The impact of tubing sizes

Okay I know this is a little off topic, but here it goes.
Petra calculated that the DDC-2 is about 15% efficient. The other 85% has no place to go except to be turned into heat. Common sense says that part of that heat is going into the air, the PCB, the plastic housing, and eventually...into the air from those. Another part is going into the water. Here it doesn't matter that much because its only a few watts at most. And I'm sure plenty of people have reported that their pump's cases get a little warm during operation, meaning a good deal of the heat is NOT going into the water...

I love to see a man on a mission.
Remember, You can lead a horse water but you can't make them drink.

it would be interesting what the figures are like with the 10W pump you initially used

0.1C difference doesn't seem all that much if you jump to Iwaki :confused: I am not sure what you mean there though????? Is that a proportional measure or actual difference?

Quote:

---

Originally Posted by serialk11r

Okay I know this is a little off topic, but here it goes.
Petra calculated that the DDC-2 is about 15% efficient. The other 85% has no place to go except to be turned into heat. Common sense says that part of that heat is going into the air, the PCB, the plastic housing, and eventually...into the air from those. Another part is going into the water. Here it doesn't matter that much because its only a few watts at most. And I'm sure plenty of people have reported that their pump's cases get a little warm during operation, meaning a good deal of the heat is NOT going into the water...

---

Yeah, I use 16% efficiencies in my calculations when mapping from hydraulic power (restriction x flow) to estimated required pump power (power draw). I then multiply that value by 0.9 to arrive at the pump heat dump directly into the water. This is all in the estimation stage.

A DDC2 is 18-20W. At a 90% water heat dump, 2W is going into the bits not touched by water. Hook up a common DC electrical resistor to a DC power supply, and feed 2W of power through it. Wait 1 minute and touch it, and let me know if it doesn't dang near burn your finger off. That's the sort of heat that you're feeling from the bottom of a DDC. Blow some air over it and it'll feel cool all the time. Most pumps sit on their bottoms with no air-flow. Give anything small 2W of heat load and sit it where it gets no air-flow, and it WILL get quite warm.

Yer typical high-performance GPU ram chip generates around 2W of heat when under load. Touch one without a heatsink on it after its been working away for a while and you'll just about scorch your skin.

Remember, temperature != heat.

Yeah but what I'm saying is the housing is plastic...and all plastics are poor conductors of heat. Anyways I understand this is all in the estimation stage and stuff.
Well anyways enough off topicness :P

Quote:

---

Originally Posted by dinos22

it would be interesting what the figures are like with the 10W pump you initially used

---

Add ~0.3C to the CPU, ~0.5C to the GPU pretty much across the board when compared to the DDC2.

Quote:

---

0.1C difference doesn't seem all that much if you jump to Iwaki :confused: I am not sure what you mean there though????? Is that a proportional measure or actual difference?

---

Your CPU and GPU will be ~0.1C cooler when using an Iwaki RD30 as opposed to a top-modded DDC2, is what I meant.

Great work! So even as heat load and an addition of a block, the results still vary very little (say that 10x!) I would have thought that with the addition of both more restriction and the addition of a higher heat load, the larger ID tubing would have shone though. Apparently I was mistaken!

Even if it is theoretical data + fudge, I still think that there is enough accuracy in the program that the real-world tests would come to within, say, 3-5% of the theoretical. Great work, my friend :D

Quote:

---

Originally Posted by serialk11r

Yeah but what I'm saying is the housing is plastic...and all plastics are poor conductors of heat. Anyways I understand this is all in the estimation stage and stuff.
Well anyways enough off topicness :P

---

Plastic conducts heat poorly, and thats just the problem. The air in the pump PCB section will be getting hot 'cos the heat can't escape 'cos of the plastic surrounds. Eventually the plastic surrounds will warm up a fair bit due to the temperature of the air inside the pump. If you insulate something well enough, it doesn't matter if you give it 0.1W of heat-load, it will continue to get hotter and hotter if the heat can't escape to somewhere else.

Quote:

---

Originally Posted by Cathar

Add ~0.3C to the CPU, ~0.5C to the GPU pretty much across the board when compared to the DDC2.



Your CPU and GPU will be ~0.1C cooler when using an Iwaki RD30 as opposed to a top-modded DDC2, is what I meant.

---

With such a small heat gain from the Iwaki the RD30 is looking even more appealing from a reliability standpoint.

Great data Cathar! I *do have to wonder, however, why this kind of testing wasn't done by companies (I'm aware that Swiftech has kept 3/8" around, but they still haven't advertised the difference, or lack there of...), who spend a lot money developing and selling this kind of technology...makes ya' wonder ;)

Anywho, this is GREAT news for the 'small' future of computers; as more and more people start using SFF cases, tighter bend radii will be needed, and now that we *know there isn't much of a difference, it'll make the transition that much smoother :)

Quote:

---

Originally Posted by Cathar

Add ~0.3C to the CPU, ~0.5C to the GPU pretty much across the board when compared to the DDC2.



Your CPU and GPU will be ~0.1C cooler when using an Iwaki RD30 as opposed to a top-modded DDC2, is what I meant.

---

i guess the only reason you'd want an Iwaki is for reliability (so they say) lol

those differences are so minuscule it's mind boggling

i am learning today :)

I thought about going to an rd30..but after contemplating going through the hassle of getting a meanwell converter and mounting it..etc...the ddc2 was more appealing,smaller,and a helluva lot easier to deal with..

