Water cooling a laptop - making the water block.

Since I'm studying at a uni away from home that means I'm away from my gaming PC. For games in the uni dorm I have a SONY VAIO laptop that I found at the dumpster. With an i5-480m, a HD5650 and 8 gigs of RAM the laptop is usable for many games at low to medium settings. The cooling in the laptop is also very good. With max load neither my CPU nor GPU go over 75°C. Knowing that I still decided to water cool this thing. Why you ask? Just for kicks, to be honest. There are a few people who have water cooled their laptops but their implementation was quite mediocre in terms of implementation, performance as well as looks.
The first thing I made for my super custom water cooling loop was the water block for my CPU. As luck would have it I found two bits of 0.8mm copper sheet at the local flea market, which made this project possible.

For starters I took the i3-370m processor that originally came in the laptop and started making measurements along with sketches for my water block.

My first idea was to make this a straight in line job. I ended up lengthening this design by 10mm to clear the capacitors and went ahead with making it.
I glued a piece of squared paper to my copper plate to aid with markings, marked it and cut two identical pieces out with my angle grinder.

The angle grinder made quick work of the cut but left a bit of a burr, which was pretty easily taken care of with the help of a file.
To cut the corners I used a pair of these long handle snips.

I then cleaned up both pieces with some 600 grit wet&dry and made marks for what I'll be doing next.

Both of these pieces would make the top and bottom of my water block. Of course, if I kept the inner chamber smooth I wouldn't get much performance due to low surface area, for that I have come up with a neat solution.
You see I want to add water cooling without removing the built in cooling. For that I'll have to make my water block thin enough to fit between my CPU and the heat spreader. This is where my neat solution came into play. To get more surface area I'll be soldering bits of 1mm copper wire every 2mm to the copper plates. If you don't get what I'm saying just stick around and you'll see what I mean.

 Both copper plates will be held together by the copper wire that I'll be soldering to them. This copper wire will also act as a thermal interface between the CPU and the factory heat spreader.
To prepare my copper wire I first stripped it with some 150 grit sandpaper.

And I then cut it into 20mm long pieces.

Next step was to solder the bits of wire to one of my copper plates. This was easier said than done. Because the copper piece is rather big it had to be kept preheated with my hot air station set at lowest air speed at 180°C. After some patient soldering the bits of wire were soldered in one by one onto my copper plate.

The other copper plate was tinned with some solder as reserve for when I'll be soldering it to the bottom plate.

And thus both copper pieces were ready to be soldered together.

To solder them together I made a contraption out of two heatsinks, some cardboard, bits of toilet paper and a clamp. The heatsinks were to keep the plates straight, and toilet paper was there to keep my copper from being marked by the surface of the heatsinks. Cardboard was there to act as a spring to maintain the pressure during soldering.

The whole shebang was heated with a big butane torch until the solder melted and the assembly was cooled down to get my final product.

The "core" of my water block was done and was quite thin coming up at 2.53mm thick.

Since there ended up being a few channels blocked by solder I drilled through all of them with a 1mm drill.

The surfaces of my water block were sanded flat by gluing a piece of 600 grit sandpaper glued to a heatspreader from a PS3. Judging by the surface finish on the heat spreader it was fly cut so it should be reasonably flat.
First rub revealed some unevenness in the surface of my water block. This was quite minor and easily removed with some more sanding.

After I sanded the surface flat both sides were soldered shut with the help of some more copper wire and the water block was given a final sand. After that I started cutting out channels where my brass tubes will be soldered to.

To this I soldered a bit of 6mm OD brass tube that I got from a telescopic antenna.

This here is where I stopped because when I cut the hole in the other end I realized that not only am I getting too much restriction but the tube does not fit so I scrapped this water block and went to make a new and improved one.
When I went to cut out the pieces for my new water block I found out that my angle grinder had gone cactus so I had to cut the copper out by hand. I found a new hacksaw blade but couldn't find a hacksaw so I had to cut the copper out by holding the blade with my bare hands, which was not pleasant to say the least.

Just as before I cut and soldered bits of copper wire to the plate, this time the bits were 25mm long with copper wire running along the sides.

The top and bottom copper plates were soldered together by the same method as before.

 

There were some channels that got blocked by solder but unfortunately while drilling them out my drill bit broke inside so the water block had to be separated to rework it a bit. Minimal damage was done so the water block was quickly reassembled and given a hot acetone bath.

A vibrator motor was taped to the site to keep things agitated. One thing I don't like about cleaning flux out with acetone is that there is always this hard white residue left and it's a pain to get out since neither water nor acetone would get rid of it, I even tried vinegar to no avail.
After I finally cleaned the whole assembly out I sanded the faces flat and prepared some brass tube for soldering by filing the chrome plating away with a diamond file and making a slit with a pair of scissors.

I made two tubes in total this way and they were both soldered to the open sides of the water block core, both facing in opposite directions to avoid bumping into components on the motherboard.

To plug the open ends I cut out two small bits of copper to be soldered to the ends. The copper end caps were cut out with a pair of long handle snips, which left a sort of steep knife-like edge, which made positioning the copper end caps on the end of the brass tubes much easier.

The rest of the flux was cleaned off with acetone. Oh and before you ask I'm using acetone because alcohol is not readily available and is a pain to get a hold of.
After the water block was cleaned and dried it was finished.

This time the water block turned out even thinner at 2.42mm.

And here it is next to the i3-370m and the previous water block attempt.

The flow in this one feels much better. One thing I'm not too sure about is corrosion. Since I have solder and copper in water I might get some galvanic corrosion happening. To help against that I will be filling my loop with some red antifreeze. As for the water pump - that will be coming next. I have a few water pumps ordered but I can't be bothered to wait for them so for the time being I'll be making one. As for the radiator - I already have one, and believe me this beast will give a whole new meaning to the word "overkill".

Now for some quick calculations. The pump I've ordered says it's a 80-100l/h so let's say that with all of the restrictions in the loop we get 36l/h. That comes out to 1ml/s. Since it takes 1 joule to raise the temperature of one milliliter of water by 1 degree Celsius and the fact that we have 1ml/s flow lands the speculated theoretical performance of this water cooled loop smack bang at 1°C/W. So I can expect my temps to go down from 70-75°C to around 50-45°C full load.

Well, that's it for now, thanks for reading.