Want advice on making a waterblock

[lemmyslender]

Currently, I have an old Pentium HS enclosed in acrylic that I am using as a waterblock. Obviously, the performance is lacking However, it is better than the stock HSF that I was using. Temps are in the 39-41 range running [email protected] (ambient is 19-23) per the mb sensor.

I want to make a more conventional type waterblock. I'd also like to keep the design fairly simple. Attached is a picture of my idea.

The channels are 1/8" wide x 9/32" deep. Inlet and outlet would be 1/2". Overall size is 2" x 3" x 3/8" thick. Haven't decided yet whether to solder a top on or use lexan (which would require some more holes to hold it on.

My dad has a machine shop and the capability to mill this for me. I've already run the design by him, but not planning on starting immediately (he's thinking X-mas present). So there is time to modify the design.

Other info:

1/2" ID Tubing = .196 in^2
Channels 1/8x9/32x6 = .211 in^2
So the flow should be OK through the block (hose barbs would be the smallest part anyway)

Material is copper (of course)

Any ideas appreciated!

 

[nikhsub1]

Doesen't look too bad to me. It looks better suited toward a CPU with a heatspreader than a bare core CPU, but overall, I like it!

 

[JML]

I think it would be best to change the dimensions from 2x3x3/8
to 2x3.2x3/8 will give you that little bit of extra room for the mounting holes

 

[Paxmax]

Yeah! I like the design very much. But then.. I'm a sucker for parallell channels... makes the lowest resistance, best heat exchange.

 

[Cheesy Peas]

looks good, im drawin some designs up for a waterblock im hoping to make soon

 

[lemmyslender]

Thanks for the input guys. I'll think about making it a little longer, but it should be OK, there is enough room for the mounting holes.

 

[BabyEater]

Thinking of doing something similar except i plan to keep the channels about 1.5mm apart going down the midddle and 4mm * 5mm as i only have a drill not a milling machine.

 

[lemmyslender]

I am trying to keep some mass in the middle, and get more surface area. I like the slight angles, so that there aren't any flat edges to interupt the water flow. Also trying to keep it simple to make the milling process easier.

Unfortunately, my Dads' machince shop is about 5.5 hours away, otherwise I'd go over at night or on a weekend and do it myself , that also means I won't get it right when it's done, I'll have to wait till my parents come out to visit, or I go visit them after it's finished.

 

[NeoMoses]

I've made lots of blocks that were 2" X 3". The mounting holes fit fine. That looks like a good design. I had a similar one on the drawing board a few weeks ago. Lemme know how it performs.

 

[BabyEater]

The reason i was trying to keep the channels close together was that i don't think the heat spreads out to the sides too much but stays more in the middle so i was trying to concentrate my channels there.

 

[lemmyslender]

Looks good. I was trying to position my channels to get (hopefully) close to even flow through them, with low resistance. The heat shouldn't spread out too far, I just wanted a little mass in the middle.

On another note, any thoughts on using a lexan top with an o-ring? Anybody tried that sucessfully?

 

[JFettig]

it looks real good, how big is the bottom peice? 1/4inch i find is really good, then mill like .07inches from the base etc... or rite around .1inches if the inlet was in center .05inches would be more optimal but anything in that area works real well

 

[lemmyslender]

Thanks J-

The block should be 3/8" thick with channels 9/32" deep which leaves 3/32" (0.094 ) So that should be just about right.

Gotta remember to remind Dad to get a good finish on the bottom. Maybe a little sanding on the mill table.

PS - Like the look of the Type-S block.

 

[Frodo Baggins]

I don't know jack about water block building, but i've seen the more effective designs today, and i've seen them with like an inlet and two outlets, or whatever with more then 2 inlet/outlets....

would his deisgn be more effectiveif he got some better flow and got that heated watert out quicker?

 

[The Overclocker]

I don't know jack about water block building, but i've seen the more effective designs today, and i've seen them with like an inlet and two outlets, or whatever with more then 2 inlet/outlets....

would his deisgn be more effectiveif he got some better flow and got that heated watert out quicker?

um... im not sure, his design seems a bit like cathars, except the channels are larger (much easier to make) and it would be very hard to have single input going into all of the channels, infact they only way to do that would be to have the same design as cathers, which really distroys the point of making your own.

i dont think the point of dual outlets is getting water out quicker. the water is still flowing at the same speed - and due to the thermal capacity of water removing water quickly is pointless.

the middle inlet used by most people is to premote turbulance - the same effect can be had by using a more powerful pump - the reason other people use to have a middle inlet is to place cold water on the core - this is not ture as the temprature increase is very small

 

[lemmyslender]

Overclocker has it right. I think it would be difficult to design an effective center inlet block with channels.

I wanted to try to keep the machining aspect simple, which the larger channels accomplish.

I had drawn up a spiral type center inlet design that had 1 inlet and 4 outlets, a "quad-spiral", but I think that you would loose too much flow bringing the four outlets together again, not to mention the extra resistance due to the extra tubing and fittings needed. Plus, I think it would be pretty hard to machine. It did have a heck of a lot of area for heat transfer though (compared to regular spiral designs)