After building a 20kV power supply i decided to play a joke on some friends so i hooked the PSU up to a door handle but it didn't work that well, the whole door turned out to be painted in a sligthly conductive paint. What did happen though was that things started to electrostatically stick to the door.
I did some more testing and eventually built the final board seen here.
The idea is that papers are placed against a grounded plane inside an electrostatic field of about 200kV/m, in my case this is achieved by having a concrete wall (ground) with several wires carrying about 20kV running top to bottom about 10cm from the wall. This charges the paper with a strong electric charge causing them to stick to the wall. This is the same effect that lets you stick a balloon to a wall after rubbing it on your hair.
The power supply is built from a simple CCFL driver used in many TFT screens today, this particular one can handle quite a lot of current if properly cooled. I am running mine at about 8V and a maximum of 3A. To this is attached a Villard cascade voltage multiplier. This gives me a peak voltage of about 40kV. The high voltage output is then connected to the wires using a 60kV 30Mohm resistor for each wire to limit the current output which means the peak voltage on the wires is around 20kV with a peak current of around 1mA excluding the discharge current of the small capacitance in the wire.
One problem when playing with high voltage is that small capacitances starts to be quite noticeable, in an earlier test i tried having just one resistor for all the wires, but the stored charge became large enough to really hurt so that's why i switched to one resistor per wire. You will notice that i touch the wires several times in the video, it's equivalent to getting an electrostatic shock from walking on a thick rug.
A problem specific to this use of high voltage is that dust sticks to the board, so it gets quite dirty in a fairly short time. Another problem is that air becomes more conductive with higher moisture, at about 50% relative humidity this board stops working so it's not that good during the summer.
To power the CCFL driver i use a 12V 3.3A switching PSU, and a current limited buck regulator(Eagle CAD files) to limit the current and voltage to a maximum of 3A and a maximum of 8V.
Please excuse my bad english in the video, i don't get enough practice.
Even though this should be completely safe, i will not take any responsibility for any harm caused to anyone attempting to recreate this. The above should only be attempted in a well ventilated area to prevent build up of ozone. Also while the current in the wires is limited by the resistors, the current directly out from the voltage multiplier could be quite high.