rocketfish all internal huge loop

[Waterlogged]

why the AP-15s vs the yate highs the yates are rated at a higher CFM the ap-15s are ~58CFM and the yates are 88CFM

For your case, I'd suggest the 2150's when they become available. I don't have access to actual flow numbers of a SH so I'll take what I do have available to me and that would be SM (Petra's SM's to be precise). A AP-15 is rated (as you have stated) @ 58CFM (that's open air), when it gets on a rad like the MCR, it drops to ~35/36CFM. The SM is open air rated @ 70.5CFM but when it's on the same MCR, it only pushes ~37CFM...now here's the real kicker, the GT is ~ 8dB lower. It's all about static pressure and the GT has far more than the Yates ever will and that is what's needed to push through stacked rads. The 2150 RPM GT's are rated for 69CFM (open air) so should work a bit better. If noise truly isn't a concern, the newer model GT's (3K, 4.2K and 5.4K) may be of interest to you.


Fan on rad numbers taken from this topic.

http://www.xtremesystems.org/forums/...d.php?t=261778

[cdawall]

For your case, I'd suggest the 2150's when they become available. I don't have access to actual flow numbers of a SH so I'll take what I do have available to me and that would be SM (Petra's SM's to be precise). A AP-15 is rated (as you have stated) @ 58CFM (that's open air), when it gets on a rad like the MCR, it drops to ~35/36CFM. The SM is open air rated @ 70.5CFM but when it's on the same MCR, it only pushes ~37CFM...now here's the real kicker, the GT is ~ 8dB lower. It's all about static pressure and the GT has far more than the Yates ever will and that is what's needed to push through stacked rads. The 2150 RPM GT's are rated for 69CFM (open air) so should work a bit better. If noise truly isn't a concern, the newer model GT's (3K, 4.2K and 5.4K) may be of interest to you.


Fan on rad numbers taken from this topic.

http://www.xtremesystems.org/forums/...d.php?t=261778

good call and thanks for the link however looking through there the mediums dont do to bad either pretty average overall

[PiLsY]

This will work a lot better with really low ambients as well. As a benching rig with screaming fans that runs outside I'm with you keeping it as cheap and portable as possible.

I think people are failing to realise that this isnt the kind of daily use rig you normally see on here.

WL has it spot on with his pic - intake rad should always be on the outside of the sandwich.

Forget GT's and all that - find somewhere online that breaks old servers and get some 7k rpm sanyo denki 120mm. I have 3 here - they are truly immense. 3.3a each though so you need to be careful powering them. They'll cost you very little and do the job perfectly. I can get you a model number but theyre usually just listed by size, pin type, manufacturer and rpm. If you need it post up and ill grab a pic. Im pretty sure these came from a Dell.

[RCG_Bex]

While I agree with all but one thing Sideroxylon has just stated, I also don't think the MCR stackers have been properly tested yet.

In the pic below, the cold air would be moving from left to right. With AP-15's or the 2150's, I think would work reasonably well.






It's on, just not for everyone. I had a major problem a long while back with a bunch of ppl with 0 posts PM'ing me asking the most inane questions that it got to the point that when I was done answering them, I was too tired to view the rest of the forum. Most everyone that has access to me is judicious in PM usage. I'll give you a trial membership.

This is exactly how I planned on doing mine. That way my CPU gets the coolest water. But like this...

Hot water Cold water
\/ /\
>>> FAN >>> RAD >>> SHROUD >>> FAN >>> SHROUD >>> RAD >>> FAN >>>
>>> FAN >>> RAD >>> SHROUD >>> FAN >>> SHROUD >>> RAD >>> FAN >>>

I'd like to try some slim shrouds if I can but we'll see...

[cdawall]

This will work a lot better with really low ambients as well. As a benching rig with screaming fans that runs outside I'm with you keeping it as cheap and portable as possible.

I think people are failing to realise that this isnt the kind of daily use rig you normally see on here.

WL has it spot on with his pic - intake rad should always be on the outside of the sandwich.

