Building Our Own CPU Water Block?!?

[AlexTheGreatish]

How hard can making your own water block be? Actually not too hard if you have a CNC Router... Also really hard when you have a Linus...

 

CAD: https://lmg.gg/WaterBlock

 

Buy Single Flute Bit.
On Amazon: http://geni.us/nAxR

 

Buy CPU Water Block
On Amazon: http://geni.us/QQAVw
On Newegg: http://geni.us/9ADQ

 

 

[aisle9]

This feels like a "what could possibly go wrong" moment.

[Imbellis]

I love these videoes that you work on Alex... That's probably because I'm going into MechE, but I'm sure lots of the audience likes it too regardless.

[aezakmi]

now these are the vids I like

[Kroon]

If you want the acrylic to become clear after machining just flame treat it.  It's easy and you get instant glass effect.

[SolarNova]

1 hour ago, AlexTheGreatish said:

How hard can making your own water block be? Actually not too hard if you have a CNC Router... Also really hard when you have a Linus...

 

CAD: https://lmg.gg/WaterBlock

 

Buy Single Flute Bit.
On Amazon: http://geni.us/nAxR

 

Buy CPU Water Block
On Amazon: http://geni.us/QQAVw
On Newegg: http://geni.us/9ADQ

 

 

 

Have you considered using a basic copper heatsink as the base of the waterblock?

You'd have to shave down the fins to a more resonable height, and trim around the edge to give the plexi top a surface to mate with, but it would get around the issue of trying to cut ur own micro fins / channels.

 

[Robotics_]

Now that you are CNCing your own water blocks, whats stopping you from watercooling directly to the CPU's heatsink? Like a hole in the waterblock, allowing water direct contact to the CPU heatsink. Idk if this is feasible, but could improve temps and remove the need for thermal paste!

[tjd10684]

 

1 hour ago, Kroon said:

If you want the acrylic to become clear after machining just flame treat it.  It's easy and you get instant glass effect.

Its very rarely a good idea to flame polish acrylic. Melting the surface introduces stress into the material that WILL cause cracking eventually. You would be much better off polishing with Novus polish with a hard felt cloth. If you must flame polish then you need to anneal the part to even out the induced stresses. 

[Terryv]

Here's a tip for when using taps.

 

When creating new threads, you should first use a taper tap. These cut more gradually and thus make it easier to get going, then finish with a bottom tap. There are also plug taps, bit those aren't needed here.

[Arika]

haven't watched the video yet, but based on their air cooler DIY videos, i fear for the life of the motherboard and CPU

[RobbinM]

1 hour ago, SolarNova said:

 

Have you considered using a basic copper heatsink as the base of the waterblock?

You'd have to shave down the fins to a more resonable height, and trim around the edge to give the plexi top a surface to mate with, but it would get around the issue of trying to cut ur own micro fins / channels.

 

This. But you can make it more simple by using the original heatsinks as your block base.

Trimming the height may be the most difficult part of it. I think I would use a belt sander in the direction of the 'slices' or how do you call those things in English, finns? Anybody?

 

 

The loop at the back is not scaled properly, but you get the idea....

 

This is a Xeon socket 771 1U cooler by the way.

Great video guys, I loved it. 

[tjd10684]

16 minutes ago, Terryv said:

Here's a tip for when using taps.

 

When creating new threads, you should first use a taper tap. These cut more gradually and thus make it easier to get going, then finish with a bottom tap. There are also plug taps, bit those aren't needed here.

This is correct but from the specifications of G1/4 threads used in water cooling fittings the threads have a not as standard 55deg included angle rather than the more typical 60 deg found in ISO and ANSI machine thread specs. Because of this its possible that they were not able to find a full set of taps like you have shown. Also when tapping a hole all the way through piece of stock its typically not necessary to bottom tap. Just run the tap all the way though to the end of the threads on the tap. As long as the material is thinner than the full formed thread section of the tap you will have full form threads all the way through the material.

 

From the video though it looks like they were using a plug tap (that's perfectly fine) they just needed to give it 3-4 more turns to get the full thread to the back side of the acrylic. 

[Terryv]

1 minute ago, tjd10684 said:

This is correct but from the specifications of G1/4 threads used in water cooling fittings the threads have a not as standard 55deg included angle rather than the more typical 60 deg found in ISO and ANSI machine thread specs. Because of this its possible that they were not able to find a full set of taps like you have shown. Also when tapping a hole all the way through piece of stock its typically not necessary to bottom tap. Just run the tap all the way though to the end of the threads on the tap. As long as the material is thinner than the full formed thread section of the tap you will have full form threads all the way through the material.

 

From the video though it looks like they were using a plug tap (that's perfectly fine) they just needed to give it 3-4 more turns to get the full thread to the back side of the acrylic. 

Thanks, I didn't know that about g1/4 threads.

 

If you open with a taper tap, then you should also bottom tap. The taper tap doesn't cut threads as deep and screws sometimes don't fit properly.

