Well I found keeping all of my connections the same size was a pain in the ass. You get one component that 3/8", a couple 1/2", and one 3/4". So I put adaptors on everything and connected 3/4" tubing. I am running a Swifty 600 pump (1/2"), a maze 4 block (3/8"), and a chiller (3/4") connections. That said I supose that some of the reductions in size are messing with my flow but such is life. As is often the case with water cooling, I started out small at 3/8" and worked my way big to 3/4".

I am going to try and hook up some quick connects between the PC and the chiller. Thinking of trying to use 3/4" air hose connectors. Not to sure how this will work out but worth a try. I found some the other day at the Ace Hardware. I might want to take this dam thing to a LAN someday. Without quick connects trying to lug the chiller and pc together would be quite a pain.

Well that proves absolutly nothing about 3/4" being any better as per your first post.

3/4" is hard to clear air bubbles out of

Yeah 3/4" is big stuff. Just what would the minimum flow rate have to be to just clear air out of such a system pH?

Ok. I've been doing some digging. The kit I originally had in mind has also been submitted to a magazine for a shootout with other commercial offerings which will be published this time next month...
I got this month's issue yesterday and whaddya know? There's a review in there of the Swiftech 22501 Extreme kit with a retail price just 50p more than the MSRP of ours. It also uses a twin 120mm rad. :)

Assuming the 6002 would be a replacement for the MCW5002, what's wrong with setting these two kits against each other? ([Edit] With your proposed block substitution, of course)
It appears to cover all aspects that would set us on the level playing field I'd hoped for - complete kit, same price point, twin fan rad, fill & bleed sytem (equivalent to our res approach).
Should quantify whether our approach is "adequate" and yours is "performance oriented" quite well. :cool:

Surely there would be no advantage of using 3/4" tubing if the pump is only at 1/2"? Maybe if the pump had a larger connection, then it may even increase the flow/speed (I still don't understand all these terms perfectly) going in to the blocks but otherwise it might even slow it?

3/4" tube gives lower friction losses - the water is moving more slowly and there's a higher ratio of water to tube surface, so less friction. I doubt you'd see much real difference though.

pay attention to pH
you will not clear the lines easily (if at all)
I use 5/8, and it takes ~ 2gpm to clear

I switched back to 1/2" on test loop for that reason.

Has anyone here used clips like on the heatkiller before? It gives me "the fear"

Try it with two springs each side then :eek:

http://www.wizarddesigns.co.uk/images/chillDew.jpg

Only time I've ever seen the benefit of shims...

I played with it some using a dead mobo and CPU and eventually could get it to go on. There's a bit of an art to it. I don't know how comfortable I'd be doing it w/ mobo installed in case and hoses mounted on wb though. A pity as I'd really like to test it. Eventually I'll get bold I guess...

Hi guys, i am from germany.
I don't posted here up to now i think. But read that page for a looong time :-) . I have to say that it is also for convenience. These plug and play systems from Festo are easy to maintain and safe. I build my waterblock myself and also gfx and northbridge block. I use silicon tube with 10 mm inner diameter. Don't know what's that in inch. But it was really complicated to get these with an metric thread. I use hose clamps and had "some" accidents with an flooded PC.
Luckily nothing was damage (up to now). With these Festo connectors, there isn't that high of an chance for waterdroplets in PC.

Don't know how good my english is, but i also made an article that might be interessting for some ;-)

--------> *with some nice PICS me thinks* <-----

:D

Also did i use an spacer up to now in all mywc'ed PC's (2) - and i didn't broke an CPU ever because of mounting my waterblock.

MfG QT

welcome Qerryterry

your English is fine, appreciate the effort to post in another language - something impossible for many of us

welcome

thanks for that page, itll kill some time. i always like reading work logs :D

also 10mm is about 3/8 inch, which is what smaller US systems use

But think of that -> I use 10 mm inner diameter silicon.
These Festo what us "germans" use are 8 or 10 mm outer diameter. Plastic there because of little blades in these fittings which hold that tubing.

I like worklogs too - thats why i wrote one hehe...

MfG QT
( good to hear that i am understandable :shrug: )

Hmm, belive me I tried to read all EIGHT pages but failed miserably...
and went tot he last one and guess what? Same thing, willy waving and nothing else.
Have we got rules yet or still battling them out?

About diameter of tubing...
Coolant musn't loose any velocity while in transit between pieces of WC setup. The ideal conditions for such behaviour is laminar flow situation in tubing. One thing to remeber though, flow characteristics move towards turbulent behaviour with increasing vessels diameter and velocity. To repehrase, with set velocity increse in tubing diameter increases Re number. Keeping the same velocity (means lowering pumps output) and decreasing vessels diameter lowers Re number pushing flow characteristics towards welcomed laminar flow :). Look at equations, use imaginary numbers to test behaviour of results to see it plainly:).

Conclusion -> from theoretical point of view there's a sweet spot for pipind diameter for every system (pressure drop after each unit) and in perfect system piping diameter would change after each 'consumer unit' to obtain flow as close to laminar as possible:)

Hunh? If you increase the diameter the velocity goes down - unless you're also increasing flow.

