroyed by galvanic corrosion in a long-term run.
Or mineral oil
@@BiologyIsHot The problem with mineral oil for cooling applications is that it has less than 1/3 the specific heat as water, which means you need ~3X as much flow for the same performance.
@Teardown Dan its not gonna necessarily need 3x the flow, but 3x the liquid capacity to have the same thermal mass - thermal conductivity of mineral oil is pretty poor compared to water at about a 1/4, probably won't affect the CPU cooler side as much as there's ample surface area but the effectiveness of the rad could be an issue.
@@arlobubble3748 Increasing the total "thermal mass" by increasing volume only changes how long it takes for the loop to come up to temperature. It does nothing for actually moving heat from A to B. If you have a fluid with worse specific heat, you need more flow to achieve the same heat transfer rate because the fluid passing through the hot plate is warming up 3-4X as fast so you need 3-4X the flow to keep hot plate temperature the same, same goes for a radiator of a given size.
@@teardowndan5364 physics! 🥳👍
I would suggest to use a nickel-plated server heatsink like the Supermicro SNK-P0048PS. By using this you might be able to direct the waterflow well-balanced through all of the fins.
idk perhaps try a painted cooler or one that doesn't have mixed exposed metals .To prevent erosion in the fluid
It just corrodes because he let the water out..
Put some glycol in there?
@@MarbleTL or
Mineral oil
probably unavoidable, paint or not the heat pipes are gonna be in contact with aluminum fins in water, these aren't designed for water
use one of the noctua chromax coolers. the heat pipes are also painted, it should hold up well. or one of the black bequiet coolers.
I saw this on my feed and clicked immediately! Definitely do an updated version.
Version 2 definitely needs a closer fit to the cooler fins and a block on top so water can't flow over the top. It also needs a distribution tube on the inlet side so water gets evenly spread across the fins. Last thing, use a corrosion inhibitor. Get some 50/50 premix automotive antifreeze. Test that, then add some Redline WaterWetter and test again. Also test plain water with and without WaterWetter.
I checked the date on this to see how long "one year ago" was. This was a wonderful surprise to see this show up on my feed!
who would have thought that putting aluminium and copper in same loop would cause corrosion :P
To be fair the air cooler has been sitting in a enclosed box with stagnate water keeping it humid and a tiny bit of air flow allowing for a heavy amount of corrosion unlike if it was just in water (Though the design needs a lot of work to get better flow)
@@SilvaDreams a heatsink isn't an air cooler, it's a passive cooler, becoming an air cooler when you throw fans on it
@@patrickpenguin8587 it's not even an air cooler, it literally just makes air hotter.
@@jocerv43 was I supposed to laugh?
Yeah with the mixed metals he should run car antifreeze mix as it has anti corrosion additives, and it would stop the scum growing.
It's worth revisiting. One change I'd make, is having the bigger sides against the fins so that the water is forced through the cooler, as apposed to having probably most of it running past it. You know, path of least resistance stuff.
th-cam.com/video/u6zsHqdNRQ8/w-d-xo.html
You should try it wich a nickelplated or painted cooler, so it doesn't corrode. And maybe you could try to modify a fan to make it waterresistant and put in on the cooler ( just to look cool). I would also try to minimize the gaps on the side to foce the water though to fins.
pretty sure the industrial noctua fans ar water proof.
@@SpecialEllio they are not, maybe they are a bit more water resistant, but not meant to be submerged
@@heygek2769 well no but maybe you can force them to run by forcing power through them from a power supply? would be a lot easier than creating a water proof fan.
You could try a waterproof fan or mini fan, and maybe put the inlet on the bottom side of the cooler? Or split the tubes so you have more inlets at the bottom to separate the flow through the fins?
I think if he make the Water com out from the bottom part of the finstack it would force the water to travel from top to bottom and it would necessarily go through the fins. Also if he makes both inlet and outlet come from teh side of the enclosure then it probabyl would fit in the case with side panel on.
