The Problem

I have been experimenting with conductive ink recipes, and failing monumentally. I washed all my failures into a bottle (sounds like a blues song! - care to write it?) with the intent of recovering the graphite at some point and throwing out the excess binder.

The solution

It eventually occurred to me that what I needed was a centrifuge. After looking at several designs, I decided to cut the bottom off a plastic bottle and try to use that. It actually worked out extremely well and I'm very pleased with it.

Supplies

PET drinks bottle. It needs to have an odd number, or multiple of an odd number, of bumps on the bottom, as spinning an odd number of tubes should have less trouble with vibration. I used a flavoured water brand one, with ten bumps.
DC motor. I used one from an old printer
Sponge packaging
PSU for motor
Scissors
Sharp knife
Soldering iron or other plastic-penetrating device.

Step 1: Preparation and Assembly

This centrifuge is very easy to make, as follows:

Use scissors to cut the bottom off the bottle. Leave it about an inch tall and trim it reasonably level. This will become the rotor.

Use your soldering iron to carefully melt through the "pip" in the centre of it, to make a hole which is a tight fit on the motor spindle. Go from both sides to reduce errors. Fit it to the motor and run it to ensure it's central.

Melt holes through the curved part of each bump. You want the tubes to stick out sideways, but to go down at bit as well. Make sure you get an odd number of holes. I used a bottle with ten bumps, so melted through alternate ones, giving me five holes.

I pushed the barrel of the soldering iron through each one, then opened it out further with the bit. Patience gives a nice rounded hole, try to rush and it will be uneven. Get them as level with each other as possible.

The motor I used had wires already attached. If yours doesn't, connect some wires to it before doing more.

I used two blocks of sponge packaging saved long ago from something, finally they came in useful!

I cut a large hole in the bigger piece, a bit smaller than the motor. Push the motor into this spindle first, from underneath, which will give it more support than the other way.. The other piece has a smaller hole in it to prevent the other end of the spindle rubbing against it.

I used a hole-saw to make the holes, but a long thin knife would work just as well.

Bring the wires out to one side, and glue the blocks together.

Fit the rotor onto the motor spindle, and power the motor to ensure it runs true. I found it worked fine with friction alone holding the rotor for the first few runs, but it became looser with use. I eventually fitted a bit of rubber to the spindle and glued the rotor to it.

Step 2: Operation and Caveats

Juicifuge works really well, but has some natural limitations:

Fill some centrifuge tubes with whatever you want to separate. You need to load all the slots if you want to run it at any significant speed.

Juicifuge is very light, so you need to place some objects to stop it walking away. A better plan would be to glue it something heavy.

Try it at low voltage first, to check the balance. Increase the voltage until you get to either maximum speed, or the highest speed at which it's stable. I deliberately used a slower motor for mine. It's running happily at around 3400 RPM, at about 24V.

Turn off the power, check the tubes!

Caveats:

This is obviously a very quick and simple design. The rotor doesn't have any significant mass, so it can't run unbalanced except at low speed. If you have only one sample, I suggest you put the same amount of water in tubes to put in the other holes.

It's very light and doesn't have anything to hold it down, consequently it likes to wander around when it's running. It's a very explorative contraption, so needs to be fenced in.

For the sake of simplicity, I didn't include a bowl. I didn't have any problems over the course of 20 to 30 runs, but any drips on the rotor potentially could make an interesting design on your wall...