D-Tek Fuzion Nozzle Kit Pressure Drop Test Results

Printable View

12-07-2007, 11:33 AM
Martinm210

D-Tek Fuzion Nozzle Kit Pressure Drop Test Results
So the latest thing with the Fuzion is to get the nozzle kit, yet everyone asks what nozzle is best? The answer obviously depends on your particular setup. But the Fuzion has always been known to be very free flowing, with little pressure drop. This is great for multiple block loops because the block without a nozzle won't effect flow rates too much for other blocks and your radiator in the same loop.

But...what happens when you add in the commercial washer and a nozzle from the kit?

Let's see...

Measurement
First up is inspection and measuring of the nozzles:
You get 5 nozzles to choose from:

3.6mm , 4.4mm, 5.5mm, 6.3mm, Quad 7.3mm x 2.8mm split
http://img99.imageshack.us/img99/926...nozzle1gp6.jpg

Testing Sample (3.6mm pegs manometer)
Here is a quick sample of in action testing of the 3.6mm nozzle. Extremly restrictive!!. I can barely even get over 1 GPM and max out my manometer. My testing equipement consists of a Dwyer Series 475 Mark III model with a pressure range of 0.0-199.9 IN WC. Accuracy is .5 to 1.5% depending on temperature. My flow meter is a King Instruments Acrylic 7520 series (250mm scale) with stainless parts 14" length .5-5.0 GPM. I can read it fairly well with the large size down to .01GPM and it has an accuracy of 2% of full scale.
http://img99.imageshack.us/img99/360...nozzle2rf7.jpg

And now onto each nozzle with washer results:
http://img99.imageshack.us/img99/856...nozzle3mr1.jpg

http://img99.imageshack.us/img99/833...nozzle4mx7.jpg

http://img99.imageshack.us/img99/711...nozzle5rw7.jpg

http://img99.imageshack.us/img99/100...nozzle6se5.jpg

http://img99.imageshack.us/img99/226...nozzle7kw5.jpg

And the summary chart showing each nozzle curve as well as the stock plus washer, stock without washer, and D-Tek's own published curve from the D-Tek site. The only thing I'd like to note is my Fuzion was previously bowed with an o-ring, so the dark blue (no washer) curve is possibly going to be lower than a straight from the box fuzion because the bowing likely has left the gap between the top and middle chambers a little larger allowing more water to bypass the block. The remainder of the curves should be representative of what you might expect.

http://img99.imageshack.us/img99/303...summaryto3.jpg

Summary
- The nozzles with washer are more restrictive than expected. The 3.6mm will put any pump to a crawl (D5 and DDC will be lucky to see 1GPM), up to the Quad nozzle which only has moderate losses. Overall though a good range of nozzles to choose from.
- There is a suprising amount of water that bypasses the block without a washer. I would recommend anyone with a fuzion to at least run a washer in place. It won't bow the block and is extremely easy to install. All of the nozzles are very easy to install and remove as well.
- Pressure drop (as expected) is relatively proportional to the nozzle size with the Quad nozzle being the least restrictive and 3.6mm nozzle bing the most restrictive.
Recommendation

CPU only loop - Install the most restrictive nozzle that gives you the best possible temps. High powered pumps like the Iwakis may favor the 3.6 or 4.4mm nozzle. More mainstream pumps may favor one of the slightly larger nozzles. It may turn out the 4.4mm nozzle is best for Iwaki/CPU setups, and the 5.5 is best for D5/DDC CPU only setups.

CPU & GPU or Multiblock Loops - You need to prioritize where you want to cool more, the CPU, the other blocks, or a balanced combination. Putting a nozzle in place will likely improve CPU temps, but it will increase temps on your other blocks and reduce efficiency of your radiator. Everyone will have a different priority on what they want to cool more. You may just want to use the washer by itself without a nozzle.

Just be aware of this substantial pressure drop increase with these nozzles along with the washer. It doesn't mean you shouldn't use them, just be aware how restrictive they are....probably best to experiement a little and see what happens. I would recommend a washer as a bare minimum, it really should be included as part of the boxed fuzion IMHO.

Hope this gives some perspective on what the nozzles do to flow rate.

I'll plan on adding these equations into my flow rate estimator so you can experiement with it...:D

Martin
12-07-2007, 11:40 AM
Marcng

hey martin, thanks for the info there!!
also, great photography too.

just for aesthetics' sake, tint the images behind the graphs, they will look better and the graphs will be more clear.
12-07-2007, 11:43 AM
Martinm210

Quote:

Originally Posted by Marcng

hey martin, thanks for the info there!!

just for aesthetics' sake, tint the images behind the graphs, they will look better and more clear.

