Which goes first? the Fan or the Radiator?

[kinghong1970]

due to limited space constraints, and some personality disorder of mine, i am really contemplating putting the radiator in the exhaust fan position... but have a problem in that i have to place the fan behind the radiator...

pic...


1) is it ok? i mean, is it ok for the fan to suck air through the radiator instead of pushing air through it?

2) it seems that the space between the CPU and back I/O panel is going to have very minimal air flow... should i be concerned?

pic...


thanks for your help guys.

[ibmkg]

Depends on you, whether you prefer space/portability vs. performance/cooling.

You never mentioned the FAN make. As far as I know, Computer FANs are designed to blow air and not suck air. If it is to be used for suction then some clearance is required. This can be explained practically :

Take a PC fan. When it blows air, take your hand and place it near the back of fan (from where it sucks air). Place another hand at the front (where it blows air). Now as soon as you remove your hand from back, you would immediately feel an increase in ‘air blow’. You can place your hand at the back at various positions to find optimal distance.

The 'suck through' method is used in cars and the only reason being to utilize air flow that is coming from outside and hitting the rad due to cars motion.

BTW, your setup looks attractive.

[kinghong1970]

thanks for reply ibmkg... that is a very interesting experiment you suggested...

this is my 2nd attempt at Water Cooling and on my previous try, i did it in a different way.

this is the setup i made for my brother's almost identical system



where i placed the radiator+fan assembly at the lower front air intake. this sacrificed a HDD cage and works with a regular sized PSU.

the PSU i have needs more clearance and when i tried the original setup as pictured on this thread, the HDD cage feels uncomfortably close to even applying too much pressure on the modular PSU cable sockets.

the way i see it, the only way i can mount a Fan to blow air is to space the radiator from the case about 1/2" but it i fear that i may not have enough room between the CPU block barb to the RAD barb...

... maybe i should have gotten a slimmer radiator?... he he he

as for the fan, the current 120x25mm fan is a stock fan i have laying around, probably stock Lian-Li fan.

if it helps any, i do plan to change the fan to a higher CFM fan... perhaps above 80 CFM-ish...

i did find this that i am contemplating on buying and installing to provide more airflow in the case...

[billbartuska]

Originally Posted by ibmkg

As far as I know, Computer FANs are designed to blow air and not suck air.

They can do either just as effectively. Fans move air by creating pressure differentials. The diffreential pressure between the intake and the ouput of the fan causes air to move to the area of lower pressure. It has nothing to do with whether the fan is "sucking" or "blowing".

But, generally, there is a larger "dead spot", created from the motor blocking air flow, on the exhaust side, and a smaller "dead spot" on the intake side.

[kinghong1970]

But, generally, there is a larger "dead spot", created from the motor blocking air flow, on the exhaust side, and a smaller "dead spot" on the intake side.

then by placing the radiator in FRONT of the fan, there will be a more even flow of air through the radiator than placing the radiator BEHIND the fan?

[blue68f100]

It all gets down to effeciency. It is well documented in AC work on commercial units it's allways more efficent to suck through a coil. Sucking will give you near 100% effiency. Blowing through a coil your effiency is normally less than <90%, depending on how much air gap you have. When blowing you generate high air flow through part of coil, with other parts near zero. So you only use 70- 80% of the coil.

But the key is having a air gap whether you blow or suck. Most of these small fans have limited static pressure they can generate. I know this does not really help with your confined space. You just need to test to see which gives you the best performance.

[ibmkg]

I think there is more to it. The design is a great factor. The coil of AC and its Fan would be designed in such a way to get max performance in suction. Suction design is primarily used as we want hot air to be removed from the AC unit.

Best would be to contact and ask the manufacturer. If possible, let us (me) know too. I would stress on using spacer so that fan can suck out hor air easily from casing. Also you might want to hack casing since the air that is to be discharged would find hurdle while passing through those holes.

Question: Why are Stock CPU HSF designed to blow air?

If suction were an efficient way, then there would be no reason to blow air to HS. Suction would also yield to suck out surrounding hot air near CPU and ultimately remove it from casing via auxiliary fan.

[ibmkg]

I forgot to conclude.

