Sanity test my WC loop please

[Jimbo Mahoney]

How do you think a low airflow, low delta T radiator would be designed?

I personally think a car radiator would be better than a heatercore because:

1) It is BIG => requires less airflow

2) It is thinner than a heatercore => Good for low cfm / pressure fans.

Any thoughts?

[Butcher]

Low airflow and low delta is hard - for low delta T you need to maximise surface area, the easiest way being more fins. However for low power fans you want widely spaced fins to offer less resistance.

[Brians256]

Sounds like the best case for silence would be a set of evaporator panels from a freezer. They are aluminum sheets with tubing built into them. Make a large enough flat area out of these sheets, and it'd be good enough. Any idea on how much surface area is needed?

[bobkoure]

OK, this is ProCooling, so I can nit-pick.
That's the condensation side
(click the image for a howstuffworks link).
The gas/fluid in a condenser is hot and there's comparatively little fluid movement, so I'd guess these to be inappropriate on at least two counts: restrictive, designed for high coolant-room delta.

[edit] or did you actually mean the evaporation panels from inside a freezer? These aren't designed for fluid flow at all, although they probably aren't much more restrictive than the condenser panel on the back/bottom[/edit]

[MaxxxRacer]

Althornin. I know what the HC's are deisgned for, but paired with a good 38mm thick fan with good pressure, they will kick the but of any purpose built rad.

Due to the highly effecient design of the HC's they do not need the kind of airflow that a blower would provide to perform incredibly well in a WC environment.

But I do agree that they are a little thick. Here at procooling alot of ppl are trying to find the rad that isnt thick and performs incredibly well so that we dont need to put huge fans on it. Myself, I never tried for that. I, thanks to niksub1 found the sanace fans and found no need for a rad that prsented low airflow resistance. Now granted even undervolted these fans arnt nexus quiet, but they are close.

about the high temp differential design of the HC's. Granted yes they are designed for this, but that doesnt mean that they dont perform excelently in our low deltaT setups. And it is true that they perform excelently when given a decent amount of airflow..

[bobkoure]

Originally Posted by MaxxxRacer

Althornin. I know what the HC's are deisgned for, but paired with a good 38mm thick fan with good pressure, they will kick the but of any purpose built rad.

If you match frontal area for frontal area, I think you might be surprised.

Due to the highly effecient design of the HC's they do not need the kind of airflow that a blower would provide to perform incredibly well in a WC environment.

IMHO they are not "incredibly efficient". Have a look at Cathar's post on "radiators in air series". I suspect you might be confusing "really big frontal area" with "efficient".

[MaxxxRacer]

when i said the design was efficient, I will show you what i mean. some ppl might be surprised by the internal design of a heatercore. AFAIK, purpose built ones dont have the design like this. atleast not these little fins that increase surface area (not frontal) by orders of magnitude. anyway here is the pic. BTW, i hijacked this from WEAPON over at XS. I think he is a member here too..

In the pic the red arrows poitn to the water chanels (the thing slits) and the blue arrows point to the flow chanels for the air. notice the little fins (actually litte slats cut into the copper) that cover the air chanels.. those guys increase the exposed surface area quite a bit compared to having the channels smooth.

[Butcher]

Undervolting your fans means you're not getting great airflow. Undervolted fans lose a lot of torque so they can't push anything like as much air as a fully volted fan. It works, because HCs are massively overkill for watercooling not because they are highly efficient. A thinner design would likely give better results because of the fans commonly used in cooling.
FWIW I consider 38mm fans somewhat weak, a larger 50-60mm thick fan is a much better choice.

[MaxxxRacer]

very true on the undervolting, but they have a heck of alot more static pressure even undervolted than do the 25mm fans most ppl use..

With a fan that can push air through the core at a decent level the thicker the better... as long as the air is moving at a good rate and your ears are happy its all good..

If i used a thin core like a bip i would get considerably less performance due to the decreased surface area compared to the HC. Granted there would be more airflow at a given fan voltage, but it would not overcome the increased area of the HC unless you were considering highly undervolted 25mm thick fans. (This example i was considering that I was using denkis which are high pressure (relatively) and are 38mm...)

If I could find some 120mm fans that are quiet and are 50 to 60cm thick im all for that.. My case im designing can accomdate that... hehe...

[Butcher]

I use a single 172x50mm fan, it probably pushes more air through restriction than the dual 120x38 mm setups people often tout as "best".

