• 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.

Water Filling Issue

D

DaMaN841

Limp Gawd
2FA
Joined
Jul 30, 2001
Messages
228
I'm setting up my watercooling as we speak, and unfortunately, the bay resevoir I have cannot be the highest point in the system. Therefore, there is no way that I can fill it without any type of overflowing. The only idea I have currently, is if I find a brass fitting that inserts into the resevoir filling spot, attach some tygon tubing to it, long enough so the end of it becomes the highest point, and fill it that way. However, it cannot be perminent (maybe if I find a plug for it, it can, but that's dangerous leaving it hanging inside the case like that). I also thought of a t fitting, but I really don't want it all that sloppy. Any ideas will really help.
 
why can't you fill the system like that? Turn on your pump, and add water. As long as the water level in the res stays below the opening (it will), everything is fine.
 
Unfortunately, I cannot fill it before I put it in the case cause the radiator is a pain to take on and off, and the res really only fits one way (I didn't want to put so much extra hose that I can pull it all the way through, too much excess). Hmmm, maybe, I'll give it another shot with filling it, water just seemed to enjoy coming back at me :( I believe it's because I have the pump after the res, which is causing the pump to be pushing instead of pulling, but I'm not quite sure if that would make a difference in filling.
 
The pump goes after the res, but said pump has to pull water from the res. then it's no problam at all.
 
if you have the pump pushing water into the res then you're going to get much worse flow rates. those pumps are designed to push water, not pull it
 
Welcome to the downside of watercooling: routing your lines and components in your case is a bitch. I'm going to take my stuff out for the third time and redo it this weekend. ;)
 
It really shouldn't be much of a problem filling if you fill it, cap it (to avoid splashes), crank the pump a bit to level out the water level, then uncap, and fill again.

You may even be able to just fill and keep the pump on at the same time.

I have a bay res and it is not even the highest component in the system...I used to have a big external radiator sitting about 1/2 foot up on top of the case. It didn't pose a problem.
 
i had a very similar situation with my swiftech reservoir sitting a little below my rad and harddrive block,
what worked for me was to tilt the case back (on its 'feet') and put a little water in, then pulse the pump on, pour a little more water, pulse the pump on again.. and so on..
by the way, i also have about 2 inches of extra tubing behind my reservoir allowing me to pull it out of the drive bay enough to make the pouring easy even though the case was tilted back about 45 degrees.
(use a funnel ;-)

good luck with that :)
 
What I ended up doing, was I went to home depot and I got a nice T piece and split the line from my Rad to the Block and brought a brass nozel to the highest point. It's working wonders. However, is it true that having the pump before the Res, it's really going to make a difference?
 
DaMaN841 said:
However, is it true that having the pump before the Res, it's really going to make a difference?
Click to expand...

It's a loop... the reservoir is always both "before" and "after" the pump. It won't make a difference where it is.
 
Hehe, thanks, I won't listen to noobs from now on :) Thanks for all your help guys, I really appreciate it.
 
zer0signal667 said:
It's a loop... the reservoir is always both "before" and "after" the pump. It won't make a difference where it is.
Click to expand...

I'm not so sure about this. If the res is after the pump it could put a lot of back pressure on the pump. It is usually best to place the res before the pump to "prime" it so there are less air bubbles. This isn't always necessary, but it also makes it easier to bleed the system, and keep it quiet.

Just my $0.02
 
Yeah, res directly before pump is better, as nearly all pumps are not self priming. If they are full of air they will not move said air one bit. You've got to get some water in them, which is easy if the pump comes after and below the res.
 
Little Grabbi said:
Yeah, res directly before pump is better, as nearly all pumps are not self priming. If they are full of air they will not move said air one bit. You've got to get some water in them, which is easy if the pump comes after and below the res.
Click to expand...


Having the res positioned so that it easily primes the pump is always a benefit. I was saying that durinf regular usage, there is no "before" or "after the pump, nor is there a "pushing" or "pulling" of water by the pump.
Let's hook up a batter to a light bulb. Wire A goes from the positive terminal to the light bulb, wire B goes back to the negative termincal. Does the bulb or battery care if wire A gets switched with wire B? No, the battery still creates the same voltage potential, and the bulb still shares the voltage drop in the same proportions with the wires. Fluid cooling loops are surprisingly similar in these principles.
 
The battery analogy is good but is missing something...

