Concept Sub-50mm CPU direct mounting water cooler with Thac Nuoc CPU block. How shortest can be?

[XCreator]

In the waiting time for Sun Cool prototype come, we have a Sub-50mm CPU direct mounting water cooler design.
Idea is using short CPU block and use external pump.
This one provide a medium performance with ~50mm height and high performance with >70mm height.
See comparison chart:

 

[Thehack]

I don't think this idea will be viable.

At 50mm low performance can we expect something that exceeds the LP53? Or a C7 type design with proper fan?

At 70mm can we expect to beat a Raijintek Pallas, or NH-L12, or big shuriken 2?

I think the main benefit of an AIO is to able to move the cooler to a different location for better layouts. If this doesn't significantly beat top-down air coolers I don't see how it can be viable.

 

[XCreator]

I don't think this idea will be viable.

At 50mm low performance can we expect something that exceeds the LP53? Or a C7 type design with proper fan?

At 70mm can we expect to beat a Raijintek Pallas, or NH-L12, or big shuriken 2?

I think the main benefit of an AIO is to able to move the cooler to a different location for better layouts. If this doesn't significantly beat top-down air coolers I don't see how it can be viable.

This one is easy to fabricate.
At ~50mm: with 92x15mm delta fan and 120x27mm copper radiator (total 55mm) we can expect normal 120mm AIO aluminum rad performance.
At ~70mm with pump base block 120x13mm fan and 27mm copper radiator we can expect 107-120% normal 120mm AIO aluminum rad performance.
(Copper rad provide 7-20% more performance than AL rad).
We get dust problem with Heat pipe Top-down air cooler.
If we have much space to place AIO, it's good, but in SFF world, to place equipment, we have to find each cm3 space.

 

[Thehack]

This one is easy to fabricate.
At ~50mm: with 92x15mm delta fan and 120x27mm copper radiator (total 55mm) we can expect 120mm AIO aluminum rad performance.
At ~70mm with pump base block 120x13mm fan and 27mm copper radiator we can expect 107-120% 120mm AIO aluminum rad performance.
(Copper rad provide 7-20% more performance than AL rad).
We get dust problem with Heat pipe Top-down air cooler.
If we have much space to place AIO, it's good, but in SFF world, to place equipment, we have to find each cm3 space.

Please excuse my very direct language, this does not mean I'm rude or dismissive as I'm just discussing the topic.

1. We know that a 92mm AIO is $80 not including the bracket, so almost $100 all in all. This is a currently available AIO. If we make a copper 120mm RAD then at low production runs and a very niche application it will be very expensive.

2. I don't think it will perform anywhere near a 120mm RAD. A 120mm RAD is accompanied by a 120mm fan. A slim 92mm fan in a very restrictive location would be terrible at cooling the RAD. Even with it being copper, adding forced air convention greatly increases the performance. A 92mm fan has half the airflow coverage as a 120mm, meaning the RAD has half the airflow required to cool the fins.

3. With that in mind, I don't think you can beat an LP53, or NH-L12 considering the performance per noise level. The 92mm AIO is about equivalent to a Big Shuriken 2, with the added. Benefit that it has a higher heat capacity thanks to being an AIO.

 

[XCreator]

Please excuse my very direct language, this does not mean I'm rude or dismissive as I'm just discussing the topic.

1. We know that a 92mm AIO is $80 not including the bracket, so almost $100 all in all. This is a currently available AIO. If we make a copper 120mm RAD then at low production runs and a very niche application it will be very expensive.

2. I don't think it will perform anywhere near a 120mm RAD. A 120mm RAD is accompanied by a 120mm fan. A slim 92mm fan in a very restrictive location would be terrible at cooling the RAD. Even with it being copper, adding forced air convention greatly increases the performance. A 92mm fan has half the airflow coverage as a 120mm, meaning the RAD has half the airflow required to cool the fins.

3. With that in mind, I don't think you can beat an LP53, or NH-L12 considering the performance per noise level. The 92mm AIO is about equivalent to a Big Shuriken 2, with the added. Benefit that it has a higher heat capacity thanks to being an AIO.

1. The Thac Nuoc CPU block is same almost metal CPU block, you can use it as normal CPU block in custom water cooling system, We just going to make this block, we will consider to make a AIO base on this. The price will be cheaper if the product is more popular.
2. See that one to compare Fan air flow:
http://www.delta.com.tw/product/cp/...asp?act=sch_rsl&tp=co&fantp=&size=&th=&vol=12
Heat transfer is not linear with air flow.
3. We are not going to make this one to be very high performance. Just arm to medium performance and dust less. Our sun cool solution is the design for extremely high performance cooling system.

 

[Thehack]

1. The Thac Nuoc CPU block is same almost metal CPU block, you can use it as normal CPU block in custom water cooling system, We just going to make this block, we will consider to make a AIO base on this. The price will be cheaper if the product is more popular.
2. See that one to compare Fan air flow:
http://www.delta.com.tw/product/cp/...asp?act=sch_rsl&tp=co&fantp=&size=&th=&vol=12
Heat transfer is not linear with air flow.
3. We are not going to make this one to be very high performance. Just arm to medium performance and dust less. Our sun cool solution is the design for extremely high performance cooling system.

