27 fpi is even better. You'll have to re-measure the fin and channel width ;)

What's the fin height? What parts do (did) you plan on cutting?

.
i have cut the width to fit my p4 hold down thats all.
i dont have anything more accurate to measure the fins but they are a little thicker than paper.

fins are 7mm before i lower them.

i intend to make another block the same as my last. 30x30 fin area 4mm high to cover a p4 ihs, but i am sure if i was to try and cut between the fins into the base i would loose most of the fins in the process. so it looks like a long day on the belt sander to get the base down to 1.5mm

i will also use the same squashed copper pipe for an inlet.

so here we have a block using the thermaltake base.

i have left the base at 4mm for the moment and lowerd the fins to 3mm. After testing the inlet jet i noticed that the water was not touching the top of the fins so took 1mm of them.

http://homepage.ntlworld.com/lee_smith/tt3.jpg
http://homepage.ntlworld.com/lee_smith/ttblock1.jpg

once the silicon sets i will do some tests. Then i plan take 1mm off the base and test again. I will continue this until i get to 1.5mm.

I also ran some water flow tests with this base and with an akasa base.

i have a 900l/h aquarium pump i use for testing which took 30 seconds to fill a 5l container This worked out to 600 l/h with the pump and 1/2 tube.

The akasa base that took 45 seconds or 396 l/h

the new thermaltake base was 45 seconds too.

so the extra fins dont seem to slow the flow anymore.

If the inlet is simular to the one on the block you sent me, I think it could do with being much thinner, 1~1.5mm to start then try upping it 0.25~0.5mm a step to find your optimum. As it is it's only 'spreading' out the flow, not 'jetting' it, which is probly why it perfoms on par with the maze3, you're not actualy getting any impingment...

Ask Cathar how wide the channel/jet was/is on the WW. It'd have to be a bit thinner on yours to make up for the extra width...

PS, I was'nt joking about buying the Maze3 if you're interested?. Half what I paid for it 5 months ago(£45) and the block you sent me would be fine, or better still some 5mm copper bar (I was planning on cutting the maze up to make another block but it's only got a 3mm base plate :D )...

hi maddogme,
the jet is 2mm at the base of the inlet. On the one you have it was only squashed to the width of the fin area.

WRT the maze 3 it depends on when i am working again as i only have so much money thats got to last me for as long as possible.
If it's no too long they i will take it.

i want some copper too so i can try and make a copper top for one of these.So far i am having no luck.

had anyone tried a 1mm copper sheet with a multi jet inlet ?

i think the reason i am not getting impingement is the inlet is a smooth transistion from round to thin jet. This will increase velocity and widen the inlet to cover the whole p4 IHS but as it's smooth the only turbulance created is when it hits the base plate. The thin fins i have dont help either.

If you look at WW the inlet is round down to the middle plate and there is a small gap above the middle jet plate. So the water will hit most of the middle plate creating turbulance. Also the fins are square on the top so again the water will hit the top of the fins and turbulance will be created.

At least after doing some research yesterday thats what i think is happening.

Would anyone have any thoughts on this.

Lee

It's kind of hard to tell, but maybe your nozzle might need some tweaking.

In WW, it's a small slit, about 3 mm wise.

The only things I can see that would hinder performance are:
-baseplate thickness
-Flowrate

What pump are you using, and what kind of flow rate are you getting with it?

Your nozzle appears to be working fine otherwise.

thanks bb2k,

here is the nozzel i am using

http://homepage.ntlworld.com/lee_smith/v3-1.jpg
http://homepage.ntlworld.com/lee_smith/v3-4.jpg
http://homepage.ntlworld.com/lee_smith/nozzel.jpg

at the nozzel end it's a 2mm slit. The main reason for this is to cover the width of the fin base.

i agree with the base plate thickness as my best block so far is the one i took the base down to 2.5mm. this is 1C lower than a maze 3 and also a block i made with a base plate thickness of 3.5mm.

in my test stup i have a aquabee 1000 l/h with 1.5m head pressure.

http://homepage.ntlworld.com/lee_smith/rig.jpg

in my actual pc i have a ehiem 1048 which i dont think will be able to cope with the impingement style blocks.

Your pump is good. Did you see the deal on the Johnson pump ;)

I think that the design suffers from the cut in the middle of the fins. The fins are most effective right over the core area.

To the contrary, in actual use the Eheim 1048 is quite a decent little performer with an impingement style block. I can barely measure the CPU temperature difference between the 1048 and the 1250 when just using the pump, White Water, and a radiator.

Thats why i wanted the thernaltake base . With the 27 fin per inch and no gap in the middle i expect it to perform a lot better.

On my first test with the base at 4mm it is working the same as the maze 3. So once i get the base thinner i shoule be looking at 2-3C better.


Cathar / ben thanks for the info on the pump. I knew the 1048 was good when i bought it.

what do you think about this.

it's made from scrap acrylic so no comments and how lit looks :)

i have a 1mm copper sheet for a base.

http://homepage.ntlworld.com/lee_smith/pipes1.jpg
http://homepage.ntlworld.com/lee_smith/pipes2.jpg

it's based ben's idea with the copper tubes that he had at the very beginning on the new design idea thread.

