So I got into water cooling recently and my initial thought was that you basically take advantage of evaporating water too cool the system.
What I've noticed is everyone just uses closed containers and use radiators that fan off the heat. This basically just makes water cooling just air cooling but with a heat sink that's far away from the source and doesn't actually take advantage of water's main advantage. Obviously its better to use distilled water for evaporation and coolant just has many drawbacks regardless of a closed or open system. But I'm wondering why is it that no one seems to talk about evaporation in water cooling? I'm running a system that doesn't need a radiator since the open container cools itself down very quickly. Honestly instead of water cooling you might as well just use air cooling since a radiator is exactly that and cooling is still limited by air flow around the radiator. You basically just made a complicated air cool system instead of just running a heat sink with a fan on it.
Also you obviously just have the reservoir outside the case and for extra cooling you can run a fan over the water surface for very low temps.
>>54844363
build a bong cooler
>>54844372
Someone posted an LTT video where they did just that on here recently. Seems like a cool idea except for the space and the constant noise of dripping water making you need to piss every 10 minutes.
>>54844541
Just use a catheter, hook up the output to the bong tower and bam, you got an automatic water refiller.
>>54844541
I never hear the dripping water. But anyway yeah I just don't understand the point of a closed water cooling system it's basically just a more complicated air cooling system. Why even bother with it just get a big heat sink the size the radiator would be it's essentially the same thing. Heat distribution is more even in a radiator but that still doesn't change the fact you need a damn fan anyway why even bother with water pumps and tubes.
>>54844363
How safe is this? Can someone easily pull one of the tubes too hard by accident and spill water everywhere?
>>54844866
>Can someone easily pull one of the tubes too hard by accident and spill water everywhere?
Yes, if that person is a clumsy retard who shouldn't be near expensive equipment.
>>54844770
made me lol
>Watercooling
Enjoy you're broken computer
>>54845438
>enjoy you're broken computer
Wow, in one post you've demonstrated that you know nothing about grammar or liquid cooling.
>>54845467
>Liquid inside my electrical appliance
What could go wrong?!
If the main idea of water cooling is to use liquid to remove heat from the die constantly in hopes that a radiator can cool it fast enough to stay ahead of the rising temperature of the water. Is it possible that the water, over a long period of time and at max load, could heat up enough to cause damage to the system?
>>54844363
A tall radiator is more effective when considering the space required for effective evaporation for cooler temps(40 - 60 C).
Heck, I think having a barrel as your reservoir would work better. I considered using my inside swimming pool as a res but the massive pump was too loud....
>>54845638
Yes they can leak and cause water damage running normally. I'd imagine under those conditions the pump would blow off.
>>54845638
Yes. In a scenario where you don't have enough radiator sure. For instance if you were to try and cool a 4790k clocked to 5.0ghz on a single 120mm radiator, the cpu will still more than likely hit 95C+. Just because you're using liquid doesn't mean the thing being cooled generates any less heat.
General rule of thumb is minimum 240mm rad for the cpu, and minimum 120mm more radiator space for anything else added to the loop. I.e. a video card. This is all subjective of course. I was running an fx-9370 at 5.5ghz and a single overvolted r9 290 @ 1.4ghz on a very thick 360mm and 240mm radiator setup. If you have the room and the case to fit it, more radiator is always welcome. Enables you to run the same speeds but with almost no fan noise.
>>54845535
>>54844363
You dumb retard it's not evaporation, its putting the heat in water, moving the water to a fan, and taking it out, simple as that.
Water had high heat capacity meaning it can store alot of energy before it boils
Kill yourself
Does anyone have any issues with a front mounted radiator sucking air into the case through the radiator? I'm wondering if this will cause excessive temps on the ram or vrm or anything. The CPU and GPU would be watercooled.
>>54845821
You have no idea what the fuck you're talking about.
>I'd imagine
Opinion discarded. You do know that half of the pumps, fittings, and tubing used in PC liquid cooling is stuff that would be used for other applications? Such as aquariums? I have the same pump for the last 5 years now that is capable of moving 3 gallons per minute at low speed. It has never had a single issue. You get what you pay for. Pay for quality, you get quality.
