As an Aircon engineer, I'd suggest a few ways to make it work better. 1) Maximize the surface area per volume of fill material, you want as much surface area per unit volume to make it compact. This is probably some tournament worthy of another season. 2) Use a spray nozzles, the water droplets should be small to enhance heat transfer, use a mist eliminator if needed to avoid water droplets leaving through the fan. 3) Typical commercial cooling tower run at water-to-air ratio of 1.2 ~ 1.6 kg water /kg air. Try to build a relationships between a) Pump speed to water flow b) Fan Speed to airflow With the relationship known you can play around with water-to-air ratio to maximize your cooling performance.
I think the main problem will be the pump pressure, the use of a nozzle will be kinda hard i guess. As for better surface area in such a small space i would think about some kind of diagonal stair ramp similar to a weir, with the air coming througt the stairs. You could add some guidance slopes on the top to evenly distribute the waterflow. I realy like the apliance of practical physiks in PC builds, but dont forget that an open watersystem can damage your hardware due to corrosion.
Yeah I'm scratching my head thinking why he just didn't grab some random showerhead. And he could make the water stream hit stainless mesh, which is called strainer in the kitchen utensil section. Less cool, More practical. Also, air should go upstream to help the chimney effect. He could even use secondary aquarium pump in the reservoir to pump it through garden sprinkler nozzle if fine mist is needed for heat loss effect through evaporation.
Yep, pretty much everything said here. That along with getting rid of any of those convergers he has in the stack. Introducing things designed to reduce the surface area and thus cooling capacity just hinders overall performance.
The best way is to use a heat exchanger so that the cooling loop stays closed then a decent pump can be used to circulate water through the heat exchanger and cooling tower.
Pump water in the top. Then run it through a distillation column. Push room temperature air in the bottom. Remember To Filter Output because rooms are dusty! Unless you go for a HEPA or Paper filter with fan forcing.
That sounds like peltier cooling with extra steps. Its also a humidifier at this point. The cooled surfaces would remove more heat then the water to air cooling would.
@@LuxGamer16 I mean we aren't discussing efficient designs here. Off the shelf products do that. This feels like an exploration of unusual approaches. And peltier electric cooling has basically nothing to do with anything discussed here. So not sure where you're getting that from.
Well adhesion to cool surfaces would be condensation but for that to really work in any reasonable capacity you would kinda need a closed system, also you then need a separate way to cool down that surface for it to condense on.
@@connorjohnson4402 I meant more that when the mist dumps it's heat to the air, then it could condense on a surface. That surface being cooled, or at least prevented from heating up much, would probably help.
You should use these mist nozzles for garden hoses / shower heads. They should create tiny droplets and therefore a lot more surface area for evaporation. Also, to work with filters next time, you can try to have the fans at the bottom pushing up, with pressure-optimized fans, and just have filter material in front of the fans.
If the droplets are too fine then instead of evaporating they just blow away. It would definitely be more efficient at removing heat but also run out of water far quicker.
Just doing a shower head would create airflow going down. No need for a fan. You also could get a decorative fountain and just hook you’re pc up to that. Yes to filter but a water filter.
To expand on this, he could use one of these with a larger local res for thermal capacity in case the other (evaporative) cooling loop goes out or runs out of water; that could buy some time but makes the system less portable. With some QDs he could easily replace the exchanger with a radiator, however.
Brining that dust in your loop is the first problem I saw as well, however a liquid to liquid heat exchanger is basically a radiator, so that would defeat the point of the entire system....
@@henkheijmen Not if hes running dual loop; the point of the exchanger would be to keep the PC water clean while the water that does the actual cooling can be dirty without much worry. The evaporative loop would still handle all the cooling of the PC, it would just be doing it via an exchanger. Kind of like how nuclear reactors work with their dual loop setup; the turbines aren't driven with radioactive water from the core, after all.
If you were getting no sleep before, leave it to the community to find a way for you to get less. Great video, you really put in the yards with your work and I for one appreciate it.
I'd love to see this idea expanded on yet again. More fill density (I get what you mean about the aesthetic side, but I think you could definitely cram more in before it starts to look like a solid lump), spray nozzles (which may require a stronger or secondary pump), and maybe additional fans for improved airflow (could do 4 fans arranged to form a cube that exhausts out of the sides). Also if you didn't do it already, probably a good idea to seal all the gaps with silicone, just to ensure that the fans are pulling all of the air from the bottom, not though any small gaps. Honestly, I think this is an interesting enough idea to explore further. It'd be cool to see what it could achieve with all you've learned from previous versions.
ya I'm thinking the same way, I have worked on industrial towers and there is definitely a lot more room for fill. the fill is the important part. make 3 gaps for rgb and so you can see the water drops, the rest alternating fill. Also im wondering if a different fan style is required. Something like what is used in vacuums.
Make your connecting rings between the tubes also contain the rails on which the infill sits. Basically run three or four legs up and down from the connector rigs with ledges for the fill plates to sit on. You can also make the connector rings a fill plate in and of themselves. That way you have the connector ring is a fill plate Plus a fill plate both above and below each ring held in by those legends so they can't go anywhere.
I had a passively water-cooled gaming PC back in 2005ish. I think the cooler I used was the Zalman Reserator 1 V2, IIRC. It was basically a big external tower with fins running down the side to help dissipate heat. You should see if you can get one of those for when you decide to go for round three, I bet it could be fairly easily adapted to what you're trying to do.
We used to do this back in the late 90s early 2000. We called it "bong" cooling. Works great, creates a humid atmosphere in you room. If you put a return pump(s) on the side at an angle you can get a really good vortex going that will increase the heat exchange. also try a shower head at the top. You need to make it taller and get the water from the showerhead to spray so it evaporates.
it's a smaller version of well-known technology at this point, and it does scale much bigger than normal radiators can. These things are probably going to be mandatory in a few years for top end GPUs
@@Hawk7886 AIOs are already at the limit because of case size constraints, to do significantly better you would need an external radiator like a car radiator
@@Hawk7886 AIOs are already as efficient as they can be. The only way for them to get more powerful would be for them to get physically larger. While this might be what ends up happening, if you live in a world were water is cheap then the cooling tower approach is much more efficient in terms of heat extraction and energy consumed. The water loss from the system and the need for more maintenance are definite issues though. Keep in mind that the stock system was using mulitple fans and the tower used only one and did a better job of cooling by ever so small of a margin.
Tech Ingredients channel have a good PVC dual stage evaporative cooler you might want to look into. Also, you're getting a fair bit of cooling from evaporation, so it would be interesting to know what the difference in coolant volume is both before and after testing.
@@jtillman8251 To be honest - I suspect even filling it with coarse gravel would work better than those plates. It's nothing new, called random column packing, and probably the only feasible route. Because proper structured packing isn't available at such sizes afaik. One could just print a whole column of bare gyroid infill or some such though, that might give good results as well. If he revisits the idea he should probably just get some readily available Lessing, Bialecki or Dixon rings - those are quite similar to bio reactor packing in some regards, but both are kind of optimized for very different applications.
Tech Ingredients more interesting video in PC cooling is showing that a window unit A/C cooling a heat pipe system is better than a window unit cooling a water cooler. Heat pipes are much "faster" thermally than direct water cooling due to the vapor phase transition, and thus more efficient at using the greater delta-T of an air conditioner. As for the video premise, it's aiming for failure which is sadly too common on YT these days to drive engagement. The opposite approach ala a Zalman Reservator is a much better pivot in water cooling but has the flaw that it works out of the box -- recreate the Reservator on a much grander scale - say a 6-8 ft tall tower with a 1 1/2" or 2" copper pipe or aluminum tube with some fins. In a sub-1 sq ft footprint, it would outperform the stock fan based radiators due to water volume and surface area. For bonus cooling, stick it under an A/C vent during months the A/C is in use. Heck if you used the steel tank from a scrap water heater or propane tank, the coolant volume and vertical velocity (hot going in top) would be so low you don't even need fins or higher thermal conductivity and radiative metals like copper or aluminum involved.
Should get some of the metallic filament for the discs inside. Also you can try embedding magnets in them so you can adjust the positioning from the outside using more magnets.
Perhaps using raschig rings instead of those distribution plates would work better. It massively increases the surface area. You can easily find ceramic ones used for distilling. Or even print a whole load of them.
I've seen some resin printed metal amateur radio antenna on youtube that basically resembled bare infill & looks like it'd be almost perfect for this sort of thing.
