Hey man... I am the guy who recommended you "The cooling tower" concept and boy am I amazed to see it working... You did a terrific job... BRAVO... Even I thought it was crazy to recommend someone this yet you managed to pull it all together... Originally the idea came from work as I'm an engineer in a power plant... We use massive cooling towers to cool our two turbines... Glad to see it working in this scale...
The only real downside to this design is that since it's an open loop that replies on evaporation cooling to function, you are going to lose water over time. I mean, not really a big deal if you keep an eye on the water level in the reservoir and fill it up accordingly, but it's still a thing. As for a case design, I'd imagine machining the case to allow the reservoir to stand embedded in the top so that the cooling tower is separated from the internal electrical bits. But here's the real question: Who's ready to start building home-made evap cooled power plants out of their old cases :P
It’s a cool idea, and proof of concept. But this thing will get so incredibly nasty dirty that it will end up plugging the CPU cooling block in no time. We refer to our towers as air washers in the industry. They literally take all the dirt out of the air. Not to mention it’ll probably run itself dry in a couple hours without continuous water make up. It’ll also become a breeding ground for all kinds of nasty and dangerous bacteria and bugs if it doesn’t have any water treatment.
Yeah... There are several problems. It also raises the humidity of the room it's in which... You wouldn't want that around electronics. And you generally want to use distilled water around electronics, which means using normal tap isn't the best idea. Now, I think on a larger scale this could work, if you pumped the humid air outside, that would be a big plus... Not quite sure how to fix the dirt problem... Filters would help, but also decrease airflow... The growing stuff problem would be somewhatly combatted by using distilled water, you could also treat it somehow... Yeah, everything's pointing to it needing periodic maintenance, even in the best cases. Maybe looking at AC units would help though, they don't tend to use evaporative cooling. (But I think that might just lead you back to using the standard closed loop, reservoir setup.)
As somebody who works on industrial cooling towers, a few pointers. 1: Make the basin conical from the pump inlet to the side holes and make it deeper. You will lose less water that way and allow for more water pressure before the pump (it should run better.) 2: Design the bottom end side walls to have louvers rather than holes. They just need to be \\// shaped, not mobile, but they will allow air in while also keeping more water in than just open holes. 3: Make it square and use two fans. The efficiency isn't much different per fan and you can then build in supports and mount it to something so it is more stable. 4: You will be constantly fighting water evaporation. My system normally needs enough water to completely refill the system every few days. Would make a neat dual purpose humidifier in the winter though. The real issue is you are going to "scale up" the system with crap in your water over time and destroy the pump/tubing/fan. I really do not recommend using a system like this for a prolonged period of time, you will destroy your components. Neat thought experiment though.
@@vinade2100 Actually wouldn't really help, evaporation causes it to lose so much more energy than direct cooling that it's irrelvant. It would just be a bigger heatsink with only a tiny increase in cooling capacity while increasing complexity and production cost.
if your pc outputs 200 watts of heat then this design would end up losing 0.3 liters of water per hour which is about a third meaning that you would need to have a larger reservoir for your water and only feed in distiled water.
Functionally this is the smallest evaporation cooler I've seen slapped onto a cooling loop. So, not shocked it works, am somewhat surprised at the performance to size...
When people were starting out with water cooling in the Athlon days, I have seen some really goofy setups with evap coolers which were HUGE, like 6 foot long, 8" wide sewage pipes fed by a shower head, just to see how far down they could get the temps.
@@Stoney3K if you lived in a dry area they were great too since the drier the air the more effective they work. You could get REALLY low temps if you sized them right. the reason for making it so big i think was to get as low of temps as possible and also having a lot of mass to make sure you couldnt heat soak the setup and to give more time between topping off the res.
When watercooling was getting big back in the late 90's, I experimented with something like this. It does work, as long as you can handle the drawbacks (water loss, humidity, cat attraction). A few things you could do to possibility optimize performance is to swap the drip tower to have mist nozzles, they'll get the water into a much smaller size so they spread more effectively. You could also swap the 3d printed slats for Evap cooler media, it's usually a honeycomb matrix which spreads pretty effectively.
you just literally described a RTO system except with water. i guess you could call it a RTH like the cooling towers that bed media and have induction loops with negative air that humidify then dehumidify from gas to solid with ammonia/glycol to cool the water to right at freezing. i'm a millwright and have built many and that's an awesome idea sir.
Oh, prediction: works, no throttling, about as good as a 120mm aio, depending on the pump speed (and pressure) 😁 And just a little mess. *edit: watched the video, wow.... I did not think that it is that good. Just wow.
@@jacobrzeszewski6527 Finer nozzle and a different fan (e.g. Noctua NF-A14 IPPC-2000 IP67) could easily double the cooling power. The only problem I see are the nozzles, because they need more pressure (> 3 bar)
@@SuppenDfg shouldn't be a big problem, the pump can definitively do that. There have to be enough to not impede flow overall or stall the pump. And too many will result in no spray at all...
@@SuppenDfg Required nozzle pressure is just a function of how big the hole is and how much water the pump moves per minutes. Everything else is flavor. Point is, just get finer nozzles, the REALLY fine ones are usually called "Misters", as they produce more of a mist than a spray of droplets, that would actually be better for cooling as each tiny water droplet is its own surface area for evaporation. Although he will need to protect the fan from the mist getting up there and making rain on his room.
I would just make the tower out of wood (rectangular) and paint it to seal it from the water and have it have an expanded reservoir. Sponge or air filter for infill (your 3d printed ones would last longer) and forced draft to protect fan from high humidity/splashing. Wouldn't take days to make.
You are missing huge cooling potential with nozzles not spraying water into fine mist! If you would have nozzles that spays water into fine mist, it will cool even better (much much better). You can also use nozzles bought from the store, because it sprays water better :) Please make another better cooling tower, the performance will be astonishing! :) Love you channel, keep up the great work.
Huh. A friend of mine interviewed at a company where their data center was on-site, at least some their machines were water cooled, and they had a fountain in the lobby that was part of the water cooling system. Seems like the that design was likely also inspired by the same type of design.
This kind of cooling is used in some data centers indeed, I remember an article about a new data center a few years ago that used a very similar system.
Yup, a cooling tower, especially a draft cooling tower, is more efficient than traditional HVAC, so they get used industrially all the time. You don't tend to find them in places with high humidity however, because they're much less efficient when they can't evaporate water. In a data center, the cooling tower is cooling water that cools the air that the air handlers use. They're not water cooling the servers directly unless they're running something custom.
Yeah I've worked in a place where the server racks were force fed the cooled air from central water cooling towers. Definitely industrial scale cooling
i love your willingness to experiment, and how you show the full results of trial and error. failure is part of the process, and not a negative thing. you're learning more by doing these experiments than most pay to learn from textbooks. keep up the great work! love seeing your chronicled adventures in cooling. i think the larger lip around the bottom would be helpful, and maybe some other way to re-capture the evaporated liquid! :D
LOL so true! The other way would be to close the air loop. But that would require a large air cooler to act as a recondensor to collect the water back out of the air loop and reuse it.
@@Bearthedancingman The point is though, that if you recondense the water it will transfer the thermal energy back into the loop, causing the evaporative cooling to have no effect. It only works because it is an open system.
Kyle? I remember looking for porche radiators. After all the workcooling stuff you used to post. It really got me into of hardware. Do you think you could make a video of the techniques of the bad old days?
Sometimes “almost as good” is as good as you can hope for. Then it becomes a happy surprise when it turns out to be “better”! And a triple-fan radiator is no laughing matter either! This is genuinely a pioneering success in the PC cooling sector, a proof-of-concept of something so ridiculous that it *should* fail and a new road to map! *BUT* we need to push this idea further! Combine the radiator with this cooling tower, such that the water first goes through the triple-fan radiator *before* being funneled into the tower!
With a cold water connection that automatically tops up the supply and flow sensors at the input and output to turn off the pump when an overflow may occur, it might work reliably enough.
@@starbase218 U could Put one or maybe two into an old fishtank, so its a closed system, the water evaporates and then it starts to condensate at the glas and flows Back in the system. Problem here is U need a huge amount of water, the inlet is at the top of the tank, but the outlet needs to pick up the water which condensated and flowed down the glas. So there needs at least 1 inch water at the bottom. A real proof of concept could take 2 hours and More until the water is heated up. You could fasten it and putting preheated water in it and measure the watertemp and if it stabilises while Stresstesting at a certain temp it shows the system is working. There is another point. When water condensate it sets heat free and the question is, is the outer surface of the glas big enough to transfer it to the Air of the room. Because thats essentially how cooling works. You Transfer the heat from the source to the air, whether its a aircooler Just with air or a watercooler combining water and air. And sorry for my Kind of Bad english, i am from Germany and its Not my mother tongue 😂
As for evaporative cooling, the warmer the water is as it goes through the tower, the more effective the tower is. It would be best to pipe from the tower, to the radiator, then back to the cpu. One HUGE downside to these systems is bacterial growth. Open loops near body temperature are a perfect breeding ground for nasty slime and smells. Not to mention life threatening in some instances.
Im new to your channel, just found it yesterday, great stuff. I love the free and open expression of ideas and experimenting. I would love to see a series on improving a standard 20x20 box fan, with blade improvement, and even inserts to improve air flow. We all use these box fans, but face it the tech on it is ancient. Just a suggestion.
I love how you test all the wildest stuffs. This reminds me of Tech Ingredients Revolutionary Air Conditioner video, but he actually uses it to cool his house. maybe you can use that to get some better ideas on new designs, or even reach out maybe get some new ideas what could be tried :)
That look at about 6:12 to make sure the camera is still working is great 🤣 Also, really surprised by the results! If you were to take actual spray nozzles and put them in instead of the pla printed ones, you'd get even better results!
I've notice bang good do some really fine nozzles cheap designed for fitting to a gazebo/parasol to cool you in the garden. Imagine a few of those used instead to make an ultra fine mist, imagine the surface area involved there & how much cooling that could provide.
Of course it's working. This system is an a/c that can go below ambient temperatures. You could improve it by controlling the humidity levels with salts.
Yeah, i have used 3d pen to touch up alot of prints, i also printed a lvl3 helmet from PUBG in many parts, and "welded" it together with a 3d pen. Works really good :-P
Depending the materials, you could also do welds with heat (soldering iron) or chemical (e.g. ABS -> Acetone). Both work great if you're going for functional strength over beauty.
@@kylek29 I only used it on PLA. I find that hotglue is not sticking all that well, i mean it sticks, but it will easely rip off.. 🤔 Might be that i used a bad quality tho.
I've been using 3D pens my whole life..is the alternative to which you're seemingly all accustomed a 2D pen? Better yet, 4D? Please link to suppliers either way, the former for amusement/novelty; latter for sheer superpower status! Kidding aside, UV-reactive photopolymer pens are the best tool for nerds like me since the 3D printer.
This is one of the coolest things that I have ever seen, and easily the most creative and interesting alternative cooling solutions that I have ever seen. I've watched quite a few of your other videos lately and you're doing more interesting and creative stuff than even LTT, and they have an entire metal working shop, laser cutters, CNC routers, etc. Fucking kudos man.
You can 3d print a large diameter concave or cone shaped lid that floats a few inches above the fan that collects the water droplets around its rim and redirects the water down a tube and back into the reservoir. This will help reduce the added humidity in the room and reduce the amount of times you need to top it off with new water .
@@SuppenDfg I think he is only talking about catching the liquid droplets that have not evaporated, but slipped through without phase changing. The video did touch on this when he said his design didn't have a 'drift eliminator'. Some coarse foam below the fan would do the trick too
@@SuppenDfg The heat removed from the system to evaporate the water must itself be lost from the water in order for it to condensate and be returned to the system, so it wouldn't be counterproductive, it would just reduce the need for water to be added. The efficiency could be even further boosted by using a condensation tower (essentially mirror the cooling tower across the fan) instead of just adding a simple hat to it.
@@GremlinSciences The vapor will loose its heat partially to the surrounding air, but also to the water droplets building up on the "hat" so that when the droplets finally fall that water will be above ambient temperature. I'm not sure how much heat will go to the air and how much to the water, so it's quite possible that it still works, but imho it should be less efficient.
This is definitely a great idea. I could easily see an improved version of this with finer misters, correct seals, and a way to have it detect when it needs more water to add in more could be amazing.
@@danieljenkins4213 seems easier to implement that my thoughts of an anduino board with a program to take in the humidity and watwr temps to calculate when it needs more in. I knew it was overly complex and had to be an easier way.
I don't see a lot of comments about the water hardness as time goes on. The same reason why people with fish tanks don't just endlessly add water: Evaporation removes water but leaves a lot of the mineral deposits. You'd need to completely change out water to prevent massive calcium buildup over time. Using distilled water would help I guess, but I think it's still a thing.
I am an engineer and stumbled on your video. Im impressed with your build. You have proven that open (evaporative) cooling is far superior to closed cooling. This is why datacenters use the same principle. You have humidified the air. Now you will have to worry about legionella in your computer room.
One problem I see with this: It will introduce dust from the air into the water ... pumped around the loop, it could clog small canals in the water-blocks.
This is an actual problem with water towers. There needs to be a sump outflow to remove dirty water to prevent the accumulation of dirt in the system. So the amount of clean water added back needs to replace both the evaporated water and the dirty water lost.
Yeah needs a filter over the fan or an inline filter like for fuel/oil in cars. Otherwise you’ll need to dump the loop every month or so and occasionally clean the block (I had my loop in a bucket, eventually there was so much hair because I shed about like a cat, that the blocks needed to be cleaned every half year, but it was an open bucket next to my feet, lol!)
@@jakegarrett8109 Nobody wants to disassemble, clean and reassemble their entire system every 4-6 months. And I doubt an air-filter will be effective enough, unless it is so tight, that air-flow is heavily impeded as well, lowering the overall evaporation effect. But hey, I'm just guessing ... havn't tried :)
I enjoyed the video! I appreciate how they always take us along for the trial and error process, including failures and mistakes too. That corsair jab got me good too lol...Keep at it!
I've done this... One of the best things I've ever made. However, depending on your design (mine is based on 180x180mm interlocking panels), be prepared for an extremely long amount of print time. My case, which was designed around an ATX Mobo took over a month to print. Some of the more complicated panels (ie the power supply mount) took almost 20h alone. The average print time for a panel was 9h. All told I had 24 panels to print, and it took roughly 225h total to print. If you really skeletonize the design you could certainly bring down the time but be careful, it still has to support the weight of your parts.
@@jonmcentire I would have loved a 3D printer for when I made my brother a PC case, his tho is wood and metal, looks pretty awesome if I may "humbly" say so. Adding to the complexity, my tools at disposal were only hand tools and the simple power tools you can have at, well, a small apartment (like a jigsaw and drill. Rest all hand done). The back plate where the PCIe slots, fans, power supply and I/O shield mount to (it's a cube style case by the way) were quite the task to cut with those tools let me tell ya. In the end, the result turned out incredibly precise, no need for feet leveling even for example. Still quite proud of the end result. There's something very different about a completely custom design that is just hard to match by spending any money for something already made.
Please, please, please do some follow up videos with improved nozzle spray, variations in infill, rates of water consumption, and...series and parallel setup of towers! Really love the channel, thanks for bringing us along on the explorations.
You should add a "catch" system to recover the water lost out of the top due to evaporation. I'm imagining something similar to a old-fashion moonshine still positioned above the fan.
@starshipeleven Just need to mirror the baffles from the cooling tower to slow down the steam and catch the condensation. The droplets will still be fairly warm, but they can just be ran back through the tower or let flow through a length of pipe before being fed back into the system. Water is very good at stripping away heat and maintaining temperature, but that only actually applies until it evaporates. Once water evaporates, it's no better than regular air at both, force it it move or change direction and it will lose a large chunk of it's thermal energy.
This was actually pretty cool! (Bad pun.) Cool your computer with a humidifier. LOL For a proof-of-concept model it did surprisingly well. I can't wait to see more refinements. Besides the obvi, an automatic feeder tank to maintain water volume would be nifty. I also can't help but wonder how much the plastic itself is a hindrance, where some stone might help. Maybe print some molds to fill with cement to make the tower part?
dude imagine if you made one out of some decent thermo conductive material like aluminum or similar, with tight tolerances and proper orings and gaskets, I would honestly think about swapping to test bench to run one permanently.
It could be made of gold and wouldn't make any difference this cooler works by evaporating the water with the air flow through it. About the only thing that could improve this is a larger surface area mixing zone. A honey comb design maybe, or pack it full of coarse stainless steel (no rust) or plastic brillo pads... you want as much interaction between the water and the air as possible. Probably one thing that would have to be included at some point is a set of air filters to keep room dust out of the interface material, and ultimately out of the cooling water so it doesn't gum up the pump and cooler(s)
This used to be a thing 15 years ago. Except we didn't print our towers but used drain pipes instead. Back in the days the were called "bong coolers" because they looked like a bong. I used to have one and it performed really well but as you've also pointed out water loss was a huge pain in the ass, I ruined my pump because I forgot to refill it. That being said I still enjoyed your video! You've gained yourself a new subscriber! Kudos mate!
I saw that video series and noticed one big problem, he was forcing all the water through the computers where he would have been better off by running at the end of loop a reduced cross connection to the return so that it was a full flow and the computers pulled water that was potentially cooler by the free flow design and not possibly blowing up computers.
Congrats on the PC/Personal Room Humidifier. I used to work In a Thermal Electric Power Plant and worked around all kind of towers. Id suggest adding multiple tiers of shallow convex and concave trays to disperse the water evenly as it cascades down to decrease the flow from the Nozzle to the reservoir. Making the tower as tall as a PC would allow for more trays. One more thing find a pattern for a transitional segment that goes from round to square with round corners and stretch it out to bout 18" tall that'll make the design structurally aerodynamic. Try catching the evaporated water and condense it!
okay so i found it really dope BUT your evaporation is still quite limited and the evaporation actually will cool more air than water. I think that replacing your plastic grill with some cheap radiators (something not too dense) would allow for way better thermals : The radiator being the hot element and the water droplets being in direct contact with it , it should pretty much instantly evaporate thus cooling the water passing through the radiator. For even better evaporation you could make the spray nozzles even a bit smaller and use a pump with more pressure. The smaller the droplets , the more the heat to mass ratio will be big and the more cooling you should get out of it. This design should also allow for a much cleaner setup as you would have two separate water loops (yeah adding more complexity but it could be really worth the try imho) : one "classic" system loop and one "water tower" loop that cools a radiator. That way if there is a mess to be made it will be far away from your expensive system.
Are your sure about that. My gut feeling tells me that the water is "cooled" more than the air at least if the droplets arent superfine and evaporate almost completely. For the reason why: If water and air would have a temperature difference there would be a heat exchange until the temperatures are balanced, but to get there the water would have to pass on much more thermal energy to the evaporation process than the air because of it's bigger thermal capacity. I might be wrong and if that's the case I would love if someone could explain to me why.
Awesome! I have seen some huge bong coolers, but they did not do well with water distribution. An easy modification for your setup is to not remove the radiator. Use the closed air heat transfer to bring the loop water to near room temp, then run it into the tower. you will have cooler temps, and much less evaporation to top-off. Some performance parameters relating to cooling towers would be cool as well, the main being what is the temp of the water compared to 'wet-bulb' temperature (the coldest possible using evaporation) This would give an idea about if the tower is over or under sized. The wet-bulb temperature is a function of dry-bulb (regular/normal) temperature and the relative humidity. These style coolers are very efficient in Arizona where it's dry :) To determine the wet-bulb it's easiest to look at a psychrometric chart, or use one of the many calculators google links to.
It's essentially a swamp cooler. I used to use those all the time in Richland. They work great in dry climates. This is a really fun concept. I'd like to see an entire case built around this design. I bet it would turn out to be an extremely cool and eye catching case.
Nice job man very impressive :) you got yourself a subscribe. Although you gotta vent that humid air somewhere, The more humidity goes up, they less cooling you get through evaporation. Eventually you run the risk of making your room too humid, or making your air conditioner work harder.
Love the video, but I can see a couple problems down the line with this setup. 1 if you ever run out of water, you might damage your computer. 2 you can pick up dust and other debris in your loop which may eventually clog up the system leading to failure. Thought occurs, make 2 loops. Primary loop can go to a cheap $30 120mm radiator that's mounted inside the cooling tower, and connects to your CPU/GPU/whatever else. This is a closed loop with no danger of running out of coolant, or getting external debris. Secondary loop is the sprinkler/tower, reservoir, and another pump. Water is dropped from the sprinkler as normal over the diffuser plates and it drops down into the reservoir below the Colosseum. From there, theres a few different ideas of whether to mount the radiator. 1 spot would be just above the Colosseum, below the diffuser plates (maybe omit the diffuser plates?). Second spot would be just below the Colosseum at the mouth of the reservoir. 3rd would be at the bottom of the reservoir, such that any water being pumped up into the tower, needs to be drained through the radiator (and thus have good fluid flow through it). Personally I don't own a watercolor setup. But depending on the next couple months (and what they decide the second stimulus check is going to be), I might splurge it on a computer with watercooling, and get a 120mm radiator on the side to experiment with.
I have a pro-tip for you instead of hot glue that I use all the time with my 3d prints that need some touch ups, additions, etc; get a 3D printing pen from amazon or something, and use that to weld the two 3d printed parts together. That way you have a strong, waterproof joint. Make sure to use basically the hottest temperature possible, because you don't want the heat going too deep, or you'll warp the part, or break through the skin into the infill, so keep kind of a light tough. Also, don't extrude hardly any extra plastic into the joint, or you'll not weld it, you'll basically just glue the two pieces together very poorly. just fuse the existing plastic together. The only downside is that only edges can be welded together, so two large faces butted against each other will only be held at the edges.
There's a good reason why Cooling Towers have been used for High-Rise Buildings for over 80 years, they work really well. They are not without drawbacks though, there Cooling Capacity is directly related to the Relative Humidity of the Air, so they work really well in a Dry Climate, and just mehhh in a Humid Climate. They also concentrate water born minerals at a ridiculous rate , which must eventually be flushed out, and/or, handled Chemically.
If you can wrap the air inlet port with some desiccant or something you will get more efficiently out of it. Evaporation is the means of heat transfer. So, the drier the air coming in, the more moisture it will except. End result is more cooling. Room may get pretty humid after a while which will cause efficiency loss obviously. I really like what you are doing here. Keep up the good work!
Bong water cooling was done years ago with water cooling setups on systems. The ones I recall were about 6 feet in height. As I recall, a shower head, a water pump and a fan mounted in roughly 6 inch pvc. Evaporation was one of the main issues.
Congrats on it working. I got excited for a minute but then I realized this would never work for me due to humidity already being insanely high in the summer. I struggle to keep my room below 70%, with something like this adding even more water to the air it would be horrible for all my electronics.
I love your channel. You're like a lovely blend of bits of J2C and 3D Printing Nerd plus a willingness to look at all the round wheels in the world and think a little bit mad-scientist in the best possible ways. Also I think you'll have a great time with your Prusa printer. I've had many happy years with mine (I have the original MK2) and only just run in to my first serious issue - some combination of nozzle-clog and the extruder gear chewing filament - so tomorrow I will be doing a tiny disassemble (removing hobbed extruder gear to clean it) and maybe swapping to a new nozzle if I have a spare.
one major reason it worked so well for you : your ambiant humidity is low. Evaporation cooling highly depends on ambiant humidity to work : the lower your ambiant, the better the cooling; as the higher the ambiant, the less water can evaporate since ambiant air is already saturated I live in Paris near a canal, and it's humid all the time. It can tell you that small evaporation towers aren't used much here, as their efficiency is really bad ;)
The concept reminds me of the old Zalman Reserator 1. The Zalman dissipated heat via fins on the reservoir, but as an added bonus it would evaporate water inside the reservoir too, you could tell because condensed water would be attached to the lid of the reservoir. It was designed to provide purely passive water cooling, but I remember finding some tubing that would slip over the reserator and attaching a fan to the top of the tube. The fan ran dead slow, but provided a much more efficient method of removing the warm air away from the reserator fins and dropped the CPU and GPU (one water loop across both) by 10 degrees C.
This is brilliant... WOW lol. I don't know how you don't have more subs and this channel isn't bigger. You could have just created (with community assistance) a new category of PC cooling.
You should make it like a hybrid cooler, here is why: The evaporation is now cooling the inner parts of the cooling tower. This in turn will also cool the water flowing through it, but you're essentialy cooling parts you do not want to cool. You want to go back to the "water cooling an air cooler", but not in a closed water loop this time, you just want some mist inside the air cooler. I see 2 tricky parts with this approach: 1. How do you protect the cpu and other electronics in the case against leaking water 2. How do you create a mist inside the air cooler My idea is to create a box around the fins and fans of an air cooler, with an opening in the front and back for the air intake and exhaust air, these opening should not go all the way to the bottom in order to keep the water inside the box. It should probably be open on the top as well in order to place the fan(s) and a drain at the bottom seams useful. The mist should ideally be generated before the intake fan, this way the fan will help in distributing the mist through the fins of the cooler. A high fan speed should help preventing the mist from going inside the case, but you probably want to have pulses here: - spray some water with an high fan speed - stop spraying, lower the fan speed and let it evaporate - repeat Might even be interresting to place 2 intake fans on each end and let the air escap through the top. If you can get this working, I think it should cool better than the water cooled air cooler as it's now an air cooler helped with evaporation Not my original idea though, I've seen this big jaeggi coolers at a datacenter which basicly do this (but than outside and not having to deal with fitting in a computer case).
That's really cool. But I have to mention, that the cooling effect depends on how does the air in your room is. When it's wet, the cooling will get worse
I once tried a similar, but larger scale, version of this. I used about 4 feet of pvc pipe with a shower head at the top, a pvc angle joint at the bottom to mount the fan out of the way of the water flow (counterflow design), all draining into a 5 gal bucket which was the reservoir.
Pretty cool. The only issue with the concept is that unlike traditional water cooling where you just heat up the room because of the radiators, you're now adding quite a bit of humidity to the air on top of all that heat. Running something like this could get very uncomfortable unless you have good airflow/AC in the room. It also wouldn't work very well in humid climates. Evaporative cooling is fantastic for deserts, but your efficiency drops significantly as the humidity in the air increases
I really don't think this was"bad design" (certainly not horrible), just a small oversight in planning. The Your printed design already has ample venting, only needed a small step-back-and-observe to figure out a reasonable compromise of pressure escape route without too great a risk of catastrophic leakage. I think you did great man. If I can make a small quality-of-life suggestion for these kinds of projects; make yourself of maybe ask EK for, some multiport reservoir tops. They tend to make experiments and mods just that bit extra comfortable. The dream for me, would be something like a handful of EKWBs large/tall POM multiport tops, completed with the reservoir end threading, o-ring and o-ring groove - and thats it. No ports drilled and tapped at all, just ready for DIY custom ports.
i feel like with some improvements this design would make an awesome desk cooler for some builds. have this sit outside the case and the tubes running into it. there are some special setups that this would be amazing. especially if it was machined instead of printed. imagine having an in-desk pc with this (made out of aluminum) sitting on the corner like a fountain. the only issue i could see is things getting into the water since its open. could cause some issues.
Choose materials carefully, copper or silver plating on surfaces with water contact will provide some antimicrobial effects, and further to that, distilled water with a germacide will also help. Except against dust, you'll still need to open it up and wash/pressure wash the parts and change your fluids, might even want a heat exchanger just so the open loop fluids with dust and contamination never reaches the CPU cooler block on it's own closed loop. Intakes could have dust filters for air, but you don't want to limit the air flow too much All and all a complicated mess but definitely a fun project, so long as you test it away from a PC most of the time XD
I wish I had your creativity and inventiveness - you have a really unique channel in the PC/Tech space. Great work dude! Although I think the market for this invention may be small... :)
You could do cooler which is made from radiators connected in cross shape for 4 rads or in line for 2 rads. They would be cooled by rotating whole tower and radiators would get wind to cool them (it would look similar to this from top -o- for 2 radiators )
Hey! Hi Major Hardware. For your issue related to the water evaporing and needing to refill overtime: This may seem dumb but making a system using a toilet valve so when the water reach his lvl it will stop filling back in. Which would just fix your evaporation issue. Also the reason i believe this tower is doing so well is because the water is in direct contact with the rushing air, so the heat exchange is much more efficient than through an alluminium radiator
good proof of concept, The one thing that the power plants use is a heat exchanger to keep the power generating water or in their case steam separate from the cooling tower water. You could do an even further by using a closed loop like Jayztwocents did with his recent dry ice setup with the copper pipe in a bucket of cooling water. And BTW I say on the first go through your video of the quick pic change on your screen while you were talking about the results. I may use this idea with minor tweaks, thanks.
Would be super cool if this was integrated into a case where the hot exhaust air going out the top is what evaporates the water. I expect that would make this more efficient and you would still have a radiator on the front intake. The temperature it can theoretically get down to would be the wet bulb temperature, and based on the readings you have that would put the minimum temp around 17C. According to some online calculators, warmer air will increase that wet bulb temp by 2-4 degrees, but considering the water is also going to constantly be around 30-50 degrees warmer than that Idk how that would effect the change in temperature.
Hot Glue and PLA actually go very, very well together. Helpful tip: If you make a mistake and need to remove the glue, spray isopropyl alcohol and slowly pick at the glue and it will lift off cleanly.
You know how some people have little pump driven waterfalls in their garden because they like the sound and look of it? I can totally imagine someone doing this in their home, but instead of a waterfall they have like a little miniature industrial area with a WORKING cooling tower! kinda like how people have miniature train landscapes, but instead it's industrial haha
I have an Idea for an expansion to this project: The biggest problem I see with this setup (it´s a prototype, so design doesn´t count) is that you are losing so much water because of the evaporation. My Idea is that you could make the top part above the fins higher and maybe just put metal fins ( preferably copper) in there, so the water can at least partially recondense on them and flow back into the system, wich could solve the issue or at least give you more time before you have to refill (a bigger reservoir might also help there). For a prototype you could for example just make it thinner and higher and put an AMD stock heatsink a couple centimeters below the top fan, so the watervaporpasses through it and can partially recondense on it, or you could even stack multiple heatsinks. If you actually test this and it works I might even try to use it in a future build for myself :)
Well, we use this kind of cooling tower to cool down our water to condensate vapor in order to create some vacuum in our sugar factory. It works great for big water flow. For some smaller water flow, we have some kinds of rads+fans. You can even add water pulverisation on the rad to gain even more cooling potential (would love to see this one). You find the same concept just on another scale in PC as in industry. Pretty cool to see it applied to computer.
Jsyk PLA is really easy to weld together. Heat up an iron wire or nail and basically just squish it together. It welds and holds; I've used it in many of my prints. Also something else is using filament as welding rod; catch the end on fire with a lighter and let the plastic drip rapidly onto your joint. It melts surrounding plastic to itself. Once you like where you're at, wet the plastic (I use my finger and a little water) to cool and set the plastic. Way better than hot glue.
I would love to see you build a float valve to install in the sump (basin) that you could hook to a bottle of water that when your water level drops from evaporation, the float opens and let's more water in to keep your pump primed and the system running smoothly
Working as refrigeration mechanic in radiator based plants and cooling tower based plants, i not surprised that it worked, the evaporation can cool the water to below ambient, though you built it nicely.
Zalman once made like a tall floor standing reservoir with fins on the sides. It was made to cool computers passively. The design could be improved for sure but the concept is solid. Even after a week you will see that the mineral content of the water in the loop will increase a lot if you keep using this. Minerals will start to clump together like in any humidifier.
Phase change is one of the most effective methods of heat exchange. It's how we lose heat with sweat and it's how air conditioning works. In my opinion it would work better with finer mist/better spray of the water, a steeper funnel base, and maybe a different fan setup. I'd like to see more of a pressure fan than one set for flow as a lower pressure inside the tower could help shed more heat due to partial pressure evaporation. Thermodynamics is one hell of a topic to dive into.
One thing to keep in mind for anyone who decides to use this is that cooling towers that don't operate hot enough to sterilize themselves are breeding ground for mold and bacteria, so extra sanitation is required. Also, in a typical cooling tower, the cooling fluid is separated from the tower water either by a liquid-liquid heat exchanger or by putting the primary loop's radiator in the drip path so you still get cooling even if the tower dries up, the water is only an efficiency booster.
You should try channeling the air in from the top like top air-feed dripping atomizers do. Would prevent the chance of spillage and allow refinements on the air pressure with an adjustable ring of some sort.
Not a bit surprised. Evaporation is the best cooling you can archive. There is a reason why the powerplants dealing with giga Watts use exactly this solution. It is messy but effective. When you put a pot filled with water on your stove even a small pot can drain the energy from a 2kW cooking plate with ease as long as you provide water. You need 2259W to evaporate 1g/sec at 100°C(boiling Water). The estimated evaporation is about 180ml water per hour to cool your computer. (based on 100W power dissipation) Would be interesting to see whether this number is near your practical experience. Great video BTW.
Years ago I met up at Lanwar with a group of regulars of a CS server I played on. One of the guys brought a cooling tower made of a 5 gallon bucket, 4" PVC, pond pump, fan and a shower head. Granted this was Pentium 4, Athlon XP days but while gaming his CPU was barely over ambient. Was pretty cool I've always want to give it a try.
I have 2 suggestions to perhaps make this a little more efficient and safe (?? lol) ..... 1. Instead of using a large bore on top of the reservoir as you did with the ring you built, I would just replicate the original cap but with a secondary hole and a "straw" that ran above the spray line. Less chance of major spillage if knocked over. 2. Having the fan pulling air and moisture through its blades and motor system will lead to electrical failure of the fan eventually. What about a smaller fan driving an air foil system at the top [a la dy$on] ? You could then also put a condensate cap above it to help recapture some of the evaporated water. It makes the print much more complex, but could save power and seriously increase the air volume. The added potential cooling capacity may allow for a GPU water block as well ... As you pointed out, the funnel needs taller sides to reduce splash or perhaps increase the diameter of the funnel beyond the cooling tower to make adding water easier. Love your projects !
so THIS is how quality content looks like! thanks for your videos! just started wahtching you today, and totally impressed me, doing something related to what I love (computers) but being creative and also completely different to the rest of YT channels,
Hey there, HVAC Tech here, just stumbled upon this and am not through with watching BUT using the water that runs through the cooling tower to then pump through your system is bound to result in gunk buildup over time. It'd be less efficient but using a secondary loop to cool water which in turn cools the actual cooling water helps with that issue. The whole idea also doubles as a humidifier but gets less effective the higher the humidity in your air.
Hey man... I am the guy who recommended you "The cooling tower" concept and boy am I amazed to see it working... You did a terrific job... BRAVO... Even I thought it was crazy to recommend someone this yet you managed to pull it all together... Originally the idea came from work as I'm an engineer in a power plant... We use massive cooling towers to cool our two turbines... Glad to see it working in this scale...
The only real downside to this design is that since it's an open loop that replies on evaporation cooling to function, you are going to lose water over time. I mean, not really a big deal if you keep an eye on the water level in the reservoir and fill it up accordingly, but it's still a thing. As for a case design, I'd imagine machining the case to allow the reservoir to stand embedded in the top so that the cooling tower is separated from the internal electrical bits.
But here's the real question: Who's ready to start building home-made evap cooled power plants out of their old cases :P
@@razieldolomite698 Honestly, resupplying the water could easily be automated with an arduino and a bigger source reservoir, if desired.
It’s a cool idea, and proof of concept.
But this thing will get so incredibly nasty dirty that it will end up plugging the CPU cooling block in no time. We refer to our towers as air washers in the industry. They literally take all the dirt out of the air. Not to mention it’ll probably run itself dry in a couple hours without continuous water make up. It’ll also become a breeding ground for all kinds of nasty and dangerous bacteria and bugs if it doesn’t have any water treatment.
@@1967friend huh... For some reason it never clicked than an open loop would be exposed to the crud floating in the air. Seems obvious in hindsight.
Yeah... There are several problems. It also raises the humidity of the room it's in which... You wouldn't want that around electronics. And you generally want to use distilled water around electronics, which means using normal tap isn't the best idea.
Now, I think on a larger scale this could work, if you pumped the humid air outside, that would be a big plus...
Not quite sure how to fix the dirt problem... Filters would help, but also decrease airflow... The growing stuff problem would be somewhatly combatted by using distilled water, you could also treat it somehow...
Yeah, everything's pointing to it needing periodic maintenance, even in the best cases. Maybe looking at AC units would help though, they don't tend to use evaporative cooling.
(But I think that might just lead you back to using the standard closed loop, reservoir setup.)
As somebody who works on industrial cooling towers, a few pointers.
1: Make the basin conical from the pump inlet to the side holes and make it deeper. You will lose less water that way and allow for more water pressure before the pump (it should run better.)
2: Design the bottom end side walls to have louvers rather than holes. They just need to be \\// shaped, not mobile, but they will allow air in while also keeping more water in than just open holes.
3: Make it square and use two fans. The efficiency isn't much different per fan and you can then build in supports and mount it to something so it is more stable.
4: You will be constantly fighting water evaporation. My system normally needs enough water to completely refill the system every few days. Would make a neat dual purpose humidifier in the winter though. The real issue is you are going to "scale up" the system with crap in your water over time and destroy the pump/tubing/fan.
I really do not recommend using a system like this for a prolonged period of time, you will destroy your components. Neat thought experiment though.
please make a refined version cuz it was dope
I wonder how it would performed in Metal. Then the metal would act as a radiator too. Could be more efficient.