Yes, I remember.

The design also included some "cups" drilled into the base, to maximize the inpingement effect, both against the baseplate, and the cup wall.

When I get my endmills, I'll make it. A drill bit would leave a small angle, or cone, which will be OK.

WOW
 

Lee, you are a busy individual! I also need to check this thread much more often.

Congrats on beating the Maze by a full C with this latest design.

As you seem to creat blocks at a rather impressive pace I'm going to set aside the item I was about to send you in favor of something even better I just got.

Check your PM's.

BE

wicked BE, your a star. how thick is is 1/2 inch again ?

well as i am no longer working i have loads of time to make load of mess and like to experiment and make stuff.

this is my list of blocks in order of performance.

all tested with a 1.7 p4 with room temps 20-21 most at 20.5. each time i would mount the block 3 - 4 times.

1. 2.5mm akasa base with 4mm fins 30C load.
2. 4mm thermaltake base with 3mm fins 31 load
3. maze 3 31.0 load
4. 3.5mm akasa base with 2.5mm dimples /tubes 31.5 load
5. akasa 3.5mm base with 4mm fins 31.5C
6. 1mm base with many copper tube inlet 35 load as the base is not flat.

i used sandra burn in test as it records temps into a file and have the actual results saved if anyone wants to see them.

i think the thermaltake with a very thing base will be the best out of this lot.
i have tested some of the blocks on amd but due to a leak thats now not working too well but i found simular results.

I think that the design with the little copper inlets is going to have a major flaw. The pressure of the water is going to come in, and is going to have higher pressure pushing down the first half or so of the inlets, and less pressure on the end. I think it may not be getting enough water to the far side of the inlets. If you had your main inlet right on top, you'd probably have better success. The only problem with it directly on top, is that most of the pressure will be going through the center holes, and not the outside ones.

With your current copper tubes, I think that you could get better results if you used smaller tubes at the beginning, where the main inlet comes in, and bigger ones out on the end. It would force equal pressure across the plate. If you decide to go on the top down approach, you should use smaller diameter inlets in the center, and bigger ones on the outside. That way pressure goes across the entire plate and gives equal pressure to each inlet.

But you don't have to listen to me. Your designs are yielding quite impressive results. Good job!

Actually, given the huge pressure drop across the tubes, I don't think the flow through the different tubes will be affected much.

when i tested the tubes they all seemed to provide equal amounts of flow but that was just a visual nothing to prove if they are or not.

still having problems with the base so i might have to do it again.
problem is it's so thin and with it being a flat piece of copper sheet there is nothing to support it or make it rigid.

Heh, love the stats man. Just back from town.

I agree with Ben 100% the pressure will equalize it's self between those little tubes. You could perhaps try to countersink the inlet ends of those little tubes to improve the flow into the individual tubes. Don't think I'd open the outlet end, IMO that's the reverse of what you'd want.

I'd suggest taking that Thermaltake block down to 2.5mm at least for the next test. That block can do much better.

Note:

Lexan in route, MR-10 grade .5".:dome: Standard shiped this time, please PM on arrivial.

I would think that the best way to "observe" it, would be to run it upside down, in air: all the jets should hit the same height.

thanks bb2k i will try that.

He sure is fast with these blocks!. Is there a way of 'supporting' the baseplate when you take it down to less than 2mm?. I like the idea of 'cups' because you can keep a thicker bottom plate but take the cups bottom down to less than 1mm to BP...


PS, I got the polycarb Blackeagle!, thanx dude! you're a star :dome: . We should start a BE Lexan appreciation thread! :cool: ...

Yes, this is one of the strengths of the cupped design. You can honeycomb the base down to very thin levels, and still have good structural integrity for high mounting pressures.

Hi,

MadDogMe,

How many days did it take for the crossing? I'm curious as to time differance vs. shipping costs. Yours and Lee's were sent at differant postal rates.

Lee,

Any chance you've made further tests with that Thermaltake block? Alterations and results, if you have?

I bought a new set of drill bits tonight. Also bought 4 extra 3/32" bits, to allow for losses.

I'm looking for a Akasa 1U cooler #AK 350. This is a pin fin all copper block & pins and base are one piece of copper. Others I've checked on so far arn't one piece, or not all copper even. If you happen to know a place selling them please let me know.

All the best guys. And thanks for the kind thoughts.

edit:

The AK 350 is a alternative to a Rotor type block. Looking at Lee's work has given me some ideas.:D But as the Gfx I have in my head will have inlet & outlet at right angles to each other I like the idea of a copper pin fin cooler for a starting point. I'd cut away part of the pins, like Lee did his fins. Then drill small dimples in between the pins to get more cooling surface down into the base a bit more. Will most likly still need to grind off part of the base.

Think I need to move the bench mount sander ahead of the band saw on my wish list.:eek: