New Water Block Very Soon !!!!

[morphling1]

Ok, fellas this is what I've been doing this afternoon.



That's how my new block is going to look, this was made on lathe with drill press with hand wheels of course. For now it's pretty rough so the next step for me is to grind it more smoth with dremel , but not too much I don't wanna lose that kind of surface that is very good for heat transfer. Next I'll solder copper top to tha block and then put 15mm plexy block on that where the two outlets combine back to one.

[WebMasta33]

Nice block.... Do you have a IN the center and out the two ends? Or just in one end, out the other???

But about that woman on the mouse pad... she looks like she's sitting on the toilet :shrug:

[morphling1]

Yes, there is center inlet and to outlest that will merge into on in plexy part of the block, I'll have that ready tommorow.
And yes that woman is sitting on a toilet

[morphling1]

I just finished grinding,the chanells, next for me is soldering the top.

[bigben2k]

Very nice! It should perform very nicely!

Are you going for a plexi or copper top?

[morphling1]

Uf, I'm beat, I was in my workshop for 4 hours. It's tiring to focus on controling your mill with hand wheels and do that with few error as possible step by step until you reached final depth. Also it's tough to make prototype where you have to make changes as you see trouble. But anyway I'm done, I have all the parts here.


Now I'll just have to solder copper top on block and then screw plexy block on that, but thats the easy part.
C-ya later.

[jag_e_fattig]

wow, maybe it's just me, but i've never seen that kind of design on a plexi top before. it looks nice, hope it performs well

[bigben2k]

Eeeh... I don't get it.

You've got 3 holes on the copper plate, and two on the plexi. I can see a channel for the gasket, but is that another channel in the plexi? How does it all flow together?

[Cyco-Dude]

yes, that does look unusual....i hope it works like you think it will LOL!

oh, i see how it works. ben, look at the plexi top. the water enters thru the center hole, and goes thru the block. it will then exit thru the 2 holes in the copper block, then go thru that 'L' shaped channel in the plexi top, and merge at that one hole. its a parrallel design. this makes me wonder tho, i thought waterblocks work best when the water has more velocity? spliting it into 2 channels would lower the velocity of the water thru the block, thus degrading performance (so im told). we'll have to see

morph, it might help if you posted a pic of the bottom of the plexi top. if it where me, id just have the plexi top w/ 2 outlets, that way you can see the inside of the block

[bigben2k]

I see! thanks CD!

As for the design, well, parallel designs that split the waterflow work just fine, it's just a little more plumbing. It works well because it sometimes allows for better flow (i.e. a bigger pump).

In this case, Morphling1 eliminated the extra plumbing by linking the outlets through the plexi.

I would expect that the flow would be very nice within the block, but would be most restricted because it has to bend 90 degrees twice, for each outlet, then go through a channel in plexi that may be smaller than the corresponding channel in the Cu block.

But then again, the channels are so wide within the block, that it should compensate nicely for anything. Most of the heat will be in the center, and those long posts will help a lot.

One advantage of spiral designs is that the channels leave a big fat corner to collect extra heat.


Hold up... do I see that the block intake in the plexi is formed in a cone ("V") shape? Morphling1, you shouldn't restrict the flow like that, it's really not necessary! I know you're trying to shoot the water as far down as possible, but it shouldn't be a problem at all, really!

I like the big kick-as* barbs...

[WebMasta33]

^^ Yes, I finally saw that channel in the plext as well. Very nice.

[Cyco-Dude]

lol, it could be an optical illusion...plexi can do that methinks once morph posts a pic od the side and / or bottom it'll show up clearer....

[morphling1]

Where to begin, if the water splits two way and the combine cross section is the same as one big chanell, the velocity would pretty much stay the same (it all about area) , and at the same time my lenght of the chanell are that way a lot shorter.
Next thing you saw correctly ,I do have a nozzle inlet on pleksy block and if the narrow part is only very short it doesn't restrict flow too much and at the same time it speed up the flow impacting in to the center of the block.
Here are pics. of my finished block.





I must say that this is only a prototype, and I will change few thing when going cnc way. So all clever suggestions are welcome.

[bigben2k]

Ok, here's a clever suggestion, try making another plexi top, without the center jet.

Also, try making a plexi block, with 2 outlets.

Otherwise, you did a fabulous job!!! This is certainly the most original copper/plexi-top combo I've seen!

And it's soooo shiny!!!

[morphling1]

Yes I was thinking of that too, because it's very easy to mill pleksy , and I can make few different tops, and see who the winner is. Good one Big

[bigben2k]

[Brad]

I love the top, and the overall design rocks. I'm guessing it's socket A

[morphling1]

Thanks, yes this is Amd socket A block. I just tested this block with air preasure under water and no leaks, now I'm encouraged and will use full plexy top for final block.
And good news for P4 users, I dug out intel specs. for heatsink mounting and I will now also make P4 version of that block, and also it will bi 50x50 TEC thick copper plate ready

Now on to testing.

[morphling1]

Ok, initial testing done. When I first install waterblock on mobo, I ran out of artic alumna paste, so I put it to little on, but I countinued anyway and the results with now 103W Duron (1110@2,3V) was 11 °C above water temp. And now I found out I got some as2 left, so wb off, as2 on, wb in, results 2°C drop instantly
But the damdest thing is I still didn't recieve my rma-ed axp (2months) so I can't show you how perform with 30% bigger core. I had one xp for few hours to test and I got 1920MHz at 2,11V (~120W) and the temp was lower than with that hot little Duron.

[Haddy]

haha u made amdmd.com's bit and peices......

http://www.amdmb.com/index.php#News-4439

too bad itsto the overclockers.com forums

[morphling1]

Yeah, I mean how could they miss this thread

One more pic. of block installed and runnig.

[Brad]

are you happy with an 11C temp over ambient, or do you think thats a bit high compared to your previous block

[morphling1]

Yeah I'm very happy with that, you must know that I went from old Duron 1080@2,11V to 1110@2.3V and that put a lot more heat out. I pretty confident that this block perform on paar or even better than most of the blocks on the market. But to be sure, I'll send one block to pHaestus for testing, and then all of you will get the picture of its performanse.

[bigben2k]

I had a lucid moment yesterday... and dreamed up this block:

In short, it's a 3/8 or 1/2 barb center mount, surrounded by a 1 3/4 barb mount. A little tube in a big tube.

The tubes would connect to a res, a few inches above the wb, (a 2" cube) where the big tube discharges, and the small tube passes RIGHT THROUGH that res, to be connected to the rad. the cube res would have an outlet.

The circle in the graph is a fin pattern.

pros: very little flow resistance

cons: heavy, tall, not flexible

The theory here is that better performance can be achieved with a higher flow, rather than trying to design an intricate channel and fin pattern. It also "steals" the corner masses, from the spiral block.

What do you think Morphling?

[Brad]

if you were clever enough, you could do that, and have it barely taller than morphlings block, at it wouldn't be inconvenient at all