And could it be sized down to work equally as well as the control test he did while being less size intrusive
@@vinade2100 Actually wouldn't really help, evaporation causes it to lose so much more energy than direct cooling that it's irrelvant. It would just be a bigger heatsink with only a tiny increase in cooling capacity while increasing complexity and production cost.
Nuclear power plant inspired of course.
if your pc outputs 200 watts of heat then this design would end up losing 0.3 liters of water per hour which is about a third meaning that you would need to have a larger reservoir for your water and only feed in distiled water.
This reminds me of the old homemade evaporative "bong" coolers, that overclockers built in the early 2000's.
You should check them out!
This was exactly my first thought.
those are what I think of when I hear cooling tower under the topic of PC water cooling.
Functionally this is the smallest evaporation cooler I've seen slapped onto a cooling loop. So, not shocked it works, am somewhat surprised at the performance to size...
When people were starting out with water cooling in the Athlon days, I have seen some really goofy setups with evap coolers which were HUGE, like 6 foot long, 8" wide sewage pipes fed by a shower head, just to see how far down they could get the temps.
@@Stoney3K ah, yes. The "Bong" series coolers. 😊👍
@@Stoney3K if you lived in a dry area they were great too since the drier the air the more effective they work. You could get REALLY low temps if you sized them right.
the reason for making it so big i think was to get as low of temps as possible and also having a lot of mass to make sure you couldnt heat soak the setup and to give more time between topping off the res.
Didn't Linus build something like this but completely screw it up?
@@KiraSlith He did it and it worked well, but he shat all over it for being inconvenient. What a weirdo.
Noctua Meeting: see we should have to hire him as we still could afford to hire him.
Noctua Meeting next year: "so how close are we now to having him buy us out?"
the corsair burn at the end lol
i like that you caught it
@@MajorHardware 🤣
Just need a dig at Jay over his comments on your channel next time 😉
When watercooling was getting big back in the late 90's, I experimented with something like this. It does work, as long as you can handle the drawbacks (water loss, humidity, cat attraction).
A few things you could do to possibility optimize performance is to swap the drip tower to have mist nozzles, they'll get the water into a much smaller size so they spread more effectively. You could also swap the 3d printed slats for Evap cooler media, it's usually a honeycomb matrix which spreads pretty effectively.
you just literally described a RTO system except with water. i guess you could call it a RTH like the cooling towers that bed media and have induction loops with negative air that humidify then dehumidify from gas to solid with ammonia/glycol to cool the water to right at freezing. i'm a millwright and have built many and that's an awesome idea sir.
Wait, cat attraction isn't a plus?
Misting requires more pump pressure, maybe too much.
@@keantoken6433 Perhaps upgrading to a fountain pump? Though that means an independent power supply.
@@keantoken6433 A LOT more pump pressure. IIRC, our system ran at 1,200PSI. Though we were using fogger nozzles.
Oh, prediction: works, no throttling, about as good as a 120mm aio, depending on the pump speed (and pressure) 😁
And just a little mess.
*edit: watched the video, wow.... I did not think that it is that good. Just wow.
Imagine with optimized sprayer nozzle.
@@jacobrzeszewski6527 Finer nozzle and a different fan (e.g. Noctua NF-A14 IPPC-2000 IP67) could easily double the cooling power. The only problem I see are the nozzles, because they need more pressure (> 3 bar)
@@SuppenDfg shouldn't be a big problem, the pump can definitively do that. There have to be enough to not impede flow overall or stall the pump. And too many will result in no spray at all...
@@SuppenDfg Required nozzle pressure is just a function of how big the hole is and how much water the pump moves per minutes. Everything else is flavor.
Point is, just get finer nozzles, the REALLY fine ones are usually called "Misters", as they produce more of a mist than a spray of droplets, that would actually be better for cooling as each tiny water droplet is its own surface area for evaporation. Although he will need to protect the fan from the mist getting up there and making rain on his room.
I would just make the tower out of wood (rectangular) and paint it to seal it from the water and have it have an expanded reservoir. Sponge or air filter for infill (your 3d printed ones would last longer) and forced draft to protect fan from high humidity/splashing. Wouldn't take days to make.
You are missing huge cooling potential with nozzles not spraying water into fine mist! If you would have nozzles that spays water into fine mist, it will cool even better (much much better). You can also use nozzles bought from the store, because it sprays water better :) Please make another better cooling tower, the performance will be astonishing! :) Love you channel, keep up the great work.
Huh. A friend of mine interviewed at a company where their data center was on-site, at least some their machines were water cooled, and they had a fountain in the lobby that was part of the water cooling system. Seems like the that design was likely also inspired by the same type of design.
This kind of cooling is used in some data centers indeed, I remember an article about a new data center a few years ago that used a very similar system.
Yup, a cooling tower, especially a draft cooling tower, is more efficient than traditional HVAC, so they get used industrially all the time. You don't tend to find them in places with high humidity however, because they're much less efficient when they can't evaporate water.
In a data center, the cooling tower is cooling water that cools the air that the air handlers use. They're not water cooling the servers directly unless they're running something custom.
@@jttech44 time.com/5814276/google-data-centers-water/
Yeah I've worked in a place where the server racks were force fed the cooled air from central water cooling towers. Definitely industrial scale cooling
There are data centers cooled entirely with evaporative cooling.
i love your willingness to experiment, and how you show the full results of trial and error. failure is part of the process, and not a negative thing. you're learning more by doing these experiments than most pay to learn from textbooks. keep up the great work! love seeing your chronicled adventures in cooling. i think the larger lip around the bottom would be helpful, and maybe some other way to re-capture the evaporated liquid! :D
"I guess Corsair does this." LOL Nice one.
I love how you went into this full of doubt and instead managed to make the the best size/performance bong cooler ever
Time to break out some PVC from the hardware store and make a BIG one.
Then using a boxfan instead of a A12X25
and have a sensor that tops the loop ? and win.
I would really advise dumping the exhaust out the window in that case because if you don't, it will rain from your *ceiling* sooner or later.
LOL so true!
The other way would be to close the air loop. But that would require a large air cooler to act as a recondensor to collect the water back out of the air loop and reuse it.
@@Bearthedancingman The point is though, that if you recondense the water it will transfer the thermal energy back into the loop, causing the evaporative cooling to have no effect.
It only works because it is an open system.
"Thats where Clouds are made, Guranteed!" Awesome thinking as a kid!
well, I mean he was partly correct so kudos to his kid self.
Hehe, I remember building "bong" coolers literally 20 years ago. We were actually discussing these just a month ago on HardForum.
Kyle? I remember looking for porche radiators. After all the workcooling stuff you used to post. It really got me into of hardware. Do you think you could make a video of the techniques of the bad old days?
Nothing like some shower heads, some pvc and delta server fans
Legit hardforums got me into overclocking and water cooling when i was 13 almost 20 years ago ;)
I remember the HardOCP tv videos, That's what made me delid my 7700k and oc it to 5.1
Who the fuck are you and what is hardforum
Sometimes “almost as good” is as good as you can hope for. Then it becomes a happy surprise when it turns out to be “better”! And a triple-fan radiator is no laughing matter either!
This is genuinely a pioneering success in the PC cooling sector, a proof-of-concept of something so ridiculous that it *should* fail and a new road to map!
*BUT* we need to push this idea further! Combine the radiator with this cooling tower, such that the water first goes through the triple-fan radiator *before* being funneled into the tower!
With a cold water connection that automatically tops up the supply and flow sensors at the input and output to turn off the pump when an overflow may occur, it might work reliably enough.
@@starbase218 U could Put one or maybe two into an old fishtank, so its a closed system, the water evaporates and then it starts to condensate at the glas and flows Back in the system. Problem here is U need a huge amount of water, the inlet is at the top of the tank, but the outlet needs to pick up the water which condensated and flowed down the glas. So there needs at least 1 inch water at the bottom. A real proof of concept could take 2 hours and More until the water is heated up. You could fasten it and putting preheated water in it and measure the watertemp and if it stabilises while Stresstesting at a certain temp it shows the system is working. There is another point. When water condensate it sets heat free and the question is, is the outer surface of the glas big enough to transfer it to the Air of the room. Because thats essentially how cooling works. You Transfer the heat from the source to the air, whether its a aircooler Just with air or a watercooler combining water and air.
And sorry for my Kind of Bad english, i am from Germany and its Not my mother tongue 😂
Or maybe this could be the cooling solution for massive datacenters?
As for evaporative cooling, the warmer the water is as it goes through the tower, the more effective the tower is. It would be best to pipe from the tower, to the radiator, then back to the cpu. One HUGE downside to these systems is bacterial growth. Open loops near body temperature are a perfect breeding ground for nasty slime and smells. Not to mention life threatening in some instances.
Did I spy a Hitchhikers guide reference?
*edit* My mind is legitimately blown by the results.
I honestly didn't think this would work this well at this scale
Im new to your channel, just found it yesterday, great stuff. I love the free and open expression of ideas and experimenting. I would love to see a series on improving a standard 20x20 box fan, with blade improvement, and even inserts to improve air flow. We all use these box fans, but face it the tech on it is ancient. Just a suggestion.
I love how you test all the wildest stuffs. This reminds me of Tech Ingredients Revolutionary Air Conditioner video, but he actually uses it to cool his house. maybe you can use that to get some better ideas on new designs, or even reach out maybe get some new ideas what could be tried :)
Yeah, that's science, man. Think, try, rethink, retry. Some people just think better or try harder than others.
@@blackpete there really is no failure in science. only data.
Thanks for this. Never seen those vids before.
whats interesting though is this cooler design is old school as hell.
