In general there are two ways of making circuit boards, either chemical or mechanical. The chemical way always includes masking the traces on the copper with a protective layer and then dissolving the exposed copper in a solution. The mechanical way is to mill through the copper.
The purpose of this project is to add another mechanical way that relies on the technique of electric discharge machining. The goal is to create a cheap way to make high quality circuit boards at home with a high success rate and low effort.
Today I tried to test the Circuit I build. I redesigned it to use only two instead of four secondary coils, halving the voltage to 90V. I also proceeded to adding the N-channel MOSFET as a low side switch to the circuit and pulsed it with a 40kHz 10% duty cycle square wave coming from a 555.
The result was rather disappointing, because I didn't get any spark in the water.
I think I have to read more articles regarding the topic an redesign the circuit.
Today I managed to encourage myself to begin building the power supply and not worry about blowing the fuses of our hackerspace again. The supply basically is right now a 2:1 transformer followed by a full bridge rectifier and a big capacitor. I think that the final design will have (a) switch(es) to set how many secondary coils in series shall be used.
Right now the circuit outputs about 180 V DC without a load. That's a multiple of the Voltage other peoples used in their design and I will see wehter I need to adapt to a lower voltage.
Last Friday the transformers for the test circuit arrived. I have not tested them jet. Right now I'm spending time with my friends and family. But as soon as the new semester begins (in a week) I will continue with the project.
The test circuit will mainly be the two transformers charging a big capacitor over a rectifier. I'm considering to build in a switch to choose between using 3 or 4 secondary coils of the transformers to have two voltage levels. The capacitor will discharge itself over the air gap (work piece and electrode). The discharge will be controlled by a MOSFET wich I intend in the end to also use not only as a switch for the discharge, but also as PWM control.
Right now still the problem of a short circuit exist - in the case of the work piece and the electrode touching and fusing together. The one book solved this by adding a power resistor that matched the voltage/current rating of the transformers.
I had the idea after I read this blog entry on Hackaday. After watching the video I did some research on other people who build EDMs. I found some papers regarding the topic and the PDF of the book that the blog entry is based on, but no concrete build and design instructions. So the first phase will be designing the spark circuit so that the work piece and electrodes don't fuse together and everything is ready to be put into a CNC machine. In theory we only need to remove single spots in a 35μm thick layer of copper, therefore we only need a relatively low power device. But we still need to break down the electrolyte and therefore 60 to 100 VDC. Also we must be able to control when to spark. The use of switches/relais is non-practical beacuse of the noise and the high probability of spark inside the switch and therefore power dissipation.
Looking good so far!