@Major Hardware for me the whole point of this Vapor cooling is - simplicity and only problem is sound. Your approach fix what is working and do nothing with problems. You have basically 2 solutions for PC Vapor cooling -> A. Use shower towel on a simple horizontal PVC tube with holes to spread the water by soaking and use couple of fans for cooling, B. Use simple shower head to spread the water on fully open air. Your more industrial approach have place in a very large scale, when volume of water in the system and whole scale of the system should be optimized to a reasonable scale, but you still have a PC water cooling that makes you want to pee all the time. Also there is C. option, to vaporize and condense water with sound waves, like Ultra sound, this is how electric air humidifiers work.
Might as well just build a wall fountain, At least esthetically it would be a lot nicer but might need to make it fairly large to be effective, and you'd have to deal with dust and keeping track of the water level. For an alternative material to put in the column you could use the sheets of material that go in swamp coolers, its designed to do this exact thing and you ca buy it at the hardware store.
put a water level sensor in and monitor with arduino/ rpi to open a solenoid connected to a tap or larger reservoir to maintain water levels.. wouldn't want the system burning out when you forget to turn it of before a weekend away haha
Garden irrigation and cooling systems have a "misting" head that forms a fine spray. Combine that with the fan at the top, and you get something similar to a jet engine. Another option would have a water fountain at the base, and a series of condensation plates in the column. Water from the plates is recycled to the PC - looking a lot like a fractional distillation setup.
this actually could work for normal setup if you use a water to water heat exchanger to keep the evaporating water supply separate. Could be nice, like one of those desk fountains very tempting....
Definitely need *very* good airflow for this for sure. But, like, more airflows healthier anyway (assuming where you live isn’t horrifically polluted, and even then HEPA filters are magic), so🤷🏼♂️
I kept looking at this and thinking "why not just use a rain wall instead"? You could easily put a rain wall in an enclosure (like a fish tank or plexiglass box) and then add a filtered intake fan and a fan for exhaust. You'd get plenty of evaporative cooling, filtered air and a lot less parts.
You might be able to just pack the entire tube with fibrous material that would allow a huge amount of surface area and slow flow. Just maybe put a "shower head" at the top to distribute the paths through the material.
Not fibrous material, it'll just clog up. Use distillation / scrubbing column fill. Distillation columns are responsible for industry worth trillions, you better believe they've been optimized as much as humanly possible.
This reminds me somewhat of a bong chiller I made about 15 years ago that I had outside and plumbed into my house. The cooling results from a 8 foot bong chiller are phenomenal. :)
Hey love this concept has some merit for a TEC project I'm working on. Just a thought re your surface to water / air ratio, you might like to try a bunch of small plastic balls and just fill the towers with them. At a stab I would suggest approximately 15mm to 25mm diameter. Should aid the whole cooling process quite a bit 🙂.
I made one of these 20 years ago, but I went for attached flow to the walls of the pipe. I accomplished it by lining the walls of the pipe with one layer of cotton towels. Then I had a smaller pipe with a fan that went inside the big pipe, that way it could pull air down into the pipe and back up the smaller tube. It was quiet and cooled my AMD CPU. Downside is it used quite a bit of water and stuff would start to grown in the water, you might have to address those issues somehow.
Humidifiers will either be topped-off regularly or will be plumbed in so they can be refilled automatically. To prevent growth you can mix some biocide (e.g. copper sulfate) in with the water. Filter media in the water path also helps to collect any contaminants. Any option system should be emptied, flushed, and refilled periodically. Growth tends to happen more in standing water. If you can keep the water agitated (such as the churning of the water you see in the bottom of his reservoir), that should reduce the amount of growth. I have an open water fountain for my cats to drink from. It can run for several weeks without any obvious growth. It does contain a filter which is just cotton batting and activated charcoal.
That setup looks like it would benefit from quick disconnect fittings. That would also allow you to remove it from the loop and still use it as a bong. That way when you hook it back up to the PC, the bong water will circulate through the system and give you HIGH frame rates.
Love this unconventional water cooling.... could you try stackable plate HEX instead of radiators? I believe you'll need 2 loops (hot/cold) for the setup
Ah, this takes me back. I made a similar evaporative cooler with PVC plumbing pipe and scotch brite pads to hold the water and allow evaporation. I had it cooling a dual-CPU Athlon XP system. Had to refill it every day - it worked fairly well but wasn't that practical given how much water it went through. Yours is a lot more aesthetically-pleasing!
Use a power station cooling tower as a model, spray heads at the top with upflowing air. If the droplets are small enough, you don't need anything in the column. And just filter the water with an in line filter. You can also put a grid just above the bottom of the pool to encourage sedimentation
These used to be called "bong coolers" and were used for cpu overclocking back in the pentium 3 and early Athlon days. With a 50-60w cpu, the shower head and pvc pipe type could cool the water below ambient and thus allow for what in those days was extreme overclocking. They were also relatively cheap back when dedicated PC radiators were rare and expensive. I think by about 2002 or so CPUs used enough power that they were no longer used much.
Coolingwater specialist here: to improve your design the only thing you need to do is increase the surface area. You could do this by printing a circular lamellafilter, which fits in the full length of your clear acrylic tube. More surface area equals more evaporation, upto a certain point of course.
You can also filter the water by adding a 'resting pool' similar to Moors filtration technology used in Andalusia back in 15th (not sure when was that) century. It is basically a relatively large reservoir that has a certain depth and relatively high intake and outlet spots (for example 2cm radius ports work on top of a 14cm deep 25-30cm long pool). Idea is if you put your water flow thru a very deep reservoir, all the particles brought by the water will seep down and you can collect cleaned water from a high spot. Ideally you'll need a cleaning mechanism/scheme for the bottom end, if you make it open top like the original it will help with cooling and finally if you make it just right you can use it as a master waterfill/drainage reservoir as well. It'll take a lot of space, it'll be heavy and potentially messy but an absolute spectacle if you're up to tackle :)
Used to play with these in the early 2000's. What you want is as small dropplets as possible spread as uniformly as possible and you want them to free fall against the airflow as long as possible. Designs back then where otherwise quite simple and similar, just a collecting reservoir on the bottom with pipe to the PC and pretty long and wide tube (I used to have 1,5m 300mm PVC tube as the "rain tube"). The top part is the difference, we didn't go for direct airflow because we used the very top to diffuse the water (1-4 pipes from the PC pushing water to small PVC pipes at the very top with holes on the under side and under them a diffusing layer that was just a PVC cap with a ton of small holes through which the water just barely dropped down to the pipe). I used to have 2 fans in their own pipes separating from the rain pipe to upwards to the sides so water couldn't be pulled into the fans directly. The true art came from tuning the fans. In the best case the airflow needed to be just rightly distributed so the dropplets don't get flown around in the pipe and hit each other and grow in size, so as stable and directly down to up as possible. The second part was the force of the airflow, you wanted just enough that the dropplets don't fall straight down but they fall slower because the airflow because the longer the dropplets fall, the colder they get. I would recall some people who put way too much time to it managing to get the water returning to the PC being subambient temperatures. But the whole thing fell to the history once someone found out you could just use a small intercooler and strap fans to it instead of this huge rain tower and we got closer to the modern watercooling (we were still cooler because having the cooler within a case is such a limitation, just drill couple holes to the back, get rubber grommets for them and bring the cooling outside, you can even dunk the cooler to a fishtank and water cool your water cooler).
In the early 00's we just used 120mm PVC pipe with a Y piece on the bottom and an shower head on top, have the fan blow in air through the free arm of the split piece and you could achieve sub ambient water temperatures when running without load.
Loved the experiment! Also the constructive criticism of the flaws discovered on your first design . How about if you lowered the bowl down in the duct ? I think the neck of the fish bowl would be small enough for air flow .
One way or another, you have inspired me...But 15 years ago I had a similar idea....mine is simpler though...But it is based on genius physicists' discoveries...But I stopped pursuing the idea...I will make it happen this time because we haven't overcome such a thing. I commend you for trying for making something not just talking. I didn't exactly measure parameters, but I achieved a dramatic difference, just life as a networks admin and a horrible divorce got me off track...All to say don't let anything stop you from improving your design, everybody has unique ideas and you are on the right track, cheers.
I dont know why I like these somewhat overly complicated, intricate and/or unique water cooling ideas. Whether its due to the fact I want my pc to stand out a bit or just because its different I really want to go like way over the top with it one day and have a large tower like this that could sit next to the desk and look like its purely a decorative fountain. Although pumping could become an issue.