Forget GT's and all that - find somewhere online that breaks old servers and get some 7k rpm sanyo denki 120mm. I have 3 here - they are truly immense. 3.3a each though so you need to be careful powering them. They'll cost you very little and do the job perfectly. I can get you a model number but theyre usually just listed by size, pin type, manufacturer and rpm. If you need it post up and ill grab a pic. Im pretty sure these came from a Dell.

that would be great i have a good group of ultrakazes but those are nothing compared to sanyo's i have 3 4500RPM 120x25mm fans that have some kick to i was going to put those on the backside of the sandwich.

i am glad that someone caught on that this isn't a 24/7 rig lol. it just kinda moves and i bench with it last chips i pushed on it was 3150 on some sempron s754 chips.

[NaeKuh]

sigh.

Guys u want to space the middle fan out so it has room to pull.

Spacing the fan out allows the dead zone in the middle of the fan to be smaller.
By doing so, you get the maximum amount of surface area though the rad, instead of the typical dead zone in the middle.

This is why i have been telling people if your going to stack, sandwitch it between shrouds and a high powered 38.

In the pic below, the cold air would be moving from left to right. With AP-15's or the 2150's, I think would work reasonably well.

it doesnt matter.

Your under the assumption that the water inside rad 2 will be significantly higher then the water inside rad 1, when the truth is, the delta air which the hotter rad spits out will not be higher then 1C at most.

Which direction you have is moot, because of the rules of closed loops.

You just wont get enough heat in the system for you to have to merit a specific order like that, unless your flow rate is very very low.

[RCG_Bex]

@ Naekuh - well this is my question - because I had already through of shrouds in the middle (I reliased If orgot to post it in my diagram last night - far too tired! But I'd seen your post on this before!!!), but how big would the shrouds have to be minimum? Could I get away with 10mm shrouds? Or would it have to be more?

~Bex

[Alexandr0s]

Well, assuming Martin's test of radiator shrouds goes the same on stacked radiators, I'd get a 20mm shroud on both sides minimum. Probably easiest would be to use 120x120x25mm fans as shrouds.

[Waterlogged]

it doesnt matter.

Your under the assumption that the water inside rad 2 will be significantly higher then the water inside rad 1, when the truth is, the delta air which the hotter rad spits out will not be higher then 1C at most.

Which direction you have is moot, because of the rules of closed loops.

You just wont get enough heat in the system for you to have to merit a specific order like that, unless your flow rate is very very low.

NaeKuh, that's rule #1 in my book...I haven't forgotten at all. Take a look at the OP's requirements again and you'll see it's a perfect fit setup for what he wants. Not all of us are using MM cases that have 3000 cubic feet and this is a viable method for fitting extra rads in the space available.

Maybe it's time for you to do a micro build.

[cdawall]

NaeKuh, that's rule #1 in my book...I haven't forgotten at all. Take a look at the OP's requirements again and you'll see it's a perfect fit setup for what he wants. Not all of us are using MM cases that have 3000 cubic feet and this is a viable method for fitting extra rads in the space available.

Maybe it's time for you to do a micro build.

thank you for sticking with the original design for me.

as for your rad setup i will not be able to do that for the top ones due to room but the lower ones it should be easier to do that lol.

also no way i will be able to stuff 20mm shrouds anywhere the case isn't that wide lol

[RCG_Bex]

Yeah I have the same issue. If I chucked the outer 38mm's for 25mm's then I could put bigger shrouds in, but I think the biggest I could manage currently is 10mm

[Aaronharmon]

I think all this talk against rad stacking is conjecture and nonsense. for one thing the arguments against it don't make much sense.

1) These are low FPI rads sandwiched between 3 fans. You do not need a leaf blower or a dyson to get air moving through there.

2) The other argument was that you are blowing heated air across the second rad. IMHO, BS, plain and simple. The delta t between the incoming and outgoing air is so small that the impact would be insignificant.

There are arguments against heated air that are valid. I wouldn't use the air coming off an oc'd gtx 480, but as long as the rads are in a good cool spot, I say stack away. At least until someone coughs up the data proving otherwise.

Aaron

[Sideroxylon]

I think all this talk against rad stacking is conjecture and nonsense. for one thing the arguments against it don't make much sense.

1) These are low FPI rads sandwiched between 3 fans. You do not need a leaf blower or a dyson to get air moving through there.

2) The other argument was that you are blowing heated air across the second rad. IMHO, BS, plain and simple. The delta t between the incoming and outgoing air is so small that the impact would be insignificant.

There are arguments against heated air that are valid. I wouldn't use the air coming off an oc'd gtx 480, but as long as the rads are in a good cool spot, I say stack away. At least until someone coughs up the data proving otherwise.