Quote:

---

Originally Posted by nikhsub1

that I was considering moving to smaller tubing with push fittings as I'm bored, I think it is neater and will not impact performance much at all. Now to find the proper tubing, g 1/4 and g 3/8 push style fittings - hmmmm.

---

You just made me smile!

This is great data provided by Cathar and should bring a sense of calm regarding the "high flow / low flow" fights on multiple forums. :up:

Quote:

---

Originally Posted by TheJollyFellow

Great data Cathar! I *do have to wonder, however, why this kind of testing wasn't done by companies (I'm aware that Swiftech has kept 3/8" around, but they still haven't advertised the difference, or lack there of...), who spend a lot money developing and selling this kind of technology...makes ya' wonder ;)

Anywho, this is GREAT news for the 'small' future of computers; as more and more people start using SFF cases, tighter bend radii will be needed, and now that we *know there isn't much of a difference, it'll make the transition that much smoother :)

---

Tis a very slippery slope what you ask. First, people as a whole do not trust mfgr data - this is just fact. If company xxx came in here and said what Stew said, they would likely be laughed out of here and the thread closed in short order. There is one mfgr in particular (won't mention names) that presents VERY good data that is often balked at - I have been guilty of this myself. This company has been very discouraged at times of this attitude (and rightly so IMO) and has seriously considered releasing no data whatsoever (would be a mistake IMO). So you see, the mfgr's make their wares and it is up to (IMO) the unbiased users with a brain (see Cathar for example) to decipher the crap from the good.

Quote:

---

Originally Posted by Cathar

The PQ curve for the pump is replicated from the Laing site.

---

hold the phone. this means the PA120.2 and GTX pressure drops weren't factored in....so all the data would be different...right?

Quote:

---

Originally Posted by sick_g4m3r

hold the phone. this means the PA120.2 and GTX pressure drops weren't factored in....so all the data would be different...right?

---

:confused: You asked me that exact question before, and the answer I gave was that the radiator and water-block pressure drops are factored into the system PQ curves that intersect with the pump PQ curve.

Quote:

---

Originally Posted by Cathar

The separate curves for the system include the pressure drop of the radiator, the waterblock, the tubing, and the fittings. Where the curves intersect with the pump curve is the flow-rate that we will see.

---

yeah i asked that, and you replied you copied it exactly from the website, but it is the tubing curves that you altered, not the DDC curve. the DDC curve also must include the pressure and flow drops to obtain the correct data

Quote:

---

Originally Posted by sick_g4m3r

yeah i asked that, and you replied you copied it exactly from the website, but it is the tubing curves that you altered, not the DDC curve

---

Why would I want or need to alter the DDC curve?

Quote:

---

Originally Posted by Cathar

Why would I want or need to alter the DDC curve?

---

Because those figures will not be the same with the parts used

Quote:

---

Originally Posted by sick_g4m3r

Because those figures will not be the same with the parts used

---

The DDC curve operates against some level of restriction. I factored the current 3/8" fittings on the pump into the system restriction curves. When I modelled the different fitting types, I can determine the impact in one of two ways:

1) Alter the pump's PQ curve based upon the fittings on it
2) Factor the fitting impact into the system's PQ curve, and intersect that with a constant pump PQ curve.

There may be very marginal variations between the two, but as I said, the nature of the shapes of the curves is such that any variations impacts everything roughly equally.

It's just easier to do #2.

ok i guess i see now. thanks for explaining yourself

Quote:

---

Originally Posted by serialk11r

Petra calculated that the DDC-2 is about 15% efficient.

---

...mmmkay... was this some random guesstimate that I happened to have mentioned over AIM or something (likely based on testing that Lee did)? Because, up until just now (since you brought it up), I hadn't run the calculations with my own data (which has already been published in XS).

After doing a little graphing and some rough calculations, I came to the following (all at peak pumping power):

Stock DDC-2: ~17% efficient (~3W max pumping power, ~18W total consumed)
Stock DDC-3.2: ~13% efficient (~2.1W max pumping power, ~16.32W total consumed)
DDC-2 w/DDCT-01s top: ~21% efficient (~4.25W max pumping power, ~20.64W total consumed)
DDC-3.2 w/DDCT-01s top: ~19% efficient (~4.2W max pumping power, ~22.32W total consumed)

Quote:

---

Originally Posted by sick_g4m3r

hold the phone. this means the PA120.2 and GTX pressure drops weren't factored in....so all the data would be different...right?

---

The pressure drop incurred by the components used doesn't change the overall performance characteristics of a pump (there's a reason why you see intersecting curves in the first graph that Stew posted... The red curve represents how the pump performs (flow) at all head pressures attainable by the pump, the other curves represent the pressure drop of the cooling components used vs. flowrate, and the point of intersection represents the head pressure at which the pump operates and flowrate the pump attains given the components attached to it).

Hopefully that'll clear things up a little for you.


Anyway, interesting information, Stew... :up:

Quote:

---

Originally Posted by Petra

The pressure drop incurred by the components used doesn't change the overall performance characteristics of a pump (there's a reason why you see intersecting curves in the first graph that Stew posted... The red curve represents how the pump performs (flow) at all head pressures attainable by the pump, the other curves represent the pressure drop of the cooling components used vs. flowrate, and the point of intersection represents the head pressure at which the pump operates and flowrate the pump attains given the components attached to it).

Hopefully that'll clear things up a little for you.

---

there we go! thats just what i needed, thanks a ton. and again, amazing info cathar

Yeah the point at which push comes to shove, as it were.

Petra you brought that efficiency thing up about 4 months ago or so, when I made my rant thread ;)