Ideas for improvements of my slot waterblock

I posted my completed article on the slot jet waterblock. But since I have been thinking of ways to improve the block. I was thinking of thinnung the base first. I could cut a hole in the existing 1/4" (~6mm) base and then solder on a very thin piece of copper 1/32" (.8mm). Or I could just replace the 1/4" base wiht some 1/8" (~3mm) copper I have. What do you think?

Oh yeah, I almost forgot. I made another inlet slot thingy and I tested the water flow. I had to actually sand the inside of the slot to get a smoother stream out of the slot. I will definitely do that to the block when I get it apart. I may also resolder in a new slot inlet that is longer so it gets closer to the base. Any ideas on a good slot outlet to base distance? Right now the slot outlet is 1/4" away fromthe base.

O

I had the same question before.
You can see what replies I recieved HERE .

But that was for a slightly different H20 block too. -kinda.
as can be seen HERE .

- as in the PM I sent you earlier.

A thinner base (3mm) would be the way to go.
(mine started out as 5mm then I cut into it)

-If you read some of Cathars posts on the subject.
& he has some excellant performing H20 blocks.

Keep at it though.

G/L & keep us posted.

Things I am going to try

Here are the ideas I have so far. First I'll try the 1/8" (~3mm) base, Idea 2. If I can get it to work. I'll probably use a second gasket to attach the base.



O

Go with #'s 1 + 3 together.

Be careful of the mounting pressure so it doesn't bend the 'New"
baseplate.

& maybe go with the same diameter "Inlet" all the way down to the baseplate. - so the flow "velosity" doesn't slow down when it enters the main block. - pressure loss / drop

- High Volume / High Pressure
-Not High Volume / Low Pressure.

Ah, know what would be cool Owen? If u could leave the base as is, but directly under the jet mill it out so it's much thinner just over the area above the die, U could leave it roughed up there, to get some turbulance and extra surface area over the die, yet it would be thinner...

BTW I dig your block, very very nice work man.

J.

How would you keep a 1/32" baseplate from collapsing under the pressure of mounting the CPU?

WarriorII,

Yeah, that's the conclusion I came to also after reading yet another paper on this. The smaller the ratio of slot width to 'opening inthe base the better cooling. I'll take it apart and drill out a slot the width of the core. I already made another slot-nozzle that should be long enough to reach down close to the opening.

Since 87,

I measured the piece I want to use and it is 1mm thick which is a little more than 1/32". It seems pretty stiff to me. It shouldn't have to take all teh force because I will only make a slot like hole in the base.

I'll take pics to explain it better when I do it.

O

If the base of your block is to thin it will flex when mounting it on the motherboard, which would result in less than optimal contact with the die. Just my $.02

1/32" base seems awefully thin.

Owenator said:

I will only make a slot like hole in the base.

Click to expand...

Oh, I see.

I wouldn't make your nozzle no more than 2mm above the base and that would be pretty close to optimal. The further you get from the base the performance tends to drop as well. There's more to it still than just that if you want to tweak the details out of it. You will need to keep the water that has already picked up heat from circulating back into your impingement also. Think "submerged" when working out your nozzle. That's the biggest key. Not only does the water hit the base, but also some of that water will swirl back up into your impingement and heat your cool supply of water. Submerged water impingement behaves very similiar to air impingement.

OK Hear's what I did

I decided while thinking about this and looking at the block that for tonight I'd just add a milled out channel (juliendogg's idea).



While I had my system torn down I took a few action photos.

Here is the flat part of the jet.



Here's the thin part.



O

Nice action shots!

You might have to try various depths, to get the nozzle height over the baseplate just right.

did that mod help temps?

Better temps!

jAY said:

did that mod help temps?

Click to expand...

I did a stress test this afternoon. The idle temps dropped a lot, from 38C to 32C. Load went from 42C down to 40C, not as significnt a drop but still better. The room temp is the same as before 23C. I'm working on a longer inlet that will go down into the new channel to 3mm or so above the base. I won't get to that tonight. I'll update the thread when I make any more improvements.

O

Wow, smooth flow. I like it.

I was hoping you'd figure it out before you milled anything in the base plate.

It nreally needs some kind of ruffness there where the water will be hitting it.
QUOTE from Juliendogg:

U could leave it roughed up there, to get some turbulance and extra surface area over the die, yet it would be thinner...

Click to expand...

I have to agree. But that's my .02 & it's your block.
AND it's is REALLY REALLY NICE.

Maybe doable without having to undo the block completely, if you use a dremel....

that CAD design looks really nice. How bout dimples on the center part?? Should be fairly easy like rotor blocks.

Would it really be possible using a dremel? I think it will just turn his block into crap as dremel doesnt like to do clean cuts... Plus it seems like hes gonna need STEADY a$$ hands....

dimples only really work well, if there is lateral flow going over them, else they just presurize themselves and prettymuch ain't worth their volume in Copper.
STEADY hands.... Indeed yes, but do not lose faith so fast...

stick a mandrel and cutting disk in the chuck of your drill press and use the cross slide vice similar to a horisontal milling machine, thats how I cut a lot of my first tinychannels. Set the RPM as high as it goes and take small small cuts.

Owen, it would be pretty dang good if you did that and didnt have that channel in there. but with the channel, maybe cut the grooves in the bottom and the edges of the channel, not as far out as #rotors drawing, but deeper.

-edit-
I forgot to add that a 1/32" base is way too thin without something supporting it like fins or something. I think the minimum you can go with that one is 3/32", I bet 1/16" would even bend.

Jon

Jon,

I am thinking of using the big one of these two cutters in my drill press like you said.


The existing channel is about 3/16" deep so that only leaves about a 1/16" base. The base was 1/4" to start. Idon't think the base has caved in, but I'll check when I disassemble it next.

The piece of copper I want to use for the 1/32" base seems pretty darn rigid. Here is a thickness shot of it next to a dime.

.

If I get some time tonight I'll try it out. Other wise ther's always the four day weekend.

Oh, yeah. I was originally planning on beveling out the edges of the milled slot but #Rotor's idea is interesting. I've sprayed this nozzle against the bottom of my kitchen sink and the water flows mainly perpindicular to the slot so it would flow through the grooves nicely. I don't think I'll get much better heat transfer though from them with this design.

Part of this was an experiment to make it not like the white water or casade. I wanted to explore the 'jet' part of jet impingement. As it is I can do the channel cuts, test it out and then grind them out and test some more. What I'll probably end up doing next is drilling a 1 1/2" hole (like the one I used to make the cavity) in the base and covering it with the 1/32" piece of copper. Then I would bevel out the sharp edge from the hole. I might later even switching to a new 'untouched' 1/8" copper basebase. It's nice to have extra copper bar lying around!

I love the fiddling around aspect of this!

O