Is it worth the trouble to vent radiator output outside the case?

[Ruiner]

Most (well designed) internal watercooling setups have the radiator drawing air from the front or top of the case.

Assuming the goal of a quiet setup (i.e. one slow turning 120mm exhaust fan and perhaps a fanless PSU), is the waste heat of the radiator a big issue? The dT over ambient air isn't huge, but that heat is still going into the case. It would mostly affect NB, mosfet, PSU, HDD and vRAM cooling.

I'm kicking around some ideas for my next rig. One is to mount my fan/BIP2 combo to the top of a mid tower case so that ducted fresh air moves through filtered openings in the drive bays (or case sides), through the rad, then the fans, then out through blowholes. I've found the airflow noise on the intake side of my BIP is louder than the fan side.

Another would be to do the same thing pointing down through the case floor, with ducted intake air coming through a filtered front/lower intake. This would be the quieter solution but the longish BIP2 might interfere with the mobo.

Another solution would be to mount a pa160 in the traditional bottom front, place an acoustically dampened barrier 2" behind the fan, and vent the exhaust through the case bottom or sides.


Thoughts? Flames?
Any direct experience?

[Cathar]

A little bit of logic is required here. In a typical air-cooled system all the heat is being dumped into the case anyway, and the only way that gets removed is via exhaust fans. Exhausting the radiator into the case pretty much gives the same effect. The hottest bits are still being cooled well by the water-cooling, and everything else is about the same as it was with regular air-cooling.

Far more important to draw in fresh cool air through the radiator in terms of achieving better temperatures on the hot bits that generally affect cooling performance the most. In my experience the memory and mosfets and northbridge don't really care that much if they're 4C warmer than at some other time, unlike a CPU which in general does care to a noticable degree.

So yeah, exhausting into the case is fine IMO. If you want to achieving cooler in-case temps then exhaust the air a little better, but overall the ambient+2-4C air coming out the radiator isn't really going to affect anything much at all.

[HammerSandwich]

My quiet build had an initial problem with the PSU fan ramping up. I never took internal air temps, but using minimal airflow certainly gave high air temps exiting the rad. The solution was to divide the case into two thermal zones, so the PSU doesn't breathe the heated air. Details here.

[GlassMan]

Note that warm air exhausted at the bottom will tend to surround and be drawn back into the case.
The equilibrium temp of the water will be determined by the temp of the incoming air to the radiator. The temp of the none watercooled components will be determined by the air temp in the case. Cooler is better. (Everything else being equal).
Cathar , no doubt, has the priorities correct, getting the coldest air possible to the radiator. But if you care to try some thinking, modding and ducting, you can get the air temps in the case as low as you can as well.
Tagan (through the psu venting) and seasonic (bottom fan venting) are very quiet psu's that are more powerful ans will cost less than a fanless one (which you wil have to drive air through some how to keep cool). I'm assuming some overclocking, since a reserator will do the job at stock for a basically silent computer.

[Ruiner]

My current plan is the 3000slk: http://www.newegg.com/Product/Showim...+Case+-+Retail

A pa160 wil fit nicely after removing the lower drive bays. A passively cooled Antec Phantom 350 will power it, with a blow hole directly over the psu to help cool it. I may put a vertical partition in the case so that exhaust from the pa160 flows straight up and out one of the 5.5" bays.

The 120mm exhaust will draw fresh air from the side panel to cool mosfets, NB, vram and such.

[Joe]

http://www.procooling.com/articles/h...uber_lan_b.php

My lil testament to venting the radiator out of the case.

To me it seems rather counter to the purpose if you vent the hot case air through the radiator or vent the radiator into the case.

[jaydee]

Originally Posted by Joe

http://www.procooling.com/articles/h...uber_lan_b.php

My lil testament to venting the radiator out of the case.

To me it seems rather counter to the purpose if you vent the hot case air through the radiator or vent the radiator into the case.

Pretty sure I never seen that article. Not bad at all. That is a very similar concept I want to do but make the PC about half the size. Isn't it interesting back then that was a "mini" PC? Today that is considered mid sized.

I see Chip screwed you on that block and made excuses. LOL. Anyone know what he is up to today?

[Ruiner]

Wow, old school, Joe. I used to have one of those radiators, mounted exactly like you did here: http://www.procooling.com/articles/h...v2_-_joe__.php , elbows and all.

I'm looking to do a 'smallish' tower, but still want a full atx board....the matx offerings still lack OC options in bios for some reason.
This case http://forums.silentpcreview.com/viewtopic.php?t=19618 is a full 2" shallower than the 3000slk. That translates to ~4" shorter and shallower than my current FongKai beast, or the new p180.

It will be tough getting my 1048 in there, though.

edit: pics of the slk2650 loaded: http://forums.silentpcreview.com/vie...=186888#186888

[stev]

Ruiner,

The Swiftech H2O and the Corsair COOL both hang the rad off of the back of the case to eliminate both issues dealing with air in or out of the rad for the case.

Just a quick thought!

[Brians256]

I will say that putting the fan inside the case lowers the noise compared to an external rad.

[Fillman]

My quiet computing kit consists of LR Storm G4, BIP2, case is an Antec Sonata:



Key noise reduction elements:
- Removal of stock 120 exhaust fan, true ATX LOL.
- Replacing stock power supply with Enermax Noisetaker
- Pumping duties via submerged pump.
- Low RPM fans (papst) on the rad (natch)

While the PS is not fanless it is certainly quieter then the stock solution and maintains exhaust flow in the absence of the case fan, which I believe offers overall greater benefit noise wise. Enermax did a nice job with the Noisetaker series IMO.

I prefer submerged pumps for silence, even with my ear within a couple inches I cant tell its on due to noise isolation offered by the res.

I used to run nosiy fans and pushed the oc's with no observable benefit of the extra cpu cycles. Today I much prefer using the same gear for the noise benefit. Silence is golden.

[ricecrispi]

Fillman
Noisetaker is above average silentwise compared to most psu but you can do better.
I would run a 120mm low rpm exhaust fan because they will reduce temps into your psu and that will probably lower noise from the psu. Im pretty sure enermaxs run off thermosistors and hotter they get, the faster the fan speed, and louder they get. I would get nexus or coolmaster fan for the exhaust fan. I would also add artic cooler or NVsilencer on the vidcard. Total shoud cost $40 l and make the setup really quiet.


I'm with cathar and believe exhuast air is good inside the case as long as you have correct airflow in and out. If you duct and isolate the exhaust away from the mobo components they wont get any air flow to remove heat from mosfets etc. Those components need air to cool and lack of airflow will lower performance of those components. So if you do isolate the rad exhaust you stilll need additional air cooling or another intake for these components.