Sure, just having fun with my new testing toys..:D

Yeah, I was fooling around in excel, your right, I should tint the images a touch.:up:
12-07-2007, 12:20 PM
nikhsub1

Good info Martin! I run the Dual DDC's with alpha tops (18w) in series with the 4.5mm nozzle, MCW60, MCW30, PA 160, PA 120.3 and my res. Flow rates seem to still be 'good'. At a guess I'd have to say ~1.5 - 1.6 GPM or so.
12-07-2007, 12:37 PM
Darkatom

I'm amazed at the difference a washer alone will make.
12-07-2007, 12:51 PM
Xilikon

Great results, it mean that I should ditch the 4.5mm nozzle and get a bigger one or the quad nozzle since I have just a DDC3.2 with Petra's top. With the 4.5mm nozzle, the water is barely moving in the microreservoir.

Anyway, I'm running the 4.5mm nozzle just as a test to see if this one work good on a quad core versus the quad nozzle, as promised. I have almost a week worth of CoreTemp log and i'll work the min/max/average per core.

About the washer, it seems that it's mandatory to have good performances so D-Tek should include it with all the Fuzion cpu blocks to fix a design issue.
12-07-2007, 12:54 PM
toejam

That seems weird they would include that washer as part of the package.
12-07-2007, 01:01 PM
Vargher Warg

Great test! :clap:

Thanks for your effort and the test results, and I'm looking forward to see it implemented in your flow rate estimator. :up:

So waddaya reckon? I'm going to have a cpu+gpu loop with Fuzion and
MCW60 - driven by a DDC 10w pump w/Alphacool top, cooling my B3 Q6600 & Ultra 8800

So should I at least use the washer that comes with the nozzle kit?
I was thinking to use the Quad nozzle, but that is maybe to much for my weak pump?

Sorry if I sound too noobish, I'm totally new to the watercooling stuff!

Thanks!
12-07-2007, 01:41 PM
AndrewZorn

Quote:

Originally Posted by Vargher Warg

Great test! :clap:

Thanks for your effort and the test results, and I'm looking forward to see it implemented in your flow rate estimator. :up:

So waddaya reckon? I'm going to have a cpu+gpu loop with Fuzion and
MCW60 - driven by a DDC 10w pump w/Alphacool top, cooling my B3 Q6600 & Ultra 8800

So should I at least use the washer that comes with the nozzle kit?
I was thinking to use the Quad nozzle, but that is maybe to much for my weak pump?

Sorry if I sound too noobish, I'm totally new to the watercooling stuff!

Thanks!

I was wondering almost the same thing, I'll be using a 655 pump and a 120x3 radiator myself. I don't want to sacrifice my GPU temp by like 10*C or anything just to get another 1*C lower on the CPU... what's a guessing tradeoff, using the Quad nozzle?

What I don't get is the washer... it comes with the nozzle kit? I don't see it in pictures or anything, I don't know what it is or where it goes.

D-Tek should be including these nozzles and washers with the block. $7+shipping for small molded plastic pieces... if I can be convinced that a CPU+GPU loop doesn't want the nozzled, I'd be so happy to not pay for all that extra crap.

EDIT: and wow, this is already #1 Google result for "fuzion nozzles"...
12-07-2007, 01:51 PM
Petra

Just ran a quick, dirty test with the retail 4.4mm nozzle and washer in place and, yeah, you're right, the pressure drop with the retail nozzles and that washer in place is a lot higher than the prototype nozzles without washer. Looking good, Martin :up:

Interestingly, though, I'm seeing about 6.2psig at 1.5GPM on my analog gauges and my manometer is rather useless right now because the battery is dead. *sigh* Nothing ever works right when I need it to... :rolleyes: ;)

Oh well, I'm gonna be ripping the whole thing apart and changing a bunch of stuff again anyway...
12-07-2007, 02:11 PM
Xilikon

AndrewZorn, the washer is included in every nozzles kit. Even if you buy it just for the washer, it's a worthwhile investment for sure.
12-07-2007, 02:59 PM
MorGo7h

Thats some nice work over there!
I cant get the nozzles here in Belgium :( And if I order them from the UK I have to pay more shipping costs than the nozzles cost :p
12-07-2007, 03:17 PM
Talonman

Outstanding report Martin!

Thanks for the post. :up:
12-07-2007, 03:56 PM
Martinm210

Thanks all, I still have some work to do on my set rig and calibration checks to make sure, but you can at least compare the nozzles to the washer only curve for a relative difference..:up:

Quote:

Originally Posted by nikhsub1

Good info Martin! I run the Dual DDC's with alpha tops (18w) in series with the 4.5mm nozzle, MCW60, MCW30, PA 160, PA 120.3 and my res. Flow rates seem to still be 'good'. At a guess I'd have to say ~1.5 - 1.6 GPM or so.

If those are DDC2's I wouldn't be suprised, they are great pressure pumps and running two in series would be ideal for a nice tight nozzle.

I just got my spare D5 from Petra's yesterday, and I also have a spare heatercore sitting here. So it would probably help to run one more comparison of actual flow rate using the D5, just to get that feel for it..