If you want the current setup (rad inside the case), then you have no choice other than to suck hot air out.

Be warned that even if you have 100% efficient airflow through the rad, the case temperature would be higher than the room resulting in the loop temperature to be a degree or two higher. The principal of water-cooling is to remove the heat buildup away from CPU and hence from casing efficiently.

From what I see, your rad is a Heater Core. If so heater cores are designed for air to be blown through it.

My conclusion could be wrong. Wait until a pro replies to my post and removes any misconceptions I have.

Nice hardware you got there.

[jaydee]

Suck, blow flip a coin. Tested both ways didn't get a measurable difference with .1C accurate equipment.

Adding shroud to space the fan from the rad not only helps with the dead spot in the middle of the fan but also seems to reduce noise.

[kinghong1970]

Suck, blow flip a coin. Tested both ways didn't get a measurable difference with .1C accurate equipment.

Adding shroud to space the fan from the rad not only helps with the dead spot in the middle of the fan but also seems to reduce noise.

now, that is something i was contemplating yesterday...

if i had a fan-shroud that is only 1/2" thick, then i can clear more space inside the radiator, if only by 1/2" and also, i can mount the fan outside of the case...

anyone know if there is a 1/2" fan shroud?

If you want the current setup (rad inside the case), then you have no choice other than to suck hot air out.

Be warned that even if you have 100% efficient airflow through the rad, the case temperature would be higher than the room resulting in the loop temperature to be a degree or two higher. The principal of water-cooling is to remove the heat buildup away from CPU and hence from casing efficiently.

yes, i know this is not ideal, i will be pulling air that has already been "warmed" by the MB and RAM... if the radiator mount was at the lower front of the case, then the air would be fresh/cooler... but again, my restraints...

From what I see, your rad is a Heater Core. If so heater cores are designed for air to be blown through it.

ok, dumb-a$$ question here... what is the main difference between a heater core and a radiator? and to the best of my knowledge, the HWLabs Black Ice GTX Extreme 120 is not a heater core, rather a radiator...

but again, i'm a noob, what do i know...

btw, i really appreciate the help and opinion guys... this is indeed a learning experience...

[billbartuska]

Originally Posted by kinghong1970

then by placing the radiator in FRONT of the fan, there will be a more even flow of air through the radiator than placing the radiator BEHIND the fan?

What are you talking about. front/behind??? it's pull or push...from either side.

[billbartuska]

Originally Posted by kinghong1970

anyone know if there is a 1/2" fan shroud?

...what is the main difference between a heater core and a radiator?

HWLabs Black Ice GTX Extreme 120 is not a heater core, rather a radiator...

but again, i'm a noob, what do i know...

There are no dumb questions....

As for shrouds, some places sell them, but they're pretty easy to make. Then you get exactly what you need.

People use automobile heater cores (and pretty much anything else they can get their hands on) for WC radiators. They work well, (Chevette/Pontiac heater cores, that is), but not as good as a purpose built watercooling radiator.

Here's some rad info to get you started:
http://www.thermochill.com/PATesting/
http://www.xtremesystems.org/forums/...ad.php?t=77260

And some more info:
http://www.overclockers.com/articles1088/
http://www.xtremesystems.org/forums/...d.php?t=137832
http://www.silentpcreview.com/article739-page1.html

[kinghong1970]

thank you very much billbartuska. the sites you linked were very enlightening...

i have found one vendor for shroud, but now i am a bit concerned about my radiator choice and the Swiftech GTX.

i keep hearing/reading about the GTX mixing metals in their block...
as for coolant, i am not trying to set a world record and hence i made an aesthetic choice of going with the Primochill banded coolant in red color.

will my choice of coolant, radiator and block have adverse reaction in the future?

[ibmkg]

If I were in your boots I would use spacers.

Use bolts and spacers (Al or Cu tubing as sleeve) between rad and fan. The gap could be sealed off using hot glue gun (glue sticks appropriately cut to size and glued, using some LEDs in between would make it look cool) or sponge nicely cut or packing material.

Look no futher on 'how to make shroud', we do have talnented people here:

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

(By #Rotor)

Then again if you did find a nice shroud, buy it and use it.