>>54844363
What will happen if you put a little bit of LN or helium. In you loop?
>>54845903
Well the whole point of liquid cooling is to remove as much heat as possible from the case and components. Just turn the fans around on the radiator. Negative pressure airflow. I had more air being forced out of my case than coming in and even after playing with air flow and fan orientation for hours, the max temp difference I saw in a 19C room was a fluctuation of max 1.5C.
>>54845535
There really is no talking to retards like you is there?
Yet another question. Is there benefit to running 2 pumps in series in a setup that goes: res,pump,radiator,cpu,res,pump,radiator,gpu
>>54846104
None at all. A loop actually benefits from a single pump running a bit slower as it keeps the weather inside the radiator a little longer.
>>54845862
what happened to it ?
>>54845862
uh, sir I don't think you understand physics and evaporation at all
>>54846205
that's not true at all you just need to look at the whole thing as one big thermal system
the more temp transfer the better
obviously you probably don't need 2 pumps since the difference will be almost nothing but in general you want quick water movement for faster heat transfer
a single small pump is probably already past optimal efficiency anyway
>>54846678
So a system that goes:
res,pump,radiator,cpu,radiator,gpu
would be better?
>>54846745
Well I don't feel like you need a radiator at all with evaporation cooling plus a small fan but if not then I would go with that setup.
The radiator before the cpu and gpu would ensure the coolest water going in. But honestly I'd just get one large radiator for the whole system. Remember the entire water system will pretty much be the same temperature throughout as long as you have decent water flow. The main bottleneck of any thermal system is heat dissipation and that is factored by the radiator.
>>54846104
The primary reason for dual pumps is redundancy, so you have a backup if one ever fails.
The only other reason to have multiple pumps is if your loop is exceptionally restrictive (i.e., four or more waterblocks).
>>54844866
it's not water
it's not conductive
>>54847470
distilled water is not conductive either
>>54844363
The advantages of watercooling are that 1.) you can have a lot more fin area than standard air cooling, especially for GPUs, and 2.) because of its heat capacity, water provides a sort of buffer against temperature spikes.
Evaporative water cooling can get you lower temperatures than conventional water cooling (below ambient, even), but it takes up more space and since it's an open system and the water is literally evaporating, it will need topping off and the water will get dirty over time.
Water has a higher thermal conductivity than air, so it can absorb more heat from the heat sink quicker, when compared to air. That is the main advantage, and the although you still have to cool the water down using air cooling (fans) you have a much bigger heatsink(the radiator), and more fans to do it with. Critiques welcome.
>>54847679
Yeah the water is evaporating but it's such a small amount I have yet needed to add more water after about 2 weeks. Also I have a metal mesh covering it and the water looks just as clean as it always has.
Besides has anyone actually tested their water to see if they pick up ions from the metal in the system? I've heard that no matter what you need to change your water every few months to prevent ionization buildup.
>>54847803
>Water has a higher thermal conductivity than air
True, but that's not very relevant to the comparison.
Air cooling works like this:
Heat source -> copper|aluminum heatsink -> air
Water cooling goes:
Heat source -> water -> copper|brass|aluminum radiator -> air
So with water cooling there's actually an extra step, and in fact this increases the thermal resistance (i.e., it's worse). Water cooling makes up for it by allowing for much greater surface area in the radiator -> air step, and more uniform heat distribution through the radiator.
>>54848016
Have you tried stress testing your loop for longer periods of time, like several hours? If you have a lot of water in the loop (as may be the case with evaporative cooling) it can take a long time to reach thermal equilibrium. It may be that the water's heat capacity is masking the actual performance of the system.
>>54848124
yeah I've been gaming for a while and many hours and my temps stay a at max around 60-65 C. My entire system is cheap the only things I spent money on was the cooling blocks a pump and zip ties. I have a good amount of surface area for evaporation about a foot squared I think that is the key to these low temps.