Quote:

Originally Posted by Xilikon

Great results, it mean that I should ditch the 4.5mm nozzle and get a bigger one or the quad nozzle since I have just a DDC3.2 with Petra's top. With the 4.5mm nozzle, the water is barely moving in the microreservoir.

Anyway, I'm running the 4.5mm nozzle just as a test to see if this one work good on a quad core versus the quad nozzle, as promised. I have almost a week worth of CoreTemp log and i'll work the min/max/average per core.

About the washer, it seems that it's mandatory to have good performances so D-Tek should include it with all the Fuzion cpu blocks to fix a design issue.

I don't think you should necessarily ditch it, you may get better temps with it...it's just that you should be aware of what the reduced flow rates may do to other blocks. More than anything I would just recommend that people try a couple of different ones and see what works best. They are very restrictive so there is going to be a perfect one that is likely different for several different setups.

Yeah, I think at least the washer should come with the fuzion. My washer only curve seems to match D-Tek's posted curve, so it has me wondering if that's what they tested with. Either that or there could be added drop from the test rig that created it.

Quote:

Originally Posted by Vargher Warg

Great test! :clap:

Thanks for your effort and the test results, and I'm looking forward to see it implemented in your flow rate estimator. :up:

So waddaya reckon? I'm going to have a cpu+gpu loop with Fuzion and
MCW60 - driven by a DDC 10w pump w/Alphacool top, cooling my B3 Q6600 & Ultra 8800

So should I at least use the washer that comes with the nozzle kit?
I was thinking to use the Quad nozzle, but that is maybe to much for my weak pump?

Sorry if I sound too noobish, I'm totally new to the watercooling stuff!

Thanks!

Yeah, I'd probably start with just the washer, maybe try a couple of the other larger nozzles too:up:

Quote:

Originally Posted by Petra

Just ran a quick, dirty test with the retail 4.4mm nozzle and washer in place and, yeah, you're right, the pressure drop with the retail nozzles and that washer in place is a lot higher than the prototype nozzles without washer. Looking good, Martin :up:

Interestingly, though, I'm seeing about 6.2psig at 1.5GPM on my analog gauges and my manometer is rather useless right now because the battery is dead. *sigh* Nothing ever works right when I need it to... :rolleyes: ;)

The other thing I never accounted for in the estimator either is the added viscosity with something like a more typical 5% ethelene glycol mixture.

Oh well, I'm gonna be ripping the whole thing apart and changing a bunch of stuff again anyway...

Thanks Alex,

I appreciate the quick test, that gives me something solid to compare against. Your results are a little lower than what I got but relatively close at least within one nozzle deviation...still closer to the 4.4 than the 5.5 results I ran.

I've been curious if water temperature and visocity changes would have any influence on a test run. I've been using cold tap water for my testing which is probably close to 10C (snow on the ground at the moment). I should make a quick run with some temperature variations just to see if it does make any difference at all. It would be pretty easy to splice in another T with a temperature probe for information. I know my manometer specs notes accuracy is only within 1.5% on the colder side, and .5% when more room temperature.:shrug: Worth a test run anyway...:up:
12-07-2007, 04:13 PM
madmaxx

Martinm210 again thnx for your time and effort that you put into getting some answers and definitive #'s for us to go buy. your info will be well used.

i applaud you :clap: :clap: :clap:
12-07-2007, 05:09 PM
NaeKuh

dayam another great review martin...
12-07-2007, 05:35 PM
disruptfam

awesome martin has anyone tested the 4.5 and the quad nozzle on a quad??

which is better?
12-07-2007, 06:02 PM
eva2000

Great stuff... you had me scared there first few graph background photos i saw, i thought I had my nozzle installed the wrong way round.. until i saw the last photo background graph hehe
12-07-2007, 06:27 PM
Xilikon

Quote:

Originally Posted by disruptfam

awesome martin has anyone tested the 4.5 and the quad nozzle on a quad??

which is better?

I did... Matter of fact, I just swapped the nozzles 2 hours ago (took a bit of time since the damn T line made the filling and bleeding a big mess !!! I cannot wait to receive a new microres back from NCIX). Right now, it seems the quad nozzle gave me 2-3 degrees better versus the 4.5mm nozzle along with less variations between the cores. I'll know better when I will let it run another week so we have enough data to compare both because the bleeding isn't 100% done.

IMHO, on a quad core, either a 6.5mm nozzle or a quad nozzle would be a better bet since the 4.5 mm nozzle will point to a small area which fall between 2 cores so the cooling area is not enough.

My testing methodology isn't very rigorous (I cannot bear to do remounts and nozzles changes with a T line a limited time) but I like the temps (average 35,36,31,33 on all cores vs 37-36-32-34 with 4.5mm).
12-07-2007, 06:40 PM
AAjax

Op, thanks for the post :D
12-07-2007, 06:44 PM
leo_bsb

great review martin! Thanks for your time.
We just need that the greatest tool of planning WC, the excel sheet, to be updated with your data. ;)
12-07-2007, 07:39 PM
SiGfever

Nice test. I have a Dwyer manometer and was not aware that it was rated for water, did you check with the factory about its use?

