It's Done and it Works

BillR said:
If you are indeed right then AC, Koolance et al and all the HVAC manufactures in the world can sure save a ton of money by firing all their engineers now can’t they? :rolleyes:
Click to expand...


I think I covered this in an earlier post. I quote:

"Designers usually set a certain goal they wish to meet, at which point the product is brought to market. Modders and the aftermarket crowd can increase performance of said product by modding it beyond original goals of designer."
 
Erasmus354 said:
More flow is NEVER a bad thing, why cant you get that through your head? Also, the guy has a Cuplex XT, an impingement style block, which as TN has reported AC admits benefits well from increased flow. Any more sly remarks or are you willing to accept that you were wrong?
Click to expand...

Please don't quote me out of context. What I said was that certain Aqua Computer blocks will benefit from an increased flow (Cuplex Pro and XT in particular), but other blocks they make will not. What I didn't say is that I was also told that if I hooked up a big old BIB high flow pump I would see a decrease in performance. So I think everyone is right here depending on your particular slant on life. Which is really all I have ever wanted to point out in all my arguments here. :D

I am also glad to see you guys finally admit that "Practicallity is not the point, bragging rights is, and some people wish to have the best water cooling solution." (copied from another thread in the big ring tonight). So I was right about you guys being Top Fuelers ;)
 
plywood99 said:
I think I covered this in an earlier post. I quote:

"Designers usually set a certain goal they wish to meet, at which point the product is brought to market. Modders and the aftermarket crowd can increase performance of said product by modding it beyond original goals of designer."
Click to expand...


I won’t argue that, it’s a valid point.

My point was and is I’ve been water-cooling since before there were commercially available parts. I have a milling machine and a lath. I made my own water block, bought a high flow pump and used a heater core from a car.

Wow, my CPU ran cooler then on air. Major accomplishment.

What I did find while designing and re-designing was, if I actually restricted the water flow (I used forceps to compress the tubing to various degrees) to the block I could decrease my CPU temp. After that I made restriction nozzles on a trial and error basis to find the “sweet spot”.

It was a design engineer from the York Air Conditioning Company that turned me on to the idea. I sort of figured with all his degrees in heat transfer etc it was just possible he knew more then me. He might know more then you, who knows.
 
BillR said:
I won’t argue that, it’s a valid point.

My point was and is I’ve been water-cooling since before there were commercially available parts. I have a milling machine and a lath. I made my own water block, bought a high flow pump and used a heater core from a car.

Wow, my CPU ran cooler then on air. Major accomplishment.

What I did find while designing and re-designing was, if I actually restricted the water flow (I used forceps to compress the tubing to various degrees) to the block I could decrease my CPU temp. After that I made restriction nozzles on a trial and error basis to find the “sweet spot”.

It was a design engineer from the York Air Conditioning Company that turned me on to the idea. I sort of figured with all his degrees in heat transfer etc it was just possible he knew more then me. He might know more then you, who knows.
Click to expand...


Yeah made my own block too, quite a few actually, and yes I used the same jet inpingement principle to design it.

But what you are talking about with restriction nozzles, AKA, jet inpingement, is different from a total loop with a diameter of your restriction nozzle, or small diameter tubing if you wish. At that point friction takes over ans slows down the flow.

Jet inpingement actually increases velocity of water striking the block.

To get what I'm saying measure the water flow from a 1 inch piece of 1/4 inch tubing. Now measure the flow from a 6 foot piece of 1/4 inch tubing. Do you think they will be the same?
No they will not. Friction from the 6 foot length now overcomes, so to speak, the velocity increase of the smaller diameter tubing....
 
plywood99 said:
Yeah made my own block too, quite a few actually, and yes I used the same jet inpingement principle to design it.

But what you are talking about with restriction nozzles, AKA, jet inpingement, is different from a total loop with a diameter of your restriction nozzle, or small diameter tubing if you wish. At that point friction takes over ans slows down the flow.

Jet inpingement actually increases velocity of water striking the block.

To get what I'm saying measure the water flow from a 1 inch piece of 1/4 inch tubing. Now measure the flow from a 6 foot piece of 1/4 inch tubing. Do you think they will be the same?
No they will not. Friction from the 6 foot length now overcomes, so to speak, the velocity increase of the smaller diameter tubing....
Click to expand...

