Tempratures climbing after addition of 6800 NV-68

[MaxxxRacer]

ok given that information, the pressure drop will be the same or the deltaP (p is pressure) will be the same no matter what the flow is (given it isnt like .0001gp) Well if it is earlier in the loop then the pressure drop will be the same but it will start at different points. lets say the pressure drop is at 4. just a random number 4. so the pressure is 10 before going into block 1. after block one the pressure is 6. now lets say it goes through the exact same type of block right after the first one. now the pressure is 2. voodoo based on what you said that would seem to be correct. and that has been my point the entire time.

[MaxxxRacer]

about the batteries the single battery will have the same voltage no matter where u put it in the loop of batteries, but if you measure just that battery in the loop with all of hte other batteries then it will be the voltage of that one battery. u have to measure at the end of the series of batteries to get the voltage of them all...

[Althornin]

Originally Posted by MaxxxRacer

ok given that information, the pressure drop will be the same or the deltaP (p is pressure) will be the same no matter what the flow is (given it isnt like .0001gp) Well if it is earlier in the loop then the pressure drop will be the same but it will start at different points. lets say the pressure drop is at 4. just a random number 4. so the pressure is 10 before going into block 1. after block one the pressure is 6. now lets say it goes through the exact same type of block right after the first one. now the pressure is 2. voodoo based on what you said that would seem to be correct. and that has been my point the entire time.

NO.
based on what he said, what you just said is wrong, just as I and redleader are telling you.

If you are talking head loss - because flow rate is determined by TOTAL head loss through entire loop, and thus flow rate is constant through entire loop - and head loss of a given component is related to the flow through it (more flow = more head loss), IT DOES NOT MATTER WHERE IN A LOOP THE COMPONENT IS PLACED. FLOW IS THE SAME, AS IS HEAD LOSS.

I honestly don't see how you are thinking. Are you thinking that "pressure" (ie, possible pump head, or height it can pump) is somehow "used up" by a device, and less is available for what is "behind it" in the loop. What we are (and have been) trying to tell you is that flow is determined by total head loss in the entire loop, and because flow cannot vary in the loop (how can your waterblock flow 4 liters/min and you radiator flow 2 liters/min in a serial loop!!??!) component order matters only so much as water can heat up across any given component - and given that a CPU puts out say, 100 W of heat, and you are flowing at a reasonable rate, the temperature differential in any given loop is insignificant (maybe .5C).

Basically, you telling someone that:

Having the rad after the pump (before it hits the cpu) is not doing you any favors. its just killing the pressure that the rbx is getting.

which makes no sense, and then getting all smug and pretentious about "correcting me" and talking about "it's physics" makes me kinda pissed off.

I mean, your above statement makes NO SENSE. the only way the rad reduces pressure given to the RBX is that in a closed loop system, water pressure (not head, or head loss) goes from positive near the pump outlet to negative by the pump inlet, with a gradient inbetween. But, as you (and I, and others) have said, water is incompressable, generally speaking, so what the hell are you talking about?

Repeat after me:
flow rates are the same throught a loop.
head loss of a componenet is based upon flow rate.
Because flow rate is the same no matter where you are in a loop, head loss must be the same no matter the order of componenets.

See?

[MaxxxRacer]

I know flow rate is the same through the entire loop. I have never contested that. In the end if the components are in series then the flow rate will be the same.

All im talking about and have been talkign about is pressure not flow rate. there is some other component to this that I dont quite understand, but the pressure will be higher before a highly restrictive block than after it exits. Im not talking about flow, just the water pressure itself. The Psi before the block will be higher than after the block. Thats all that I am saying. Sadly it would be really hard to measure this.

Even though water is not compressable the psi force that water is exerting does change. In a submarine when you go deeper the psi changes. the water doesnt compress, but it tries to compress everything else. Now do you see how the pressure can drop as the water flows through the system.