For mixing of metals, till date I have not witnessed galvanic corrosion myself. It is possible but in long term. Using a high coolant to water ratio and maintaining the rig voltage to zero (by properly grounding) would slow down the process. In any case, it would take years.

[Brians256]

If the coolant doesn't have anti-corrosion additives (pH buffers, etc...) and it is a copper and aluminum system, it can happen quite a bit sooner than years. The worst case that I've seen (second hand) is a system with a copper block AND an aluminum block where the aluminum was anodized except for a pinhole spots. Those spots became pits in months.

He's right, though, in that if you use a proper amount of cheap anti-freeze you won't see any corrosion. You might lose a tenth of a degree Celsius or two, but the peace of mind is priceless.

[billbartuska]

Originally Posted by kinghong1970

i have found one vendor for shroud, but now i am a bit concerned about my radiator choice and the Swiftech GTX.

I found these acrylic thingeys at The Container Store (http://www.containerstore.com/browse...3&PRODID=60168)

A little cutting and you have a shroud...and a pretty good looking clear one at that.

As for coolant...colors????!!!! ....I guess each has his own.
Distilled water, not tap, not deionized, not anything else.
5% Antifreeze, 10% if mixed metals.
A few drops on non alcohol based Iodine.

[bobkoure]

Originally Posted by ibmkg

Computer FANs are designed to blow air and not suck air.

Hunh?
I go away for a while and come back to this bit of information? I think I don't belong here anymore.

BTW, mostly you just need to make sure there aren't obstructions within a half-chord or so of the tops of the blades (you can just use the front-edge to back-edge length as "chord"). Google for "coanda effect" to see why. No problems with obstructions on the underside (that's why the struts are there - and why stuff gets attached to the undersides of airplane wings - if anything gets attached at all, of course).
I did a number of experiments with various fans and a heater core and found that pull was better for noise efficiency (lower noise for same cooling) BUT the only thing to take from that is that "it varies depending on fan and radiator", often stated as YMMV
Oh - and you can easily make a spacer if you need to move the fan away from the rad - just hollow out a "dead" 120mm fan and stack 'em...

[ibmkg]

Apologies for the post. I thought since 'suck air' requires clearance, I concluded that could be due to design. Since blow air does not require any clearance (or little when compared).

[Brians256]

Nope. I've had marginally better results sucking air from a rad. But, I'll pick whatever is easiest to implement, these days.

[ElMoIsEviL]

I know I'm late but if it makes you feel any better I get better performance using the Black Ice GT Stealth (120 & 240) by sucking air through rather then blowing air through.

Not sure why.. might have something to do with the fin density. With my BIP and BIX there was no measurable difference.

[billbartuska]

On the "push" side of the fan the airflow tends to move in a spiral. The motor also blocks some of the airflow. This doesn't happen on the "pull" side though.

Depending on the fan design, and how it is mounted, this may or may not effect performance. Generally, a shroud lessens these effects.

[ElMoIsEviL]

Originally Posted by billbartuska

On the "push" side of the fan the airflow tends to move in a spiral. The motor also blocks some of the airflow. This doesn't happen on the "pull" side though.

Depending on the fan design, and how it is mounted, this may or may not effect performance. Generally, a shroud lessens these effects.

That makes sense then. Higher Fin density means more fins stuck in the dead spot. Low fin density means less.

Because it's like a 5C - 10C difference I'm seeing (using Watercooling and TEC's here) with the BIGTS 120 + 240 vs no difference at all measurable on my old BIX2 + BIX setup.

[billbartuska]

Originally Posted by ElMoIsEviL

Higher Fin density means more fins stuck in the dead spot. Low fin density means less.

No.

"more fins stuck in the dead spot" = more fins not stuck in the dead spot (all the other fins!).

Your new rad just performs better with your fans...that's all.

[ElMoIsEviL]

Originally Posted by billbartuska

No.

"more fins stuck in the dead spot" = more fins not stuck in the dead spot (all the other fins!).

Your new rad just performs better with your fans...that's all.

Hmm I guess so. They look sexy too (the rads)

[billbartuska]

Originally Posted by ElMoIsEviL

They look sexy too (the rads)

Pickies?