The key point here - hot media/cold air surface area size. You need to increase it at max possible level. So here is some strategies to reach it: 1. Make thin streams of water through small (up to 1...2mm nozzles). But it will may need pump with high enough compression (between PC and nozzles). 2. Make freefall water droplets. This will require more nozzles to achieve required flow. 3. (Most promise) Make your own radiator for controlled surface size at wide range of flow - push a water through an array of plastic tubes for drinks. It`s easy to calculate requried number of tubes for your application. Coz these tube has a very thin walls (0.15 - 0,30 microns) - high enough result heat conduction. Because it`s tube - easy streams and air control.
Missed (mist) opportunity!! 🤣 You could have put a fan inside, let the water strike the middle or the blades and it will be flung and misted , greatly increasing your evaporation and thus your cooling! Now you can have a fan blade showdown of water flinging and misting designs!!!! Probably a good idea to silicone conformal coat the center electronics.... And mount the fan hub up to keep some water out of the bearing But also be aware of the indoor humidity. As relative humidity rises during use, evaporation will become less and effective heat removal will reduce. So over time this will do a worse job unless you can keep your indoor humidity low and stable with outdoor air exchange.
Wow! Your own cooling tower ! If I may, there is two kind of cooling tower and yours is a little bit in the middle in the wrong way. You have free fall cooling tower and packed cooling tower. Free fall cooling tower have less drag (air easily pass through), but the surface between air and water is not so great. The packed cooling tower need strong fan as there is a lot of drag. But the cooling is also better inside as the surface between air and water is better. As you use a PC fan, I recommand you to: - redesign the head to have a better spraying of water. What you want is a shower of small droplets. - remove all packing inside. - replace the fan for a high pressure one. Beware that the cooling effect depend of the relative humidity of the room. Pretty project through. Have fun !
You should do a video on cleaning your loop, trying various things to flush it, and using a boroscope camera to get in up close to the parts we can't normally see, like down the fitting hole of a radiator. For instance, one of my ideas is to pour drain cleaner into the radiator, let it sit overnight, then flush with water until clear. Possibly other things as well, like Isopropyl alcohol and automotive paint solvents. Then move on to stuff that should be more compatible with the plastic and acrylic bits of your loop, once you think the radiator can't be cleaned any more.
Definitely don't use drain cleaner! Common drain cleaner is made of sodium hydroxide which will readily attack copper metal. Newer ones made with different chemicals may be safe for copper pipes - check the documentation carefully. Most acids, such as hydrochloric acid or vinegar (acetic acid) are safe.
You actually inspired me with the original video to build a similar system, I decided to use a sort of "rain" shower head I had lying around and 2 clear 5 gallon buckets for mine though.
AQUARIUM TOUR PLEASE! Is it saltwater? fresh water? cichlids? marine? i have been watching your vids and had NO idea you where a fish guy too! i must see! Please! Also, Cool tower cooler man!
I remember people making coolers kinda similar to this 15-20 years ago with the water getting slightly cooler than ambient. It was constructed from just plain old grey pipe and shower head at the top and a 45 degree piece in the middle facing downwards with the fan blowing air through it on the water dropping past it all the way from the top to the bottom without any middle grills or anything.
I very long time ago, I used I "Y" fitting on 4" pipe. 1.5' going down into a bucket and 3.5' going up in the air. Using a rain type of fitting for a kitchen sink at the top of the pipe and 3 92mm case fans stacked on top of each other connected at the "Y" fitting I was able to leverage evaporative cooling. This was all sitting in a 5 gallon bucket with a fountain pump cooling my Duron 900. Another way to increase your surface area is to use scotch bright pads(maybe you can find white ones), I actually enjoyed the rain cooler. Cool to see the same concept used twenty years later.
Also, filter the water, not the air. Being an aquarium keeper, you already know how to do so effectively, and you can't choke out a water filter by making it wet.
Fun idea. Use a window screen mesh to add surface area for the separation disks. Just mount them on top of the current stack disks. The finder mesh you use the more air pressure you will need. Possibly 2 fans on the top to pull out.
Very cool. I feel like this has potential to be quite effective with further refinement. It would be especially nice in places with dry air, since it would act almost like a desktop humidifier, and achieve better cooling via enhanced evaporation.
Tech Ingredients does a few videos on water-tower evaporative cooling. Almost the same idea with fans blowing air from the bottom up, into a shower head.
For a very long time I suffered from the same ideas and their implementation. 1 to increase the area from which evaporation takes place, I used a bunch of different straws for cocktails, in my case 6 mm showed the best result 2 geotextile was used to filter the air - it is very similar to the material of vacuum cleaner bags 3 I used sponges for aquariums to filter water, it also made the sound of falling water drops quieter if the water level was lower than the top of the sponge 4 I used a channel fan and in order to avoid contact with water, the air supply was from below 5 since the circuit is open, the tank needs a huge one so that all the water can drain there, well, you need a powerful pump so that you can pump everything back, or you need a check valve(near exit) that creates a decent resistance in the circuit
edge lighting the tube from the top down might look good, or having a lit up central acrylic tube with bubble (take apart a cheap garden bubble solar light)
You should print spacers to go between the elements to keep them flat/level. It would be awesome to see at least the top distribution plate have a rotating element, powered by the weight of the falling water. That would take some design work, but maybe a hydraulic bearing system would give an load capability on the bearing proportional to the water flow since it would power the bearing. I've seen a bearing/motor design in higher pressure systems, so I don't know how effective it would be here. But it might be worth a look.
Imagine hooking up a water system to your entire house, where you can plug in many devices with quick fittings so you can water cool any device anywhere in your home using a big fountain type bubbler like this? That would be such a cool thing to see someone do.
One way you can help with spreading the water is to have a gradual increase in pipe diameter when reaching the top of the tower. Second, you could use styrophome on the filter area so that even if it gets wet, it doesn't clog. Third, you could slow down the pump significantly because you gain all that extra volume in the tower. Basically it looked like you hade more water entering the chokepoints than leaving the choke points at the same time. Also you could try to change the inlet to the center of the fish bowl instead of on the side.
Instead of mesh discs, you could print "Pall rings" for random packing with a top and bottom retainer grate. Pall rings are used in commercial-scale for distilling and similar processes. Pall rings or other random/dumped packing would be a great way to distribute that water nicely. Combine that with a counter flow air path (air goes up, water falls down), you could probably get some more cooling mileage out of basically the same materials. you could use a y shaped branch to inject air down one leg of the y to go up the other leg of the y, in which the water is falling down. Then you wouldn't need to worry about getting the fan wet, but you could still get upwards/counterflow air. This liquid down, gas up counterflow is how it's done for the most part with chemical engineering.
I was thinking about doing something similar, before I realized I could use any of 3 legs in my old radiant heat lines to sink heat into the slab of concrete my house is built on. It was abandoned, in favor of baseboards years ago but the piping still exists! I used to work in a chlor-alkali plant, we did have chillers and a cooling tower but what you've got there is more similar to the chlorine stripper. We used small cylindrical packing material that had plenty of surface area each, packed into a column with air in from the bottom an liquid from the top, with ventilation. There was one disc at the bottom to keep the packing in. Airflow was priority, as was surface area for chemical stripping. These would, for your column have about a 1.5" diameter and length, easily 3D printed. Food for thought. I have also worked in both water and wastewater, so I have XP with flow and blinding. Maybe turn down flow and bump up air. The full column of packing material with a plate on the bottom with 1" holes will keep the packing in and allow for airflow without blinding (adjust water flow). I had also considered adding a float and a 1/4" fill line for makeup water with a float and a valve, you will lose water to evaporation (hopefully!). Looks promising, and cooling can be increased with a taller column/more packing easily. It opens the cooling loop(heating loop) as well so it's fairly ideal, just add water and air! You could even add ice to the basin to increase cooling for a bit too!
Excellent work man. I’m impressed. I’d rather use that than radiators any day. I mean hell, I’m running 2-360’s with push/pull configs so 12 fans on the rads alone lol. The sound of a waterfall is much more preferable and calming than 12 fans lol.
This video made me curious if you were the guy who put an EVO 212 or whatever it was in an acrylic case and water cooled it. And apparently yes, you were. Man it's cool to see how far you've come since then
I was just about to research how to do a more efficient evaporative cooler for a project of mine. I plan on integrating your fill design with my cooling towers. Instead of an upper fan try using an aquarium air pump in a lower section of the tower. This way you have a positive displacement system, relatively speaking compared to a fan. If you plan on sticking with a fan suction design, use a fan designed for radiators.