Aaron

Did you bother to click any of the links showing the difference between stacked rads and independent rads? They clearly show that 2 independent rads receiving fresh cool air perform better than the stacked rads.

I'm not sure where you get your info regarding the temp of air coming out of a rad. If your rad is doing it's job it should be close to your water temp. As an example, with an ambient of 20c my water temp is 24c and air coming out of the rad is 23-24c. Being that water temps are very dependent on the temp of air being fed to the rad that's a significant temp swing. The performance of the second rad receiving that heated air will be reduced in comparison to the first rad receiving fresh cool air.

It's not so much a question of whether stacked rads work or not but their efficiency when compared to individual rads.

[pluck]

I think all this talk against rad stacking is conjecture and nonsense. for one thing the arguments against it don't make much sense.

1) These are low FPI rads sandwiched between 3 fans. You do not need a leaf blower or a dyson to get air moving through there.

FPI is only part of the issue.

What determines the best fan speed for a rad is surface area per fan. By creating a rad sandwich you double the surface area per fan thereby increasing the optimal fan speed.

[NaeKuh]

NaeKuh, that's rule #1 in my book...I haven't forgotten at all. Take a look at the OP's requirements again and you'll see it's a perfect fit setup for what he wants. Not all of us are using MM cases that have 3000 cubic feet and this is a viable method for fitting extra rads in the space available.

Maybe it's time for you to do a micro build.

Dude WL we had this arguement a while back ago.
You growing senile on me?

Your model suggest using a fan @ each rad with a total of 3 fans.

3 x 25mm = 75mm

Mine uses 2 25mm + 1 x 38mm = 88

Do we need to go about this again?

And he's not stacking monsters.

@ Naekuh - well this is my question - because I had already through of shrouds in the middle (I reliased If orgot to post it in my diagram last night - far too tired! But I'd seen your post on this before!!!), but how big would the shrouds have to be minimum? Could I get away with 10mm shrouds? Or would it have to be more?

~Bex

U dont need that big of a gap.

Guys.. its simple... you stack the fan on the rad u get a big honking dead spot in the middle.
you stack a fan on the other side, the dead spot is still in the middle.
You stack a rad again, and once again your carrying the DAYAM DEAD SPOT TO THE OTHER RAD.



Because you have a dead spot in both rads, your efficiency is dropped more then if it was a single RAD.

The gap in the middle does not need to be big. San aces for example only need a slight space out to get the full vortex on it blades to kill the dead spot.

If you truely have no room, a Tribewerk in the middle would be better then how u guys have it modeled so far.

the main point in stacking rads for the 100th time ONLY WORKS when you mitigate that dead center the fan brings in on BOTH STACKED RADS.

Otherwise u guys dont understand the function in stacking rads, and should not go and attempt it.

FPI is only part of the issue.

What determines the best fan speed for a rad is surface area per fan. By creating a rad sandwich you double the surface area per fan thereby increasing the optimal fan speed.

EXACTLY! and what do you do by carrying over the dead spot on both rads? :P

[Sideroxylon]

Dude WL we had this arguement a while back ago.
You growing senile on me?

Your model suggest using a fan @ each rad with a total of 3 fans.

3 x 25mm = 75mm

Mine uses 2 25mm + 1 x 38mm = 88

Do we need to go about this again?

And he's not stacking monsters.



U dont need that big of a gap.

Guys.. its simple... you stack the fan on the rad u get a big honking dead spot in the middle.
you stack a fan on the other side, the dead spot is still in the middle.
You stack a rad again, and once again your carrying the DAYAM DEAD SPOT TO THE OTHER RAD.

Because you have a dead spot in both rads, your efficiency is dropped more then if it was a single RAD.

The gap in the middle does not need to be big. San aces for example only need a slight space out to get the full vortex on it blades to kill the dead spot.

If you truely have no room, a Tribewerk in the middle would be better then how u guys have it modeled so far.

the main point in stacking rads for the 100th time ONLY WORKS when you mitigate that dead center the fan brings in on BOTH STACKED RADS.

LOL, you guys are getting Naekuh's panties in a bunch.

The OP has already pretty much stated he's not interested in an optimized loop but seems more concerned with just filling a case with rads regardless of efficiency and actual performance gains/needs. He already has about 4X too much rad for his current setup.