I know it seems odd as the flow stays the same at any given point through a closed loop (my sytem isnt entirely closed as i have a res)

here is a point to think about. when the water gets to the res in the system the pump is no longer forcing the water out of the res, but is sucking the water out of it. if it was pushing it through the res rather than sucking it out, the res would be completely full all the time. if i open the top on my res the res doenst overflow and that is becuase the water that is flowing out of the res is being SUCKED out by the pump. NOT being PUSHED out. Granted since i have my res above the pump gravity would take effect if i unhooked the pump, but u dont exactly get alot of flow using gravity.

[Delirious]

Originally Posted by MaxxxRacer

here is a point to think about. when the water gets to the res in the system the pump is no longer forcing the water out of the res, but is sucking the water out of it. if it was pushing it through the res rather than sucking it out, the res would be completely full all the time. if i open the top on my res the res doenst overflow and that is becuase the water that is flowing out of the res is being SUCKED out by the pump. NOT being PUSHED out. Granted since i have my res above the pump gravity would take effect if i unhooked the pump, but u dont exactly get alot of flow using gravity.

Pumps dont suck, they create an area of low pressure with the impeller that draws more water into it, thus the need to prime them. If they could suck then u wouldnt need to feed them water before you started it to get it to work.

Read the part about Centrifugal Pumps cause that is what a pond (made for comp) pump is.
http://www.powerandmotoryacht.com/ma...mer/index.html

[Voodoo Rufus]

but the pressure will be higher before a highly restrictive block than after it exits

That is true.

Say you put a static pressure monitor right after the pump, and another one on the outlet of your block. In that case, the pressure on the outlet would change as you changed the blocks location in the setup. But if you placed both monitors before and after the block, it will never change for a given flow rate.

I thought that pumps are not good at pulling, but still can pull a bit once you have liquid in them. Hence the reason MR sees no air getting pulled in when his res is open. I had the same effect when I opene my res, and it was on the bottom of the case too. It's not just gravity pulling the water down. Correct me if I'm wrong though.

[MaxxxRacer]

omg.. that is semantics.... they do suck up water. the reason that they dont suck without the water into them because the impeller is not made to push air but rather water...

but you are correct on the not sucking and rather creating a vaccum. but guess what.. thats what a vaccum does!!! like the ones that clean the floors of your house.

the pumps are bad at sucking. but the pumps do suck up water well enough to feed themselves from a resivour and not get any cavitation. That is considering that your tubing is big enough to not restrict flow enough to starve the pump. IE using 3/8 inch tubing would cause horrible cavitation in a watercooilng loop.

[redleader]

Head is the height that the water can be pumped to vertically. pump with 26feet of head can pump up to the exit point of hte tubing 26 feet above the height that hte pump is at... obvioulsy it can pump more than 26 feet if the tubing is at the same height relative to the pump. it could pump much much furter.. but stil there is a point when the resistance of the water in the tube will cuase the pump to stop flowing even if the tubing is horizontal. this is just because the water would be so heavy that the pump could not start moving it.

Yes, but it has units of pressure, hence my point that you should think of it as pressure and not distance.

not that this maters but water isnt compressable. if you take water that is 10 thousand feet below sea level (not that there are any places that deep i think) and bring it to sea level it wont get any "bigger". But this doesnt apply to water cooling.

All things are compressible, water included. Water at the bottom of the ocean is actually denser then at the top because its under a lot of pressure. My point was that we can ignore this in most liquids and solids because they don't compress noticably at the pressures we deal in.

ok given that information, the pressure drop will be the same or the deltaP (p is pressure) will be the same no matter what the flow is (given it isnt like .0001gp) Well if it is earlier in the loop then the pressure drop will be the same but it will start at different points. lets say the pressure drop is at 4. just a random number 4. so the pressure is 10 before going into block 1. after block one the pressure is 6. now lets say it goes through the exact same type of block right after the first one. now the pressure is 2. voodoo based on what you said that would seem to be correct. and that has been my point the entire time.