Was thinking about the CPU cooling. I was wondering if mist nozzle spray onto a finned block of sorts could put the phase change on CPU block... Granted the mental picture of everything is rather exotic. Imagine a smoke stack out the top of the case. 😂
In process engineering we would use a filler material to distribute the water like little glass sections of pipe or metal saddle shapes or washed pistachio shells. The ultimate fillers are Raschig Rings. The trick is to maximize interaction without blocking the flows of water and air. And be carefull to not contaminate your water if you try any material.
@@connorjohnson4402 The swamp cooler material is, afaik, meant for low water flow and large front area rather than depth of material. It might work fine, but I'd not be surprised if it is less than ideal.
An idea for a future version of the bong tower would be to use spacers for the fill instead of adhesive. You would lose some ID for airflow, but everything would stay put!
Love the innovation. You asked for a challenge? MAKE a PC case using multiple 500mm radiators stuck together (possibly welded?) for the top, side, front and bottom, so four out of the 6 case sides. Then get a regular rear "metal"/3D printed rear plate with cutouts for IO. And finally hang the motherboard mount on cables under tension, like the EVGA E1 PC case. I would imagine two length ways 500mm radiators to form the top, the same for the bottom and front, then five 500mm radiators to form the side of the case. A total of ELEVEN 500mm radiators, for what could be enough passive cooling to require no fans, just the pump to move the water through all the radiators. Alternatively, you could either find, or 3D print, a distro plate to form the front of the case, so only top+side+bottom were radiator walls, reducing the number of needed 500mm radiators to nine.
G'day Major, Along with the Fan Showdown I love these "caraaazzzy ideas" videos, it is Awesome the way you get everyone involved in the channel content, rather than just telling us "I did this". WOW! compared to a regular loop it is also really 'cool' & I guess if you like those calming fountains it would be great knowing your fountain also has a PC Purpose, but me even for the short running time 😬I had to stop the video to go for a bathroom stop😂, 🤔some thoughts to help longevity shorter clear tower tubes (same overall height but 6 or 9 instead of 3) with the Fill pieces integrated into the joiners, maybe a straight sided bowl at the bottom to hold the same amount of water but lower the Water level to prevent splashing the air filter.
A great follow-up would be a Tesla fountain, which many people gloss over before realizing its intent as a cooling tower for heat engines/boilers and as an efficient means for a fountain to heat the surroundings, or take heat energy from it and cool a closed system.
This was really cool, I love wonky stuff like this. Imagine a large sized one of these cooling a whole room of computers. You'd have to be dedicated to the water trickle sound though, unless you found some way to mitigate it where the water drops into the base below (Where I assume most the sound comes from)
From what a heard, there are Google data centers where that’s pretty much the case. Evaporative cooling can be cheaper then running air through the system (especially if your data center is in a desert, where the air is very hot). Course, there’s the issue of being in a *desert*, and the whole “water is precious in the desert” thing, *but*, it does exist!
If you wanted a passive external cooler, have you tried converting a Oil Heater and moving the oil through the water block and out of the case into a large Oil-filled Radiator Heater? Have the cool Oil flow back into the water block? Obviously you would get rid of the heating element since your CPU/GPU will provide the hot oil. A QD (Quick Detach) fittings on the outside of the case should make it easy enough to move around. And there are small Oil-Filled Heaters that have their own Caster Wheels
I once had the idea of using one of these epoxy river/ waterfall- desks as part of my watercooling loop. But it probably wouldnt have looked pretty for long with hot water steaming up. I did build an evopartive cooler from a small jar that had a showerhead inside and a fan on top. Worked good as a temporary setup but I had to fill up the water level once a day.
The fill is to break up the water into droplets of a certain size that maximizes a ratio of surface area to volume, so that the evaporation from a droplet cools that droplet maximally. Evaporating a film from a solid surface does nothing in a system like this. There needs to be droplets returning to the sump. The fill also slows its travel time from top to bottom, giving more time for evaporation.
Friend, take a look in "bong cooler". They place the fan in the bottom so the water vapor dont runs through the fan, which will surely reduce its life span.
A "rain" type shower head is what we used to use back in the 90's 🤣. Just a long 6 foot 8" PVC pipe ("reclaimed" from the junk yard for $5-10) with the shower head at the top, fan above, pump and all down below. Literally 30+ years ago this was how you beat the "heater core" people to ambient temperatures.
Creating stilts for your water plates and indentations for those stilts to seat into would make it so _zero_ glue is necessary, and if you use TPU to create gaskets you can have a _loose_ interference fit using the properties of shore hardness to your advantage so the water doesn't end up sneaking through the sides of the plate. A concavity flush with the plate and TPU seal would force the flow where you want it. If you make the seals be slightly _convex_ on both sides with a rather long and gentle fillet you can have the outside-facing seal press against the tube, while the plate is held into place. Material spread would keep the plate in place _under friction_ but the stilts would still be necessary for absolute placement so water flow does not displace them. The stilts should take advantage of the TPU material so they also sort of press-fit into place. The result would be a self-holding structure able to retain water with minimal loss - _if any_ - along the sides.
Did something like this around the year 2000.. Its the easiest way to get water below ambient temp. 3" pvc on top of a 5 gal bucket. Had a y flange at the bottom of the pipe by the bucket. Had one pusher fan. At the top I had a shower head. The pump was in the bucket. Just the air passing over the shower of water was enough to pull the heat out no need for all the in fill if the initial dispersion is good. After that I went to a water fountain setup to actively cool it down to around 35-40deg. Back around when the celeron 300a was around with a block I made that was. It was crude but worked like a champ.
A cooling tower works by evaporative cooling. You need a ton of surface area for the water to interact with the air flowing through it to get any cooling. You also need a spray system instead of just dumping water into it and trying to distribute it with a shower screen.
Long time follower, first time commenter 😅, after watching this video I had some suggestions for vas improvement for both air flow and water distribution. Have you considered trying the “print on fabric” technique with a mesh or micro fishnet? You could easily have stacked rings of various mesh and netting all from a simple trip to the screening section of a hardware store. All consider adding leg to your filters to act as guides and supports to build your tower. Cheers and great success!!!
Saw someone do this exact thing with white PVC piping and some fans like 20 years ago. I think his conclusion was, it works great for cooling, but falling water is loud, and it makes your room super humid.
As an Aircon engineer, I'd suggest a few ways to make it work better.
1) Maximize the surface area per volume of fill material, you want as much surface area per unit volume to make it compact.
This is probably some tournament worthy of another season.
2) Use a spray nozzles, the water droplets should be small to enhance heat transfer, use a mist eliminator if needed to avoid water droplets leaving through the fan.
3) Typical commercial cooling tower run at water-to-air ratio of 1.2 ~ 1.6 kg water /kg air.
Try to build a relationships between
a) Pump speed to water flow
b) Fan Speed to airflow
With the relationship known you can play around with water-to-air ratio to maximize your cooling performance.
I think the main problem will be the pump pressure, the use of a nozzle will be kinda hard i guess.
As for better surface area in such a small space i would think about some kind of diagonal stair ramp similar to a weir, with the air coming througt the stairs.
You could add some guidance slopes on the top to evenly distribute the waterflow.
I realy like the apliance of practical physiks in PC builds, but dont forget that an open watersystem can damage your hardware due to corrosion.
Yeah I'm scratching my head thinking why he just didn't grab some random showerhead. And he could make the water stream hit stainless mesh, which is called strainer in the kitchen utensil section. Less cool, More practical. Also, air should go upstream to help the chimney effect.
He could even use secondary aquarium pump in the reservoir to pump it through garden sprinkler nozzle if fine mist is needed for heat loss effect through evaporation.
Yep, pretty much everything said here. That along with getting rid of any of those convergers he has in the stack. Introducing things designed to reduce the surface area and thus cooling capacity just hinders overall performance.
The best way is to use a heat exchanger so that the cooling loop stays closed then a decent pump can be used to circulate water through the heat exchanger and cooling tower.
Wow cool
The bong tower
Mh smoking thru it. Now i'm interested. ;)
Beat me to it
The big bong
But it's not cracking
Not funny
Honestly, this is cool enough to be its own series
COOLER SHOWDOWN S6E99
Bog cooler showdown.
Yes, it is!
Design one that you can station outdoors to reduce the heat in your workspace during the summer.
I want this
Another idea might be to use an ultrasonic evaporator to force atomization and then recover moisture through adhesion to cooled surfaces.
Pump water in the top. Then run it through a distillation column.
Push room temperature air in the bottom. Remember To Filter Output because rooms are dusty!
Unless you go for a HEPA or Paper filter with fan forcing.
That sounds like peltier cooling with extra steps. Its also a humidifier at this point. The cooled surfaces would remove more heat then the water to air cooling would.