[RobbinM]

If they come in a set of three, taper is usually not marked, plug is marked with 1 ring and bottom is marked with 2 rings.

[unijab]

31 minutes ago, Terryv said:

Here's a tip for when using taps.

 

When creating new threads, you should first use a taper tap. These cut more gradually and thus make it easier to get going, then finish with a bottom tap. There are also plug taps, bit those aren't needed here.

 

 

 

If the threads on the fittings are based on piping thread.. they are tapered intentionally.

 

 

 

 

 

[tjd10684]

4 minutes ago, unijab said:

 

 

 

If the threads on the fittings are based on piping thread.. they are tapered intentionally.

 

 

 

 

 

That is only sometimes true G1/4 threads are based on BSPP (British Standard Pipe Parallel) making it a straight thread similar to a machine thread. There is also an ANSI standard called NPS threads that are straight as well. You find them on fittings that rely on a method of sealing other than just the thread interface itself. 

[Velcade]

Nice to see Henkel products in the wild!  454 baby!!!

 

 

[RobbinM]

1 hour ago, tjd10684 said:

 

Its very rarely a good idea to flame polish acrylic. Melting the surface introduces stress into the material that WILL cause cracking eventually. You would be much better off polishing with Novus polish with a hard felt cloth. If you must flame polish then you need to anneal the part to even out the induced stresses. 

When polishing, start with sanding. If you start with the final step you'll end up with a shiny piece of material that isn't flat.

Depending on how coarse (do I spell this right?) the piece is, you can start with 320-600-1000 wet sandpaper. Depending on how fine your polishing paste is, you might want to take it up to 2000 grid. Then polish it in two steps with 5000 and finally 12000 for that shiny effect. I use this method on headlights very frequently in my shop.

 

When I have to make something flat, like a cylinder head, I tape the sanding paper to a piece of glass. No double sided tape, just duct tape on the edges.

Then I move the cylinder head on this surface in figures of 8.

[Computernaut]

"We don't need to drain the loop." Yeah, heard that one before; last time Linus got water all over his motherboard. I can't believe the editors didn't add a piece of that video there.

Is it just me or do all the videos with Alex's DIY projects turn into Linus cringing sessions?

[dan42c]

My ¢2: For the red, you would probably be better off using something more along the lines of a cold plate. A cold plate would be lower profile and could have no seams inside, so no leaks. You also have the option of a press fit in a grove, soldering to a flat plate or thermal epoxy to a partial grove, each with advantages and disadvantages. Typical cold plate:

 

 

My first thought would be to use a thin walled stainless tube, given the space constraint (also mcmaster has a lot of options), pressed into an aluminum block(s) (probably with a little thermal epoxy, as press fits can be a bit tricky).

[corrado33]

3 hours ago, Arika S said:

haven't watched the video yet, but based on their air cooler DIY videos, i fear for the life of the motherboard and CPU

It wasn't nearly as bad as the last few videos, but very unimaginative. They basically made an extraordinarily simple water block that managed not to leak and got shitty temps out of it. Also they just CNCed everything, so they very likely just downloaded the plans online.

 

I just don't get it.... would it have been hard to cut slots in the middle at half depth (compared to the o-ring groove) for more surface area? What about literally scratching it with a razer blade. It is only copper after all. The lack of imagination out of that crew is.... stunning. You could have done so.... much.... interesting.... stuff. If I had access to that equipment I'd be able to build you dozen different waterblock ideas just off the top of my head. 

 

For examples: Variables you could change in a controlled manner. 

 

1. Surface area exposed
2. Volume of water chamber above copper
3. Effects of flow (sideways water entry or top water entry?)
4. Designing channels that promote good (or bad) water flow. 
5. Copper plate thickness. What's the limit before you get concave bending due to the pressure?
6. How about a spiral design cut into the copper, with another copper sheet on top. Water goes in through the middle, has to travel out through the spiral, then over top of the top copper cover, then out of the plastic housing? 

 

So, so many ideas. Is it just lack of imagination?  Or lack of skill in CAD? Or what?

[Fasauceome]

5 hours ago, Robotics_ said:

Now that you are CNCing your own water blocks, whats stopping you from watercooling directly to the CPU's heatsink? Like a hole in the waterblock, allowing water direct contact to the CPU heatsink. Idk if this is feasible, but could improve temps and remove the need for thermal paste!

Actually people have designed a waterblock that replaces the IHS (integrated heat spreader, not quite a heat sink) and LTT reviewed it

https://youtu.be/n4O_Dq3xRJA

 

[Ezio Auditore]

Would a gasket maker work instead of an o-ring?

[Computer King]

I built my own water cooler in high school using a vertical milling machine. (Non CNC)  Similar design.  The copper I used was a 2 pound bullion bar from Provident Metals.  The back plate I cut out from plate aluminum.  To seal it I used Permatex waterpump and thermostat RTV silocone.  I have been using the block in my system for years now without any issues.

 

[Gmoneyinthebank]

It hurts to watch that video.  Please hire a technical adviser.