@@Sc1Z you mean elaborate ways to collect data, lol.
This is just so damn cool. The way you combine computers, 3D printing, and engineering is one of my favorite things
That look at about 6:12 to make sure the camera is still working is great 🤣
Also, really surprised by the results! If you were to take actual spray nozzles and put them in instead of the pla printed ones, you'd get even better results!
I've notice bang good do some really fine nozzles cheap designed for fitting to a gazebo/parasol to cool you in the garden. Imagine a few of those used instead to make an ultra fine mist, imagine the surface area involved there & how much cooling that could provide.
Of course it's working. This system is an a/c that can go below ambient temperatures. You could improve it by controlling the humidity levels with salts.
14:44 you should buy a 3D pen for fix leak or emergency.
Yeah, i have used 3d pen to touch up alot of prints, i also printed a lvl3 helmet from PUBG in many parts, and "welded" it together with a 3d pen. Works really good :-P
Depending the materials, you could also do welds with heat (soldering iron) or chemical (e.g. ABS -> Acetone). Both work great if you're going for functional strength over beauty.
@@kylek29 or hotglue
@@kylek29 I only used it on PLA. I find that hotglue is not sticking all that well, i mean it sticks, but it will easely rip off.. 🤔 Might be that i used a bad quality tho.
I've been using 3D pens my whole life..is the alternative to which you're seemingly all accustomed a 2D pen? Better yet, 4D? Please link to suppliers either way, the former for amusement/novelty; latter for sheer superpower status!
Kidding aside, UV-reactive photopolymer pens are the best tool for nerds like me since the 3D printer.
This is one of the coolest things that I have ever seen, and easily the most creative and interesting alternative cooling solutions that I have ever seen. I've watched quite a few of your other videos lately and you're doing more interesting and creative stuff than even LTT, and they have an entire metal working shop, laser cutters, CNC routers, etc.
Fucking kudos man.
You can 3d print a large diameter concave or cone shaped lid that floats a few inches above the fan that collects the water droplets around its rim and redirects the water down a tube and back into the reservoir. This will help reduce the added humidity in the room and reduce the amount of times you need to top it off with new water .
That would be counterproductive, the heat that was absorbed via evaporation would be put back into the loop.
@@SuppenDfg I think he is only talking about catching the liquid droplets that have not evaporated, but slipped through without phase changing. The video did touch on this when he said his design didn't have a 'drift eliminator'. Some coarse foam below the fan would do the trick too
Conformal coat the electronics and steel in the fan to prevent corrosion.
@@SuppenDfg The heat removed from the system to evaporate the water must itself be lost from the water in order for it to condensate and be returned to the system, so it wouldn't be counterproductive, it would just reduce the need for water to be added.
The efficiency could be even further boosted by using a condensation tower (essentially mirror the cooling tower across the fan) instead of just adding a simple hat to it.
@@GremlinSciences The vapor will loose its heat partially to the surrounding air, but also to the water droplets building up on the "hat" so that when the droplets finally fall that water will be above ambient temperature. I'm not sure how much heat will go to the air and how much to the water, so it's quite possible that it still works, but imho it should be less efficient.
This is the best video of yours I've seen yet, great I'd love more cooling tower experiments!
This is definitely a great idea. I could easily see an improved version of this with finer misters, correct seals, and a way to have it detect when it needs more water to add in more could be amazing.
an elevated standby "tank" with a float valve should do the trick
@@danieljenkins4213 seems easier to implement that my thoughts of an anduino board with a program to take in the humidity and watwr temps to calculate when it needs more in. I knew it was overly complex and had to be an easier way.
@@Speedlution would still have to refill the secondary tank, just less often
It'd also need a UV sterilizer to keep algee and Bacteria at Bay. Might as well go full UV reactive dye at that point 👍
I don't see a lot of comments about the water hardness as time goes on. The same reason why people with fish tanks don't just endlessly add water: Evaporation removes water but leaves a lot of the mineral deposits. You'd need to completely change out water to prevent massive calcium buildup over time. Using distilled water would help I guess, but I think it's still a thing.
I am an engineer and stumbled on your video. Im impressed with your build. You have proven that open (evaporative) cooling is far superior to closed cooling. This is why datacenters use the same principle. You have humidified the air. Now you will have to worry about legionella in your computer room.
One problem I see with this:
It will introduce dust from the air into the water ... pumped around the loop, it could clog small canals in the water-blocks.
This is an actual problem with water towers. There needs to be a sump outflow to remove dirty water to prevent the accumulation of dirt in the system. So the amount of clean water added back needs to replace both the evaporated water and the dirty water lost.
Yeah needs a filter over the fan or an inline filter like for fuel/oil in cars. Otherwise you’ll need to dump the loop every month or so and occasionally clean the block (I had my loop in a bucket, eventually there was so much hair because I shed about like a cat, that the blocks needed to be cleaned every half year, but it was an open bucket next to my feet, lol!)
@@jakegarrett8109 Nobody wants to disassemble, clean and reassemble their entire system every 4-6 months.
And I doubt an air-filter will be effective enough, unless it is so tight, that air-flow is heavily impeded as well, lowering the overall evaporation effect.
But hey, I'm just guessing ... havn't tried :)
maybe there could be two loops connected with a water-water heat exchanger
maybe there could be two loops connected with a water-water heat exchanger
Awesome design ! Super impressed by the results, it's amazing ... Congrats mate
20:35
“But I guess Corsair does it” Ouuuuf, shots fired 😂
I enjoyed the video! I appreciate how they always take us along for the trial and error process, including failures and mistakes too. That corsair jab got me good too lol...Keep at it!
I think it’s soon time to make a custom pc case with the 3D printer...
I've done this... One of the best things I've ever made. However, depending on your design (mine is based on 180x180mm interlocking panels), be prepared for an extremely long amount of print time. My case, which was designed around an ATX Mobo took over a month to print. Some of the more complicated panels (ie the power supply mount) took almost 20h alone. The average print time for a panel was 9h. All told I had 24 panels to print, and it took roughly 225h total to print. If you really skeletonize the design you could certainly bring down the time but be careful, it still has to support the weight of your parts.
It could even integrate this cooling design into the case!
Maybe a space optimized sub 7L case and itx desgin
@@jonmcentire I would have loved a 3D printer for when I made my brother a PC case, his tho is wood and metal, looks pretty awesome if I may "humbly" say so. Adding to the complexity, my tools at disposal were only hand tools and the simple power tools you can have at, well, a small apartment (like a jigsaw and drill. Rest all hand done).
The back plate where the PCIe slots, fans, power supply and I/O shield mount to (it's a cube style case by the way) were quite the task to cut with those tools let me tell ya.
In the end, the result turned out incredibly precise, no need for feet leveling even for example. Still quite proud of the end result. There's something very different about a completely custom design that is just hard to match by spending any money for something already made.
Please, please, please do some follow up videos with improved nozzle spray, variations in infill, rates of water consumption, and...series and parallel setup of towers! Really love the channel, thanks for bringing us along on the explorations.
You should add a "catch" system to recover the water lost out of the top due to evaporation. I'm imagining something similar to a old-fashion moonshine still positioned above the fan.
The point is to NOT catch water back as it would get the heat removed from the loop through evaporation back in the system.
@starshipeleven Just need to mirror the baffles from the cooling tower to slow down the steam and catch the condensation. The droplets will still be fairly warm, but they can just be ran back through the tower or let flow through a length of pipe before being fed back into the system.
Water is very good at stripping away heat and maintaining temperature, but that only actually applies until it evaporates. Once water evaporates, it's no better than regular air at both, force it it move or change direction and it will lose a large chunk of it's thermal energy.
This is an amazing channel, this is the most elegant cooling tower I've see for a PC application. I would love to see this design refined
It's basically a variation of a "bong" cooler.
This was actually pretty cool! (Bad pun.) Cool your computer with a humidifier. LOL For a proof-of-concept model it did surprisingly well. I can't wait to see more refinements. Besides the obvi, an automatic feeder tank to maintain water volume would be nifty. I also can't help but wonder how much the plastic itself is a hindrance, where some stone might help. Maybe print some molds to fill with cement to make the tower part?
dude imagine if you made one out of some decent thermo conductive material like aluminum or similar, with tight tolerances and proper orings and gaskets, I would honestly think about swapping to test bench to run one permanently.
It could be made of gold and wouldn't make any difference this cooler works by evaporating the water with the air flow through it. About the only thing that could improve this is a larger surface area mixing zone. A honey comb design maybe, or pack it full of coarse stainless steel (no rust) or plastic brillo pads... you want as much interaction between the water and the air as possible.
Probably one thing that would have to be included at some point is a set of air filters to keep room dust out of the interface material, and ultimately out of the cooling water so it doesn't gum up the pump and cooler(s)
This used to be a thing 15 years ago.
Except we didn't print our towers but used drain pipes instead.
Back in the days the were called "bong coolers" because they looked like a bong.
I used to have one and it performed really well but as you've also pointed out water loss was a huge pain in the ass, I ruined my pump because I forgot to refill it.
That being said I still enjoyed your video!
You've gained yourself a new subscriber!
Kudos mate!
I'd like you to check out Linus' whole room water cooling set up he tried, and see if you can do better.
Probably. That was a mess.
Probably not. I don't think that he has 10 workstations :P
One problem with that setup was the pipe was long it dissipated a lot of heat in to the room
@@jeremytrees7266 well, ya. linus isnt the brightest candle on the cake XD
I saw that video series and noticed one big problem, he was forcing all the water through the computers where he would have been better off by running at the end of loop a reduced cross connection to the return so that it was a full flow and the computers pulled water that was potentially cooler by the free flow design and not possibly blowing up computers.