@@LuxGamer16 I mean we aren't discussing efficient designs here. Off the shelf products do that. This feels like an exploration of unusual approaches. And peltier electric cooling has basically nothing to do with anything discussed here. So not sure where you're getting that from.
Well adhesion to cool surfaces would be condensation but for that to really work in any reasonable capacity you would kinda need a closed system, also you then need a separate way to cool down that surface for it to condense on.
@@connorjohnson4402 I meant more that when the mist dumps it's heat to the air, then it could condense on a surface. That surface being cooled, or at least prevented from heating up much, would probably help.
You should use these mist nozzles for garden hoses / shower heads. They should create tiny droplets and therefore a lot more surface area for evaporation. Also, to work with filters next time, you can try to have the fans at the bottom pushing up, with pressure-optimized fans, and just have filter material in front of the fans.
this would require a lot more water pressure would it not?
@@ApolloUldaman yes, but I don't see why you wouldn't be able to run a separate pump that's capable of delivering that pressure.
If the droplets are too fine then instead of evaporating they just blow away. It would definitely be more efficient at removing heat but also run out of water far quicker.
Just doing a shower head would create airflow going down. No need for a fan. You also could get a decorative fountain and just hook you’re pc up to that. Yes to filter but a water filter.
@@LyK0sa the goal is mostly passive cooling
A liquid to liquid heat exchanger would be a really smart investment on this system so you don't bring that dust into the processor.
Or what industrial systems use: a silt trap
To expand on this, he could use one of these with a larger local res for thermal capacity in case the other (evaporative) cooling loop goes out or runs out of water; that could buy some time but makes the system less portable. With some QDs he could easily replace the exchanger with a radiator, however.
Brining that dust in your loop is the first problem I saw as well, however a liquid to liquid heat exchanger is basically a radiator, so that would defeat the point of the entire system....
@@henkheijmen exactly
@@henkheijmen Not if hes running dual loop; the point of the exchanger would be to keep the PC water clean while the water that does the actual cooling can be dirty without much worry. The evaporative loop would still handle all the cooling of the PC, it would just be doing it via an exchanger. Kind of like how nuclear reactors work with their dual loop setup; the turbines aren't driven with radioactive water from the core, after all.
If you were getting no sleep before, leave it to the community to find a way for you to get less.
Great video, you really put in the yards with your work and I for one appreciate it.
I'd love to see this idea expanded on yet again. More fill density (I get what you mean about the aesthetic side, but I think you could definitely cram more in before it starts to look like a solid lump), spray nozzles (which may require a stronger or secondary pump), and maybe additional fans for improved airflow (could do 4 fans arranged to form a cube that exhausts out of the sides).
Also if you didn't do it already, probably a good idea to seal all the gaps with silicone, just to ensure that the fans are pulling all of the air from the bottom, not though any small gaps.
Honestly, I think this is an interesting enough idea to explore further. It'd be cool to see what it could achieve with all you've learned from previous versions.
ya I'm thinking the same way, I have worked on industrial towers and there is definitely a lot more room for fill. the fill is the important part. make 3 gaps for rgb and so you can see the water drops, the rest alternating fill. Also im wondering if a different fan style is required. Something like what is used in vacuums.
Ever sense your first video over this tower cooler. It's stayed in my head for a very long time. Glad you have done another one
Make your connecting rings between the tubes also contain the rails on which the infill sits. Basically run three or four legs up and down from the connector rigs with ledges for the fill plates to sit on. You can also make the connector rings a fill plate in and of themselves. That way you have the connector ring is a fill plate Plus a fill plate both above and below each ring held in by those legends so they can't go anywhere.
Yeah, posts separate each filter.
I had a passively water-cooled gaming PC back in 2005ish. I think the cooler I used was the Zalman Reserator 1 V2, IIRC. It was basically a big external tower with fins running down the side to help dissipate heat. You should see if you can get one of those for when you decide to go for round three, I bet it could be fairly easily adapted to what you're trying to do.
Poor Zalman RIP.
The reserator was a cool piece of hardware.
@@epiccollision They no longer exist ?
I used to love their coolers.
can i just have like a 5 hour video of the water tower trickling? it's so soothing 😌
We used to do this back in the late 90s early 2000. We called it "bong" cooling. Works great, creates a humid atmosphere in you room. If you put a return pump(s) on the side at an angle you can get a really good vortex going that will increase the heat exchange. also try a shower head at the top. You need to make it taller and get the water from the showerhead to spray so it evaporates.
The simple fact that it worked is amazing to me. I'm sure with some work it could become a zen piece of furniture while fueling your epic gaming
it's a smaller version of well-known technology at this point, and it does scale much bigger than normal radiators can. These things are probably going to be mandatory in a few years for top end GPUs
@@marcogenovesi8570 uh... Not likely. You'd see more powerful AIOs before anything like this.
@@Hawk7886 AIOs are already at the limit because of case size constraints, to do significantly better you would need an external radiator like a car radiator
@@Hawk7886 AIOs are already as efficient as they can be. The only way for them to get more powerful would be for them to get physically larger. While this might be what ends up happening, if you live in a world were water is cheap then the cooling tower approach is much more efficient in terms of heat extraction and energy consumed. The water loss from the system and the need for more maintenance are definite issues though. Keep in mind that the stock system was using mulitple fans and the tower used only one and did a better job of cooling by ever so small of a margin.
@@T3ddyRuxp1n they're not even close to being "as efficient as can be" dude.
Tech Ingredients channel have a good PVC dual stage evaporative cooler you might want to look into.
Also, you're getting a fair bit of cooling from evaporation, so it would be interesting to know what the difference in coolant volume is both before and after testing.
I was also thinking of their solar air conditioner system as a basis for this sort of loop.
The bio-balls the guy in that video was using would probably work really well in this.
@@jtillman8251
To be honest - I suspect even filling it with coarse gravel would work better than those plates.
It's nothing new, called random column packing, and probably the only feasible route. Because proper structured packing isn't available at such sizes afaik.
One could just print a whole column of bare gyroid infill or some such though, that might give good results as well.
If he revisits the idea he should probably just get some readily available Lessing, Bialecki or Dixon rings - those are quite similar to bio reactor packing in some regards, but both are kind of optimized for very different applications.
Tech Ingredients more interesting video in PC cooling is showing that a window unit A/C cooling a heat pipe system is better than a window unit cooling a water cooler. Heat pipes are much "faster" thermally than direct water cooling due to the vapor phase transition, and thus more efficient at using the greater delta-T of an air conditioner.
As for the video premise, it's aiming for failure which is sadly too common on YT these days to drive engagement. The opposite approach ala a Zalman Reservator is a much better pivot in water cooling but has the flaw that it works out of the box -- recreate the Reservator on a much grander scale - say a 6-8 ft tall tower with a 1 1/2" or 2" copper pipe or aluminum tube with some fins. In a sub-1 sq ft footprint, it would outperform the stock fan based radiators due to water volume and surface area. For bonus cooling, stick it under an A/C vent during months the A/C is in use. Heck if you used the steel tank from a scrap water heater or propane tank, the coolant volume and vertical velocity (hot going in top) would be so low you don't even need fins or higher thermal conductivity and radiative metals like copper or aluminum involved.
Should get some of the metallic filament for the discs inside. Also you can try embedding magnets in them so you can adjust the positioning from the outside using more magnets.
Big brain!
Window screen
Not sure, but would metal-filled filament pose a corrosion risk in the loop with mixed metals?
Or just add some pockets to add magnets to both the inside fill and than add magnets on the outside like you said
@@xomm definitely, especially with how brittle those filaments are. They're basically metal powder mixed into plastic
Perhaps using raschig rings instead of those distribution plates would work better. It massively increases the surface area. You can easily find ceramic ones used for distilling. Or even print a whole load of them.
I've seen some resin printed metal amateur radio antenna on youtube that basically resembled bare infill & looks like it'd be almost perfect for this sort of thing.
@@NM-wd7kx One big slug of ramen infill would probably work great tbh
Might as well upgrade to Białecki rings
He would need to use a much more powerful fan for that
@Major Hardware for me the whole point of this Vapor cooling is - simplicity and only problem is sound. Your approach fix what is working and do nothing with problems. You have basically 2 solutions for PC Vapor cooling -> A. Use shower towel on a simple horizontal PVC tube with holes to spread the water by soaking and use couple of fans for cooling, B. Use simple shower head to spread the water on fully open air. Your more industrial approach have place in a very large scale, when volume of water in the system and whole scale of the system should be optimized to a reasonable scale, but you still have a PC water cooling that makes you want to pee all the time. Also there is C. option, to vaporize and condense water with sound waves, like Ultra sound, this is how electric air humidifiers work.