Understood. But again, while you are indeed increasing velocity through impingement you are still decreasing your overall total mass flow.

By your own explanation, an impinged line can’t flow the mass of coolant a non-impinged line can.
 
BillR said:
Understood. But again, while you are indeed increasing velocity through impingement you are still decreasing your overall total mass flow.

By your own explanation, an impinged line can’t flow the mass of coolant a non-impinged line can.
Click to expand...


Agreed, but that would not perhaps make the whole system a low flow system. All modern blocks that I know of restrict flow to a degree. That does not make them low flow.

You can see my system in my sig. I would most certainly not classify it as a low flow system even though my diy cpu block and my modded gpu block are quite restrictive, the gpu block less so of course.

Pics of system here:

My system

I think we are discussing things from different perspectives.
 
Top Nurse said:
Please don't quote me out of context. What I said was that certain Aqua Computer blocks will benefit from an increased flow (Cuplex Pro and XT in particular), but other blocks they make will not. What I didn't say is that I was also told that if I hooked up a big old BIB high flow pump I would see a decrease in performance. So I think everyone is right here depending on your particular slant on life. Which is really all I have ever wanted to point out in all my arguments here. :D

I am also glad to see you guys finally admit that "Practicallity is not the point, bragging rights is, and some people wish to have the best water cooling solution." (copied from another thread in the big ring tonight). So I was right about you guys being Top Fuelers ;)
Click to expand...

And how exactly did I quote you out of context? You just backed up my quote again. I said the guy has a Cuplex XT, a block which benefits well from increased flow....that is all. :rolleyes:
 
plywood99 said:
Agreed, but that would not perhaps make the whole system a low flow system. All modern blocks that I know of restrict flow to a degree. That does not make them low flow.

You can see my system in my sig. I would most certainly not classify it as a low flow system even though my diy cpu block and my modded gpu block are quite restrictive, the gpu block less so of course.

Pics of system here:

My system

I think we are discussing things from different perspectives.
Click to expand...

Again, no argument on your ideas, as you restated them.

Your original inference was that more is better which may or may not be true based on what the designer of the water block had in mind when they started out.

My contention is that some blocks are indeed designed to work best on high mass flow. Other blocks are designed to work on lower, slower flow.
 
Erasmus354 said:
And how exactly did I quote you out of context? You just backed up my quote again. I said the guy has a Cuplex XT, a block which benefits well from increased flow....that is all. :rolleyes:
Click to expand...

It does benefit from increased flow, which is why I increased the flow rate in Aquasuite. By increasing the frequency of the pump the flow increases from 36-39 l/h to 50+l/h. I am not 100% sure, but it appears that the CPU temp dropped 2-3C.
 
My patience is wearing thin with these arguments popping up all over the boards. BE CIVIL, DO NOT FLAME. I'll be back...
 
zer0signal667 said:
My patience is wearing thin with these arguments popping up all over the boards. BE CIVIL, DO NOT FLAME. I'll be back...
Click to expand...

Amen to that Brother! Can I be a Mod for a day..an hour..OK, just 15 minutes please???

Seek, I just wanted to say good job on your build. I really like your watercooling componets and I gotta say "me likes that case." :D Did you say you included a flow meter? I also have a V2000 and I'm trying to figure out where I am going to put mine?

edit: More pics please. :)
 
ShoNuff said:
Amen to that Brother! Can I be a Mod for a day..an hour..OK, just 15 minutes please...

Seek, I just wanted to say good job on your build. I really like your watercooling componets and I gotta say "me likes that case." :D Did you say you included a flow meter? I also have a V2000 and I'm trying to figure out where I am going to put mine?

edit: More pics please. :)
Click to expand...

Flow meter is here...early pic. Rad is above the PSU.

Picture440.jpg
 
That’s still a nice clean install; I really like the way you used proper fittings to keep runs as smooth as possible. Makes for neatness in every meaning.

Nice job :)
 
theseeker said:
It does benefit from increased flow, which is why I increased the flow rate in Aquasuite. By increasing the frequency of the pump the flow increases from 36-39 l/h to 50+l/h. I am not 100% sure, but it appears that the CPU temp dropped 2-3C.
Click to expand...