BUT, it's not possible for a home-made fan to be waterproof. He would need to separate the motor and, well let's not get too much into this, you would really need a lot of engineering and tools to make an electric motor waterproof for submersion.
Honestly, I commend you for being so calm with that thing in your system without at least putting paper towels on anything electronic nearby.
I think you need internal baffles, something to force the water to flow through the fins
This was the project that made me sub \o/
I can't believe it was that long ago
Yes this, or make the casing tight over the heatsink.
Never expected to see a follow-up. What a treat this is. :) Imagine if you'd dried it out before storing. Also I'm really surprised the epoxy held since it's not only the seal, but also mechanically holding the entire block to the heatsink. Amazing.
epoxy can hold a lot more than 1-2 kg of water...
Maybe try with one of Noctua's Low-profile coolers? It surely would fit in a case afterwards :]
It would be really cool to revisit the watercooled air cooler. A couple of suggestions:
It could be fun to see how a big aircooler performs. I think something along the lines of Deepcool Assassin 3, or a Noctua NH D15, although fabrication probably will be difficult.
Second suggestion would be to move the inlet and outlet, in a diagonel pattern. Inlet in the bottom and outlet at the top. That way the waterflow would move across the finstack, perhaps improving cooling.
Use mineral oil in the loop, so you can still put a fan directly on the cooler
Wouldn't that eat away any plastic in the loop, such as the the box around the air coolers and the tubing?
@@LeonisYT Look up mineral oil PC.Those are completely submerged in it and that doesn't seem to be a problem
Or just car/motorbike cooling fluid, thats cheap and with a anti-rust mixture.
@@xxxSevenforxxx Well if the coolants are also not electrically conductive, then yes they should work as well, but using mineral oil to cool PC components is known to be safe and also not that expensive
@@xxxSevenforxxx didn't Jay try that? And beer??
Two ideas for future projects.
#1: reverse flow to see if you get better or worse cooling.
(used to work at a hydronic heating company and we used inverted flow in our water heaters and gained %20 more heat transfer.)
#2: use a non conducive liquid for coolant and leave the fan mounted to the cooler for more even water flow through the cooler.
Would oil not work better for heat transference also it would reduce the corrosion.
Great video. Keep up the good work.
MrNlce30 Water has a greater specific heat capacity than oil so no.
Only real benefits of the oil would be in corrosion and no real risk if there is a leak, since it isn’t a solvent like water.
if you care about corrotion long term then use what cars use, that is destilled water with glycol and aditives.
oil is not perfect and it also has lower thermal resistance than water so water will always win, so your best bet would be adding stuff to the water to stop the corrotion.
If you are coming from the "oil cooler pc" kind of idea, thats only because its nearly imposible to keep destilled water destilled when you throw stuff in it and also when it stops being destilled, your whole pc is toast. Moneral oil may not be as good of a coolant but its waaaaaay safer for cooling live electronics.
In a water loop, the coolant is isolated from the electronics so it doesnt really matter that its non conductive, so you are better off with a better coolant like water.
Mercury would be a better coolant than water, but its super heavy, toxic and it eats copper and aluminum (it alloys on contact) so it wouldnt last long.
Its all about which is the best and more convenient coolant for each situation.
Oil has a tendency to ruin rubbers and some silicones aswell.
I mean the best coolant would be liquid ammonia, but it isn’t practical for usage.
I'd love to see a revisit of this. I have no suggestions for the cooler to use but the box can be made smaller. The top of the box could be right against the tops of the heat pipes. The sides of the box could be very close to the sides of the cooler. Inlet on one side and outlet on the other. You would then literally be forcing water through the fins. Maybe even have more than one inlet and outlet. Possibly 2 or 3 of each, spread over the height of the cooler. Water line distribution blocks are easy enough to find and aren't very expensive.
"Getting air in the loop" *Return line on the top* FACEPALM
True. Although I think that is the best way since hot water will naturally rise to the top.
Water in the loop will all be the same temp. This is why it makes no difference what order parts are in the loop. Air in the line can cause air locking in the pump and inefficient heat transfer with in the rad
Aint that how its always supposed to be? Fill from the bottom and take from the top?