[pdf27]

Originally Posted by MaxxxRacer

...some 120mm fans that are quiet and are 50 to 60cm thick im all for that.. My case im designing can accomdate that... hehe...

big case!

[HammerSandwich]

Originally Posted by bobkoure

Have a look at Cathar's post on "radiators in air series".

Start here.

[Brians256]

Originally Posted by bobkoure

OK, this is ProCooling, so I can nit-pick.
That's the condensation side
(click the image for a howstuffworks link).
The gas/fluid in a condenser is hot and there's comparatively little fluid movement, so I'd guess these to be inappropriate on at least two counts: restrictive, designed for high coolant-room delta.

[edit] or did you actually mean the evaporation panels from inside a freezer? These aren't designed for fluid flow at all, although they probably aren't much more restrictive than the condenser panel on the back/bottom[/edit]

I am very familiar with HVAC, so I did mean the evap panels. The reason is that they are flat and could be used as a subsitute for case panels. The condensor unit would probably be more effective but there are a couple issues.

Condensor units are usually very large and are single pass. The evap panels are smaller and could be hung on the sides of the case frame and then run in parallel for lower head loss.

Condensor units are heavy and made of copper. The evaporators are typically lightweight aluminum, which fits well with the style people usually use in their cases and the need to move the cases around.

The problems I see are:
1) Evap panels have small ID tubing and may be too restrictive.
2) Where do you get these panels?
3) How easy is it to attach these panels so that they replace the existing case panels?
4) Is there enough surface area to yield a decent C/W?

There are probably more issues than that.

[BillA]

Originally Posted by Brians256

Sounds like the best case for silence would be a set of evaporator panels from a freezer. They are aluminum sheets with tubing built into them. Make a large enough flat area out of these sheets, and it'd be good enough. Any idea on how much surface area is needed?

Brian, not clear to me
you thinking of these panels as passive radiators ?
probably would help to perforate them, no ?

[lolito_fr]

And add some fins...



Review

[BillA]

be interesting to test (formally, i.e. bench)
very slick looking

[Brians256]

Originally Posted by unregistered

Brian, not clear to me
you thinking of these panels as passive radiators ?
probably would help to perforate them, no ?

Yes, passive would be best. Perforation would just improve the surface area, of course. As lolito_fr shows in the post after you, bill, adding fins would be even better. However, fins would be hard to deal with all over unless some handles were added.

I have no idea what the C/W vs LPM of a single panel would be in air when there is no fan. You can't say that the air is un-moving, of course, so I have no idea (as you note) how to reliably test something like this. Do you test it in a chamber? If so, how do you keep air movement to a minimum? What is the testing goal? Do you test it in a room? Under a desk? How close do you allow items to each side?

lolito_fr, where is that case from?

<edit> OK, just saw the review under your picture. </edit>

[lolito_fr]

Some more pics here for your amusement http://www.pc-look.com/boutik/Plus_128003-015.php
have a pike at the top one, Brian

The fins would probably do most of the work. If you look at the surface area, in this case the base of each rad = 0.1m² and the fins approx 0.25m²

Total area roughly 0.7m²
Considering H = 5-10w/m²K
then you have an air side C/W of 0.14-0.29 °C/W

No idea what the liquid side C/W would behave like. For the rad I would think it may even be negligeable!?

[Jag]

Why not a well built radiator and a integration scheme like this one?


Passive means too much real estate and low performance, even considering the noise factor.

[Brians256]

Originally Posted by Jag

Why not a well built radiator and a integration scheme like this one?


Passive means too much real estate and low performance, even considering the noise factor.

If that scheme is quiet as the passive scheme (where there are still a couple of fans for case ventilation (perhaps PSU too?), then by all means sure.

The point of the quiet crowd is good enough performance, not best. For them, you might or might not be able to overclock, but you surely aren't going to be hearing their computer (or not easily).

[pdf27]

Originally Posted by Brians256

The point of the quiet crowd is good enough performance, not best. For them, you might or might not be able to overclock, but you surely aren't going to be hearing their computer (or not easily).

Yeah. The whole reason for hanging around here is so I can figure out the best way to overclock silently. The noise limit for me is a suspended quiet HDD (Seagate Barracuda IV or Samsung Spinpoint N) - anything audible over whichever of those I fit at idle is unacceptable to me. After that, it's just a matter of tweaking the design to get the maximum overclock out of it (not particularly for any reason, just because I can!)