Hardcore cpu block project! (56k warning)

Cadman420 (he tends to read never really post on here) and I started working on a project since we have all the professional tools to work with at are shop. We made are own water blocks out of copper. But not just any plain water blocks super uber blocks! lol. Ran tons of test on them everything from placing it on industrial heatplates to see how much heat it can move to putting a 120 PSI into them for 12 hours to make sure they will not leak period.

Below i'll keep updating it with pictures etc to show your are projects, the test we run on them etc.

3 dimensional model

This is what we started with

Brazing

After Brazing

Finished

Pressure test

Test picture

Temp of surface plate


Temp of liquid only reached 98 degrees while sitting on a 228 degree surface temp.

This is the first prototype cpu water block that's going on my board (DFI lanparty in sig).

Normal cpu water blocks don't exceed 3/8 copper. Our block is 2" x 1 1/2" x 1/4" with 1 1/2" diameter by 3/4" with 3/4" chamber with 2" x 2" x 1/4" face plate and 1/2" copper pipe. This allows for zero water constriction. The way the pipes are designed it's self purging of air bubbles. The block weights in a0.8 LBS.

Cadman's currently designing and making a water block for my south bridge.

More pic's and info coming soon to other projects were working on.

looks very cool... il be keeping an eye on this one, good luck!!

Lol! This is awesome. Subscribed.

dang it looks awesome.

Subscribed. I'm interested on how well it ends up working for ya!

awesome i'm glad you guys like it. Like i said were working on some other "prototype" designs to test other things. But this first prototype works insanely awesome as it is. The cup warmer that we first tested on runs at i believe 140 or 160F. We let it warm up all the way with in around less then a minute the thing was cold, COLD COLD. It was to the point where the cup warmer couldn't even refresh itself fast enough to feed the block any heat.

So we put it on that industry heat plate with i believe 2 inch thick steel. Even then it was rippin it up.

Probably later this week or beginning of next week the first prototype will be going into my current rig. Were working on a plate that will bolt down to the block and then block to my board to hold it also engraving it. Not sure what i want to engrave in it yet i'm thinking of the monster energy logo hehe seeing that i drink 3 a day usually at least.

very nice to see people get their hands dirty.


The part I like about your design the most is it follows the KISS rationale. "Keep it simple, stupid". Nice job on cleaning the block after brazing. That proabably took more time than the brazing itself.

I do have a few questions regarding the test setup:

1. What kind of radiator are you using? How many fans? etc... While your water temperature appears good relative to the load, it could also mean there is a poor interface at the block and heatplate. When you turn off the pump, does the block heat up very fast?

2. What is the heat output of the plates? The best way to test for performance is to know how many Watts you are cooling, and use the temperature difference to calculate heat transfer.

3. Are you using the IR thermometer to measure water temperature? I have one similar at work, and there are a few things which do not play well. Liquids can often give wrong readings unless they are stirred well to limit reflection. This applies to any shiny surface as well.

you might have too much copper. You want the least amount of thermal mass as possible. Then, the heat will get into the water..you just want the copper to move the heat not store it.

NOXqzs
1.) The radiator is a single 120 MM fan on a coolermaster radiator. There is a resivor that can't be seen in the picture. This is were I inserted a digital thermometer for the water temp. The surface contact is copper on aluminum, I didn't use any thermal paste on this run for the test. And yes once I found a suffient heat supplier that could recover fast the block did heat up.
2.) This question will be better answered on my next test. I have to lap the surface yet, and add more sensors, also I will be monitoring temp, closer. I also have to shorten the hose length to get a more accurate test. This was something crude that I threw together. to give me a better understanding for my next design. If you would like, I will have rumbl3 post good picks of the surface to surface contact, and better pics of the temp readings. I have about 30 pictures or so above is just some what of a quick slide show.
3.) No, I am using the IR gun to monitor the surface temp of the aluminum plate. If you look closely I have painted the surface with a flat black under the block. This was to eliminate the reflection as much as possible.
I would like to add that even tho I don't normally post on here, I do read this forum on a daily basis and took many of the discussion into consideration when designing this prototype. I believe this answers your question, except for #2, and there is a good point there. So I will have better pictures of the sensors for rumbl3 when he updates this post.

