WC tubing size makes little difference!!!!

BristolBulldog · 18 Jun 2007 at 09:43

The thermaltake kit is very good. The amazing thing is their CPU block in the 745 kit. I changed this to a swiftech apogee extreme and i got the same temps (well, around a 1c drop max actually). the weight of the block is excellent.

What is terrible is their chipset, and vga block. the mounting is terrible. Also their piping age hardens very fast, and i ended up with many pipe fractures. For the cost the performance is very very good, and within a few degree's of 1/2" higher flow systems.

weescott · 18 Jun 2007 at 09:58

BristolBulldog said:
The thermaltake kit is very good. The amazing thing is their CPU block in the 745 kit. I changed this to a swiftech apogee extreme and i got the same temps (well, around a 1c drop max actually). the weight of the block is excellent.

What is terrible is their chipset, and vga block. the mounting is terrible. Also their piping age hardens very fast, and i ended up with many pipe fractures. For the cost the performance is very very good, and within a few degree's of 1/2" higher flow systems.

Now THAT is a contradiction if ever I saw one!

LMAO

WoZZeR · 18 Jun 2007 at 10:14

I use 8/10mm throughout, my GFX blocks being Aquacomputer have G1/8th threads and they cool fine.

Finally vindicated.

Marci · 18 Jun 2007 at 12:14

He had an extremely nasty bike accident that just about killed him.

Heh - he had one crash that nearly killed him, more or less recovered, then had another that laid him back up again more recently...

gurusan · 18 Jun 2007 at 12:15

this is hardly surprising...if 2 setups are using the same pump then smaller tubing will result in higher flow velocity which is the important thing. If the pump is flowing more than the 3/8th tube can handle then that's when it becomes a restriction.

Basically the same as in a vehicle exhaust system....smaller diameter piping will make more torque down low in the RPM range...but once the vehicle starts pumping much more exhaust in higher RPMs the smaller piping will become a restriction and larger is needed.

..but in the low RPMs if you use larger piping you'll lose exhaust velocity and consequently power

BristolBulldog · 18 Jun 2007 at 13:27

weescott said:
Now THAT is a contradiction if ever I saw one!

LMAO

erm. ok. Probably right there. I'll clarify.

The temps you get from the thermatake kit are very very good. It made me dissapointed in my 1/2" kit because i was thinking "thermaltake= cheap, and low bore size. so 1/2" and higher rate pumps etc will get me much better temps. It didnt.

On the other hand, there are certain things i really didn't like about it, as mentioned above. Better now?

weescott · 18 Jun 2007 at 14:05

BristolBulldog said:
The temps you get from the thermatake kit are very very good.

...and then you switch on the PC!

But seriously. I know what you you are trying to say and indeed Thermaltake kits will be OK for some people.

drunkenmaster · 19 Jun 2007 at 15:35

don't know why anyone surprised, firstly, big pump tests have shown very very VERY little temp change with differing flow rate before. also, its water, pumping 4 or 6 litres through a tiny tiny waterblock will mean that either way the water stays in the block for a very short amount of time and doesn't heat up much at all.

the thing here that creates a slight unknown and problem is it was all testing with the same pump, one thats not exactly the highest pressure pump around aswell. so maybe it simply doesn't have the juice to really push more water through wider bore tubing. you need to do the same test on a D5, and a aquaextreme whatever the heck its called, the one that does 10 metre head pressure or close to that. will the results change that much, i really don't know.

think of it like testing a 8800gts and gtx with a 2Ghz single core cpu, they'd perform very closely in 3dmark because both would be hugely cpu limited. test it with a quad core aswell to shows a very different picture.

i still don't think there would be huge amount of difference in temps anyway.

once you get past a couple litres a minute the water just doesn't have time to heat up really.

i guess you also have to test a quad core overclocked a crossfire 2900xt's and see if the temps still hold up, the difference in tubing, assuming pump>rad>cpu>gpu>gpu>res wouldn't be that big in the cpu, but what about by the 2nd gpu?

Amonlym · 19 Jun 2007 at 16:34

swapped from a koolance 2PC601 (1/4" iirc) with the cpu305 block to 1/2" tdx etcetc. the temp drop... was near as makes no difference equal to the reduction in water temperature. the water blocks/tubing/having gpu+nb in loop made no notable change.

granted, going from 38c water at load to 21c water at load makes a big difference. but thats not the fault of the tubing or block. Just rad + fans were not as capable.

thats about what i expected anyway so..

MikeTimbers · 19 Jun 2007 at 16:48

I know it's a long read but please ensure you read the whole thread. There is some fantastic information in there and some awful misinformation here!

chaparral · 19 Jun 2007 at 18:50

But these tests by Cathar where only done with about 100w of heat....Which is not really very much heat at all....

I would really like to see these same tests done with about 300 to 400watts of heat..

Raikiri · 19 Jun 2007 at 19:17

chaparral said:
But these tests by Cathar where only done with about 100w of heat....Which is not really very much heat at all....

I would really like to see these same tests done with about 300 to 400watts of heat..

Estimation for an 8800GTX (150w) and 120w CPU:

1/4" QF : CPU = 62.33C, GPU = 52.28C
5/16" barbs: CPU = 62.31C, GPU = 52.27C
5/16" QF: CPU = 61.48C, GPU = 51.39C
3/8" barbs: CPU = 61.56C, GPU = 51.47C
3/8" QF: CPU = 61.13, GPU = 51.01C
7/16" barbs: CPU = 61.03C, GPU = 50.91C
1/2" barbs: CPU = 61.03C, GPU = 50.91C

It's somewhere later on in the thread, I think he plans to test some more stuff.

chaparral · 19 Jun 2007 at 19:23

Raikiri said:
Estimation for an 8800GTX (150w) and 120w CPU:

1/4" QF : CPU = 62.33C, GPU = 52.28C
5/16" barbs: CPU = 62.31C, GPU = 52.27C
5/16" QF: CPU = 61.48C, GPU = 51.39C
3/8" barbs: CPU = 61.56C, GPU = 51.47C
3/8" QF: CPU = 61.13, GPU = 51.01C
7/16" barbs: CPU = 61.03C, GPU = 50.91C
1/2" barbs: CPU = 61.03C, GPU = 50.91C

It's somewhere later on in the thread, I think he plans to test some more stuff.

Thanks...I didn't see this test....

I hope he does a test with CPU / GPU /NB / SB in one loop....As this is what i been looking at doing with my system..

As it be much more easy with 3/8" tubing...
But then i have to change all my 1/2" barbs ,hose clips & tubing

MikeTimbers · 19 Jun 2007 at 20:51

No you don't. 3/8" or 7/16" would still fit over 1/2". The point is that many 1/2" barbs have only 3/8" real internal diameter so it really doesn't matter if you use 3/8" tubing.

The Halk · 19 Jun 2007 at 21:48

drunkenmaster said:
don't know why anyone surprised, firstly, big pump tests have shown very very VERY little temp change with differing flow rate before. also, its water, pumping 4 or 6 litres through a tiny tiny waterblock will mean that either way the water stays in the block for a very short amount of time and doesn't heat up much at all.

I hope you don't feel I'm nitpicking here. Just a little correction.

No matter what speed the water is going, as long as it is moving, it spends the same amount of time in every position in the loop. So the length of time each particle of water is in a block doesn't matter.

What does matter is the turbulence, the faster the water goes, the more impacts it has with the block, so the more heat is transferred.

Everybody accepted that there were diminishing gains - at some point. But there wasn't a concensus about when that started to happen or to what degree.

What Cathar has shown is that it happens much earlier than was widely believed.

There is no downside to flow at all. The faster the flow is the cooler the block will be.

But we cannot increase flow without increasing the heat dump into the water - i.e. the more work the pump does, the more heat it produces, so the more the water heats up from that.

At a certain point it is worse to increase the power of the pump, because the extra cooling performance from more flow is offset by the extra heat dumped into the water. But before you reach this point the gains from extra flow have already diminished to the point where it's almost pointless.

That's what Cathar was saying, as best as I can understand it.