You don't seem to get that its all relative. When you say that the pressure is X, you're saying its X higher then something else. Could be the pump inlet, could be atmosphere, could be vaccum. Lets extend your example:

Lets say a block is right after a pump that produces 5 PSI of head above atmosphere. The drop across the block is X-3, where X is the high side pressure. Thus after the block the fluid is 2 PSI higher then our reference (5-3 =2). So far so good.

Now lets move the block to be after the res. As you said, an open res is at atmospheric pressure, which is zero in our example (since I speced the pump head in psi above atm). In this case the high side of the block (X) is at 0 PSI. And the low side is still X-3 since we agree that flow cannot vary in a series loop. That means that the low side is 0-3 = -3 PSI above the atm, or more commonly it is 3 PSI below atm.

Now do you see why absolute numbers are unimportant? We can construct arbitary values at each point, and it doesn't really matter provided the drops are constant. Lets say we take the loop to the bottom of the ocean. Now X might be 10000, and X-3 = 9997 PSI. But the difference is still 3, and the flow is the same (ignore that the water is a few percent denser and thus acts a little different in the block probably shifting the pressure drop slightly.)

Is there some ambiguity still why it doesn't matter where the pump is? If so be specific what part of my explination you disagree with or do not understand.

[MaxxxRacer]

I did look up the compressibility of water and I'm sorry for saying that it isnt compressible. From my research I found that it is actually compressible by 1%. Not much, but its still something. Apparently this happens in insanely powerful pumps like for dams and stuff..

I realize that the pressures are relative. I stated that. I didnt directly say relative but its implied in the text.

Let me explain this again. If you had the block after the res, the pumps head pressure would NOT be pumping water through the block. in actualitity the vaccum created by water flowing through the pump would in effect drag water through the water block.

If there is any component after the res and before the pump this component will cause cavitation in the pump (given that the component is restrictive) becuase the pump will be starved of water. this is so becuse the pump is having to suck water into itself from the res. it is just taking the water from the res. the water in the res isnt be forced into the block. its being pulled in.

So this proves that order does in fact matter. granted this wasnt exactly my origional point (origional point regarded the block and the rad). Here is a simple way for you to test my pont. See the tube leading into your pump. sqeeze down on it a bit and see what happens. it will fuxor the pump. By that i mean that you will start to hear cavitation and then see bubbles running through a close loop. if you squeeze down on the tube leading out of hte pump you will see that the pump will do fine. nthign willl happen except for slowing down flow due to restriction.

When you said that having the block after the res in your example would give a psi of -3. Well in a way you are correct. the pressure causing the flow throuh the block would infact be coming from after the block. and not before it. This pressure being negative pressure. With negative pressure you get suction. this suction would pull the water through the water block. Here is some testable evidence to the negative pressure.

Why is htere cavitation in pumps. Becuae the pump is not getting enough water and it is pumping so much water out of the pump that there is a negative pressure build up inside the pump. when you have negative presure in water it starts to release bubbles (cavitation). The reason is because when you decrese the pressure of a liquid in decreases the boiling temperature and in effect you are boiling the water at room tempeature. cool sh!t huh. a little chemistry there for you. its cool stuff. but anyway.

The above example proves that the absolute numbers are actually important and not just something that can be disregarded.

Myself i dont have much tubing to spare and really hate messing with my system. But if you want to try putting your cpu block after your res. see what happens. I GARUNTEE you that you will get worse performance with the block.

As you mentioned the water being denser where the psi is higher isnt the issue. what is the issue is that with modern blocks, they dont rely on flow so much as they rely on jetting the water into the block. with higher psi behind the block then there will be stronger jets. if the water is being pushed lazily through the block then there wont be the jetting effect and thus the blocks true performance is not realized.

The reason that you think that it is entirely about flow is that to obtain these higher pressures that give you this "jetting" effect inevitably have to be higher flow.

now for the old dnager den blocks for the gpu and the zchip NB block. those blocks rely entirely on flow and are relatively un restrictive. The maze4 is also one of these blocks that relys entirely on flow and not the "jetting" effect that i was talking about.

the jetting effect is present in the WW, Cascade, TDX, RBX, Storm, Radiical WHX, X4PRO X3M, Asetek Antarctica, and others that i dotn know exist.