Might as well just build a wall fountain, At least esthetically it would be a lot nicer but might need to make it fairly large to be effective, and you'd have to deal with dust and keeping track of the water level. For an alternative material to put in the column you could use the sheets of material that go in swamp coolers, its designed to do this exact thing and you ca buy it at the hardware store.
Line the wall with peltier coolers lol
put a water level sensor in and monitor with arduino/ rpi to open a solenoid connected to a tap or larger reservoir to maintain water levels.. wouldn't want the system burning out when you forget to turn it of before a weekend away haha
Garden irrigation and cooling systems have a "misting" head that forms a fine spray. Combine that with the fan at the top, and you get something similar to a jet engine.
Another option would have a water fountain at the base, and a series of condensation plates in the column. Water from the plates is recycled to the PC - looking a lot like a fractional distillation setup.
this actually could work for normal setup if you use a water to water heat exchanger to keep the evaporating water supply separate. Could be nice, like one of those desk fountains very tempting....
Although something like this would be a fun addition to a homelab cluster of multiple motherboards
Or it could make for a very moldy room very quickly
Definitely need *very* good airflow for this for sure. But, like, more airflows healthier anyway (assuming where you live isn’t horrifically polluted, and even then HEPA filters are magic), so🤷🏼♂️
I kept looking at this and thinking "why not just use a rain wall instead"? You could easily put a rain wall in an enclosure (like a fish tank or plexiglass box) and then add a filtered intake fan and a fan for exhaust. You'd get plenty of evaporative cooling, filtered air and a lot less parts.
You might be able to just pack the entire tube with fibrous material that would allow a huge amount of surface area and slow flow. Just maybe put a "shower head" at the top to distribute the paths through the material.
Not fibrous material, it'll just clog up. Use distillation / scrubbing column fill. Distillation columns are responsible for industry worth trillions, you better believe they've been optimized as much as humanly possible.
@@Steamrick Could also you the material they use in swamp coolers, it does a similar thing and you can easily buy it at the hardware store
This reminds me somewhat of a bong chiller I made about 15 years ago that I had outside and plumbed into my house. The cooling results from a 8 foot bong chiller are phenomenal. :)
Hey love this concept has some merit for a TEC project I'm working on. Just a thought re your surface to water / air ratio, you might like to try a bunch of small plastic balls and just fill the towers with them. At a stab I would suggest approximately 15mm to 25mm diameter. Should aid the whole cooling process quite a bit 🙂.
Bio balls would be perfect for this application! Small plastic balls designed with high surface area in mind.
I made one of these 20 years ago, but I went for attached flow to the walls of the pipe. I accomplished it by lining the walls of the pipe with one layer of cotton towels. Then I had a smaller pipe with a fan that went inside the big pipe, that way it could pull air down into the pipe and back up the smaller tube. It was quiet and cooled my AMD CPU. Downside is it used quite a bit of water and stuff would start to grown in the water, you might have to address those issues somehow.
Humidifiers will either be topped-off regularly or will be plumbed in so they can be refilled automatically. To prevent growth you can mix some biocide (e.g. copper sulfate) in with the water. Filter media in the water path also helps to collect any contaminants. Any option system should be emptied, flushed, and refilled periodically. Growth tends to happen more in standing water. If you can keep the water agitated (such as the churning of the water you see in the bottom of his reservoir), that should reduce the amount of growth. I have an open water fountain for my cats to drink from. It can run for several weeks without any obvious growth. It does contain a filter which is just cotton batting and activated charcoal.
@@reverse_engineered Thanks for the solutions you presented. I only ran it for a month or two, it was a very effective humidifier.
That setup looks like it would benefit from quick disconnect fittings. That would also allow you to remove it from the loop and still use it as a bong. That way when you hook it back up to the PC, the bong water will circulate through the system and give you HIGH frame rates.
Way to think outside the box,... or case! I really like the idea of hearing flowing water versus flowing air, very zen 😇
Love this unconventional water cooling.... could you try stackable plate HEX instead of radiators? I believe you'll need 2 loops (hot/cold) for the setup
Ah, this takes me back. I made a similar evaporative cooler with PVC plumbing pipe and scotch brite pads to hold the water and allow evaporation. I had it cooling a dual-CPU Athlon XP system. Had to refill it every day - it worked fairly well but wasn't that practical given how much water it went through. Yours is a lot more aesthetically-pleasing!
XP or MP? I didn't think the XPs worked as duals.
Use a power station cooling tower as a model, spray heads at the top with upflowing air. If the droplets are small enough, you don't need anything in the column.
And just filter the water with an in line filter. You can also put a grid just above the bottom of the pool to encourage sedimentation
This was very thoughtfully constructed and a ton of fun to watch, thank you!
These used to be called "bong coolers" and were used for cpu overclocking back in the pentium 3 and early Athlon days. With a 50-60w cpu, the shower head and pvc pipe type could cool the water below ambient and thus allow for what in those days was extreme overclocking. They were also relatively cheap back when dedicated PC radiators were rare and expensive. I think by about 2002 or so CPUs used enough power that they were no longer used much.
Coolingwater specialist here: to improve your design the only thing you need to do is increase the surface area. You could do this by printing a circular lamellafilter, which fits in the full length of your clear acrylic tube. More surface area equals more evaporation, upto a certain point of course.
You can also filter the water by adding a 'resting pool' similar to Moors filtration technology used in Andalusia back in 15th (not sure when was that) century. It is basically a relatively large reservoir that has a certain depth and relatively high intake and outlet spots (for example 2cm radius ports work on top of a 14cm deep 25-30cm long pool). Idea is if you put your water flow thru a very deep reservoir, all the particles brought by the water will seep down and you can collect cleaned water from a high spot. Ideally you'll need a cleaning mechanism/scheme for the bottom end, if you make it open top like the original it will help with cooling and finally if you make it just right you can use it as a master waterfill/drainage reservoir as well.
It'll take a lot of space, it'll be heavy and potentially messy but an absolute spectacle if you're up to tackle :)
Used to play with these in the early 2000's. What you want is as small dropplets as possible spread as uniformly as possible and you want them to free fall against the airflow as long as possible.
Designs back then where otherwise quite simple and similar, just a collecting reservoir on the bottom with pipe to the PC and pretty long and wide tube (I used to have 1,5m 300mm PVC tube as the "rain tube"). The top part is the difference, we didn't go for direct airflow because we used the very top to diffuse the water (1-4 pipes from the PC pushing water to small PVC pipes at the very top with holes on the under side and under them a diffusing layer that was just a PVC cap with a ton of small holes through which the water just barely dropped down to the pipe). I used to have 2 fans in their own pipes separating from the rain pipe to upwards to the sides so water couldn't be pulled into the fans directly.
The true art came from tuning the fans. In the best case the airflow needed to be just rightly distributed so the dropplets don't get flown around in the pipe and hit each other and grow in size, so as stable and directly down to up as possible. The second part was the force of the airflow, you wanted just enough that the dropplets don't fall straight down but they fall slower because the airflow because the longer the dropplets fall, the colder they get.
I would recall some people who put way too much time to it managing to get the water returning to the PC being subambient temperatures. But the whole thing fell to the history once someone found out you could just use a small intercooler and strap fans to it instead of this huge rain tower and we got closer to the modern watercooling (we were still cooler because having the cooler within a case is such a limitation, just drill couple holes to the back, get rubber grommets for them and bring the cooling outside, you can even dunk the cooler to a fishtank and water cool your water cooler).
In the early 00's we just used 120mm PVC pipe with a Y piece on the bottom and an shower head on top, have the fan blow in air through the free arm of the split piece and you could achieve sub ambient water temperatures when running without load.
This is pretty cool. Hope to see a v3 some day.
Loved the experiment! Also the constructive criticism of the flaws discovered on your first design . How about if you lowered the bowl down in the duct ? I think the neck of the fish bowl would be small enough for air flow .
I hate how much I like this, more please.
One way or another, you have inspired me...But 15 years ago I had a similar idea....mine is simpler though...But it is based on genius physicists' discoveries...But I stopped pursuing the idea...I will make it happen this time because we haven't overcome such a thing.
I commend you for trying for making something not just talking. I didn't exactly measure parameters, but I achieved a dramatic difference, just life as a networks admin and a horrible divorce got me off track...All to say don't let anything stop you from improving your design, everybody has unique ideas and you are on the right track, cheers.