At least thats how I know all of my coolers. Like, liquid heat exchanger thingies, not water coolers for yo PC, i mean the ones for cooling down say the vapor from a destillation
@@tatzecom water coolers rely on the physical contact of water. Air that gets trapped in the pump can also cause an imbalance and ruin the pump over time or worse cause air lock. When you take from the top the nossle needs to be fully submerged or else water gets into the loop this is why it is more pratical to take water from bottom
@@n3m37h That's what the second reservoir is for. The top one bleeds from the top, the bottom one takes from the bottom, no risk of air.
Yes a v2 would be great. After reading some of the other comments. Many have said what i was thinking too. But, here are some extra ideas.
Mentioned already and agree with:
1. A coated/painted cooler
2. using the fans from the air cooler inside the water to act as a pump.
some other ideas:
1. using a downdraft style cooler to fit in the case better
2. maybe use mineral oil instead of water
3. i like the idea of a smaller air cooler used with something that puts out more heat. ( we know this works. I don't find the d-15 with water to be that interesting. )
Amazing how fast time pass!
I'd try pouring some vinegar into the loop in order to try to internally clean it, and to remove some corrosion in there.
If you can get the air out to prevent further oxidation, having a bit of vinegar in the loop permanently might even be a good idea.
I don't understand how you only have 34.9k subscribers, the quality and frequency of your content is deserving of millions of subs. It boggles my mind how so many undeserving(in my opinion) channels can have millions upon millions of subs, yet a channel like yours has so few. I sincerely hope that someday soon your channel will start gaining subscribers in a big, nay YUGE way; I'm gunning for you brother, keep on truckin'.
Yes do a new one !! Definitely smaller and see what the results are.
The next one should be coated in rustproof coating and somehow antimicrobial? Maybe a uv bulb to hit the liquid with a light shroud?
I have a thought. Mineral oil?? It’s nonconducting and you could add a small fan for creating flow. The radiator should create enough cooling to keep the mineral oil cool theoretically. The problem would be the pump. It would have to be a very strong pump as the oil is much thicker and would need a stronger pump to circulate effectively.
This could work with motorcycle fork oil. I'ts almost as dense as water.
@MajorHardware You should run a high ratio of whist vinegar and distilled water through the system with a filter in line somewhere. You could remove much or even nearly all of that scaling. Then, you should run distilled water through it a couple of times with the filter just to make sure everything is out as much as possible before finalizing on just distilled water. Then it should be back to nearly new functionality without building up new scaling.
Also, and I'm sure someone else must have mentioned this in the comments, but your room temp was 2C higher than the original test and your CPU temp was 2C higher too. To me, that seems like it's working as efficiently as the first run through.
This is what made me subscribe a year ago :D
Me too :D
same
Same
Lol yeah me 2
i think me aswell
A 2.0 version would be awesome. As some people stated, you could have water flow through the fins which I think would be worth looking into.
2:34 if the cooler stay like that for a few week I wonder if the mb and case is strong enough
Whatever cooler or liquid you use next, you want to make sure the casing around the heatsink fits as tightly as possible. I'm guesstimating there's now a 5mm clearance around the heatsink, where water will pass though but little to no heat is transferred. This is essentially a pressure bypass. If you reduce this gap or ideally eliminate it, all the pressure from the pump can be used to push the liquid through the heatsink fins. It might be a good idea to turn to 3D printing instead of acrylic, to get that shape right.
As for a heatsink, something like a Noctua NH-D9L could be an interesting route. It's a two-tower style cooler that you could turn into two chambers; one where you push the liquid up, overflowing into the next chamber with the outlet on the bottom. Much like the radiators are built.
Do a noctua black edition or be quiet painted cooler so there's no exposed metal other than heatpipes
Normal Noctuas are already coated with nickel.
The coating isn't even as useful, it'd be better to stick with one of the old all-copper heatsinks like the thermalrights or Zalmans.
@@SCComega The point is to prevent corrosion.