:eek: you went from 0.5 lpm to a little less than 1 lpm....now THAT is a low flow setup. I still stand by my opinion that I feel you could benefit by moving to 8mm tubing, though I would hate to tear up my setup to put in new tubing after completing it, perhaps make that your next upgrade in 6 months.
 
Erasmus354 said:
:eek: you went from 0.5 lpm to a little less than 1 lpm....now THAT is a low flow setup. I still stand by my opinion that I feel you could benefit by moving to 8mm tubing, though I would hate to tear up my setup to put in new tubing after completing it, perhaps make that your next upgrade in 6 months.
Click to expand...

Well you can't really go by what the reading is in-line as the flow is always way lower than what the advertised flow from the pump is. So the only people around here that probably know that for sure is people who have flow meters in their loop. Most people who have electronic flowmeters are Innovatek and Aqua Computer users. Do you have a flow meter in your system? Can you please give us your in-line flow rates?
 
BillR said:
Your original inference was that more is better which may or may not be true based on what the designer of the water block had in mind when they started out.
Click to expand...


This is my only point of disagreement with you BillR. I have never seen or read of a block that performs worse at a higher flow. I believe such a thing as that is impossible. Now there is a point of diminished returns, as is seen in the Swiftech 6000 series blocks, where increasing flow past 1 watt of hydraulic power does very, very little to performance, but it still gets better.

perfpwr3.png


Note too in the graph the NexXxos XP block. It is designed specifically for lower flow systems but performs admirably at higher flow rates...
 
Top Nurse said:
Well you can't really go by what the reading is in-line as the flow is always way lower than what the advertised flow from the pump is. So the only people around here that probably know that for sure is people who have flow meters in their loop. Most people who have electronic flowmeters are Innovatek and Aqua Computer users. Do you have a flow meter in your system? Can you please give us your in-line flow rates?
Click to expand...
I don't have a flow meter, but I know my loop is pushing WAY more than this Ally McBeal skinny setup. *EACH* return from my RBX (or is it TDX, I never remember) pushes a solid 1/2" column of water back into the reservoir and it does it with quite a bit of force. It's kinda scary how much flow I'm getting.
 
plywood99 said:
This is my only point of disagreement with you BillR. I have never seen or read of a block that performs worse at a higher flow. I believe such a thing as that is impossible. Now there is a point of diminished returns, as is seen in the Swiftech 6000 series blocks, where increasing flow past 1 watt of hydraulic power does very, very little to performance, but it still gets better.

perfpwr3.png


Note too in the graph the NexXxos XP block. It is designed specifically for lower flow systems but performs admirably at higher flow rates...
Click to expand...

I will concede your point about diminishing returns because it applies to just about everything. What I have found notable is this review:

http://www.legitreviews.com/article.php?aid=205

In this case adding more blocks, thus more heat and resistance to flow actually lowered the CPU temp.

Agreed, it’s one review, one guy, one set of circumstances but does help to illustrate the point I was attempting to make.

As far as diminishing returns, face it, we (with water cooling) all work with “ambient” as our goal and nothing is 100% efficient.
 
Top Nurse said:
Well you can't really go by what the reading is in-line as the flow is always way lower than what the advertised flow from the pump is. So the only people around here that probably know that for sure is people who have flow meters in their loop. Most people who have electronic flowmeters are Innovatek and Aqua Computer users. Do you have a flow meter in your system? Can you please give us your in-line flow rates?
Click to expand...

I am using the AC flow sensor.The big question is this; what should the calibration be set to in Aquasuite? The Aquaero manual states 550 imp/l, using that calibration my flow is .76 l/m. I have pm'ed Shoggy for his expert opinion.
 
theseeker said:
I am using the AC flow sensor.The big question is this; what should the calibration be set to in Aquasuite? The Aquaero manual states 550 imp/l, using that calibration my flow is .76 l/m. I have pm'ed Shoggy for his expert opinion.
Click to expand...

Can't remember exactly, but I think Shoggy told me the magic number is 509.
 
Top Nurse said:
Well you can't really go by what the reading is in-line as the flow is always way lower than what the advertised flow from the pump is. So the only people around here that probably know that for sure is people who have flow meters in their loop. Most people who have electronic flowmeters are Innovatek and Aqua Computer users. Do you have a flow meter in your system? Can you please give us your in-line flow rates?
Click to expand...