[bobkoure]

Originally Posted by MaxxxRacer

... it could pump much much furter.. but stil there is a point when the resistance of the water in the tube will cuase the pump to stop flowing even if the tubing is horizontal. this is just because the water would be so heavy that the pump could not start moving it.

??
If the tubing is horizontal, the resistance is from friction. Friction goes up as velocity goes up (so it also goes up as tubing diameter goes down). It's a bit complicated because there's an interplay between laminar effects and eddy flow. I don't claim to understand that part particularly well.
However, the weight of the water doesn't come into it, except in the sense of inertia, which is an issue with flow start and stop. This is not intermittent flow so we are not dealing with start and stop.
Here's a thought experiment (well, you can do it, if you like).
Take two 100' lengths of horizontal pipe. One has 1/32" ID, the other one 1" ID. The water in the pipe weighs more in the 1" ID pipe. (calculating weight of water in each pipe left as an exercise for the student).
Start out with the water pressure at one end exactly the same for both sizes of tubing. You can do this with an essentially unlimited size reservoir with water level a set distance above the pipe, a pump, whatever - and put a measuring bucket at the far end. Let the water run until the output at the far end reaches a steady rate, whatever that rate might be.

OK: So now you're saying that there will be more flow into that bucket with the 1/32" ID tubing, because the cylinder of water being moved is lighter than the one in the 1" pipe.

Really?

[MaxxxRacer]

Im not saying that and I never did. that is obviosly incorrect. You are just twising my words around.

Its obvious when the tubing is horizontal (never is in our case) that the restriction comes from tube D and friction inside of the tube and friction between the water molecules. since its horizontal the water would flow out slowly by itself due to the fact that there is no cap on the end of the tube, but after a certain lenght, the pump itself would not be pushing the water out. it would just be flowing out itself from gravity. (after a certain head point, pumps auto-shut-off)

Now lets say the tubing is vertical. Then the weight of the water does come into play. at that point the pump is having to push the entire column of water that is in the tube up to the head (lets say there is jsut a tube that sprays off into oblivion). That is the way head is measured. And its entirely due to the weight of the water. At a certain point the water is too heavy for the impeller to push it upward.

Again, you and I both know that its stilly to say the thing Diameter tube will have higher flow. higher exit velocity when exiting the tube but no more flow. obviously less flow. hence wider is better (to a point)

[redleader]

I did look up the compressibility of water and I'm sorry for saying that it isnt compressible. From my research I found that it is actually compressible by 1%. Not much, but its still something. Apparently this happens in insanely powerful pumps like for dams and stuff..

Research harder next time. You can compress water a lot more then 1% according to 30 seconds on google.

If there is any component after the res and before the pump this component will cause cavitation in the pump (given that the component is restrictive) becuase the pump will be starved of water. this is so becuse the pump is having to suck water into itself from the res. it is just taking the water from the res. the water in the res isnt be forced into the block. its being pulled in.

This is wrong. Flow is constant in a series loop, and does not depend on componets. Putting a restriction on the inlet doens't starve the pump anymore then putting it on the outlet. Either way the effect is the same. Cavitation is due to drops in pressure induceing a phase change in the water. Pumps are supposed to have a minium amount of head at the inlet to avoid a phase change in the coolant, and this head is speced by the pump manufactorer for a given fluid and temperature.

Its also compeletely irrelevent. Stick to the issue.

When you said that having the block after the res in your example would give a psi of -3. Well in a way you are correct. the pressure causing the flow throuh the block would infact be coming from after the block. and not before it. This pressure being negative pressure. With negative pressure you get suction. this suction would pull the water through the water block. Here is some testable evidence to the negative pressure.

In what sense am I not correct? Or do you just like repeating whats already been posted for no evident reason?

The above example proves that the absolute numbers are actually important and not just something that can be disregarded.