Quote:
"SPECIFICATIONS
Service: Air and compatible combustible gases.
Wetted Materials: Consult factory."
12-07-2007, 08:54 PM
septim

very informative as usual. thanks...
12-07-2007, 09:38 PM
Martinm210

Quote:

Originally Posted by SiGfever

Nice test. I have a Dwyer manometer and was not aware that it was rated for water, did you check with the factory about its use?

Quote:
"SPECIFICATIONS
Service: Air and compatible combustible gases.
Wetted Materials: Consult factory."

Yeah, but you just need enough length of tubing between your fittings and manometer. A water column will raise up on the posotive side, but it just compresses the air that's trapped in between. It's a sealed connection, so the water pressure is transfered directly to the airpressure the manometer is measuring. You do have to keep an eye on placement though to ensure no water can get in there.

I'm currently making the T lines rise up a couple of feet then back down to the manometer. It works fine although I'm thinking of going with some 1/4" tubing. The 1/8" is a bit small to be self purging.
12-08-2007, 03:37 AM
jayhall0315

Hey Martinm,
All the testing seems very sound. Sorry I have been too busy to add more data from my own testing to back you up/help you out. Keep up the great work. I do not know if you have much of a background in the sciences but you should look into some of the better texts available on heat flow, hydrodynamics and thermodynamics...might generate even more crazy ideas for you to work with.

Cheers to your hard work,
Jay
12-08-2007, 08:35 AM
Martinm210

Quote:

Originally Posted by jayhall0315

Hey Martinm,
All the testing seems very sound. Sorry I have been too busy to add more data from my own testing to back you up/help you out. Keep up the great work. I do not know if you have much of a background in the sciences but you should look into some of the better texts available on heat flow, hydrodynamics and thermodynamics...might generate even more crazy ideas for you to work with.

Cheers to your hard work,
Jay

Thanks Jay,

I'm a civil engineer so I do have background in the hydraulics side...:up:

I also spent many years growing up in a machine shop, so that's helped alot on the fabrication/machining side.

I've had a taste of thermo from the college days, but it's been more hobby and self taught than anything. I'm always looking to learn new areas..:D
12-08-2007, 08:49 AM
shabranigdo

Nice I hope you get sent one of the new EK blocks to test, would be very interesting to see these kind of test performed on it also to see how it performs. The new EK block looks really good, but also the Fusion seems hard to beat. Decisions decisions.
12-08-2007, 09:09 AM
Martinm210

Quote:

Originally Posted by shabranigdo

Nice I hope you get sent one of the new EK blocks to test, would be very interesting to see these kind of test performed on it also to see how it performs. The new EK block looks really good, but also the Fusion seems hard to beat. Decisions decisions.

I've asked, and would really be interested in doing just that:up:

Eddy said he was waiting until they had the high flow and low flow(Extreme Performance) nozzle plates done first.
12-08-2007, 09:25 AM
Sideroxylon

Great work and info Martin...so please don't take this the wrong way. Just wondering why your putting so much time and effort into these flow tests? I've been wc'ing now for several years and have noticed that unless I turn my pump down to almost no flow that my temps stay pretty much the same throughout the different flow ranges.

I have a 50Z-DC12 pump that I control with a miniNG. The pump will run all the way down to 75% power before stalling out. My temps don't start rising until I get to the 75%-80% power range. I have no way of measuring flow besides visually checking flow in my res. At 75% power I can barely see the flow in my res. I run my pump between 80%-85% normally because I get no temp difference between there and 100% power. There is a huge difference in flow between 85%-100%. From casually flowing through the res at 85% to extremely turbulent and noisy at 100%. So, I've determined that flow is important but not nearly as much as many people make it out to be.

So, again, I don't mean to take away from all the work you've done and info you've provided, but I wonder how useful is the information to the average user without correlating flow with temps in some way?

**EDIT** I've just gone back and read your original post on your flow estimator. It makes sense that there are too many variables between all the different components people use to give rock solid temps concerning flow but I would still be very interested in seeing your temps as you tested the different nozzles. When it comes down to it I think we all just want to know if changing nozzles will lower temps or not.
12-08-2007, 02:01 PM
Martinm210

Quote:

Originally Posted by Sideroxylon

Great work and info Martin...so please don't take this the wrong way. Just wondering why your putting so much time and effort into these flow tests? I've been wc'ing now for several years and have noticed that unless I turn my pump down to almost no flow that my temps stay pretty much the same throughout the different flow ranges.

