Random (semi-)novel cooling idea (and possible video idea for LTT).

[pyrojoe34]

So I was just thinking today about a possible interesting way to get improved component cooling by combining traditional liquid cooling and evaporation cooling. I know Linus already did that crazy bong cooler but this idea would involve a lot less work/space, mostly use already existing parts, and would presumably increase the evaporation surface area to volume ratio and efficiency compared to the bong cooler.

 

1. Take a custom loop (or even AIO with long enough tubes) and move the radiator a safe distance from the PC (outside like with whole house water cooling system would probably be best).

2. Find (or find a way to seal) waterproof fans to mount on the radiator (optional, this idea should even work without fans but not nearly as well and maybe worse than just a normal loop).

3. Buy a small water mister and allow it to lightly spray water onto the radiator fins (enough to continually coat the fins but not block airflow).

 

This water would then be rapidly and constantly undergoing a phase-change by evaporation mediated by the low pressure (high velocity flow > lower pressure > increased evaporation) of the air from the fans which would in turn, draw heat from the radiator much more efficiently than just air would. In theory this evaporative cooling may even allow for a loop temperature slightly lower than ambient (a well established possibility with efficient evaporative cooling). This would obviously work best in a dry area (so not Vancouver) where relative humidity is low but you could probably test the concept even in the PNW by doing it in a room with a dehumidifier to have drier air.

 

I'd be interested to see how well this might work. @LinusTech what are your thoughts? This would be a cheap but (IMO) interesting concept to try out.

[samcool55]

Or take a rad and dump it into a bath of water + ice. Tadaa sub ambient temps on a budget!

It is a nice idea but relatively complicated imo. But still not a bad idea, i'm wondering how ghetto a posible solution might look

[pyrojoe34]

1 minute ago, samcool55 said:

Or take a rad and dump it into a bath of water + ice. Tadaa sub ambient temps on a budget!

It is a nice idea but relatively complicated imo. But still not a bad idea, i'm wondering how ghetto a posible solution might look

That requires you to constantly refill the ice as it melts. This would allow you to set it up and just let it run in perpetuity given a constant water source (with a low volume mister this would also likely take no more than a liter of water a day so much less wasteful as well).

[AnonymousGuy]

This is already a thing where people will mist their AC evaporator coils to theoretically improve cooling performance.

 

The problem is that if your water isn't filtered to the point of removing all minerals, they will eventually cake on to your radiator.  People have ruined their AC evaporator unit by doing this.

[pyrojoe34]

Just now, AnonymousGuy said:

This is already a thing where people will mist their AC evaporator coils to theoretically improve cooling performance.

 

The problem is that if your water isn't filtered to the point of removing all minerals, they will eventually cake on to your radiator.

True, although a decent RO system (many places/houses may already have one installed) could prevent mineral buildup.

[samcool55]

Just now, pyrojoe34 said:

That requires you to constantly refill the ice as it melts. This would allow you to set it up and just let it run in perpetuity given a constant water source (with a low volume mister this would also likely take no more than a liter of water a day so much less wasteful as well).

Yeah you can use ice or just a constant flow of cold water, usually tap water is colder than ambient temps so even without ice it should still be relatively efficient, just not as efficient as it would be with ice

 

Or throw in some dry ice, lasts longer and gives a nice effect as well

[iRileyx]

1 minute ago, pyrojoe34 said:

That requires you to constantly refill the ice as it melts. 

Get an ice machine (some of them re-use the water from the melted ice to make more ice), somehow rig it so the radiator can fit in there and the lid still closes and you're good to go

[AnonymousGuy]

2 minutes ago, RKRiley said:

Get an ice machine (some of them re-use the water from the melted ice to make more ice), somehow rig it so the radiator can fit in there and the lid still closes and you're good to go

Mild facepalm, but you're better off just using a water chiller than having the intermediate step of making ice

[pyrojoe34]

3 minutes ago, RKRiley said:

Get an ice machine (some of them re-use the water from the melted ice to make more ice), somehow rig it so the radiator can fit in there and the lid still closes and you're good to go

Again, feasible options but an ice machine (freezer) would add 500-1000W of extra power use. A single small mister would take a few watts at most so better cooling at almost no cost.

[Guest]

As a person who uses a small mist maker for an small office room: it will indeed get clogged up with crap within ~15 days of 24/7 usage. Even is you use some of the cleaner water supplies.