Reservoirs are generally at atmospheric pressure. If the pump is after the res, the pump inlet will be at or near atmospheric pressure as well. The more components are between the res outlet and the pump inlet, the lower the absolute pressure at the pump inlet. And the lower the inlet pressure, the greater the risks of cavitation (with its loss of pumping power, noise, increase wear). Plus, the tubing leading to the inlet is at greater risk of collaspe and/or kinking.
 
i've gotta agree with zer0signal667 analogy..

and as far as pressure is concerned, the vast majority of the reservoirs out there are air-tight.. and so there's no need to worry about flow-direction with respect to the pump...

for example, if while the pump is running i take the cap off the Bar Res i have, you can clearly hear the air rushing into it.. and then all of the problems that you (HeThatKnows) described do occur...

it is still important for the cpu block to be after the rad though...
but that's a given.. i dunno why i said it..
 
Do the math on the 'CPU has to be after the rad'' thing. Pick a flow rate and a heat load, then calculate the how much cooler the water is after the rad. I think you'll be surprised.

In any event, the air rushing into the res when you took the cap off thing is more related to the elasticity of your tubing. Even the stiff vinyl tubing will expand some when pressurized. In your case a greater length of tubing was feeling positve pressure than the length feeling negative gauge-pressure, resulting in a net increase in the volumn of the system, resulting in a net reduction in pressure. Cap the cap off, the pressure equalizes...
 
I'll do some tests with the rad cpu order, i'm curious to see how this works out because i was only basing this on concepts and not experiment

but with regard to the presure inside a closed loop and the position of the reservoir wrt the pump... it doesn't matter where the pressure difference is coming from when i unscrew the cap.. the fact that there is a constant presure through out the whole system means that the pump will not be cavitating etc...
 
Jonsey said:
Welcome to the downside of watercooling: routing your lines and components in your case is a bitch. I'm going to take my stuff out for the third time and redo it this weekend. ;)
Click to expand...

That is one of the things that ran me away from water cooling when I started out. It's like trying to see how many clowns you can fit in a volkswagon. Craziness man. You should use something easier like an Aqua-Tame. They carry them at www.deepinnerlight.com
 
james.m.flood said:
I'll do some tests with the rad cpu order, i'm curious to see how this works out because i was only basing this on concepts and not experiment
Click to expand...
Base it facts and physics - the specific heat of water is well known. If you can convert and manipulate units, you can calculate the anwers with specific heat, flow rate, and heat load.

the fact that there is a constant presure through out the whole system means that the pump will not be cavitating etc...
Click to expand...
Ahh, but there isn't a constant pressure thoughout. What makes the water flow? Pressure differences. When water flows through a length of tubing or a water block or rad or res, what does viscous friction cause? A pressure loss.

As for cavitation - think about how a centrifugal pump works. A spinning impeller flings water forward and outward. But what happens behind an impeller paddle? As it retreats, the space behind it needs to be filled with water, water that is pushed into that space by the inlet pressure. If there's not enough inlet pressure to fill it with liquid water, the pressure behind the paddles with be low enough for water-vapor bubbles to open up. As those bubbles migrate to areas of higher pressure, they collapse (creating noise, wasting pump power).
 
i fail to see how this has anything to do with specific heat capacity or why you would dream of converting a flow rate and pressure problem into a heat problem?

please educate me...
 
Two separate questions: First, is the temperature difference large enough to say that having the rad before the CPU is important. First you said that it was 'a given' that it was important, then said it was based on a concept and needed experiment. I say, look at the concept and analyse the physics.

Water enters the rad at your chosen rate, heat is exchanged to the air at a specified rate, water exits the rad with a lower temperature. I think you should calculate that temperature drop as a function of water-flow and and heat-flow. All you need is those flow rates and the specific heat of water. Put some numbers on the concept you put forth, use the physics you've studied, gain a deeper understanding of watercooling.

Second question: do the relative position of pump and res matter. The flow of a water around the loop only depends on point to point pressure difference, not asolute pressures. So the order won't affect flow. But, the operation of the pump itself does depend on absolute pressure, so its best to have the res just before the pump to maximize the pressure at the pump inlet (in industry terms, you need NPSH "net positive suction head').
 
HeThatKnows said:
First you said that it was 'a given' [...], then said it was based on a concept and needed experiment
Click to expand...
yeah, the concepts were given, but i agree that often 'given' concepts can be wrong.. which is why i'm goina experiment and see, you seem like a pretty smart fellow who wouldn't be sure abut something unless you've done your own research.

HeThatKnows said:
Second question: do the relative position of pump and res matter.
Click to expand...
I have to appologize because i was missunderstood, when i read a previous post of yours in this thread i was under the impression that you thought the Res should go After the pump, and not before..
So we were actually arguing the same point..
Sorry about that...
 
