i7 Water cooling system

I'm making a water cooling system for my friend's computer with i7 processor. The cooling block is basically 3 parts, a copper plate that's fixed to the processor, an 'injection plate' which guides the water through the copper plate and a plate on top to attach the pipes.

Here's the injection plate, just done with my CNC router ... not the milling machine!

Attachment 4280
Attachment 4279
Attachment 4281

Took quite a while to make that - mainly due to needing 7 different tools (milling cutters from 2mm to 10mm and drills) and machining on both sides. The actual cutting time wasn't too bad.

It will be anodised next, and the copper part nickel plated.

now that's quite impressive


James

nice one man... 10mm!!! looks quite clean, have you lowered your gantry ?

I think when I finish my CNC, I'll make a water Block for My GTX 460.

I'll do it with a 0.5mm bit.

http://www.ebay.co.uk/itm/0-5mm-Tung...#ht_1623wt_804

if I can that is... all copper though :)

Also, I have a Tap here for 1/4bsp.

You can borrow it if you want :)

it will cost you shipping there and back.

Thanks for the kind words. The 10mm cutter was used on the router to reduce the thickness from 12.7mm to 10mm (part is 110*110*10mm)... I used 0.8mm depth of cut.

Quote:

---

Originally Posted by jcb121

I'll do it with a 0.5mm bit.

---

I'm going to use a 0.5mm slitting saw - bought one from Arc Euro recently. It's cheaper than the slot drill and much less likely to break.

Thanks for the offer with the tap, but I think we're going to use the M14x1 tap I have as it's less restrictive.

I didn't think they did water cooling connectors in M14. you may find you struggle finding compression fittings.

and this is the only thing I can see a 4th axis being useful for! :)

Quote:

---

Originally Posted by jcb121

I didn't think they did water cooling connectors in M14. you may find you struggle finding compression fittings.

---

They don't - I'll make them.

I may well use the 4th axis for slitting the copper block.

Some progress with this - here's the block that holds the pipes with decorative engraving done with 0.8mm and 2mm tools:

Attachment 4349

Also done the copper block that sits on the processor, but no photos of that yet.

Cooling block for motherboard and graphics card to follow shortly ... need to buy some 6mm (or so) aluminium. I'll probably get this as it looks useful:

http://www.aluminiumwarehouse.co.uk/...duct_info.html

nice work.

Have you got the parts anodised, i have just started anodising my self with some great results.

i think you might have sealing issue with the o ring groove due to surface finish.

i would avoid nickel plating the copper as it will affect the thermal transfer.

Quote:

---

Originally Posted by fragger6662000

Have you got the parts anodised, i have just started anodising my self with some great results.

---

I have anodised several parts myself before with good results, however this sadly is not one of them.

This is:

Attachment 4510

Quote:

---

Originally Posted by fragger6662000

i think you might have sealing issue with the o ring groove due to surface finish.

---

I chose the correct sizes for the O-ring slots to ensure it seals properly. I will confirm this by submerging it in water and pressurising it with compressed air. If I see bubbles then either there's a problem or I'm using a much higher than operating pressure.

Quote:

---

Originally Posted by fragger6662000

i would avoid nickel plating the copper as it will affect the thermal transfer.

---

While this is true, the effect is rendered neglegible by the several micrometre thickness of the plating. This is widely accepted as the best solution to galvanic corrosion, confirmed by nickel plating throughout the majority of high ended water coolign components.

Some more pictures showing the copper block and assembled parts:Attachment 4507
Attachment 4508
Attachment 4509

I've seen it and it look really really good

James

you did well machining the copper as it does have a tendency to work harden and snatch tools. i was sure you were going to pressure test it, but o rings do need a good surface to seal. I designed something similar and it let in coolant like there was nothing there.

Making progress with this...slowly!

Cut the connectors for the water tube, nothing special here. Cut the M14x1 thread on the lathe to ensure it's not wonky.
Also nickel plated the copper which worked well, but forgot to take a photo.
Attachment 5121Attachment 5122

Plus some bits for the motherboard, started these ages ago but still not finished:

Attachment 5123

That's really quite cool Jonathan well done

out of interest how long did it take to cut the mother board components out

James and Luke yes Luke is still going strong

Looks good Jonathan
Looks like we both use the same method for mounting material and parts to the table.
First mount material, do all insides and holes then mount parts through holes and cut outside profiles. Pretty simple and good way to hold the material.

Quote:

---

Originally Posted by luke11cnc

That's really quite cool

---

That's the general idea.

Quote:

---

Originally Posted by luke11cnc

out of interest how long did it take to cut the mother board components out

---

It was so long ago I can't really remember (they're still not fully cut out as I've been doing other things). Do you mean to make the whole profile, or just cutting them out once the inner profile is complete?

Quote:

---

Originally Posted by wiatroda

Looks like we both use the same method for mounting material and parts to the table.
First mount material, do all insides and holes then mount parts through holes and cut outside profiles.

---

Almost - I generally get it to drill the holes first, then put screws in those holes and then cut the inside and outside profile. Doing it in that order means the material doesn't have to be clamped to the bed particularly well to start with.
It gets annoying when, as I have today, there's 180 parts to screw down with 2 screws in each...tried double sided tape but that left a bit of a mess and still needed one screw in each to be sure.

Quote:

---

Originally Posted by chip

i was wondering in order for them to be re-anodised would they need to be taken back to bare metal or can they just be steel wooled then re-anodised?.

---

Anodising oxidises the outer surface of the part, then the dye bonds with that so to re-anodise and dye you need to remove the previous layer of aluminium oxide. If you submerge it in sodium hydroxide (NaOH / caustic soda) the aluminium oxide layer will be dissolved. However you have to be careful since that will reduce the size of the part, especially if you leave it in too long. If there's fine threads they could disappear...

Quote:

---

Originally Posted by Jonathan

That's the general idea.



It was so long ago I can't really remember (they're still not fully cut out as I've been doing other things). Do you mean to make the whole profile, or just cutting them out once the inner profile is complete?




Almost - I generally get it to drill the holes first, then put screws in those holes and then cut the inside and outside profile. Doing it in that order means the material doesn't have to be clamped to the bed particularly well to start with.
It gets annoying when, as I have today, there's 180 parts to screw down with 2 screws in each...tried double sided tape but that left a bit of a mess and still needed one screw in each to be sure.

---

I found it is good way to mount when cutting lots of small parts. I use drywall screws + power screwdriver- quick and easy. Drywall screws have sharp threads and cross-groove which locks perfectly with screwdriver bits- no slippage thou. And obviously no need to worry to match mounting holes/grooves in machine bed.

http://astro.neutral.org/anodise.shtml - pretty useful website about DIY anodising.
Fabric dye not always work and fade with time. Commercial dye gives way better results
I must try it some day