I am well aware of the fact that pumps advertise max flow rate, ie the flow it can push with 0 restriction. However that is why they also test at varrying levels of restriction. Look at some of the pump comparisons over at procooling and you can reasonably estimate your flow rates without actually knowing the exact flow.

And anything under 1 lpm is quite low, high flow setups typically range in the 3-6 lpm range.
 
First of all, what I'm posting here isn't against AC/low flow/small tubing; it's just trying to set some things straight.

1. More flow does not decrease the time coolant has to remove heat from a loop. For a very good explanation of this, read this Overclockers.com "Watercoolking Myths Exposed" article (first myth, still read the rest of the article, it's really good).

2. For the pump inceased flow vs. pump heat vs. waterblocks debate:

  • Remember that every single waterblock will cool better with increased flow. At different points for each block, increased flow benefits will start to taper off, but that fact remains that while keeping everything else in a loop constant, greater flow in a waterblock=better performamce. Using ProCooling.com's Interactive Waterblock Test Results, you can see that blocks such as the 6002 and Innovatek Rev3 see very little real-world noticeable temperature improvements with increased flow, while blocks such as the TDX see nearly 4 times the temperature improvement with increased flow as the previously mentioned blocks.
  • Pumps add heat to a loop, that fact cannot be denied. For a good explanation read Cathar's How much pump is enough? How much is too much? post on the Pro/Forums (scroll down to the discussion/conclusion if the numbers/equations seem too difficult). If you took my suggestion to read the entire Overclockers.com article, then you would have also read some more discussion of the topic on the top of the 2nd page.
  • Combining the previous two discussions together, you will likely see temperatures increase pairing an Innovatech Rev3 with an Iwaki MD-30RZ compared to using the same block with a Laing D5 or even a DDC. This is because the block doesn't see as much increased performance from the extra flow compared to the greatly increased heat put into the loop by the much more powerful pump. A block like a G4 or G5 will see increased performance from the extra flow of the Iwaki compared to a D5, as it is designed to effectively use the increased flow (if you read the post by Cathar, you would realize that I specifically used the MD-30RZ, which will give worse overall temperatures compared to cooler MD-20RZ).
3. BillR, not to pick on you, but the review you posted does not show a direct comparison of the Exos-2 with just the cpu block vs. the Exos-2 with 4 blocks in the loop, at the same cpu clock speeds. Going off topic here, but I highly doubt that review (even if it is on a site called "Legit Reviews"), as it shows the stock A64 hsf keeping a 3500+ Winchester at 41*C load/34*C idle, with a 25.5*C ambient. This review pretty much defies the laws of physics and chmistry, while showing temperatures far below any I have personally seen or read in many other reviews.

Back on the OP's topic, congrats on your setup, it looks nice and clean. I totally understand how you feel with your h2o setup finally done and working, now the only thing left to do now that you've pimped it online is to show it off at a lan party like I did ;)
 
Erasmus354 said:
I am well aware of the fact that pumps advertise max flow rate, ie the flow it can push with 0 restriction. However that is why they also test at varrying levels of restriction. Look at some of the pump comparisons over at procooling and you can reasonably estimate your flow rates without actually knowing the exact flow.

And anything under 1 lpm is quite low, high flow setups typically range in the 3-6 lpm range.
Click to expand...

Problem: The calibration settings of my flow sensor. AC states a setting of 550 imp/l, I take issue with that setting for various reasons, too long to discuss. From my own calculations, that is my engineers calculations, he determined the actual flow to be 215.33 l/hr, that is with the frequency raised to 65hz.
 
theseeker said:
Problem: The calibration settings of my flow sensor. AC states a setting of 550 imp/l, I take issue with that setting for various reasons, too long to discuss. From my own calculations, that is my engineers calculations, he determined the actual flow to be 125.33 l/hr, that is with the frequency raised to 65hz.
Click to expand...

Not surprised ;) Part of the problem is that if you look at the flow sensor it is quite restrictive, so I think it doesn't show the flow you would get with it not being in the loop. Many people open the bore on those flowmeters or use Innovatek straight line flow meters.
 