I have a 50Z-DC12 pump that I control with a miniNG. The pump will run all the way down to 75% power before stalling out. My temps don't start rising until I get to the 75%-80% power range. I have no way of measuring flow besides visually checking flow in my res. At 75% power I can barely see the flow in my res. I run my pump between 80%-85% normally because I get no temp difference between there and 100% power. There is a huge difference in flow between 85%-100%. From casually flowing through the res at 85% to extremely turbulent and noisy at 100%. So, I've determined that flow is important but not nearly as much as many people make it out to be.

So, again, I don't mean to take away from all the work you've done and info you've provided, but I wonder how useful is the information to the average user without correlating flow with temps in some way?

**EDIT** I've just gone back and read your original post on your flow estimator. It makes sense that there are too many variables between all the different components people use to give rock solid temps concerning flow but I would still be very interested in seeing your temps as you tested the different nozzles. When it comes down to it I think we all just want to know if changing nozzles will lower temps or not.

No problem, after all my testing and flow rate experiments I've also become very aware at how resiliant alot of blocks these days are to lower flow rates, particularly the higher restriction (mini pin)types. But that's not for all blocks, alot of the large channel or smooth based blocks like the MCW-30 and Stinger V8 really shine at higher flow rates.

The differences we're talking about are small for most blocks...usually only a few degrees when you start talking temperature differences of flow rate as long as you keep it above 1 GPM.

But what's hard to determine is "How much temperture difference is important to any one person". Some people will buy another water block over 2C, and some people could care less about 5C.

I spend the time on it mostly because it interests me. I've done alot of hydraulics type stuff, and it's an area that the user has the most control over. You can only do so much with a water block, but there are unlimited ways you can setup the system, and that's probably why I like water cooling so much. The "Setup" is what makes it custom for most people.

D-Tek did a really nice job in providing a wide range of options with these nozzles. More than anything I just wanted to have a better understanding of what they do to flow rate. To complete the testing I would need to do the same on thermal testing.

Most of the testing of these nozzles has been with more powerful pumps, so with these results I just wanted to encourage users to try a couple of different nozzles. For more mainstream pumps, the 5.5 may be the better option over the 4.4mm nozzle.

Not really a big discovery here, but it's something that's a bit more quantified than a subjective "It flows well, or better than" type of run. I was under the impression the nozzles still flow really well, but once I put one in, I had a sneaking suspicion that's not the case. Testing them confirms they are much more restrictive. And by more restrictive, the smaller nozzles on a Fuzion are more restrictive than any block I currently had in the estimator, that's all I wanted to point out.:up:

I've also been experimenting a bit with making my own blocks. Thermal performance is one side and probably the most important, but pressure drop is the other side that should be noted particularly in multiple block loops. Waterblock design is a careful balance between thermal performance and restriction, and the differences are getting so tight these days, I think it's important to keep track of both sides to some degree.

I guess I just like testing and tinkering too...part of the hobby I suppose..:D
12-08-2007, 06:18 PM
Sideroxylon

I guess my question really stems from the small amount of frustration at not being able to really use all the juicy info you've given us. If only somehow your info could be merged with some temp data it would really be the best WC'ing Encyclopedia ever. We gotta get you and Nikshub working together. That would be quite the dynamic duo. :up:

Thanks for the response and all the hard work and info you've given us. It's people like you that help our hobby progress and encourage the rest of us to make better rigs.
12-09-2007, 07:48 PM
Martinm210

FYI

This data has now been integrated into the flow rate estimator. Still alot of work to do, but these seem to work and will give you a better feel for what this means regarding flow rate:

http://www.xtremesystems.org/forums/...&postcount=225

Let me know how it works for you..:up:
12-10-2007, 12:53 AM
Evil-Vincent

Hey, I currently have a stock D-Tek Fuzion on a cpu only loop, and my s939 3800X2 is running at 3.07ghz 24/7 stable, and load temps go to 57C after 9 hours of orthos.

I live in The Netherlands, and I was looking for a shop that sells the washer and the nozzle kit. I did find the nozzle kit at a shop in England, but I cannot find anything about the washer.

What washer is it, where do I mount it and where do I buy it? Thanks.
12-10-2007, 01:39 AM
Prisoner1138

Quote:

Originally Posted by evil-vincent

Hey, I currently have a stock D-Tek Fuzion on a cpu only loop, and my s939 3800X2 is running at 3.07ghz 24/7 stable, and load temps go to 57C after 9 hours of orthos.

I live in The Netherlands, and I was looking for a shop that sells the washer and the nozzle kit. I did find the nozzle kit at a shop in England, but I cannot find anything about the washer.

What washer is it, where do I mount it and where do I buy it? Thanks.

It comes with the nozzle kit.
12-10-2007, 12:23 PM
Martinm210

Yeah, the washer comes with the kit.:up: Here is a picture of it installed. It comes with an adhesive backing, so all you do is peel the backing and stick it on the top of the mid block. This washer seals the gap that exists between the middle and top block.

http://img218.imageshack.us/img218/6...8washerhs8.jpg
12-10-2007, 01:10 PM
Evil-Vincent

Okay thanks a lot. Just one more question: What nozzle do you recommend me to use on my s939 3800x2? I would assume the smallest in a cpu only loop.
12-10-2007, 01:22 PM
Martinm210

Quote:

Originally Posted by evil-vincent

Okay thanks a lot. Just one more question: What nozzle do you recommend me to use on my s939 3800x2? I would assume the smallest in a cpu only loop.