The discussion of the 'rad sandwich' should probably be taken to its own thread so you can really get Naekuh wound up.

[NaeKuh]

LOL, you guys are getting Naekuh's panties in a bunch.

[Alexandr0s]

Forgive us, oh Great NaeKuh, for we are mere mortals that know nothing of the mystical world of watercooling

[NaeKuh]

Forgive us, oh Great NaeKuh, for we are mere mortals that know nothing of the mystical world of watercooling

so wait i got my diety?

when did this happen?

[cdawall]

So what about no center fans and using a shroud for the center?

[NaeKuh]

So what about no center fans and using a shroud for the center?

the point is the dead spot which u are isolating.

You guys, maybe you dont fully understand the concept or something im saying is not making sense.

You see the air coming out a rad was tested by martin.
A MCR320 has the capacity of using roughly 60% of the air which it takes.

When using 60% of the air, it also means only 60% efficiency, because 100% efficiency would mean a super radiator.

So your taking the lost potential off the first and using it on the second.

This is the methodology of stacking radiators.

So you want the lowest possible variation from rad 1 to rad 2.
That just wont happen when you got a big dead spot on both rads.

You lose effiency off the first, hence your second will suffer.

This is what your trying to minimize.

But increasing the total surface area on which the air is pulling though the entire rad, you get a better recycling of air, which in turns allows a better performance on the second rad chamber.

You guys carrying over dead spots, is not a true stack.
Its a half rushed stack, which doesnt have much performance over a single rad layed out on itself.

And trust me, its been tested many many times, to the point where we got rid of stack rads, because of how difficult it was to apply it in working situations.

It comes back to the ancient proverb:

Do it right, or dont do it at all.

[Alexandr0s]

Naekuh's right. The main problem with stacking is that at first it looks like a good idea to fit many radiators into a cramped space, but when you try to do it right, you realize that you usually are using more room than with 2 seperate radiators, which give better performance anyway. Stacked radiators are great, but until we have fans with shrouds both below and on top of the fan, I don't know to what extent it's worth it.

And Naekuh, I'd say you're not even close to being a deity. A demigod perhaps, but deity? No way .

[RCG_Bex]

@ Naekuh : Ok so what about if you had 2x 38mm kaze 3000s (the 2nd pull with a shroud) with a 55mm tribewerk 1800 in the middle? I could manage a kaze with 10mm shrouds either side in the middle, or 7mm shrouds in the middle and 2nd pull fan... but you've clearly got the better experience / judgement on this 1

1. >>> 38mm >>> rad >>> 55mm tribewerk >>> rad >>> 38mm >>>
2. >>> 38mm >>> rad >>> 10mm shroud >>> 38mm >>> 10mm shroud >>> rad >>> 38mm >>>
3. >>> 38mm >>> 7mm shroud >>> rad >>> 7mm shroud >>> 38mm >>> 7mm shroud >>> rad >>> (possibly 7mm shroud) 38mm >>>

I would say hey, lets swap down to 25mm's but i always thought the static pressure was greater on the 38mm's but then, I've never really used shrouds correctly before... That and I also have a lot of 38mm kazes kicking around....

~Bex

***EDIT***
If I can get the koolance rads shipped then I have more room...
1. >>> 38mm >>> 7 (10 maybe) mm shroud >>> rad >>> 55mm tribewerk >>> rad >>> 7 (10 maybe) mm shroud >>> 38mm >>>
2. >>> 38mm >>> 7mm shroud >>> rad >>> 7mm shroud >>> 38mm >>> 7mm shroud >>> rad >>> 7mm shroud >>> 38mm >>>

[Alexandr0s]

Bex, the problem with the Triebwerk's is that they only have a shroud on one side of the fan. Meaning you will only solve half the problem . You'll need some sort of shroud on top of the triebwerk to remove the dead spot there as well.

[PiLsY]

That and the fact that you'll have to rent out your mum to buy them...

Arguing about the efficiency of stacks in this case is irrelevant - Its got to be portable so within the confines of that case. Stacked rads are better than a single - especially when using silly fans. If you don't have the room for extra free air rads then its irrelevant if thats a better config than stacked.

I think we can all agree that a stack is better than a single rad?

Cdawall - if you want to optimise the stack listen to Naekuh . Otherwise just stick to silly rpm fans and its all the same anyway...