Printable View
Thanks Jay,Quote:
Originally Posted by jayhall0315
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
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:Quote:
Originally Posted by shabranigdo
Eddy said he was waiting until they had the high flow and low flow(Extreme Performance) nozzle plates done first.
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.Quote:
Originally Posted by Sideroxylon
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
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.
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:
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.Quote:
Originally Posted by evil-vincent
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
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.Quote:
Originally Posted by evil-vincent
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:
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.
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.
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.Quote:
Originally Posted by Martinm210
Injection molding processes aren't exactly my specialty... so, just a guess :shrug:
Ahh..that would explain it. Thanks!:up:Quote:
Originally Posted by Petra
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:Quote:
Originally Posted by avrionis
Very nice! I bet my flow is almost nothing then.
50Z + heatercore + Fuzion GFX + Fuzion/4.5mm nozzles = no flow :rofl:
Martin! Can I use the D-Tek barb O-ring or is it too small?
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
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.Quote:
Originally Posted by avrionis
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.
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: !
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.