Suggestion maybe - Rinse entire system with distilled water, perhaps a few times. Then fill with distilled water. The corrosion may be due to minerals/salts in the tap water. May have to create new cooler block due to old block corrosion leaching bad chemicals back into system. Just a thought.
Make a collab with The Thought Emporium
and plasma coat it in ceramic😂
that would be pretty insane. So fits to both guys lol
Wouldn't a ceramic coating just trap the heat?
Ceramic is an insulator, it sucks at transfering heat.
@@HepauDK I don't know exactly, but isn't it few µm thick? Is it enough to block heat transfer?
Use PC cooling liquid. It has anti corrosive properties and more!
Make a 2.0, with a better flow design over the fins, and use the big ole noctuah cooler or the be quiet dark pro!
Would love to print this out and do it at home for an "extreme" water block!
Keep up the good work, thank you for your time!
-this should be a product i can buy right now - keep messin with it
i did found you channel because of your Water Cooled Air Cooler and just watched the first part and searched for the second part and here i am
Wow been a year already.
haissam Ali I literally watched the video yesterday it feels like 😂
Funny I just found this series. Then you update it a few days later! Nice! I was interested in this!
I've got it. Turn the best air cooler into a water cooler. It's time to go Noctua air to water. :D
Now it's been 2 years since your previous video and 1 year since you made this video and people are still finding your channel because of them.
Use a Intel box cooler ;)
They tiny, barely works stock.
Will water make it usable?
Also it's black so might take water better.
Also they free and common in many shelves.
100% want a reboot of this. Adding a corrosive inhibitor instead of just DI water would help with your corrosion. Honestly not sure where that corrosion came from, cause if it was sitting on a shelf for most of that time it wasn't galvanic corrosion, which would be the big concern from using an aluminum finned heatsink with copper radiator tubes and pump internals.
You could use a massive NH-D15 and put the water inlets on the sides instead of the top to get it to fit in a typical case. You could also use mineral oil in it, which would let you put a fan between your fin stacks on the D15 to minimize the pressure drop and get more even spread of coolant across the fins, but mineral oil also doesn't carry heat as well as water, so I'm not sure if that would help at all. Better option would be a low profile cooler like a NH-L12S or a NH-C14S or a NT06. It'll fit better in the case and give you a lower center of gravity so you'll have less strain on the motherboard when in a typical tower case. With the NH-12 or NT06 you can put the inlets on the side top or bottom, for the NH-C14S you'll want to put the inlet/outlet on the top or bottom because the sides are all kind of closed up and would be a bit restrictive. Put your top/bottom/sides all flush to the fins so you don't have much or any coolant that doesn't flow through the fins to maximize your coolant over the heat transfer surfaces. Leave about an inch of water box on each side that you can tap your inlet/outlet fittings into. That'll also give plenty of space for your flow to evenly distribute across all of your fins, since the fins will act as their own distribution plate. You can help that out by putting your inlet and outlet diagonally across from each other instead of straight across from each other, but I don't think that will make that much difference.
And for testing accuracy, I'd like to see the deltaT from coolant temperature to CPU temperature, since that's what your waterblock is driving. You can lower your water temperature by changing pump/radiator/water mass/fans, but your waterblock will keep a relatively constant delta from CPU to water for a given power output. Also, testing back pressure across the cooler would yield some useful information, since higher flow rate does help somewhat, and all PC watercooling pumps are centrifugal, so lower back pressure makes more flow makes (to a limited extent) more cooling. Parallel pumps might be useful for this as well, since you want to maintain turbulent flow between your fins, and I'm not sure you'd get high enough velocity to cause that with only one pump. You'd need a flow meter and a D/P detector to measure that stuff.
