• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

WC Routing Question

S

Saint Dymphna

n00b
Joined
Jul 1, 2003
Messages
19
Ok heres the question and big thank you in advance to any answers i may recieve.

I have 1. CPU waterblock
2. 5 1/4 bay Res
3. 2x swiftech 80mm radiators
4. A pump (swiftech)
5. 2x T junction (as opposed to Y junc) for the hoses.
6. All the tubing and fittings otherwise needed.

My question is....... can i use the T junction to hook up the rads in parallel the same as i see people using Y junctions to do the same. If i do this does it matter if one rad is higher than the other or should i try to get them on the same level. Im worried that having them at different levels will cause one to work less efficiently than the other. Or.... can i safely put (for example) one at a blowhole and one at the rear fan mount, or even put one above the other on a side panel. Ive also considered putting the back to back with the T junc above them.

Im new at the custom systems so any input is apreciated. Ive been running a exos nonstop for a few months now and it hasnt failed me so im pretty stooked on that but i want more!!!!!

My routing as is is going to be Pump-->Res-->Rad(s)-->CPU as is...... if another config works better id love to know. I thought about doing Pump-->Rad-->Res-->Rad-->CPU-->Pump..... but that seems to defeat the pupose.

Ok alot of info im askin for there. Hope someone can help.

Thanks alot,

SD
 
From what I remember, you need to follow the zen of water flow.

RES > PUMP > CPU > Y fitting > RADIATORS > Y fitting > RES

The reason the Y fittings need to be there is so that your flow stays up, pressure stays up, and you get even feed through both sides.
 
Hmmmm.... Why would i put the Rad's after the cpu? Dont i want the water as cold as possible when it hits the CPU block? Res-Pump-Rad-Cpu would work though? Wouldnt it....?

SD
 
I'm running my setup like you are talking about RES > Pump >RAD > Y fitting > CPU and GPU > back to RES. Works fine for me and keeps everything pretty dang cool.
 
Are you running 2 radiators? Maybe 1 Swiftech rad is enough? Nah! i want all the power i can get. I was looking around again and ive seen more than a few set ups with 2 radiators like mine.... one above the other but they are all using Y fittings.

SD
 
Originally posted by Saint Dymphna
Hmmmm.... Why would i put the Rad's after the cpu? Dont i want the water as cold as possible when it hits the CPU block? Res-Pump-Rad-Cpu would work though? Wouldnt it....?

SD
Click to expand...

Going Pump->reservoir->radiator->CPU->pump is quite possibly the WORST setup for watercooling.

You always want your reservoir going to the pump inlet because all flow is killed in the res. Pusing water through it would make your flowrates almost stop dead, so the way to keep flow (and pressure through the blocks) at its highest is to go reservoir -> pump inlet.

As for going "pump -> CPU -> rad" OR "pump -> rad -> CPU", there is virtually NO difference between the two. The laws of thermodynamics state that the water in your watercooling loop is the same temperature no matter where it is...not even right after the radiators. Thats just the way it works, trust me. Do whichever works better for your tubing and internal setup in your case.

For the radiators, its best to put them in parallel ( Y fitting to split flow -> Rad1 and Rad2 seperatly -> Y to combine flow back from the radiators). I would strongly recommend against using T lines to split and combine flow. They would kill your flow and create problems if you have uneven tubing lenghts. Get some proper Y fittings if you can so you can even out pressure and flowrates through the radiators.

-OMP
 
Most excelent. TYVM for the advice, its been quite awhile since my last physics class so i was unaware about the temp having to be the same. As to the using Y split ill go find some of those tomorrow and do my best to make the tubing the same length. Does it matter if the water has to rise higher for one Rad than the other? If not ill jsut put the 2 Rads in my side door where i have 2 openigns already.

Thanks again for the advice.

SD
 
Originally posted by Saint Dymphna
My routing as is is going to be Pump-->Res-->Rad(s)-->CPU as is...... if another config works better id love to know. I thought about doing Pump-->Rad-->Res-->Rad-->CPU-->Pump..... but that seems to defeat the pupose.
Click to expand...

I agree, this is the worst order of WC setup. This is so gonna kill the water flow and the pump will be screamin for help. I think the first two in the order(ALWAYS) is RES>PUMP.
 
Ah so I guess I should have my T-Line > Pump inlet > Radiator > CPU > GPU > T-line for good flow through the system? I suppose I can pull it off :)

flip
 
Originally posted by DaveX
The flow is the same at every point in the loop. Read this thread.
Click to expand...