I don't know for sure, I think alot of it depends on the pump used. There will also be a point for everyone where more cooling at the nozzle (smaller nozzles) does less improvement than the loss of flow rate over the entire base. It may also have something to do with pressure drop increasing heat dump. So going to the smallest nozzle is likely not going to be the best choice.

I've heard nikhsub1 and cathar both mention around the 4.4mm nozzle working best, but that was with an iwaki pump. I'm going to start my own thermal testing pretty soon here on a couple of them, but I'm guessing the 5.5mm-6.3 might be a good one for more mainstream pumps. But that's just a guess.

I think our boys at D-Tek know alot about these nozzles, so they probably purposefully made this range available to meet the wide range of needs out there.

If you had absolutly no idea, I'd probably start in the middle. 5.5mm nozzle..but that's just a complete guess on my part.:shrug:
12-10-2007, 02:08 PM
nikhsub1

The 4.37mm nozzle should be 4.5mm, that was the spec that it was supposed to be made at. The smallest nozzle is useless really, it is way too restrictive. When I spoke with Cathar on this issue, he said that he found ~4.7mm nozzle to be pretty ideal, this was after I told him I had great success with the prototype 4.5mm nozzle. I think in a CPU only loop the 4.5mm nozzle is the way to go - I run that nozzle with a GPU and a NB in the loop but I run dual pumps. 5.5 may be OK too, I haven't tested every configuration of these suckers.
12-10-2007, 02:52 PM
Martinm210

Ok, not sure why mine measured smaller, but I even remeasured around in multiple locations I still got about 4.37 +- .03mm so I called it what I found rounding up to the nearest .1mm. I guess measuring with digital calipers is not the most accurate way, using a set of numbered drill bits to check diamter would probalby be better.

Anyhow, the second from the smallest...:D

I'll give a couple of sizes a try and see if my single D5 has a different preference. I would also think quads may favor a slightly larger nozzle as it becomes less important to cool right in the very center and more important to cool the whole base well.
12-10-2007, 03:25 PM
Petra

Quote:

Originally Posted by Martinm210

Ok, not sure why mine measured smaller...*snip*

Mine also measured small... My guess would be that it may have something to do with the shrinkage issues that D-Tek experienced with the first run of the nozzle kits. Basically, they were having issues with the injection molding process resulting in the nozzles coming out too small (e.g. they wouldn't fit snugly in the waterblock) and that ended up being the main reason for their delayed release.

Injection molding processes aren't exactly my specialty... so, just a guess :shrug:
12-10-2007, 04:13 PM
Martinm210

Quote:

Originally Posted by Petra

Mine also measured small... My guess would be that it may have something to do with the shrinkage issues that D-Tek experienced with the first run of the nozzle kits. Basically, they were having issues with the injection molding process resulting in the nozzles coming out too small (e.g. they wouldn't fit snugly in the waterblock) and that ended up being the main reason for their delayed release.

Injection molding processes aren't exactly my specialty... so, just a guess :shrug:

Ahh..that would explain it. Thanks!:up:
01-09-2008, 10:20 AM
avrionis

Do I bow with the nozzle washer

If I use a O-ring to bow the fuzion, do I still need to put the nozzle washer in?
01-09-2008, 06:21 PM
Martinm210

Quote:

Originally Posted by avrionis

If I use a O-ring to bow the fuzion, do I still need to put the nozzle washer in?

No, the o-ring will seal the chamber, only use the washer if you're not bowing with an o-ring.:up:
01-09-2008, 07:03 PM
lowfat

Very nice! I bet my flow is almost nothing then.

50Z + heatercore + Fuzion GFX + Fuzion/4.5mm nozzles = no flow :rofl:
01-09-2008, 07:25 PM
avrionis

Martin! Can I use the D-Tek barb O-ring or is it too small?
01-09-2008, 08:00 PM
silverphoenix

Once I get my fuzion in I'll see which nozzle works best for me the quad or 4.4 It will be on it's own loop. MCR320+res+DDC3.2+Petra top
01-09-2008, 08:08 PM
Martinm210

Quote:

Originally Posted by avrionis

Martin! Can I use the D-Tek barb O-ring or is it too small?

I've heard of people using barb o-rings and it working. You can try it, but I would recommend picking the thinnest o-ring you can find. I think D-teks are a little thicker, you want one that is 1/16" thickness or thinner.

I bought a larger ring size so I had a little more room for error, 1/2" ID, 5/8" OD, x 1/16" thickness.
01-24-2008, 06:04 AM
Prodigys

Hey Martin, i saw your flowrate estimator a while back ago, but never this one.