Dude friggin epic
Congrats on the PC/Personal Room Humidifier. I used to work In a Thermal Electric Power Plant and worked around all kind of towers.
Id suggest adding multiple tiers of shallow convex and concave trays to disperse the water evenly as it cascades down to decrease the flow from the Nozzle to the reservoir. Making the tower as tall as a PC would allow for more trays.
One more thing find a pattern for a transitional segment that goes from round to square with round corners and stretch it out to bout 18" tall that'll make the design structurally aerodynamic. Try catching the evaporated water and condense it!
Looks like an offbrand Dyson 😁
When you order a Dyson on wish..
Man, shush! Dyson can hear us! They'll steal all the ideas... 😉
okay so i found it really dope BUT your evaporation is still quite limited and the evaporation actually will cool more air than water. I think that replacing your plastic grill with some cheap radiators (something not too dense) would allow for way better thermals : The radiator being the hot element and the water droplets being in direct contact with it , it should pretty much instantly evaporate thus cooling the water passing through the radiator. For even better evaporation you could make the spray nozzles even a bit smaller and use a pump with more pressure.
The smaller the droplets , the more the heat to mass ratio will be big and the more cooling you should get out of it.
This design should also allow for a much cleaner setup as you would have two separate water loops (yeah adding more complexity but it could be really worth the try imho) : one "classic" system loop and one "water tower" loop that cools a radiator. That way if there is a mess to be made it will be far away from your expensive system.
Are your sure about that. My gut feeling tells me that the water is "cooled" more than the air at least if the droplets arent superfine and evaporate almost completely. For the reason why: If water and air would have a temperature difference there would be a heat exchange until the temperatures are balanced, but to get there the water would have to pass on much more thermal energy to the evaporation process than the air because of it's bigger thermal capacity. I might be wrong and if that's the case I would love if someone could explain to me why.
20:40 for corsair drenching, neat concept!
Awesome! I have seen some huge bong coolers, but they did not do well with water distribution. An easy modification for your setup is to not remove the radiator. Use the closed air heat transfer to bring the loop water to near room temp, then run it into the tower. you will have cooler temps, and much less evaporation to top-off.
Some performance parameters relating to cooling towers would be cool as well, the main being what is the temp of the water compared to 'wet-bulb' temperature (the coldest possible using evaporation) This would give an idea about if the tower is over or under sized. The wet-bulb temperature is a function of dry-bulb (regular/normal) temperature and the relative humidity.
These style coolers are very efficient in Arizona where it's dry :) To determine the wet-bulb it's easiest to look at a psychrometric chart, or use one of the many calculators google links to.
Don't forget to bring a towel!
Love the South Park reference!
You're a towel.
It's essentially a swamp cooler. I used to use those all the time in Richland. They work great in dry climates. This is a really fun concept. I'd like to see an entire case built around this design. I bet it would turn out to be an extremely cool and eye catching case.
Tech Ingredients showed that you could make solar powered cooling, with similar findings.
th-cam.com/video/7w4rg3UcsgI/w-d-xo.html
Man, I really dig your channel. Just how much 3d printing opens your world for building, diy, testing ... Just everything
Do American non-nuclear power stations not have cooling towers?
Nope, we make our power from pure Chad Energy
Nope. They use rivers.
Some do and some don''t. Same with the nuclear plants.
Nice job man very impressive :) you got yourself a subscribe. Although you gotta vent that humid air somewhere, The more humidity goes up, they less cooling you get through evaporation. Eventually you run the risk of making your room too humid, or making your air conditioner work harder.
My expectations were like: Yeah, should be able to exceed the Intel stock cooler
Love the video, but I can see a couple problems down the line with this setup.
1 if you ever run out of water, you might damage your computer. 2 you can pick up dust and other debris in your loop which may eventually clog up the system leading to failure.
Thought occurs, make 2 loops. Primary loop can go to a cheap $30 120mm radiator that's mounted inside the cooling tower, and connects to your CPU/GPU/whatever else. This is a closed loop with no danger of running out of coolant, or getting external debris.
Secondary loop is the sprinkler/tower, reservoir, and another pump. Water is dropped from the sprinkler as normal over the diffuser plates and it drops down into the reservoir below the Colosseum.
From there, theres a few different ideas of whether to mount the radiator. 1 spot would be just above the Colosseum, below the diffuser plates (maybe omit the diffuser plates?). Second spot would be just below the Colosseum at the mouth of the reservoir. 3rd would be at the bottom of the reservoir, such that any water being pumped up into the tower, needs to be drained through the radiator (and thus have good fluid flow through it).
Personally I don't own a watercolor setup. But depending on the next couple months (and what they decide the second stimulus check is going to be), I might splurge it on a computer with watercooling, and get a 120mm radiator on the side to experiment with.
I have a pro-tip for you instead of hot glue that I use all the time with my 3d prints that need some touch ups, additions, etc; get a 3D printing pen from amazon or something, and use that to weld the two 3d printed parts together. That way you have a strong, waterproof joint.
Make sure to use basically the hottest temperature possible, because you don't want the heat going too deep, or you'll warp the part, or break through the skin into the infill, so keep kind of a light tough. Also, don't extrude hardly any extra plastic into the joint, or you'll not weld it, you'll basically just glue the two pieces together very poorly. just fuse the existing plastic together.
The only downside is that only edges can be welded together, so two large faces butted against each other will only be held at the edges.
There's a good reason why Cooling Towers have been used for High-Rise Buildings for over 80 years, they work really well.
They are not without drawbacks though, there Cooling Capacity is directly related to the Relative Humidity of the Air,
so they work really well in a Dry Climate, and just mehhh in a Humid Climate.
They also concentrate water born minerals at a ridiculous rate , which must eventually be flushed out, and/or, handled Chemically.
If you can wrap the air inlet port with some desiccant or something you will get more efficiently out of it. Evaporation is the means of heat transfer. So, the drier the air coming in, the more moisture it will except. End result is more cooling. Room may get pretty humid after a while which will cause efficiency loss obviously. I really like what you are doing here. Keep up the good work!
Bong water cooling was done years ago with water cooling setups on systems. The ones I recall were about 6 feet in height. As I recall, a shower head, a water pump and a fan mounted in roughly 6 inch pvc. Evaporation was one of the main issues.
Congrats on it working. I got excited for a minute but then I realized this would never work for me due to humidity already being insanely high in the summer. I struggle to keep my room below 70%, with something like this adding even more water to the air it would be horrible for all my electronics.
I love your channel. You're like a lovely blend of bits of J2C and 3D Printing Nerd plus a willingness to look at all the round wheels in the world and think a little bit mad-scientist in the best possible ways.
Also I think you'll have a great time with your Prusa printer. I've had many happy years with mine (I have the original MK2) and only just run in to my first serious issue - some combination of nozzle-clog and the extruder gear chewing filament - so tomorrow I will be doing a tiny disassemble (removing hobbed extruder gear to clean it) and maybe swapping to a new nozzle if I have a spare.
CONGRATS ! U finally did better than good factory equipment. I love this channel keep it up.
one major reason it worked so well for you : your ambiant humidity is low.
Evaporation cooling highly depends on ambiant humidity to work : the lower your ambiant, the better the cooling; as the higher the ambiant, the less water can evaporate since ambiant air is already saturated
I live in Paris near a canal, and it's humid all the time. It can tell you that small evaporation towers aren't used much here, as their efficiency is really bad ;)
The concept reminds me of the old Zalman Reserator 1. The Zalman dissipated heat via fins on the reservoir, but as an added bonus it would evaporate water inside the reservoir too, you could tell because condensed water would be attached to the lid of the reservoir.
It was designed to provide purely passive water cooling, but I remember finding some tubing that would slip over the reserator and attaching a fan to the top of the tube. The fan ran dead slow, but provided a much more efficient method of removing the warm air away from the reserator fins and dropped the CPU and GPU (one water loop across both) by 10 degrees C.
Well this is why im subscribed! :D Good job! Really cool to see alternative cooling solutions :P
This is brilliant... WOW lol. I don't know how you don't have more subs and this channel isn't bigger. You could have just created (with community assistance) a new category of PC cooling.
You should make it like a hybrid cooler, here is why:
The evaporation is now cooling the inner parts of the cooling tower. This in turn will also cool the water flowing through it, but you're essentialy cooling parts you do not want to cool.
You want to go back to the "water cooling an air cooler", but not in a closed water loop this time, you just want some mist inside the air cooler. I see 2 tricky parts with this approach:
1. How do you protect the cpu and other electronics in the case against leaking water
2. How do you create a mist inside the air cooler
My idea is to create a box around the fins and fans of an air cooler, with an opening in the front and back for the air intake and exhaust air, these opening should not go all the way to the bottom in order to keep the water inside the box. It should probably be open on the top as well in order to place the fan(s) and a drain at the bottom seams useful. The mist should ideally be generated before the intake fan, this way the fan will help in distributing the mist through the fins of the cooler. A high fan speed should help preventing the mist from going inside the case, but you probably want to have pulses here:
- spray some water with an high fan speed
- stop spraying, lower the fan speed and let it evaporate
- repeat
Might even be interresting to place 2 intake fans on each end and let the air escap through the top.
If you can get this working, I think it should cool better than the water cooled air cooler as it's now an air cooler helped with evaporation
Not my original idea though, I've seen this big jaeggi coolers at a datacenter which basicly do this (but than outside and not having to deal with fitting in a computer case).
A very good demonstration of how powerful water evaporation is! Would love to see a continuation of this!
