Two Radiators-Parallel or in series?

[Armeniandave]

So my MM Extended Ascension should be here soon.
I have acquired three MCR320 that I need to split between my Q9550 and my 9800GX2/9600GSO.

I was planning on using one rad for the Video cards and the other two for the CPU or perhaps the other way around.
Right now the GPU doesn't get past 52C with a MCR220, so I am not sure two rads would be needed.

So here's the real question:
If I run the rads in series then the first rad gets all the heat and the second only gets the residual but if I run them parallel then both would potentially get the same amount of heat and therefor, in theory, would be more efficient.

So what do you think I should do?

[Snyxxx]

I would always recommend running the rads in series. I have monitored my in/out radiator water temps and they only differed by ~1 degree C at most.

Also, let the GPU suffer with one rad as they are usually volt limited in regards to overclocking. Main reason for me to cool the GPU is to get rid of the fan.

Give the CPU both rads as it would benefit more from the cooling provided.

[NAekuh]

Dave that cpu at most will put out about 160-220W extreme overclocked...

and upward 300W if your doing a suicide run.

The MCR320 will handle 300W and hold a good delta...

The Gpus you have the 9800GX2 is a hot biyatch... but your 9600GS isnt... even if you F@H i honestly think you can handle both on 1 mcr320. :P

use 2 in a series right after each other if you want quieter fans...

[Xilikon]

IMHO, either one will work very well but remember that it's very very tricky to install a proper parallel setup. Remember that water always take the path of least resistance (which is also the shortest) so if you were to have tubing shorter a bit on one radiator vs another, flow will take preference and skew the results. This is why most recommend a serie setup ;)

[RedRaider]

In series....

FUCK Parallel loops...

I'll be rolling (4) PA120.3 radiators in series for my CPU only loop...:)

[Armeniandave]

I agree and if I do go parallel it will be very important to maintain the exact same distance once the coolant starts its split and then when it comes back together.

I wish someone would have already done a test to compare the two setups with all other factors remaining constant.


IMHO, either one will work very well but remember that it's very very tricky to install a proper parallel setup. Remember that water always take the path of least resistance (which is also the shortest) so if you were to have tubing shorter a bit on one radiator vs another, flow will take preference and skew the results. This is why most recommend a serie setup ;)

[Septim]

me thinks, temps wise, series or parallel, the Cs will be close... it matters little...
but radiator placement and the where drawing in of air would matter more...

i would of course be happier avoiding those Y fittings and do simple series loop....

[Armeniandave]

Septim I agree with you. Either one will work fine.
The thing is I want to do something different than
what everybody else is doing, as long as the performance
difference is minimal.

As far as airflow I have the extended Ascension, so there will be at least 7 fans intaking on the front, if not all 9.
Granted airflow probably won't be a problem..:clap1:

Right now I have 1 MCR220 on my CPU loop and my temps are great. I plan on putting the 2 MCR320s on the CPU only. So I am already way overkill either way.

[Snyxxx]

The thing is I want to do something different than
what everybody else is doing, as long as the performance
difference is minimal.



Try NO fans. That would be different than most. :p

[Armeniandave]

Try NO fans. That would be different than most. :p

I said different not idiotic!
Just think how quiet it could be with no fans.
Oh well I'll never know with this case.

[RedRaider]

I agree and if I do go parallel it will be very important to maintain the exact same distance once the coolant starts its split and then when it comes back together.

I wish someone would have already done a test to compare the two setups with all other factors remaining constant.

Test it yourself and post results...

I am also of the opinion you should always pull outside the case air in through the radiators. I am also a firm believer in case air Over Pressure.

Again, these are my opinions, which will not be swayed.:fact:

[HESmelaugh]

RRR, I have some testing on temps in (aircooled) case with lots of different fan scenarios (push, pull, mixed, etc.). If you're interested, I could post it here in the testing section.

It's not directly related to liquid cooling, though.

[Armeniandave]

I don't think it's worth testing both ways.
Most people want to run them in series
and are not interested in running them
parralel.
If I did it it would only be to make it different,
not neccesarily better.
As far as air blowing in I wouldn't try to change
your opinion.

[Eternalightwith]

Do you have any observations that you've mmade that convinced you this was the best way to go?

David

Test it yourself and post results...

I am also of the opinion you should always pull outside the case air in through the radiators. I am also a firm believer in case air Over Pressure.

Again, these are my opinions, which will not be swayed.:fact:

[Armeniandave]

Do you have any observations that you've mmade that convinced you this was the best way to go?