I have a PA120.2 with Fusion and a DDC Ultra pump, i'm now going on a Q6600 but what should i use in your opinion ?

I had my E6750 on 4Ghz with 1.53Vcore and that gave me 48°C 100% stressed. That was just Fusion stock.

And what do you mean with a washer and where does it fit then ?
Maybe a picture is more clearly to explain for me i think :p: .

Thanks and keep up the good work :up: !
01-24-2008, 06:12 AM
Xilikon

With a quad, get a quad nozzle, no question. As for the washer, it come with the kit and it should go around the center hole (the one between the middle and the top). This is to seal it from leaking in the middle chamber and help focus the flow on the base plate.
01-24-2008, 06:56 AM
Prodigys

Quote:

Originally Posted by Xilikon

With a quad, get a quad nozzle, no question. As for the washer, it come with the kit and it should go around the center hole (the one between the middle and the top). This is to seal it from leaking in the middle chamber and help focus the flow on the base plate.

I saw yes, just tested myself. Block was apart, i blowed on the intake and the air escaped trough the outtake even when i covered the back of the center whole, this i had mentioned earlier but not given any attention.

Now i found a capsule in my workshop and made it so that it sealed it, and the base was not bowed.

But the nozzle's, cant get them anywhere in the country here... Will search for a home made thingy, thanks anyway :) .
01-24-2008, 07:04 AM
Xilikon

Quote:

Originally Posted by Prodigys

I saw yes, just tested myself. Block was apart, i blowed on the intake and the air escaped trough the outtake even when i covered the back of the center whole, this i had mentioned earlier but not given any attention.

Now i found a capsule in my workshop and made it so that it sealed it, and the base was not bowed.

But the nozzle's, cant get them anywhere in the country here... Will search for a home made thingy, thanks anyway :) .

the DIY way would be to get a O-ring of the right thickness (around 1 - 1.3 mm) and of the right diameter then place it in the block (you may have to use a dab of glue to keep it in place). The O rings can be found in any hardware or plumbing stores ;)
02-10-2008, 11:28 AM
DOM.

so which one should I use with this

D-TEK FuZion CPU Waterblock - 1/2" with D-Tek FuZion Accelerator Nozzle Kit

D-Tek FuZion GFX GPU Waterblock

2X Swiftech MCW30 Chipset Water-block

Danger Den Bay Reservoir High-Density Polyethylene

PrimoChill Radiator Xtreme Series 360

Swiftech MCP655™ 12 VDC Pump

Tygon 3603 Tubing 1/2" ID 3/4" OD

loop (Res>Pump>Rad>CPU>GPU>NB>SB>Res)
02-10-2008, 12:16 PM
Conumdrum

Are you talking the nozzle? I have a E6600, kinda same setup. Reading through tons of old posts about this I saw Martin said a 5.5 with regular washer (comes in kit) for an E6600. Your milage may vary, but start with a 5.5 or none, if ya want..
02-10-2008, 12:30 PM
Martinm210

Sorry I havn't completed any thermal testing yet, starting over now so it'll be a while.

I'd probably just run a washer only or maybe a 6.3 on something that restrictive. But in the end, heat load should be your bigger concern and adding another radiator or stronger fans would probably make a bigger difference.
02-10-2008, 05:59 PM
DOM.

Quote:

Originally Posted by Conumdrum

Are you talking the nozzle? I have a E6600, kinda same setup. Reading through tons of old posts about this I saw Martin said a 5.5 with regular washer (comes in kit) for an E6600. Your milage may vary, but start with a 5.5 or none, if ya want..

well I have the 4.5 in there and it didnt really change the temps much from out of the box maybe 2-3c at most, just saw this thread and was thinking if I should change it see if it helps might do it tomorrow, do you got a pic of your loop ? 20-37 is that the cpu temp ? and what v-core are you using and room temp :D

Quote:

Originally Posted by Martinm210

Sorry I havn't completed any thermal testing yet, starting over now so it'll be a while.

I'd probably just run a washer only or maybe a 6.3 on something that restrictive. But in the end, heat load should be your bigger concern and adding another radiator or stronger fans would probably make a bigger difference.

thanks for your time nice test, but I already have 6X Thermaltake Thunderblade 120mm Blue LED Fans on the rad, idk what to do now :p:
02-10-2008, 06:19 PM
Martinm210

Hang another 360 radiator on your case somewhere...:)
Should be good for a few degrees anyhow.

I like lots of radiator and I have 2 MCR320's and one MCR220 in/on my armor...
02-10-2008, 06:23 PM
DOM.

Quote:

Originally Posted by Martinm210

Hang another 360 radiator on your case somewhere...:)
Should be good for a few degrees anyhow.