It proves that you don't understand whats being discussed.

Myself i dont have much tubing to spare and really hate messing with my system. But if you want to try putting your cpu block after your res. see what happens. I GARUNTEE you that you will get worse performance with the block.

Well you have no idea whats going, won't listen to people trying to explain it, and admit that you've never actually tried it, but hell if you "GARUNTEE" it, thats good enough for me.

[bobkoure]

Originally Posted by MaxxxRacer

Im not saying that and I never did. that is obviosly incorrect. You are just twising my words around.

I quoted EXACTLY what you said - I quoted you and then pointed out that what you were saying didn't make sense - and tried to do it in as understandable way as possible.

You clearly don't want input.
BillA uses the phrase "willful wrong-headedness". I can't think of a better one for what you are doing.

I'd suggest you listen to redleader - he's taking the time to try to help you understand this. I'm certainly done

Or are you saying I've misquoted you?

Originally Posted by MaxxxRacer

... it could pump much much furter.. but stil there is a point when the resistance of the water in the tube will cuase the pump to stop flowing even if the tubing is horizontal. this is just because the water would be so heavy that the pump could not start moving it.

Did you not write (say) this?
There are a few things I do NOT do on forums - and misquoting is one of 'em. I make a point of not even "fixing" typos or spelling errors - a quote is a quote.

[MaxxxRacer]

i give up.. i just give up.. ive discussed this with several people and they all agree.

I will say this one last time.

there is water pressure. as the water flows through a component the water pressure drops. having lower water pressure makes blocks that use the jetting effect less effective. that has been my point from the beginng. that is how this all started.

if you have a lumpy chanenl (one in existence as far as i know) and you put it after the rad there will be no change in performance (possible increase due to the rad cooling down water after pump but not measureable). the lumpy channel is not affected because its performance is based entirely on flow.

that is a condesnsed version of what i have been saying the entire time. there really isnt anything to argue with on that front.


bob, i worded the stuff that you quoted very poorly. it came out wrong and you read it differently than i meant to say. my fault.

one thing that needs to be noted. these are not closed loops. if you have a res or a t line then its not strictly defined as a closed loop. closed loop implies that there is no way to add or subtract any water from the system without unhook tubing and draining it. in a closed loop the kPa or PSI in the loop is independent from the oustide world. with a res or t line then when the water reaches that t line or res the and the pressure levels out to as if the water were sitting in a cup. My point on saying that it isnt a closed loop entirely is that most of what your saying is true and i would be wrong if the system was an entirely closed loop. what you said about cavitation would be correct if it was a completely closed loop. but its not... so things are different.

[Althornin]

Originally Posted by MaxxxRacer

i give up.. i just give up.. ive discussed this with several people and they all agree.

Great!
Agree with a bunch of clueless idiots rather than use your brain and find out the truth.

You "GARUNTEE" that performance will be worse just by changing the order of components in the loop, as if somehow, despite the fact that you agree flow will be the same, that the C/W of the waterblock will change for say, NO FREAKING REASON AT ALL.

Give up, because you are irredeemably stubborn and incorrect.

[ferdb]

I also get about 60c temperatures on my 6800 ultra at 425/1.1 under load with an MCW50 block, flow rate is around 2L/min. This isn't unreasonable since I run my system fairly warm to cut down on noise.

Originally Posted by brucoman

62c on the 6800? my 6800 GT at 410/1.1 with the same waterblock runs at 49c during 3Dmark03... perhaps need more flow still, try without the NB block, better yet get a second pump in series

MaxxRacer your understanding of pressure and flow are not correct. No matter in what order you put the CPU block in the system you are going to get the same rate of water flowing through it. The rate of water flow through the block will generate a pressure drop across the block that is related to it's flow resistance. That pressure differential across the block is the only pressure that the block can see, it knows nothing about what other pressures are in the system. If the water flow is unchanged in the system, the pressure across the block is unchanged no matter where you put it, and the water velocity in the block is unchanged.