If there is airflow, there is dust unless you're filtering. The only dust free system is a complete passive one.

The delta fan is a terrible. It's a common oem designed and is very noisy for the performance. Being a slim fan it will not have good enough static pressure optimal radiator usage.

 

[duke00]

I think you will get problems with the air flow between cpu block and fan. the pressure resistance will be there very high cause of the low space. Have you do an flow simulation ?

 

[XCreator]

If there is airflow, there is dust unless you're filtering. The only dust free system is a complete passive one.

The delta fan is a terrible. It's a common oem designed and is very noisy for the performance. Being a slim fan it will not have good enough static pressure optimal radiator usage.

I mean dust on motherboard, that dust can damage Mobo.
All system will be noise at 100%.
Normal condition with 30-50% fan speed is less noise.

 

[XCreator]

I think you will get problems with the air flow between cpu block and fan. the pressure resistance will be there very high cause of the low space. Have you do an flow simulation ?

there is 3mm distance from block to roto. Not big problem.

 

[Thehack]

I thought your name for it was weird. Then I realized that it means "waterfall." Viet is my second language.

Dust for the most part doesn't damage motherboard, unless you're in an area with metallic dust for some reason, then the computer would be the least of your concerns.

I think you are too optimistic about the performance. A lot of us have experience with building these things and the actual application of it is a bit far from the simulations and concept.

I see a lot of merit a low profile cpu block but if you put a radiator on top of it you might as well go for a big air cooler that is generally quieter and require less maintenance.

From general experience water radiator is only 10-20% more space efficient than air cool if you control for noise as well, many times it can be worse if you consider pump noise.

 

[XCreator]

I thought your name for it was weird. Then I realized that it means "waterfall." Viet is my second language.

Dust for the most part doesn't damage motherboard, unless you're in an area with metallic dust for some reason, then the computer would be the least of your concerns.

I think you are too optimistic about the performance. A lot of us have experience with building these things and the actual application of it is a bit far from the simulations and concept.

I see a lot of merit a low profile cpu block but if you put a radiator on top of it you might as well go for a big air cooler that is generally quieter and require less maintenance.

From general experience water radiator is only 10-20% more space efficient than air cool if you control for noise as well.

Please wait to see our Sun cool beat 360mm WC system. I sure about that, not simulation.

 

[theGryphon]

there is 3mm distance from block to roto. Not big problem.

Uhhh, yeah, that's a big problem. 3mm is nowhere near sufficient clearance from a "solid" surface for air to flow through. So yeah, as problems go in regards to airflow, this one's up there.

 

[Thehack]

Please wait to see our Sun cool beat 360mm WC system. I sure about that, not simulation.

I don't think so. It's simple physics.

 

[XCreator]

A 70mm height AIO in our EAXT case (under fabricating):

Stress Test with 1650 V3 25C ambient temperature (Result is in stable temperature condition)

Still waiting for some components.

 

[Thehack]

Make sure you stress test it for over 30 minutes. Water has excellent heat capacity.

 

[XCreator]

Make sure you stress test it for over 30 minutes. Water has excellent heat capacity.

Result is in stable temperature condition!
To heat 100ml water rise up 20K need 8400J, 140W heat source take 60 secs to provide this power.
The problem if you run system with 100% power in long time is all component temperature will rise, but the test case has a dynamic air flow so air inside keep refreshing any time so result is not different even you run stress test 1h.

 

[Thehack]

Result is in stable temperature condition!
To heat 100ml water rise up 20K need 8400J, 140W heat source take 60 secs to provide this power.
The problem if you run system with 100% power in long time is all component temperature will rise, but the test case has a dynamic air flow so air inside keep refreshing any time so result is not different even you run stress test 1h.

Not sure how that is relevant, I just wanted to point out that you need some time to reach equilibrium. 120mm rad cooling is excellent but the question is if it is better than an NH-L12S or raijintek pallas when considering performance vs noise. There is little point in a system when it is 50C but sounds annoying.

 

[RoLexus]

Not sure how that is relevant, I just wanted to point out that you need some time to reach equilibrium. 120mm rad cooling is excellent but the question is if it is better than an NH-L12S or raijintek pallas when considering performance vs noise. There is little point in a system when it is 50C but sounds annoying.

Fan is : Scythe Slip Stream SY1212SL12H 2000 rpm max noise level 37db. Sound is very good during testing.
I going to use 92mm rad or this kind of rad for better compatibility.

 

[Thehack]

Fan is : Scythe Slip Stream SY1212SL12H 2000 rpm max noise level 37db. Sound is very good during testing.
I going to use 92mm rad or this kind of rad for better compatibility.

I personally don't like running at more than 1200 RPM.

 

[XCreator]

No matter even CPU fan run at 50% (1285rpm) ambient temp 23C:

Heat transfer is not linear with air flow!
Even we remove CPU fan, just use three top case fans to cool (1000 rmp), temperature is good as well.