I like lots of radiator and I have 2 MCR320's and one MCR220 in/on my armor...

well I do have another PrimoChill Radiator Xtreme Series 360 but it was bent from usps and they didnt sent the postage to send it back so I kept it but where can I hang a 360 rad lol do you got a pic of your rig ?

http://i199.photobucket.com/albums/a...Picture001.jpg
02-10-2008, 06:27 PM
Martinm210

This is an old picture, but you'll get the idea on the mcr320's. The MCR220 is now in the front of the case.

When there is a will there is a way...:D

http://www.overclock.net/gallery/data/500/IMG_9462.JPG

Newer picture of the three loops:
http://img408.imageshack.us/img408/3...08emptyjh7.jpg
02-10-2008, 06:29 PM
DOM.

yeah I seen that with the rads but got my lil girl and im in her room she only 6days old and I know when she gets bigger she might rip one off lol

wow I though mine was bad before I canged my case lol lots of tubing
02-10-2008, 06:32 PM
Xilikon

Nice tubing forest in that case :p:

Nice idea on the radiators, made me think of wings or propulsors off Star Wars :)
02-16-2008, 04:34 AM
Leeghoofd

Quote:

Originally Posted by Prodigys

But the nozzle's, cant get them anywhere in the country here... Will search for a home made thingy, thanks anyway :) .

http://www.aquatuning.de

They have the nozzles mate http://www.aquatuning.de/product_inf...ozzle-Kit.html

I just ordered them myself :up: I ordered my whole Wc kits from them, pretty good and fast online shop...
02-20-2008, 07:53 AM
TheThreeDegrees

just want to say thanks for the review on the accelerator kit.
very informative.

nice setup in the case also BTW:cool:
02-27-2008, 07:03 PM
gmcg

Thanks, Martin, it is an extremely useful information in this post.
Also thanks for revealing the leakage issue on a stock D-Tek fuzion without o-ring, it helped a lot! :up:
03-05-2008, 09:19 AM
tastyratz

holy damn, I thought I was bad with my tubing, thats a crazy forest.
You have a lot of T's in there, ever thought of using Y splitters instead?
I bet going through that mess and swapping them out would net you a little.
frozen cpu sells Y splitters pretty cheap
03-05-2008, 06:10 PM
Martinm210

Quote:

Originally Posted by tastyratz

holy damn, I thought I was bad with my tubing, thats a crazy forest.
You have a lot of T's in there, ever thought of using Y splitters instead?
I bet going through that mess and swapping them out would net you a little.
frozen cpu sells Y splitters pretty cheap

These are straight through 3/4" brass Tees. I tested 5/8" copper Tees and restriction is so low, I pretty much couldn't barely even measure it with a manometer good to .1" of pressure resolution. We're talking a coulple of inches in pressure drop at peak flow rates so for all practical purposes you can consider a 3/4" T restrictionless.

3/4" Q-Pex Brass T's and Elbows are the least restrictive commercial fitting you can buy after copper. The brass elbows have a nicely machined radius internally, not quite as good as copper, but the next best thing. I got tired of making copper fittings, so I figured I'd try these brass ones out and polish them up nice.

But yeah, three loops means tubing management. I thought I was doing pretty good with the wire managment hiding everthing behind the MB....then I had to hook up the tubing.

Not done yet though, eventually I'll either buy or make some mosfet blocks to complete my forest...:D
03-05-2008, 07:05 PM
DavidNJ

Quote:

Originally Posted by Dom7184

well I do have another PrimoChill Radiator Xtreme Series 360 but it was bent from usps and they didnt sent the postage to send it back so I kept it but where can I hang a 360 rad lol do you got a pic of your rig ?

http://i199.photobucket.com/albums/a...Picture001.jpg

Question: Are you going pump->southbridge->northbridge->vga->cpu->radiator->pump? With the cpu last in the loop?
03-05-2008, 07:11 PM
taowulf

The output of his pump goes to the 3x120 in the back, then to the CPU and so on. The pump is mounted with the inlet on the top, rather than the out being on top if was mounted horizontally.
03-05-2008, 08:24 PM
disruptfam

looks like the cpu is 1st to me ????
03-06-2008, 07:57 AM
DOM.

Quote:

Originally Posted by DavidNJ

Question: Are you going pump->southbridge->northbridge->vga->cpu->radiator->pump? With the cpu last in the loop?

res>pump>rad>cpu>gpu>nb>sb>res

Quote:

Originally Posted by taowulf

The output of his pump goes to the 3x120 in the back, then to the CPU and so on. The pump is mounted with the inlet on the top, rather than the out being on top if was mounted horizontally.

Quote:

Originally Posted by disruptfam

looks like the cpu is 1st to me ????

:up: you guys where right

new pic of loop same order

http://i199.photobucket.com/albums/a...rldSerriah.jpg
11-22-2008, 08:34 PM
Gir92

Just out of curiosity, can someone link me to the washer needed to do the washer mod and where to install it please?

Thank you.

EDIT: NVM, I just saw the post saying it comes with the kit.