I dont know why I like these somewhat overly complicated, intricate and/or unique water cooling ideas. Whether its due to the fact I want my pc to stand out a bit or just because its different I really want to go like way over the top with it one day and have a large tower like this that could sit next to the desk and look like its purely a decorative fountain. Although pumping could become an issue.
The key point here - hot media/cold air surface area size. You need to increase it at max possible level. So here is some strategies to reach it:
1. Make thin streams of water through small (up to 1...2mm nozzles). But it will may need pump with high enough compression (between PC and nozzles).
2. Make freefall water droplets. This will require more nozzles to achieve required flow.
3. (Most promise) Make your own radiator for controlled surface size at wide range of flow - push a water through an array of plastic tubes for drinks. It`s easy to calculate requried number of tubes for your application. Coz these tube has a very thin walls (0.15 - 0,30 microns) - high enough result heat conduction. Because it`s tube - easy streams and air control.
Missed (mist) opportunity!! 🤣 You could have put a fan inside, let the water strike the middle or the blades and it will be flung and misted , greatly increasing your evaporation and thus your cooling! Now you can have a fan blade showdown of water flinging and misting designs!!!! Probably a good idea to silicone conformal coat the center electronics.... And mount the fan hub up to keep some water out of the bearing
But also be aware of the indoor humidity. As relative humidity rises during use, evaporation will become less and effective heat removal will reduce. So over time this will do a worse job unless you can keep your indoor humidity low and stable with outdoor air exchange.
Hey, I didn't know you had a new baby. Congratulations man!
Wow! Your own cooling tower !
If I may, there is two kind of cooling tower and yours is a little bit in the middle in the wrong way.
You have free fall cooling tower and packed cooling tower.
Free fall cooling tower have less drag (air easily pass through), but the surface between air and water is not so great.
The packed cooling tower need strong fan as there is a lot of drag. But the cooling is also better inside as the surface between air and water is better.
As you use a PC fan, I recommand you to:
- redesign the head to have a better spraying of water. What you want is a shower of small droplets.
- remove all packing inside.
- replace the fan for a high pressure one.
Beware that the cooling effect depend of the relative humidity of the room.
Pretty project through. Have fun !
i kind of want more episodes of this where you tinker with everything :D
Thanks for the update!
You should do a video on cleaning your loop, trying various things to flush it, and using a boroscope camera to get in up close to the parts we can't normally see, like down the fitting hole of a radiator.
For instance, one of my ideas is to pour drain cleaner into the radiator, let it sit overnight, then flush with water until clear. Possibly other things as well, like Isopropyl alcohol and automotive paint solvents. Then move on to stuff that should be more compatible with the plastic and acrylic bits of your loop, once you think the radiator can't be cleaned any more.
Definitely don't use drain cleaner! Common drain cleaner is made of sodium hydroxide which will readily attack copper metal. Newer ones made with different chemicals may be safe for copper pipes - check the documentation carefully. Most acids, such as hydrochloric acid or vinegar (acetic acid) are safe.
You actually inspired me with the original video to build a similar system, I decided to use a sort of "rain" shower head I had lying around and 2 clear 5 gallon buckets for mine though.
AQUARIUM TOUR PLEASE! Is it saltwater? fresh water? cichlids? marine? i have been watching your vids and had NO idea you where a fish guy too! i must see! Please!
Also, Cool tower cooler man!
It's like a swamp cooler. That's what we had when we lived in dry Nebraska as a kid instead of refrigerant cooling like normal air conditioners.
I remember people making coolers kinda similar to this 15-20 years ago with the water getting slightly cooler than ambient. It was constructed from just plain old grey pipe and shower head at the top and a 45 degree piece in the middle facing downwards with the fan blowing air through it on the water dropping past it all the way from the top to the bottom without any middle grills or anything.
I very long time ago, I used I "Y" fitting on 4" pipe. 1.5' going down into a bucket and 3.5' going up in the air. Using a rain type of fitting for a kitchen sink at the top of the pipe and 3 92mm case fans stacked on top of each other connected at the "Y" fitting I was able to leverage evaporative cooling. This was all sitting in a 5 gallon bucket with a fountain pump cooling my Duron 900.
Another way to increase your surface area is to use scotch bright pads(maybe you can find white ones), I actually enjoyed the rain cooler. Cool to see the same concept used twenty years later.
That's some clean soldering there my friend. I don't give that compliment to any but the most deserving.
Also, filter the water, not the air. Being an aquarium keeper, you already know how to do so effectively, and you can't choke out a water filter by making it wet.
Fun idea. Use a window screen mesh to add surface area for the separation disks. Just mount them on top of the current stack disks. The finder mesh you use the more air pressure you will need. Possibly 2 fans on the top to pull out.
This is the chillest cooler I've ever seen
Very cool. I feel like this has potential to be quite effective with further refinement. It would be especially nice in places with dry air, since it would act almost like a desktop humidifier, and achieve better cooling via enhanced evaporation.
Tech Ingredients does a few videos on water-tower evaporative cooling. Almost the same idea with fans blowing air from the bottom up, into a shower head.
Back in the day when we made these out of PVC pipe we used off the shelf shower heads at the top
This just needs to be a series of Build your own exotic water cooling set ups.
For a very long time I suffered from the same ideas and their implementation.
1 to increase the area from which evaporation takes place, I used a bunch of different straws for cocktails, in my case 6 mm showed the best result
2 geotextile was used to filter the air - it is very similar to the material of vacuum cleaner bags
3 I used sponges for aquariums to filter water, it also made the sound of falling water drops quieter if the water level was lower than the top of the sponge
4 I used a channel fan and in order to avoid contact with water, the air supply was from below
5 since the circuit is open, the tank needs a huge one so that all the water can drain there, well, you need a powerful pump so that you can pump everything back, or you need a check valve(near exit) that creates a decent resistance in the circuit
edge lighting the tube from the top down might look good, or having a lit up central acrylic tube with bubble (take apart a cheap garden bubble solar light)
You should print spacers to go between the elements to keep them flat/level. It would be awesome to see at least the top distribution plate have a rotating element, powered by the weight of the falling water. That would take some design work, but maybe a hydraulic bearing system would give an load capability on the bearing proportional to the water flow since it would power the bearing. I've seen a bearing/motor design in higher pressure systems, so I don't know how effective it would be here. But it might be worth a look.
Ayyy, I'm so happy you ended up doing it, thank you!
Imagine hooking up a water system to your entire house, where you can plug in many devices with quick fittings so you can water cool any device anywhere in your home using a big fountain type bubbler like this? That would be such a cool thing to see someone do.
One way you can help with spreading the water is to have a gradual increase in pipe diameter when reaching the top of the tower. Second, you could use styrophome on the filter area so that even if it gets wet, it doesn't clog. Third, you could slow down the pump significantly because you gain all that extra volume in the tower. Basically it looked like you hade more water entering the chokepoints than leaving the choke points at the same time. Also you could try to change the inlet to the center of the fish bowl instead of on the side.
Instead of mesh discs, you could print "Pall rings" for random packing with a top and bottom retainer grate. Pall rings are used in commercial-scale for distilling and similar processes. Pall rings or other random/dumped packing would be a great way to distribute that water nicely.
Combine that with a counter flow air path (air goes up, water falls down), you could probably get some more cooling mileage out of basically the same materials. you could use a y shaped branch to inject air down one leg of the y to go up the other leg of the y, in which the water is falling down. Then you wouldn't need to worry about getting the fan wet, but you could still get upwards/counterflow air.
This liquid down, gas up counterflow is how it's done for the most part with chemical engineering.
I was thinking about doing something similar, before I realized I could use any of 3 legs in my old radiant heat lines to sink heat into the slab of concrete my house is built on. It was abandoned, in favor of baseboards years ago but the piping still exists! I used to work in a chlor-alkali plant, we did have chillers and a cooling tower but what you've got there is more similar to the chlorine stripper. We used small cylindrical packing material that had plenty of surface area each, packed into a column with air in from the bottom an liquid from the top, with ventilation. There was one disc at the bottom to keep the packing in. Airflow was priority, as was surface area for chemical stripping. These would, for your column have about a 1.5" diameter and length, easily 3D printed. Food for thought. I have also worked in both water and wastewater, so I have XP with flow and blinding. Maybe turn down flow and bump up air. The full column of packing material with a plate on the bottom with 1" holes will keep the packing in and allow for airflow without blinding (adjust water flow). I had also considered adding a float and a 1/4" fill line for makeup water with a float and a valve, you will lose water to evaporation (hopefully!). Looks promising, and cooling can be increased with a taller column/more packing easily. It opens the cooling loop(heating loop) as well so it's fairly ideal, just add water and air! You could even add ice to the basin to increase cooling for a bit too!