 

It might still work good enough if you would just spray a little of water on the rad. A system like a plant sprayer, which would spray a bit every minute/5 minutes. (and keep cleaning the rad ofcourse)

[iRileyx]

1 minute ago, AnonymousGuy said:

Mild facepalm, but you're better off just using a water chiller than having the intermediate step of making ice

 

Just now, pyrojoe34 said:

Again, all decent options but an ice machine (freezer) would add 500-1000W of extra power use. A single small mister would take a few watts at most so better cooling at almost no cost.

 

I wasn't being serious with that suggestion ya know.

Who tf could actually think that'd be a good idea lol.

[pyrojoe34]

Just now, Dutch-stoner said:

As a person who uses a small mist maker for an small office room: it will indeed get clogged up with crap within ~15 days of 24/7 usage. Even is you use some of the cleaner water supplies.

 

It might still work good enough if you would just spray a little of water on the rad. A system like a plant sprayer, which would spray a bit every minute/5 minutes. (and keep cleaning the rad ofcourse)

With a decent RO system (like many people already may have in their homes) you could minimize any mineral buildup. Sure using regular tap water would be a bad idea but an RO or DI system would solve that problem.

[Guest]

I don't know what a RO or DI system is. (our drinking water is clean and all, no filters requires @ homes) But even when using demineralized water, I had lots of crap build-up.

[pyrojoe34]

Just now, Dutch-stoner said:

I don't know what a RO or DI system is. (our drinking water is clean and all, no filters requires @ homes) But even when using demineralized water, I had lots of crap build-up.

RO (reverse osmosis) systems are common in places with hard water to remove impurities in water using a semi-permeable membrane. DI (deinonizing) system is essentially the same thing but more common in labs where clean water is critical. This is often done using electrodeionization to remove any charged impurities (like calcium or other salts). There are many different "levels" of water purity (everything from nanopure to DI to tap water) so the quality of the system would have a huge impact on mineralization. The best systems will combine a few methods (like UV oxidation, ultrafiltration and dialysis) to make water that contains ~1 ppb (part per billion) of impurities in the water (basically 1 in 1 billion molecules is not water, which is about 300k-500k times less impurities than found in tap water).

 

Just in case you're interested at all...

[Guest]

It's interesting. So how would you like to mount your (VERY expencive lab+ grade) equipment to your consumer grade radiator again?

[pyrojoe34]

Just now, Dutch-stoner said:

It's interesting. So how would you like to mount your (VERY expencive lab+ grade) equipment to your consumer grade radiator again?

Well obviously you wouldn't need that high a quality water for this to work. Like I said a basic RO system someone might have should do the trick (I think they go for as little as $100 new), then just attach the mister to an existing faucet. Even someone without a system could just have a reservoir filled with cheap DI or distilled water (like 80c a gallon) which should go very slowly once you figure out the optimum flow rate.

 

I mean, this isn't meant to be a super practical idea that people will just start doing at home. Just an interesting concept that I would like to see done just to know how well it works. I know there are data centers that use evaporative cooling on an industrial scale so it's been successfully done before.

[Guest]

A maybe even cheaper but way cooler and moar durrable idea would be: small desk fridge. Break it open, and get the cold side element. Put it in a bucket, and put the rad in the bucket. Add water, and you are done. You should however need to really use the cooled computer part, because me thinks you might get issues with ice inside the rad/loop.

[pyrojoe34]

7 minutes ago, Dutch-stoner said:

A maybe even cheaper but way cooler and moar durrable idea would be: small desk fridge. Break it open, and get the cold side element. Put it in a bucket, and put the rad in the bucket. Add water, and you are done. You should however need to really use the cooled computer part, because me thinks you might get issues with ice inside the rad/loop.

That has been tried before and there are two main issues with it:

1. Fridges/freezers are meant to cool slowly and in cycles, they also cannot actually deal with very much heat and certainly not constant heat. This will probably end up making almost no difference in cooling and blowing out the compressor after a couple days.

2. Fridges/freezers use a lot of power (upwards of 500W) at which point you would be better off just getting a purpose made phase-change cooler designed for PC cooling.

 

The reason I though this idea was interesting is it would add at most a few watts of power for the pump while presumably giving massive improvement to cooling.

[Guest]

I don't have too much actual knowlage, but I do have some common sence. I agree that them liquid temps in the radiator can get below ambient. Main advantage of your setup. However, there will be an limit to how much heat it can remove. Will if be enough to get high OC on close to ambient temps. With a regular sized setup, I doubt your loop will get below ambient. Improved, but not below ambient.

 

I would personaly rather put an radiator inside an pond, lake or even better: river. (water flow) You "should" be able to get rid of massive amounts of heat in a rather efficient way. Your temps will not get below ambient water temps, but the amount of heat it can dump... Massive.