For anyone asking if building this sort of block is "worth it", I would argue that water cooling is literally never "worth it". Your return on investment for open loop water cooling is basically nothing. A decent AIO will get 95% of the thermal performance, which equates to 99% of the compute performance, at 1/8 the cost. The determining factor, to me, on whether this is "worth it" or not is it's performance against a similarly priced (sum of bill of materials for each setup) traditional water block. Even if it's equivalent, it's still "worth it" as a neat, unique mod that's just fun to play with. At no point does "practical" come into play when you are talking about open loop cooling. For practical, go with an AIO or the biggest air cooler your case can fit.
yeah doing this again can be interesting
Love the follow-up! I like the original video so much that it got me into water-cooling in general. Keep up the good work.
"air cooled water cooler"
well thats just a normal water cooler...
he did that one too =)
water cooled air cooler
water-cooled air-cooler
I Think cheaper make a box and copperplate.
Maybe paint it whit zink spay.
Build a D5 pump into the water box. Hear me out:
Use a tower cooler like you used on this one. Do all the things I said in my other comment about tightening the box around the fins to get better flow through your fins. With the cooler oriented as it is at 1:28, put it on the bottom face, closest to the motherboard so it takes suction directly from the water box. Have it discharge out directly away from the motherboard. You could either print a pump top or get a pump/top that would be easy to mount. An XTOP would work, I think. Just tap four holes into the top of it, drill a couple holes on the mounting face, throw an o-ring between the pump top and the mounting face, maybe print the mounting face with a slight indent to receive said o-ring.
It's settled. I'm buying a 3D printer after this deployment.
yes... we WANT 2.0 version of that "thing" XD
Oh an please make the opposite of this one too!!!
Try to cool the water of an open Loop with aircoolers only, no radiators 😂
Could imagine that you use 2 CPU blocks for it. One on the actual CPU and the other one just in the line and then you can mount the aircooler on the coldplate of it 😂
YES, this! I'd love to see it.
I like what you did, but I will drop some science on you. You can pick whatever cooler you want to use, just make sure that the fins that run along the widest sides are even with each other. When you re-make the 2.0 version, make sure that the long outside plastic panels are touching the fins and that the bottom panel (closest to the cpu) is right to top of the bottom fin, no gap, same for the top panel as well. The water input and output sides make sure to leave at least 3/4” space, 1” would be better, this will help in the directional flow. On the water input fitting, use a 5/16” or 3/8” diameter tube that is caped or crimped at the end and the tube runs the full length into the box, cut a straight slice in the whole length of this tube, only on one side, use a dremel cutoff fiber mesh type disc (they are a whole lot safer) this will leave you about a 7/64" or so, wide slot. Aim the slot straight at the middle of the cooler fins, the output fitting that you used before is okay. If you want to make sure to get all of the air out of the system, make to output side panel taper to the fitting. By building the box this way, it will create a directional water flow that will majorly increase the cooling efficiency.
let's go NH-D15!
let's go NH-D15!
let's go NH-D15!
let the KING evolve to its ULTRA form! 👑😎
Cringe
level 99 virgin
I'd be interested in seeing a second edition; i would recommend applying a clearcoat of paint or epoxy onto whatever radiator you use (prevents corrosion for longevity and won't affect efficacy very much) and attach the inlets and outlet the middle of the sides (the position they use now, but at the center of the skinny sides instead of both sharing the plane beyond the heat exchange plates).
if you had one of the larger 2 stack air coolers like the Noctua D-15 and half was air colled and half water ????
Zynoa123 I would love to see this too, but wonder if the tank on one of the heatsinks would block all the airflow from the other fan
I think the side that has the water cooled portion would be doing a majority of the work. The non water cooled side wouldn’t be completely useless though its effectiveness would depend on if a fan was used, placement and speed. It would be fun to play with though.
@@benjaminnyman8687 this is true but I'm sure there are a few options if it is a pull fan at the back exhaust out the back might work it would increase airflow
This project is why I subbed to you. You should totally make another one with better flow and a non corrosive liquid.
^make the new cooler force the water thru the finns. you ave a lot of space, where water can go arround the heatsink
Heat rises so as long as the cold water comes in from the bottom and the hot water exits from the top it will not matter
Definitely worth revisiting.
I have an ideas. Use oil instead of water. Less corotion :)
Who need a water tank when your cpu can swim with the fishes!