>>54844363
After I looked in to some of the specifications of heat pipes I've come to the conclusion that a closed loop cooling solution should always use heat pipes. You can bend them. You don't have to worry about them springing a leak. They have very low thermal resistance. A heat pipe running through a heat sink is just as good as a water pipe running through a heat sink(aka a radiator). Why? Because it is a water pipe running through a heat sink. Why pay for a pump if the heat will do the job for you?
>>54848271
You're limited by space with heatpipes. The thermal transfer rate drops off with length, so the heatsink has to be both close to the heat source, and has a maximum useful upper size limit. Neither of these limitations exist with watercooling, which can be useful where either extreme cooling performance is required, or limited space is available close to the heat source (as in 2-slot GPUs and SFF applications).
>>54848271
holy shit you may be right look at how fast the heat transfer is in this vid
https://www.youtube.com/watch?v=2vk5B6Gga10
>>54848355
Except a sintered wick 6mm diameter heat pipe can carry 70W 300mm with a degree or 2 difference in temps between the ends. That's with the heatsink below the heat pipe(gravity making things more difficult). Things get super easy when you're working with gravity which is easy when you're using a desktop. Heat pipes are the same shit as closed loop but with gravity and the capillary effect as your pump. You get about 10% reduction in heat carrying capability when working with the same diameters. You can buy heat pipes in different sizes and if you don't want to work with a thick pipe you can use them in parallel You can put them in series for longer runs. 2 heat pipes that can carry 70W with 1 degree difference have 2 degrees difference in series.
Oh there is one difference between a closed loop system and heat pipes. Heat pipes pipes use phase change to carry heat. The only reason to use closed loop in a computer is LOL LOOK AT HOW COOL MY BLUE LEDS LOOK WITH BLUE FOOD COLORING IN THE FLUID.
>>54848648
Water cooling is much easier to set up compared to a custom heatpipe/heatsink solution though. The fact that they're nowhere near as flexible as water cooling hoses makes upgrading more difficult as well.
>>54848648
I'm looking more into it and people are saying they need to stay within specific temperature ranges to initiate the phase change. How strict are these temperature ranges?
>>54848416
They're ridiculously good. A heat pipe has the heat carrying capacity of 250 copper cylinders the same size in parallel. So one half inch pipe is like a 8 inch solid cylinder of copper. You can even make your own heat pipes easily as long as you don't mind having to use them so the heat source is lower than the heat sink. Basically you take a heat pipe and fill it with about 1/8 water and the rest steam, solder it shut, and let it cool. Now you have water under a vacuum which is exactly what you need.
>>54848722
The wiki says 20-150C. If your system is colder than 20C it is going to warm up fast and solve your problem. If you make it to 150C your chip is going to melt so it won't matter. It's a pressure versus boiling point thing and can vary with how much water you put in the pipe. If you reach a point where all the water in the pipe is evaporated then your pipe is going to stop working. Since your pipe is closed the boiling point is going to be above 100C since the pressure is going to be higher than 1 atmosphere. Heat pipes are already used in computers without a problem.
>>54848715
They don't have to be flexible. You sandwich a heat pipe or two between two pieces of metal screwed together with some thermal compound. You bend them to go in the direction you want them to exit your case. You use the blocks of metal like you would a heat block. The only increase in difficulty is there aren't over priced kits on the market already.
>>54844847
>But anyway yeah I just don't understand the point of a closed water cooling system it's basically just a more complicated air cooling system
Stick your hand in your fridge. Now stick your hand in a bowl of water you left in the fridge for a couple hours. Do you fucking get it now?
>>54849026
It's more difficult because of all the bending you have to do. Plus if you ever have to upgrade or remove a block to troubleshoot, you have to re-bend the pipes, which can damage the wick or crack the pipe after repeated bends.
>there aren't over priced kits on the market
There are though, in the form of specialized cases:
http://atechfabrication.com/HTPC_cases.htm
http://www.hd-plex.com/hdplex-h1.s-fanless-computer-case.html
>>54844363
You realize water has to get to 100C before it evaporates right? And trying to do phase change cooling like an A/C is far more efficient with a real coolant.