Excellent work man. I’m impressed. I’d rather use that than radiators any day. I mean hell, I’m running 2-360’s with push/pull configs so 12 fans on the rads alone lol. The sound of a waterfall is much more preferable and calming than 12 fans lol.
This video made me curious if you were the guy who put an EVO 212 or whatever it was in an acrylic case and water cooled it. And apparently yes, you were. Man it's cool to see how far you've come since then
I was just about to research how to do a more efficient evaporative cooler for a project of mine. I plan on integrating your fill design with my cooling towers.
Instead of an upper fan try using an aquarium air pump in a lower section of the tower. This way you have a positive displacement system, relatively speaking compared to a fan.
If you plan on sticking with a fan suction design, use a fan designed for radiators.
Was thinking about the CPU cooling. I was wondering if mist nozzle spray onto a finned block of sorts could put the phase change on CPU block...
Granted the mental picture of everything is rather exotic. Imagine a smoke stack out the top of the case. 😂
That’s pretty cool. If there was two, especially in a gun metal like color, it would look like Dr. Frankenstein’s lab😂
that thermal dynamics module coming in clutch here
This seems like it would a) be a neat humidifier / heater in the winter and b) something my can will be unable to resist drinking from
Congratulations on the kid!!!!!!!!! Just wait for all the my old stone's stones 😁😁😁😁
Aquarium bio balls would probably work excellent to fill the tube
In process engineering we would use a filler material to distribute the water like little glass sections of pipe or metal saddle shapes or washed pistachio shells. The ultimate fillers are Raschig Rings. The trick is to maximize interaction without blocking the flows of water and air. And be carefull to not contaminate your water if you try any material.
Chemical engineer here. Use thin, low density scrub pads on each layer. Surface area will increase by an order of magnitude.
Scrubbing column fill - it's literally designed to maximize air/liquid mixing inside a column like this.
This or just go and buy the mat material they sell at home depot for swamp coolers, its designed for this exact thing
@@Steamrick I worry about air flow in that setup but if it works, it's ideal.
@@connorjohnson4402 I was not aware of it. This is the most practical and easy way to do it.
@@connorjohnson4402 The swamp cooler material is, afaik, meant for low water flow and large front area rather than depth of material. It might work fine, but I'd not be surprised if it is less than ideal.
An idea for a future version of the bong tower would be to use spacers for the fill instead of adhesive. You would lose some ID for airflow, but everything would stay put!
Love the innovation. You asked for a challenge? MAKE a PC case using multiple 500mm radiators stuck together (possibly welded?) for the top, side, front and bottom, so four out of the 6 case sides. Then get a regular rear "metal"/3D printed rear plate with cutouts for IO. And finally hang the motherboard mount on cables under tension, like the EVGA E1 PC case.
I would imagine two length ways 500mm radiators to form the top, the same for the bottom and front, then five 500mm radiators to form the side of the case. A total of ELEVEN 500mm radiators, for what could be enough passive cooling to require no fans, just the pump to move the water through all the radiators.
Alternatively, you could either find, or 3D print, a distro plate to form the front of the case, so only top+side+bottom were radiator walls, reducing the number of needed 500mm radiators to nine.
you don't know how long I've waited for this
G'day Major,
Along with the Fan Showdown I love these "caraaazzzy ideas" videos, it is Awesome the way you get everyone involved in the channel content, rather than just telling us "I did this".
WOW! compared to a regular loop it is also really 'cool' & I guess if you like those calming fountains it would be great knowing your fountain also has a PC Purpose, but me even for the short running time 😬I had to stop the video to go for a bathroom stop😂,
🤔some thoughts to help longevity shorter clear tower tubes (same overall height but 6 or 9 instead of 3) with the Fill pieces integrated into the joiners, maybe a straight sided bowl at the bottom to hold the same amount of water but lower the Water level to prevent splashing the air filter.
Congrats on the new addition!
A great follow-up would be a Tesla fountain, which many people gloss over before realizing its intent as a cooling tower for heat engines/boilers and as an efficient means for a fountain to heat the surroundings, or take heat energy from it and cool a closed system.
This was really cool, I love wonky stuff like this. Imagine a large sized one of these cooling a whole room of computers. You'd have to be dedicated to the water trickle sound though, unless you found some way to mitigate it where the water drops into the base below (Where I assume most the sound comes from)
From what a heard, there are Google data centers where that’s pretty much the case. Evaporative cooling can be cheaper then running air through the system (especially if your data center is in a desert, where the air is very hot). Course, there’s the issue of being in a *desert*, and the whole “water is precious in the desert” thing, *but*, it does exist!
If you wanted a passive external cooler, have you tried converting a Oil Heater and moving the oil through the water block and out of the case into a large Oil-filled Radiator Heater?
Have the cool Oil flow back into the water block?
Obviously you would get rid of the heating element since your CPU/GPU will provide the hot oil.
A QD (Quick Detach) fittings on the outside of the case should make it easy enough to move around.
And there are small Oil-Filled Heaters that have their own Caster Wheels
3:20 FACTS!! Not to mention chemical warfare is illegal
This is going to be an exciting series.
I once had the idea of using one of these epoxy river/ waterfall- desks as part of my watercooling loop. But it probably wouldnt have looked pretty for long with hot water steaming up. I did build an evopartive cooler from a small jar that had a showerhead inside and a fan on top. Worked good as a temporary setup but I had to fill up the water level once a day.
The fill is to break up the water into droplets of a certain size that maximizes a ratio of surface area to volume, so that the evaporation from a droplet cools that droplet maximally. Evaporating a film from a solid surface does nothing in a system like this. There needs to be droplets returning to the sump. The fill also slows its travel time from top to bottom, giving more time for evaporation.
Friend, take a look in "bong cooler".
They place the fan in the bottom so the water vapor dont runs through the fan, which will surely reduce its life span.
A "rain" type shower head is what we used to use back in the 90's 🤣. Just a long 6 foot 8" PVC pipe ("reclaimed" from the junk yard for $5-10) with the shower head at the top, fan above, pump and all down below. Literally 30+ years ago this was how you beat the "heater core" people to ambient temperatures.
7:23
You know you're on the right track when you introduce the Sripol method
If you wanted to perfectly spread the water you can ridge the fish bowl and have it spin slowly with each ridge covering a different vertical section.
Creating stilts for your water plates and indentations for those stilts to seat into would make it so _zero_ glue is necessary, and if you use TPU to create gaskets you can have a _loose_ interference fit using the properties of shore hardness to your advantage so the water doesn't end up sneaking through the sides of the plate. A concavity flush with the plate and TPU seal would force the flow where you want it.
If you make the seals be slightly _convex_ on both sides with a rather long and gentle fillet you can have the outside-facing seal press against the tube, while the plate is held into place. Material spread would keep the plate in place _under friction_ but the stilts would still be necessary for absolute placement so water flow does not displace them. The stilts should take advantage of the TPU material so they also sort of press-fit into place. The result would be a self-holding structure able to retain water with minimal loss - _if any_ - along the sides.
Did something like this around the year 2000.. Its the easiest way to get water below ambient temp. 3" pvc on top of a 5 gal bucket. Had a y flange at the bottom of the pipe by the bucket. Had one pusher fan. At the top I had a shower head. The pump was in the bucket. Just the air passing over the shower of water was enough to pull the heat out no need for all the in fill if the initial dispersion is good. After that I went to a water fountain setup to actively cool it down to around 35-40deg. Back around when the celeron 300a was around with a block I made that was. It was crude but worked like a champ.
A cooling tower works by evaporative cooling. You need a ton of surface area for the water to interact with the air flowing through it to get any cooling. You also need a spray system instead of just dumping water into it and trying to distribute it with a shower screen.
For sure. This is 2009 technology without any research done into it.
@@werdfeefs7027 older than 2009.
Long time follower, first time commenter 😅, after watching this video I had some suggestions for vas improvement for both air flow and water distribution. Have you considered trying the “print on fabric” technique with a mesh or micro fishnet? You could easily have stacked rings of various mesh and netting all from a simple trip to the screening section of a hardware store. All consider adding leg to your filters to act as guides and supports to build your tower. Cheers and great success!!!
Cool project, literally!
Saw the hot glue gun and went: "Oh, no!"
But now I have to pee...
Saw someone do this exact thing with white PVC piping and some fans like 20 years ago. I think his conclusion was, it works great for cooling, but falling water is loud, and it makes your room super humid.