Yes mineral oil :D
I seem to recall you didn’t use a radiator last time, which made the test kind of pointless. Glad to see you doing it right this time 👍🏼
I never saw the sequels, I’m gonna have to check those out, I never actually subbed until the tubing guides, that was awesome
Technically corrosion adds surface area to the fins lmao
And a layer of insulation.
Just a suggestion for the next cooler to add MoCool to the water. It is used by race cars to transfer more heat into the water, and assists with corrosion protection. I use it in both my Smart cars which have 700 cc intercooled turbocharged engines and struggle in summer to keep the temperature down. It could provide a better thermal conversion as well as giving your system longevity.
this is a year OLD WoW:
it dosn't feel that long.
ONLY a thought why didn't you just put one pie to you tape then you would be able to check it off the PC?????
That's awesome! I'd like to see the development and testing of another prototype.
wife probably tried to throw that nasty thing out. Its like homers sandwich.
This would be such a cool thing for a post-apocalyptic build
all that weight hanging off the motherboard just makes me cringe.
I mean... Is this small cooler with the water really THAT different than something like a Dark Rock 4 Pro?
I dont think that's an issue here.
I took a different twist on this about 10 years ago. I took a 1 rad water cooler and dipped the rad out the side of my case and into a large pot of water. Sometimes i put ice in too but usually left it without ice. It was the largest cooking pot i had in the house. As long as i topped it up every day (from evaporation) it stood steady no matter what voltage or multiplier i set it at. Truely the best solution if you don't mind having an open pot of water lying next to your open computer
I once made my own water block for a 780 video card. Used it for 3 years and it worked pretty good. The stupid things we do sometimes.
Definitely revisit and revise this then stress test it. The flow through the fins could definitely be optimized with at least a tighter case around the cooling fins
I'd love to see a redesign with t he hoses connecting so that it fits in a case, using more acrylic to create channels to distribute the water through the heatsink evenly with a flow like you would have with air, and definitely with some anti-corrosion additives to the water. LOL Other improvements could be a temp sensor for a mobo header and, of course, RGB lighting.
Now that you have a 3D printer I would love to see a new experiment involving this it would be pretty awesome to come up with some sort of solution that actually works that you can just put onto a air cooler please continue this project
Perhaps a version people would want to replicate in real life? For example, how about using stock intel cooler (which almost everyone has) and some radiator coolant that doesn't cause corrosion?? Hobbyist 12V water pumps are for like 3$. These can be modified to be run by chassis fan outlets, which can change rpm by changing voltage. I can buy a second-hand ford fiesta radiator for like 4$ here. Add a few $s for hoses and if this 12$ design can perform as good as high-end custom water coolers, this would be really useful as an experiment.
Odds are, a "waterblock" as large as an intel cooler and a radiator probably 10 times as large as a PC radiator would make a very silent and very cool system, maybe even without using any fans.
yessir v2.0! tighten up the sides and build a block off between inlet/outlet for force flow through the cooler! Heck with the same cooler even. I'm just as interested in seeing how much is on the table just in flow optimizing.
That be nice to see!Would be nice if there is extraction valve to remove the excessive liquid from it after use and a way to clence the cooler system after usage to prevent rusting ofc not attached to the rest the system while doing so, so u don't mess up the rest the system its awesome project build!Maybe re-attachable sides will do all that..
Yes, please revisit this. This time I think you should try to convert a factory air-cooler to water-cooling. Something round and low-ish profile. A factory Intel or AMD cooler is what I'm thinking of and this time add a little more focus on how the coolant moves in the chamber.
I hope you revisit the water-through-the-heatpipes idea in an aircooler with parallel water flow, plus a fan attached to the radiator. This would make it effectively an air cooled water block!
You should absolutely make a version 2 - however, you'll want to use a single metal type heatsink if you can find one - mixing metals is going to see accelerated corrosion unlike if you're using a single metal. You might be able to find one of the old all-copper thermalright heatsinks on ebay or the like, perhaps, would be the best bet, or you might get one of the oddball tower Zalmans that are all copper. You might also test to see whether alcohol might work better than water, given you have so much volumed mass to work with from the larger container, you don't need to worry about vaporization of the entirety of the fluid.