Duh... Nobody here was trying to make that point so I don't understand why you posted this. I can speak for myself when I said that its best to put the reservoir before the pump because your flowrates will be the highest. Its true, and if you put the reservoir anywhere else in the loop your flowrates will be lower for the entire loop.
 
Poptart, I like you.


I think ur perdy smart.


but.


Think, please, for a moment, about what you are saying. If you put the res after the pump, somehow your flow rate will decrease after that point?

if you are flowing 100 GPH before the res, then you are going to be flowing 100 GPH after the res as well. Period. If you are flowing 100 GPH before the rez, and only 90 GPH after the rez, you will be adding 10 gallons per hour to the rez and it will either explode, overflow, or be time-warped into the twilight zone.

I studied hydraulics in college as part of my schooling in horticulture. I design and install professional irrigation systems. Trust me, you cannot have different flow rates in different parts of a system.

EDIT -- (assuming the system is a closed loop, and excluding points where there may be a parallel setup dividing flow between 2 parts of the loops, where combined flow will add up to the flow rates elsewhere in the system.. wanted to prevent possible picking-apart of my thread real quick.. thanks)

You can have different pressures (you WILL have different pressures) and flow VELOCITY, but flow rate will remain equal throughout the loop.


I, personally, placed my rez before my pump becase I wanted direct pressure from the pump into the waterblock. Pulling water through the block just doesn't seem like it would accomplish what the RBX is designed for, though there is some conflict in my mind on this issue as well.

However, I felt pump-cpu-gpu-chipset-rad-res was the best way to do it which means I agree with you.

Just wanted to point out that in no way shape or form does the positioning in the loop of the reservoir affect flow rate. It will be the same regardless.
 
Astroman, thank you for that post. After doing some research and reading alot about flow rates, thermal blah blahs, etc etc. Im going to go to my case tonight and just work out where stuff should be in the case and hose it all up accordingly. I love doing this stuff. Im a total amature but its so fun and satisfying to tinker. In the end thats really all that matters. Otherwise id just run a Dell.

SD
 
Poptart, I like you.
Click to expand...
Thanks Astro, I like you too.

I think ur perdy smart.
Click to expand...
I'd like to think my physics degree is going to good use here at [H]:D

Think, please, for a moment, about what you are saying. If you put the res after the pump, somehow your flow rate will decrease after that point?
Click to expand...
Nope:D Not what I said.

if you are flowing 100 GPH before the res, then you are going to be flowing 100 GPH after the res as well. Period. If you are flowing 100 GPH before the rez, and only 90 GPH after the rez, you will be adding 10 gallons per hour to the rez and it will either explode, overflow, or be time-warped into the twilight zone.
Click to expand...
True. This is a simple law of physics that can not be broken.

I studied hydraulics in college as part of my schooling in horticulture. I design and install professional irrigation systems. Trust me, you cannot have different flow rates in different parts of a system.
Click to expand...
Yup. The flow rates will be EXACTLY the same in ALL parts of the system. See previous post when I responded to DaveX.

You can have different pressures (you WILL have different pressures) and flow VELOCITY, but flow rate will remain equal throughout the loop.
Click to expand...
Yup. Pressure would be different depending on where you place your reservoir. Which leads too.....

Just wanted to point out that in no way shape or form does the positioning in the loop of the reservoir affect flow rate. It will be the same regardless.
Click to expand...
Nope:D Let me explain:

There is no physical way flowrates can be different in different parts of the watercooling loop. Its not physically possible. The one thing that does change is pressure. This pressure is what determines the flowrate through the system. As you have pointed out, placing the reservoir after the pump kills pressure. This lower pressure makes the flowrate FOR THE ENTIRE SYSTEM lower overall. By putting the reservoir leading to the pump inlet, you minimize the pressure drop in the reservoir which leads to higher flowrates. Flowrates do not magically dissapear in the reservoir, but the pressure drop it creates does lower the overall flowrate in the watercooling loop. Thats why it is best to place the reservoir before the pump rather than right after it.

-OMP
 
Ok ok..... how about res-rad1-pump-cpu-rad2-res?

This would allow my system to be fairly clean still.......

SD
 
interesting point Poptart.

I still take issue to it, however.

It is true that when water exits a pressurized line, into open air, pressure becomes zero. However, in a sealed reservoir, meaning air and water tight, I don't see a pressure drop on any real scale.

Yes, there will be a pressure drop due to turbulence and hence frictional losses, but there are also the same drops through all the hoses and the heatercore, and the blocks as well.