relttem
I agree. I have a lot of copper on there, My next design doesn't have the solid copper face. I only put it on there for ease of construction for my prototype. But there is less mass then you think. The outer chamber is thin wall 1.5 x 3/4 copper tubing and there is an inner chamber that is also thin walled that is 3/4 x 5/8. The inner chamber acts like a fin centered, and the intake pipe goes to the inner chamber were the liquid moves outward to the outer chamber and exits at the top. Giving the liquid time to absorb the heat in the block before exiting. I also used 1/2 copper tubing so there is no water restriction what so ever. 1/2" in and 1/2" out. By having the exit on the top of the chamber this allows for air to rise to the exit point, and purging it from the block. I have better 3d models that demonstrate what is happening. For testing purposes it is on its side. I posted earlier to Noxqzs, I will have more testing coming up with better pics of the surface to surface contact.

Sorry if I sound like I'm trying to rain on your parade; I'm just curious about the design of blocks and would enjoy building one myself, but since I cant, I want to share what I pick up with you guys (if you don't mind).


If the surface that is to be cooled is 228 degrees (F or C? I'm going to assume F) and the water is 98, doesn't that mean that there is not much heat transfer from the hot surface to the block to the water (which is what you want)?

What happens if you slow the flow of liquid to give it time to 'heat up'?

Also, I was reading about flow through blocks/radiators. In theory, if you could have a square or a rectangle tube, the rectangle would more ideal. Theoretically, the water touching the walls get cooled off/heated up, but the middle does not encounter much heat transfer. So the square would have a 'dead zone' whereas a rectangle would not have as large a 'dead zone'.

On a similar note, you don't really want (or more of need) a tall water block, because the water would have cooled off the parts of the block farther from the heat source. I guess the dead zone also may play a factor with taller blocks.

On a similar note, do you guys have any intention of taking a heatsink, attaching some sides to it, and trying to pump water through it? Imagine if you could watercool a TRUE with that much surface area.

I hate to sip on the "hater-ade" or in this case the "truth-erade," because I would not doubt that the TRUE or HDT series HSF's would beat this thing.

I am no stranger to custom fabricated waterblocks, and I know the hard work you guys put into this thing, but I just do not see the reason to make blocks nowadays... unless your Cathar. Simply put, it was hard even years ago to beat the "aged" designs of block manufactures, but now the tech gap seen in current blocks (and their prices) just blows away any home brew.

Still is nice to see people using creativity and to learn the true value of the great blocks in the market today.

My Custome Watercooling

Keep us posted with your adventure

bryan d

PS - I wonder how my block would be against your home brew? Still have my block too...

I hate to be a naysayer, but I fail to see how this block is going to work well at all. You have way too much copper between the CPU and water. You have nothing built into the base at all to increase the surface area and increase turbulence. There is a reason that current designs use pins or diagonal cuts or the like. And your test with having it sit on a hot surface means nothing. The heat coming from a CPU is very tightly concentrated. The direction of your flow also doesn't direct the water into the transfer surface. Water just passing through is very inefficient. All blocks you will see have the flow hitting perpendicular to the CPU area in the waterblock.

If you want to make a decent block, look at other modern designs and try to come up with something like them.

I also really don't see the practicality of making CPU blocks anymore. They are too cheap and readily available to make it worth the time and expense of building them.

Take a look at the stuff these two guys in Germany are putting out. Their stuff is amazing. http://www.xtremesystems.org/forums/showthread.php?t=169885

Subscribed.
I'm interested on how well it ends up working Awesome Idea

I have read a lot of what people on here are calling "naysaying" I would like to point out that each and everyone of you have had excellent points on the block and that I have corrected on lasted design based on what you call "naysaying". I only have a solid model of the new block. I have taken almost all of your points into consideration in designing this lastest revision. So "NaySay" all you like, it allows me to tweak my design, and set up better experiments to post. Keep watching and let me know what you think. Your "NaySaying" is helping. I have been a reader on here for some time and your opinions matter to me. Please post

interesting!

when you are testing, you should be *reading the water temp going in to the block and the temp coming out of the block. You might even want to not even have a full loop to make sure your radiator isn't a limiting factor, i.e. pull from a spicket/large reservoir and output to a drain. Obviously, the higher your Tdiff, the better your block performs.

lol splat you just said what i was going to

Subscribed. Home made and other cheap cooling solutions are intriguing.

I know you're already done. What about lining the interior with Copper Braid Solder Wick and soldering it in?