D-tek nozzles

[RickCain]

Sharka has them in stock. I saw (and got) them on Friday and appeared to have plenty.

http://www.sharkacomputers.com/dtekfuzion.html

[Jedda]

Yeah, they're pretty much around the outlets now.

Petra's has them back in. So does Tom at Chilled PC for the UK side of things.

[crash2k]

hi guys, i'm going to put on a new rig with an E6850 and D-Tek FuZion with nozzle. which nozzle is better to cool the E6850? any advice?

thank you in advance.

[Digital Nirvana]

I think overall consensus is that the 4.5mm nozzle is the best overall right?

[peebee]

but for Quad it is the quad-Nozzle! Isn´t it?

[Digital Nirvana]

but for Quad it is the quad-Nozzle! Isn´t it?

yes, but the E6850 isn't a quad

[peebee]

Sure! I just missunderstood your sentence

I think overall consensus is that the 4.5mm nozzle is the best overall right?

Thought you mean any processor

sry

[crash2k]

so i go with 4,5mm nozzle on the dual core on the E6850 and, of course, the specific Quad-Core one for QXXXX

thank you very much guys, precious as always

[Omastar]

I got my nozzles from Petra's yesterday and installed the quad-split nozzle. After opening the block up (lost my original hex tool, haven't bothered to find another), I saw it was pretty gunky, so I cleaned 'er and reassembled.

The catch is my computer is in total disarray; I'm waiting on my 8800GT from step-up (coming Tuesday) to put my loop back together and test. Also waiting for my lapping kit from EasyPCKits.com to lap the Q6600. Then I will provide numbers.

[Ic3man]

Did a little testing:

Q6600 G0 (lapped) @ 3.6 1.45v using mx-2.

Without

Idle 40,39,36,36
Load (p95 5 mins) 55,56,51,50

Quad nozzle + seal

Idle 39,39,35,35
Load(p95 5 mins) 54,54,49,49

4.5mm + seal

Idle 36,36,34,34
Load (p95 5 mins) 51,51,47,47

Certainly seems like the 4.5mm nozzle works better than the quad nozzle in my case.

[Xilikon]

Weird... In my own case, the quad worked better than the 4.5mm one (2-3 degrees less). I guess the flow restriction is too much to make it effective. However, I have a EK FC8800 GTS in the same look unlike yours, which may account for the differences.

[Waterlogged]

What were the ambients for all 3 tests?

[peebee]

I had better results with the quadsplit Nozzle.
I had ~450l/h before and have now ~350l/h Loop is CPU only.
Temps decreased about 3°C per Core

[Ic3man]

What were the ambients for all 3 tests?

20c iirc.

[PeppeX]

This will be my first WC setup.

D-tek Fuzion
MCW60
Thermochill PA120.2
Micro Res
Swiftech MCP655/D5 Vario

I have a e6600 and a 8800GT

What nozzle should i use? Will 4.5 increase my gpu temps by alot?

[Red Evil]

This will be my first WC setup.

D-tek Fuzion
MCW60
Thermochill PA120.2
Micro Res
Swiftech MCP655/D5 Vario

I have a e6600 and a 8800GT

What nozzle should i use? Will 4.5 increase my gpu temps by alot?

Not alot but its still worth it`s price.
BTW thats exact same setup as mine. I only have the 120.3 rad.
I got a bit unlucky cos mine D5 dosnt work

[PeppeX]

So what nozzle are you using?

[avrionis]

I can believe everyone using the nozzle to drop temps. But the flow gets reduce (as indicated in the excel worksheet) from 2.3 to 1.05 for a cpu block_chipset block with a 120.3 & 655!

Is 1.05 still a good flow rate!

[tommithy]

I can believe everyone using the nozzle to drop temps. But the flow gets reduce (as indicated in the excel worksheet) from 2.3 to 1.05 for a cpu block_chipset block with a 120.3 & 655!

Is 1.05 still a good flow rate!

The real question to ask yourself is this:

Are all my temps lower even with reduced flow rate? After all lower temps is what we are after right?

[Martinm210]

Yeah, don't get too worked up with flow rates. If it's one thing I'm really learning about thermal performance is that flow rates above 1 GPM really don't make all that much difference. You really have to start getting down into the .7 GPM range before the thermal curves for blocks and radiators start dropping off.

So I agree with Tommithy's response. It doesn't matter if it gives you better temps. The other interesting thing about pumps...they dump LESS heat as you add restriction. It's the increased RPM in the pump motors that increases a pumps power consumption at any given voltage. I would have thought added resistance makes a pump work harder and dump more heat, but it's the other way around. It's the motor spinning too fast and acting as a brake.

Anyhow, in a CPU only loop with the latest high surface area blocks, flow rate becomes less and less important. 5 years ago when everyone was running low head pond pumps and large channel type blocks, flow rate was much more critical. Partly because the pumps didn't have the pressure capabilities without alot more heat and also because the thermal curves for those older blocks dropped off much more rapidly. Now with all the micropin/nozzle block you still get very turbulent flow and good mixing at lower flow rates.

My 2C.

I'm thinking the 5.5mm nozzle for dual cores and the quad nozzle for quads is a good place to start with standard pumps.

[avrionis]

Thanks guys!

When I looked at this excel worksheet assuming using a 4.5mm nozzle. It came out to only 1.05 GPM (it is at the low end of good flow) . Is this go enough! Since one day I might put a GPU WB also. Or should I add another 655 in series! Worksheet says it will increase the flow to 1.45 (it is now at the low end of great flow)

What is the optimal flow rate that we should have above 1 or above 1.5GPM?

[PeppeX]

So martin do you think this setup with a 4.5 nozzle would be good idea or should i go 5.5?

MCR320
MCW60
D-tek fuzion obviously :p
Micro Res
laing D5 speed 5
1/2" tubing

On your graph i got 1.07 flow rate with the 4.5 nozzle

[Talonman]

Yeah, don't get too worked up with flow rates. If it's one thing I'm really learning about thermal performance is that flow rates above 1 GPM really don't make all that much difference. You really have to start getting down into the .7 GPM range before the thermal curves for blocks and radiators start dropping off.

Good nugget of info to keep in mind!

Thanks for the post.

Oh, and this too:

The other interesting thing about pumps...they dump LESS heat as you add restriction. It's the increased RPM in the pump motors that increases a pumps power consumption at any given voltage. I would have thought added resistance makes a pump work harder and dump more heat, but it's the other way around. It's the motor spinning too fast and acting as a brake.

More good info.

[Martinm210]

So martin do you think this setup with a 4.5 nozzle would be good idea or should i go 5.5?

MCR320
MCW60
D-tek fuzion obviously :p
Micro Res
laing D5 speed 5
1/2" tubing

On your graph i got 1.07 flow rate with the 4.5 nozzle

I'm thinking 5.5, 6.3, or even a quad nozzle. I'm just not getting that good of performance out of my 4.5 and that's on a CPU only loop with a D5 all to itself but I'm still pretty early in my nozzle testing.

[disruptfam]

anyone tried the quad nozzle on a dual core?

martin?