This is pretty interesting, not sure if it's actually better than just buying a waterblock.
To prevent corrosion you should use a sacrificial anode - basically a blob of zinc you drop in the loop that takes all the corrosion so your aluminum doesn't. This is what salt water boaters do, you will find them in your home water heater as well.
You could also use mineral oil, the trouble is that it does not conduct heat as well as water, and it's thicker than water so it's harder to pump. But it would be corrosion free and solve the big problem with oil immersion cooling which is deterioration of the rubber and plastic parts, since only the metal pieces touch oil.
One problem with going to a bigger cooler is weight! The cooler itself is already pretty heavy and then you will add a couple of pounds of liquid to that. On a flat bench it wouldn't matter, but on a tower style setup it could be hard on the mounting apparatus.
Nice Project, but you should put a hot water tube at the middle of the box to prevent Air bubbles
... I think a project where we put cooper pipes passing through the fans, maybe Just 4 aside with some kind of joints on both sides, could be more light, secure, clean, cold, easy and cheaper than a entire box arround the cooler ...
In the next one, one side of the enclosure should be against the sides forcing water to have to flow through the fins instead of allowing it to flow around them. That would help with the volume of water and size of the cooler. Also I’d have the inlet at the top and outlet at the bottom so gravity helps out. Try to make it so that as much water flows across the fins as possible and make sure there’s no air between the fins before testing (saw the old video a bit ago and there was air in there)
Definitely would like to see a v2 of this paired with a modern i7 👍🏻
Yes, please! A new edition of it would be amazing!
I would like to see you do a build using something like a high proof alcohol and glycerin mixture instead of water. Something with a higher thermal load while being able to be chilled further without freezing or becoming more viscous. I imagine this will take some tweaking, but there should be some good example recipes on the net for "normal" liquid cooling setups. I bring this up because it might help solve your corrosion issues, while letting you dip that radiator into some ice water instead for some craziness.
Next time you retire it especially if you do a 2.0 run dry air through it until dries out to help preserve it. You can start by just setting it in the window and letting the sun heat up the inside and push out the moisture and then once you get most the moisture out put a piece of tubing across both ports to seal it.
Would be fun seeing one using a really cheap cooler. Such as a stock cooler. Perhaps a design more suited for water with longer between the fins.
If you do revisit this with another cooler i highly recommend finding that is all copper or all aluminum and make sure your radiator is the same material other wise your going to have that corrosion issue.
If you have to use mixed metals please add something to the water in the loop to prevent the corrosion otherwise you might kill that lovely 360mm rad. :O
An easy way to get a little more efficiency out of it would be put input up top and out flow at bottom to make it a counterflow chiller, also shrinking the box will up efficiency as well forcing more flow over the fins.
Get a THREADRIPPER for the next one!!! See how the temps hold up!
If you do a version 2.0, make sure to electroplate the cooler with something corrosion resistant.
As the copper aluminum are dissimilar metals, flowing water across it may cause electrolysis.
May help prevent corrosion in future builds.
I'd also try and place the fill and drain on opposite "Walls"/Corners to help draw flow across the entire surface area of the cooler.
That's my ideas.
Any ITX cooler would be a good experiment ! Nice vid!
You should run or put some calcium lime and rust product to clean the gunk and corrosion
Yeah, I remember it, just watched the video where you built it. That's the sad thing when I am "late" to a channel.
It'd be interesting to see a second version, but in it you should use a cooler with some corrosion-protective coating and minimize the free space at the sides of the heatsink, as flow around the heatsink doesn't help cooling. If I were to recommend a cooler, it would be the Noctua NH-U9S.
One thing I would suggest is using mineral oil to prevent the corrosion issue and also avoids the issue of leaks bricking the board.
The only issue I see with what I suggest is that I dont know what pump could handle the viscosity of the oil.