David

Theoretically if I divide the heat load then each radiator would only have to dump 1/2 the load.
If I run them in series then the first rad dumps
most of the heat and the second doesn't get used as much.

This is how I think it will work but I can't find any real good
tests to show either way is better. Parallel loops have to be
exactly the same length so the water will split evenly. That's not a problem at all for me.

Series would be easier as anyone can simply hook them together.
What I wanted was a scientific approach to testing the two against
each other. I do not have the equipment or the time to do a proper
test so I'll just have to try one or the other.

[RedRaider]

RRR, I have some testing on temps in (aircooled) case with lots of different fan scenarios (push, pull, mixed, etc.). If you're interested, I could post it here in the testing section.

It's not directly related to liquid cooling, though.

Post them. They are not germane to how a liquid cooled system would react to internal case airflow, but It would none the less be interesting.

[RedRaider]

Dave, parallel is a flow killer. Why would you even want to split the flow?

Try it, you can always change it if you don't get the results you want.

With no disrespect intended, I could care less for running a parallel loop in any system I ever build.

[Eternalightwith]

Yes I know all that. ;)
Just wanted to know your thoughts :D.

David

[Armeniandave]

Joel

I know that you think it is a flow killer and the temps are gonna suck but I want to see facts from somebody who tested it say that.

If you think about it once the fluid comes back together the flow has to be the same again.

I looked at some of the threads from another site and I could not really find anyone who used it and said it sucks. I just want to have facts vs. people just not liking it cause they like something else. Also I have so much overkill for what I'm doing that it really doesn't matter which way I do it, I'm still gonna have good temps.

Sadly I don't think I'll find any hard data from anybody. Right now it's just peoples opinions.

Also I have not decided which way I will go yet I am just exploring options rather than just doing what everybody else is doing. :)

[RedRaider]

Joel

I know that you think it is a flow killer and the temps are gonna suck but I want to see facts from somebody who tested it say that.

If you think about it once the fluid comes back together the flow has to be the same again.

I looked at some of the threads from another site and I could not really find anyone who used it and said it sucks. I just want to have facts vs. people just not liking it cause they like something else. Also I have so much overkill for what I'm doing that it really doesn't matter which way I do it, I'm still gonna have good temps.

Sadly I don't think I'll find any hard data from anybody. Right now it's just peoples opinions.

Also I have not decided which way I will go yet I am just exploring options rather than just doing what everybody else is doing. :)

Well your dreams have come true.:king:

I'll be using a Iwaki RD-30 and (2) PA120.3 rads, but lets see if it makes a difference.

I will be using (2) of these...

http://farm4.static.flickr.com/3283/3048290214_d68a9fe51f_o.jpg

with these...

http://farm4.static.flickr.com/3245/3151766504_9712eb5682_b.jpg

Just waiting on the BP "Y" fittings to arrive from GARY.:king:

[Armeniandave]

I don't get it Joel, are you going to test it for me??
That's what I would like to see. If it proves that one way is
alot better than the other then I would do it the better way.
That's all that I have wanted was some real facts because everyone has opinions that are not always based on facts. I know I do.

BTW those were the fittings I was going to use or the full rotary ones.

[Snyxxx]

About damn time someone made some good Y's. Nice job there.

[Kayin]

Good to see another person testing. I'm gonna test parallel blocks into same rad. Mostly cause I have some strangeass routing in my case, but also cause I just wanna have some fun in testing.

[RedRaider]

I used BP Rotary 45's and non rotary 45's...

I have (16) BP Dual Rotary 45's coming...:slobber:

http://farm4.static.flickr.com/3265/3150913841_be465b9f18_b.jpg

Yes Dave, I'm going to test it for you. I don't plan on running my loop in this fashion. We'll see if it causes problems with the RD-30. That is the pump that will be used though.

[Armeniandave]

Well thanks RED!!
I just want to make sure were talking about the same thing.
Look at the attached picture and you'll see what I'm talking about.

[smee]

Why can't I put more pictures up???
WTF is going on.

Serv-er..... oooooooove--r..... lllooooooooooad.....
prepare fo-r c c c c cc c rash. :haha:

[Chilly]

IMO, series ALL the way, or a HYBRID loop ;)

[Armeniandave]

Chilly you may be right but reds gonna try it out and well see if there really is a big difference.

I have been researching alot on the subject and most people believe series is better but
unfortunately they have not tried it. I think opinions are great but with something like this
facts would be better.

If it turns out that series is better then I'll go that way, but if it turns out there about the same
then that gives me more options.

Should be a good test.