>>54849283
>Plus if you ever have to upgrade or remove a block to troubleshoot, you have to re-bend the pipes
You just unscrew the heatsink and cpu block and slide them out of place. I can't imagine a case where you would need more than or two bends unless you have some aesthetic goal. The pipe comes out of the block running parallel to the mobo so you make a bend so it's traveling away from the mobo perpendicularly. After you clear your components you make a bend in the direction you want to exit the case. Those are neat links but I don't think the use case is comparable to the aquarium kits you put on regular pc hardware.
>>54845638
Not really, because the amount of heat removed from the water by the radiator is proportional to the temperature differential between the water and the ambient air. Hotter water = more heat removed.
>>54849550
I guess I don't completely understand what you're describing. It sounds like you're talking about something more than just a regular tower heatsink that uses heat pipes, right? Like some extra large finned heatsink structure that might fit where you'd put a radiator at the top of the case, say. I just think that due to the inflexibility of the heat pipes it'd be a lot fussier to install and remove. With a water loop, you can pretty much just unscrew the block on one component to get access to it, but if you have a massive heatsink with heat pipes, you need to remove the whole thing, and that's going to be more cumbersome.
>any year
>not building an uber bong
>>54849765
The blue is pipes the yellow is screws. 9001 hours in a ms paint clone. Once you unscrew the metal plate sandwich the pipes just slide out through the holes in the case.
I want to make a set up for my laptop but I can't think of a good place to mount the external heatsink. The only solid place that doesn't interfere with normal use is the bottom but that has less air flow and if the heat sink isn't as big as the bottom it makes for an uneven surface.
>>54849346
Your confusing vaporization and evaporation. Water vaporizes at 100C aka it boils. Evaporation is the transition of water to water vapour at temperatures below boiling.
>>54850147
I actually did water cooling for my laptop and used detachable fittings. I wish I had known about heat pipes whenever I saw heat sinks that used heat pipes I just assumed it was copper. But anyway I think it's much easier to just connect a water block with detachable fittings than mess with heat pipes.
https://www.freshwatersystems.com/p-8306-apc-nsf-valved-in-line-hose-barb-coupling-body-with-shroud-38-id-barb.aspx
>>54845535
>high school chem lab
>has to use hydrochloric acid for an experiment
>"omg it's so close to my hands, I can't do this"
>>54845919
It would freeze the water and fuck up the tubes
>>54850147
Don't you have some CAD software?
I actually had a similar idea myself, however waters transition to gas will only happen at 100 celcius, in other words the boiling point of the liquid is going to be the temperature the solvent used, using dichloromethane you can keep temps below 40 celcius, however you have to design a closed system (extremely flammable vapour), and it has to have careful design to alter pressure until an equilibrium is set up between evaporation and condensation , with an efficient enough radiator to condense the vapour's, and a large enough reservoir of solvent to prevent complete evaporation, although it's possible and would be passive, no electricity required,
I have no job at the moment, if you could collectively fund me ~£2000 I could develop a working product and give instructions of how to build it at home, no experience in industrial manufacturing so other people would be required, but base cost of materials would be about £20 per cooling for a standard GPU or CPU
>>54854095
The last part about manufacturing is if people wanted to fund more to develop a commercially available product
>>54854114
But essentially that is the entire concept of thermodynamic heat exchange systems, transferring heat to ambient surroundings, there's noting wrong with current water cooling, the only benefit you would have with the design I proposes is that temperatures would not practically exceed the boiling point of the solvent
>>54844541
You mean it's not normal to want to piss every 10 minutes?
uh oh
>>54844363
What you are talking about are heat pipes.
>>54844363
Watercooling is way more effective at getting lots of heat away from the CPU fast then just copper. That's the point of it.
>>54844372
Someone watched LTT? Dumbasses, serously... 10 better ways to do it, you mention a bong cooler.