Top Nurse said:
Not surprised ;) Part of the problem is that if you look at the flow sensor it is quite restrictive, so I think it doesn't show the flow you would get with it not being in the loop. Many people open the bore on those flowmeters or use Innovatek straight line flow meters.
Click to expand...

I agree, straight line makes more sense, as this sensor is very restrictive.

http://www.sharkacorp.com/images/aquacomputer/flow/aq-flow-pp-lg.gif

If I lower the setting to 0 imp/l my flow is 655.35 l/h, very sensitive as well.
 
ikellensbro said:
<snip>BillR, not to pick on you,
Click to expand...

Lol, I don't feel picked on. Like most others I'm here to share and learn. None of us knows it all and some of us admit it ;)
 
Top Nurse said:
Not surprised ;) Part of the problem is that if you look at the flow sensor it is quite restrictive, so I think it doesn't show the flow you would get with it not being in the loop. Many people open the bore on those flowmeters or use Innovatek straight line flow meters.
Click to expand...

Way back in another life I used to work for Mercedes and part of diagnosing some of their fuel injection systems required one to measure “flow rates” of fuel through the system. We had all sorts of meters and devices with which to accomplish this action.

Ultimately those hard headed Germans agreed the only way to get a really accurate measurement was to time how long it took to pump “X” quantity of fuel from one container to another at a given pressure, timed with a stopwatch. To clarify, any meaningful flow rate has to be accompanied by a pressure reading.

That method is still used today by many racing teams (Top Fuel) comes to mind. Although everything those racers do is totally computerized the ultimate calibration used to feed the computer the “right info” is still in many cases done by hand, the old fashioned way.

We won’t even go into the other variables, temperature, atmospheric pressure etc.

It's all part of the fun ;)
 
theseeker said:
Problem: The calibration settings of my flow sensor. AC states a setting of 550 imp/l, I take issue with that setting for various reasons, too long to discuss. From my own calculations, that is my engineers calculations, he determined the actual flow to be 215.33 l/hr, that is with the frequency raised to 65hz.
Click to expand...

Seems quite high to me, I would expect your flow to not be much closer to 2lpm than to 4lpm with the amount of 6mm tubing, right angle bends, waterblocks, flow sensors, temp sensors, radiators, and other sources of restriction in your loop. Combined with a pump that is by no means a powerhouse....
 
My first OC attempt has yielded the following settings.

MULT 13.5
DDR 2.75v
2.5/3/3/10
CPU 1.5v
=2.75 GHz
CPU temp after running 3DMark 40c, NB temp 36

I will try to post some screenshots later as I am having photobucket problems.
In the meantime, how does this sound?
 
theseeker said:
My first OC attempt has yielded the following settings.

MULT 13.5
DDR 2.75v
2.5/3/3/10
CPU 1.5v
=2.75 GHz
CPU temp after running 3DMark 40c, NB temp 36

I will try to post some screenshots later as I am having photobucket problems.
In the meantime, how does this sound?
Click to expand...

Ambient temp?

BTW, some people bore out the AC flow sensors so that the intake and output have the same ID.
 
Top Nurse said:
Ambient temp?

BTW, some people bore out the AC flow sensors so that the intake and output have the same ID.
Click to expand...

Ambient room temp? If so it is 70F
 
So your unning about 38 degrees F over ambient correct? Don't think that will be a problem, but usually the temp difference is less than that. How many coolers in that circuit?
 
Top Nurse said:
So your unning about 38 degrees F over ambient correct? Don't think that will be a problem, but usually the temp difference is less than that. How many coolers in that circuit?
Click to expand...


CoolingCycle.jpg
 
As promised.

FirstOC.jpg


This was right after I ran prime 95 for 3 hours and 3DMark. These settings are totally stable, however I want to go farther. As always all ideas and suggestions will be greatly appreciated.
BTW, this system is in my home office along with a 300gal reef tank and the chiller vents into the room. Therefore my room temps fluctuate quite a bit.
 
If I read that right that temp is at idle. That does seem a bit high. What were your load temps?
 
BillR said:
If I read that right that temp is at idle. That does seem a bit high. What were your load temps?
Click to expand...

Actually those were my temps right after running Prime95 and 3DMark. Currently my CPU temp is 30c.
I am more concerned about the OC'ing. I know you are shy, but try any way :D
 
You must log in or register to reply here.
Back
Top