That's really cool. But I have to mention, that the cooling effect depends on how does the air in your room is. When it's wet, the cooling will get worse
I once tried a similar, but larger scale, version of this. I used about 4 feet of pvc pipe with a shower head at the top, a pvc angle joint at the bottom to mount the fan out of the way of the water flow (counterflow design), all draining into a 5 gal bucket which was the reservoir.
Pretty cool. The only issue with the concept is that unlike traditional water cooling where you just heat up the room because of the radiators, you're now adding quite a bit of humidity to the air on top of all that heat. Running something like this could get very uncomfortable unless you have good airflow/AC in the room. It also wouldn't work very well in humid climates. Evaporative cooling is fantastic for deserts, but your efficiency drops significantly as the humidity in the air increases
I really don't think this was"bad design" (certainly not horrible), just a small oversight in planning. The Your printed design already has ample venting, only needed a small step-back-and-observe to figure out a reasonable compromise of pressure escape route without too great a risk of catastrophic leakage. I think you did great man.
If I can make a small quality-of-life suggestion for these kinds of projects; make yourself of maybe ask EK for, some multiport reservoir tops. They tend to make experiments and mods just that bit extra comfortable. The dream for me, would be something like a handful of EKWBs large/tall POM multiport tops, completed with the reservoir end threading, o-ring and o-ring groove - and thats it. No ports drilled and tapped at all, just ready for DIY custom ports.
i feel like with some improvements this design would make an awesome desk cooler for some builds. have this sit outside the case and the tubes running into it. there are some special setups that this would be amazing. especially if it was machined instead of printed. imagine having an in-desk pc with this (made out of aluminum) sitting on the corner like a fountain. the only issue i could see is things getting into the water since its open. could cause some issues.
Choose materials carefully, copper or silver plating on surfaces with water contact will provide some antimicrobial effects, and further to that, distilled water with a germacide will also help.
Except against dust, you'll still need to open it up and wash/pressure wash the parts and change your fluids, might even want a heat exchanger just so the open loop fluids with dust and contamination never reaches the CPU cooler block on it's own closed loop. Intakes could have dust filters for air, but you don't want to limit the air flow too much
All and all a complicated mess but definitely a fun project, so long as you test it away from a PC most of the time XD
I wish I had your creativity and inventiveness - you have a really unique channel in the PC/Tech space. Great work dude! Although I think the market for this invention may be small... :)
You could do cooler which is made from radiators connected in cross shape for 4 rads or in line for 2 rads. They would be cooled by rotating whole tower and radiators would get wind to cool them (it would look similar to this from top -o- for 2 radiators )
Hey! Hi Major Hardware. For your issue related to the water evaporing and needing to refill overtime: This may seem dumb but making a system using a toilet valve so when the water reach his lvl it will stop filling back in. Which would just fix your evaporation issue.
Also the reason i believe this tower is doing so well is because the water is in direct contact with the rushing air, so the heat exchange is much more efficient than through an alluminium radiator
good proof of concept, The one thing that the power plants use is a heat exchanger to keep the power generating water or in their case steam separate from the cooling tower water. You could do an even further by using a closed loop like Jayztwocents did with his recent dry ice setup with the copper pipe in a bucket of cooling water. And BTW I say on the first go through your video of the quick pic change on your screen while you were talking about the results. I may use this idea with minor tweaks, thanks.
Would be super cool if this was integrated into a case where the hot exhaust air going out the top is what evaporates the water. I expect that would make this more efficient and you would still have a radiator on the front intake. The temperature it can theoretically get down to would be the wet bulb temperature, and based on the readings you have that would put the minimum temp around 17C. According to some online calculators, warmer air will increase that wet bulb temp by 2-4 degrees, but considering the water is also going to constantly be around 30-50 degrees warmer than that Idk how that would effect the change in temperature.
Damn dude. Mad freaking respect for your continuing pursuits of excellence.
Hot Glue and PLA actually go very, very well together. Helpful tip: If you make a mistake and need to remove the glue, spray isopropyl alcohol and slowly pick at the glue and it will lift off cleanly.
I expected it to work but not this well, it's one fan after all. That's impressive, good job.
You know how some people have little pump driven waterfalls in their garden because they like the sound and look of it?
I can totally imagine someone doing this in their home, but instead of a waterfall they have like a little miniature industrial area with a WORKING cooling tower! kinda like how people have miniature train landscapes, but instead it's industrial haha
I have an Idea for an expansion to this project:
The biggest problem I see with this setup (it´s a prototype, so design doesn´t count) is that you are losing so much water because of the evaporation. My Idea is that you could make the top part above the fins higher and maybe just put metal fins ( preferably copper) in there, so the water can at least partially recondense on them and flow back into the system, wich could solve the issue or at least give you more time before you have to refill (a bigger reservoir might also help there). For a prototype you could for example just make it thinner and higher and put an AMD stock heatsink a couple centimeters below the top fan, so the watervaporpasses through it and can partially recondense on it, or you could even stack multiple heatsinks. If you actually test this and it works I might even try to use it in a future build for myself :)
Well, we use this kind of cooling tower to cool down our water to condensate vapor in order to create some vacuum in our sugar factory. It works great for big water flow. For some smaller water flow, we have some kinds of rads+fans. You can even add water pulverisation on the rad to gain even more cooling potential (would love to see this one). You find the same concept just on another scale in PC as in industry. Pretty cool to see it applied to computer.
Jsyk PLA is really easy to weld together. Heat up an iron wire or nail and basically just squish it together. It welds and holds; I've used it in many of my prints.
Also something else is using filament as welding rod; catch the end on fire with a lighter and let the plastic drip rapidly onto your joint. It melts surrounding plastic to itself. Once you like where you're at, wet the plastic (I use my finger and a little water) to cool and set the plastic. Way better than hot glue.
I would love to see you build a float valve to install in the sump (basin) that you could hook to a bottle of water that when your water level drops from evaporation, the float opens and let's more water in to keep your pump primed and the system running smoothly
Working as refrigeration mechanic in radiator based plants and cooling tower based plants, i not surprised that it worked, the evaporation can cool the water to below ambient, though you built it nicely.
Zalman once made like a tall floor standing reservoir with fins on the sides. It was made to cool computers passively. The design could be improved for sure but the concept is solid.
Even after a week you will see that the mineral content of the water in the loop will increase a lot if you keep using this. Minerals will start to clump together like in any humidifier.
Phase change is one of the most effective methods of heat exchange. It's how we lose heat with sweat and it's how air conditioning works. In my opinion it would work better with finer mist/better spray of the water, a steeper funnel base, and maybe a different fan setup. I'd like to see more of a pressure fan than one set for flow as a lower pressure inside the tower could help shed more heat due to partial pressure evaporation. Thermodynamics is one hell of a topic to dive into.
One thing to keep in mind for anyone who decides to use this is that cooling towers that don't operate hot enough to sterilize themselves are breeding ground for mold and bacteria, so extra sanitation is required. Also, in a typical cooling tower, the cooling fluid is separated from the tower water either by a liquid-liquid heat exchanger or by putting the primary loop's radiator in the drip path so you still get cooling even if the tower dries up, the water is only an efficiency booster.
Succeding or failing, you are the most daring and adventurous techtuber of the century...
You should try channeling the air in from the top like top air-feed dripping atomizers do.
Would prevent the chance of spillage and allow refinements on the air pressure with an adjustable ring of some sort.
Not a bit surprised. Evaporation is the best cooling you can archive. There is a reason why the powerplants dealing with giga Watts use exactly this solution. It is messy but effective. When you put a pot filled with water on your stove even a small pot can drain the energy from a 2kW cooking plate with ease as long as you provide water. You need 2259W to evaporate 1g/sec at 100°C(boiling Water). The estimated evaporation is about 180ml water per hour to cool your computer. (based on 100W power dissipation) Would be interesting to see whether this number is near your practical experience.
Great video BTW.
Years ago I met up at Lanwar with a group of regulars of a CS server I played on. One of the guys brought a cooling tower made of a 5 gallon bucket, 4" PVC, pond pump, fan and a shower head. Granted this was Pentium 4, Athlon XP days but while gaming his CPU was barely over ambient. Was pretty cool I've always want to give it a try.
I have 2 suggestions to perhaps make this a little more efficient and safe (?? lol) .....
1. Instead of using a large bore on top of the reservoir as you did with the ring you built, I would just replicate the original cap but with a secondary hole and a "straw" that ran above the spray line. Less chance of major spillage if knocked over.
2. Having the fan pulling air and moisture through its blades and motor system will lead to electrical failure of the fan eventually. What about a smaller fan driving an air foil system at the top [a la dy$on] ? You could then also put a condensate cap above it to help recapture some of the evaporated water. It makes the print much more complex, but could save power and seriously increase the air volume. The added potential cooling capacity may allow for a GPU water block as well ...
As you pointed out, the funnel needs taller sides to reduce splash or perhaps increase the diameter of the funnel beyond the cooling tower to make adding water easier.
Love your projects !
so THIS is how quality content looks like! thanks for your videos! just started wahtching you today, and totally impressed me, doing something related to what I love (computers) but being creative and also completely different to the rest of YT channels,
Hey there, HVAC Tech here, just stumbled upon this and am not through with watching BUT using the water that runs through the cooling tower to then pump through your system is bound to result in gunk buildup over time. It'd be less efficient but using a secondary loop to cool water which in turn cools the actual cooling water helps with that issue. The whole idea also doubles as a humidifier but gets less effective the higher the humidity in your air.
great job. Remember, goofy ideas were the start of some of the greatest things