Or are you saying that, for any pump, there is a diameter that is too large and that going over this limit will reduce flow?
This is a new one for me.
I would have guessed that if I'd set up an experiment with a pump and a number of pieces of tubing, all the same length and material (so same water-to-solid friction) but different IDs, and I measured flow rate through them that I would see (smaller ID to larger ID) either no change (point of diminishing returns already reached) or lower rate to higher rate, with at some point, there being no difference in flow associated with increased ID.
I understand about stall behavior in wings, but the issue there is keeping the base layer firmly "stuck" to the wing surface so as to use the coanda effect. I'm having trouble seeing that the same applies to tubing. Got some web references?
Bob

With the same pump flow rate is going to be THE SAME for each tubing diameter in open system test (tubes connected to pump, reservoir and receptacle). Velocity of fluid will vary and be lower with increased diameter.
What I described was an experiment where you can keep VELOCITY constant but flow rates are changing depending on tube diameter.
What you say is right about the point of decreasing performance.
If you looked at Re equation you could see how it is dependant:).
I am no avionics expert I am affraid and I cannot say if your analogy is right. If you had your gas as incompressible then yes, it would behave as fluid :).
Boundary layer is yet another matter:)

Hi,
I'm using German components (AquaComputer) in my rig, connected by 10mm I/D Tygon.
I've got a 2.4c Pentium o/c'ed to 3.0 Ghz, a watercooled Radeon 9700 and a WD Raptor in a HDD cooler. The diode temps are always below 30 celsius (86 farenheit), its nice and quiet and it looks good, so I'm happy (and in the end that's the only thing thats important).

When I was wondering which kit to buy I must say I did worry about the supposed 'inferiority' of the small diameter tubing, blocks etc, but I only wanted to watercool my rig to make it quieter (it wasn't overclocked then) so I thought it wouldn't matter too much.

I eventually decided on AquaComputer kit for a combination of reasons (performance, build quality, looks and ease of installation), and once I had installed the kit in my rig, I was so delighted with the results I promptly decided to overclock it. I admit I have no experience of 'American' style 1/2" kit, but I can honestly say, from a practical viewpoint, I can't see what extra benefits it would bring.

My view is that as long as your gear does a decent job, and your happy with it, that should be enough, but again this is only my personal point of view. Having said all that, I can't wait for the results of the tests !

Olly

Olly, welcome to the forums. Indeed what you propose is a "sensible" approach to water-cooling. Need to appreciate the different sub-markets that exist within water-cooling though. If all everyone wanted to do was to achieve a quiet but practical cooling solution with water, we'd all be running Eheim 1046 pumps with blocks with a few parallel channels and a semi-passive radiator sitting in the lower-front of the case whose only air-flow came from the PSU and a couple of case-fans sucking air out of the case.

It also comes down to what you want to achieve with your water-cooling. For some when overclocking hard, the difference between acceptable water-cooling and top-end water-cooling may very well be as much of a 10% overclock. Admittedly 10% may not be all that user-visible, but if it folds faster, crunches data faster, and gives higher frame-rates, then this is the sort of "reward" that these people are looking for with high-flow systems.

For some people, the ability to buy a $200-250US water-cooling system that can offer better gaming performance (through better cooling of the CPU and GPU) than a $1000US phase-change setup which only cools the CPU is their goal, and if buying a better block brings them that much closer to what a phase-change unit can do, without the $1000 price-tag, then that's also where the reward lies.

That pretty sums it up here :) Going double phase change is prohibitive due to exorbitant costs and real life performance increase is not that staggering if one checked how much next FPS, Ghz or whatever you want to measure costs compared to decent WC assy.

Welcome Olly, I am also itching to see some testing results here
:D

your ambient has to be really cold if your CPU is below 30C. what is the general ambient and how much did you spend on the kit?

He did say that "The diode temps are always below 30 celsius", rather than the cpu is always below 30deg celsius.

I'm not sure whether this is recongnition on his part that the diode temp can often be misleading or not, though I suspect this is the case.

In other words, there is nothing wrong with what he said. I don't belive he is trying to claim that the cpu is at 30deg celsius.

8-ball

Relax. He did say that those are the P4 diode temperatures, which are about as accurate as sticking a thermal probe in one's room and calling that the CPU temperature...

i get too involved in my geek discussions, sorry

i cant wait to see the results

Any update as to whether this comparative test is going ahead?

Hi, I'm from Portugal. I have been reading this forum for some time now, and am really curious about the tests mentioned in this thread. Anyway i'm a little confused about something...I have an Athlon XP-M 2500+ @2526Mhz (210.5x12) @1.87v in a A7N8X rev 1.04 with 2x256Mb TwinMos Winbond CH-5 @ 2.7v 2-3-3-11, all rock stable...till last week i used air cooling (Coolermaster Aero7+), and while in winter it gave me decent temps, now the summer has arrived, and since here in Portugal the temps have been 30ºC(86ºF)and above for the last few weeks, my temps went WAY up to like 43/44ºC (109/111ºF)idle for cpu, load i was afraid to check :cry: ...so this situation made up my mind and i went and bought a WC kit composed of:
- Innovatek HPPS 12v pump
- 1A-Cooling 5Z280V1 Radiator
- 1A-Cooling HV2 WB for cpu
- Aquacomputer TwinPlex rev 1.2 for northbridge
the complete kit (except for the chipset wb) can be seen at this thread
The kit was really EASY to install...and it is all 10/8 piping(u can check a pic of mys sys here)...So what did this change in terms of temps? they went down 15ºC!!now im at 29/30ºC (84/86ºF)idle and 36/37 full (96/98ºF) (cpuburn, prime95 only gets to 33/34ºC (91/93ºF). My room temp is about 26/27ºC (79/80ºF) (although im not sure this temps are accurate, but it IS damn hot) during the day. Now, my question is...are these temps bad? All the US-systems get THAT better performance than this? and i dont have a high-end "german/european" system, its more on the "medium" range...sorry for such a long post :rolleyes:

Your temps are not comparative. Hard to say. Also depends on what you want. 1-2C (probably the difference between the best of the 2) is not much to me but it is HUGE to some people. I just saw a guy at ocforums buy a $1,300US phase change setup to go from 2800mhz to 3000mhz. I personally would have just bought a faster CPU for $1,000 less but you get my idea on what people want.

Try a little more pressure. The block starts sounding like a high pressure sprayer.