Pump and res order aren't important, but pump>rad>block order matters. Some pumps can heat the water a little, so I go pump>res>rad>block>pump. ;)
 
james.m.flood said:
which is why i'm goina experiment and see.
Click to expand...
Do the math, do the math! In another thread, you said third year physics...it shouldn't take but ten minutes. You won't regret it (and you'll be able to demonstrate your awesome knowledge in future threads with the aid of precise numbers).

Calculate now!!!!!!!!!!!!!!!!!!!!

:D
 
uclajd said:
I go pump>res>rad>block>pump. ;)
Click to expand...
i'll only speak for myself here, but i have a feeling HeThatKnows will strongly dissagree with that....

as for myself, i disagree, because of all the reasons talked about earlier in this thread regarding the pressure entering and leaving the pump...
 
Yeah, pump>res is bad; res>pump is mucho better.

Rad>block vs block>rad, is largely irrelevant. Convenient tubing routing can easily outweigh any percieved temperature benefits.

Speaking of which... finished those calcs yet, James? ;)
 
this is vacation man ;) wait till monday
allthough, thanks to you, it's been the main topic of my mental calculations for the last 24 hrs...
when i should be thinking about food, sleep and video games ;)
 
HeThatKnows said:
Yeah, pump>res is bad; res>pump is mucho better.

Rad>block vs block>rad, is largely irrelevant. Convenient tubing routing can easily outweigh any percieved temperature benefits.

Speaking of which... finished those calcs yet, James? ;)
Click to expand...
It's not like I just started watercooling this afternoon. I've run it every conceivable way, and I get the best temps my way, which BTW is recommended by both Asetek and Swiftech. YMMV.
 
uclajd said:
Search the sites. I got hard copy manuals. ;)
Click to expand...
Swiftech has pdf's of all their manuals on their site. Can you give me a quote from your paper manual, so I can find it the right manual/section online?
 
HeThatKnows said:
Swiftech has pdf's of all their manuals on their site. Can you give me a quote from your paper manual, so I can find it the right manual/section online?
Click to expand...
Dood, I have like 400 threads out and you wanna argue this trivial matter?

The sheet included with the MCW50 shows a loop schematic with raddy after pump..

Again, my tests show that the pump heats the water a little, so I have it before the raddy.

YMMV and it's a free country, so loop it however you want.

Some people are more concerned with adhering to their position and being "right" or an "expert" than with the truth. :rolleyes:

Some are even so concerned with this recognition that they suggest omniscence in their user name.
 
uclajd said:
Dood, I have like 400 threads out and you wanna argue this trivial matter?

The sheet included with the MCW50 shows a loop schematic with raddy after pump..

Again, my tests show that the pump heats the water a little, so I have it before the raddy.

YMMV and it's a free country, so loop it however you want.

Some people are more concerned with adhering to their position and being "right" or an "expert" than with the truth. :rolleyes:

Some are even so concerned with this recognition that they suggest omniscence in their user name.
Click to expand...


For the sake of the scientific method, can you or can you not provide results based on controlled experiments? I'm not saying I agree or disagree with you, but as an engineer, I'm always interested in methodology of experimentation. My argument in the matter also makes it particularly interesting to me. Anyone else with such results is also proposed the same question.

edit: I apologize, this is off topic and should be taken to another thread if so desired.
 
When flowing through a heat exchanger (rad, water block, or whatever), water temperature will change by an amount (in degess Celcius) = 0.0143 times the heat flow (in watts) divided by the water flow rate (in liters/min). (Check my work, James!)

A pump heats up the water a little, eh? Let's take the Laing D4, which uses 18 watts. Of course not all of that ends up as heat in the water, some escapes to the air and some becomes useful hydraulic power. But lets use the whole 18 watts anyway. In my system with a heatcore, CPU block, GPU block and res, I get around 6.5 liters per minute. Lets call it an even six. Whip out the calculator...the pump is heating the water by at most 0.043° C.

I have a pair of four channel type-T thermocouple readers that I use. Even with carefully matched thermocouples and channels, I can't measure pump heat. I can barely measure the heat from the CPU (90w plus or minus 8) with its two tenths degree change.

uclajd, your tests show the pump heats the water up some. How did you measure that? To get better than hundredth-degree resolution, you'd need a thousand dollar or more platinum-RTD system.

You're right it's a free country, believe what you want. And we're all very proud you "have like 400 threads out." (Uh, what's your post count again?) But, come on! If you're gonna say stuff, have something to back it up with.

By the way, I never implied omniscience. Heck, I never specified what I know. Could be "He That Knows how to scratch his balls" or "He That Knows how to get the panties off a cheerleader". Hey it could happen! :D
 
You must log in or register to reply here.
Back
Top