I suspect the actual pressure on each side of a sealed and filled reservoir to be very close to one another, especially at the low flow rates (measured in GPH not GPM) we are using here.

In irrigation systems, where flow rates often exceed 10 gallons per minute in residential areas (much higher in commercial applications such as golf courses), pressure loss due to friction in pipes etc is huge. However, in low-flow instances such as a water cooling system, I would expect them to be Much, much less significant. Of course, running 1/4 inch tubing vs. 1/2" ID tubing is going to be a significant difference using the most powerful of pumps, because 1/4" tubing has insane pressure losses even at low flow rates.

Still, I don't imagine the pressure drop through a sealed reservoir to be as huge as you are making it out to be. I suppose this could be tested somehow, but I don't feel like placing fittings in my system to attach a guage to measure dynamic pressure while running.

hmmmmmmm.... BTW, I did re-read your post and it is true, you didn't say that flow rates would vary.. I must have misread something somewhere. You retain your reputation as having mostly intelligent postings ;) lol
 
I'm confusing myself... I didn't pay attention in class during the part about pumps etc.. lol

i knew I should have worked harder.. LOL


anyway, I realize that on one side of the res we are pushing water through and the other side we are sucking water out.


but aren't we doing the same thing everywhere else? I mean, the res is essentially a large piece of pipe that is sealed. what is the difference between using a 3 foot section of hose with a t-fitting and a fill-cap or inserting a rez?

I think if you were to submerge the pump and have the output from the res going into open air, water would still be pushed through, meaning that there would still be pressure, and I don't think the resultant pressure loss would be all that measurable.

I want to take my stuff all apart just to see what happens now..


NOT
 
Originally posted by OneMadPoptart
Duh... Nobody here was trying to make that point so I don't understand why you posted this. I can speak for myself when I said that its best to put the reservoir before the pump because your flowrates will be the highest. Its true, and if you put the reservoir anywhere else in the loop your flowrates will be lower for the entire loop.
Click to expand...

I'm saying it won't matter whether you go res-->pump or pump-->res. The difference in flow is negligable as long as you don't use restrictive tubing size from the res-->pump or pump-->res.
 
Ummmmm..... so...... I understand all thats being said here but im still curious.... If i had one Rad at the blow hole and one fan at ... lets say the intake bottom front in parallel (rediculous i know but for the sake of argument) would the flow through both be equal assuming equal lenghts of tubing. Im tempted to say no only because the water would have to travel higher to the other Rad.

I think what im going to do is put my Rads facing each other and split them with a T line where the HD's usually go.

Definatly going to post pics and temps later when i get this done.

Again though thanks for the enlightening discussion.

SD
 
if you are using 2 different radiators they will have different internal flow characteristics, and the water will, for the most part, take the path of least resistance. Therefore, flow will vary between the radiators, but that really doesn't matter.

What matters is that you will, no doubt, get superior cooling performance using the second radiator regardless of variances in flow between the two.
 
LOL..... Im doing a 24 hour run right now... though its likely that ill put it all in tonight while i watch kill bill vol1 on video. I will definatly post results. For the record i have a Koolance exos right now that runs about 54.5 (224 fsb on a 3.0c @3364)

SD
 
Speaking of res and flow rates got me thinking about this overclockers.com here. I think the one important thing to get out of this is the reducing the Net Positive Suction Head to keep the flow up (very technical explanation here). Moving the res after the pump would raise the suction head at the pump and therefore... lower flow rates (thoughout the whole system, to keep Astroman happy :D j/k).
 
That's the thing...if you have your res after your pump, there won't be much of a difference if you use un-restrictive fittings. If your pump outlet is 3/4", then use 3/4" fittings on your res intake.
 
OK WOW!.....

Lets see.. jsut turned this beast on. I was running an Exos before and i had temps of 52-54 while running F@H now..... omg

how can this be right..... 42..... wow....

Swifty block, pump, rads..... 2 80 mm minis....

I do have a tornado one one of them though.... got it turned down though..... wonder what will happen when i turn it up.

SD
 
Yup. Koolance sucks for decent cooling needs.

Now you need a real waterblock and to get that res before your pump. THEN you will see what good temps are:D
 
You want it RIGHT before the pump. Thats what we've been telling you to do the entire thread.

Unless your loop ends reservoir -> pump , then you are not getting the best flowrates and thus, cooling that is not reaching its potential. Especially when running a restrictive load like 2 radiators, you really should go pump -> blocks -> rad1 -> rad2 -> res -> pump.
 
You must log in or register to reply here.
Back
Top