G-EDM

G-EDM EVOIII

The new EVOIII board provides a single PCB containing the pulse stage and the sensing stage. It replaces the EVOII pulseboard and the EVOII sensorboard.

The EVOII Cube is a cool looking piece of Art but it is hard to manufacture. To reduce build costs and make it easier to produce the EVOIII PCB was designed. Specs and performance of both versions is the same. There is one difference: The EVOIII uses a single gate driver IC while the EVOII cube has a discrete gate driver circuit. At higher frequencies they don't switch identically and the cube needs a slightly higher duty cycle.

The board provides two isolated analog feedbacks and also isolates the PWM to drive the mosfet gates.

Here are the links to order the boards directly and build them on your own.

All three boards are needed to build a G-EDM.

Pulseboard:

https://www.pcbway.com/project/shareproject/G_EDM_Low_budget_DIY_Wire_EDM_machine_pulseboard_8975f488.html

Sensorboard:

https://www.pcbway.com/project/shareproject/G_EDM_Low_budget_DIY_Wire_EDM_machine_sensorboard_24b74366.html

Motionboard:

https://www.pcbway.com/project/shareproject/G_EDM_Low_budget_DIY_Wire_EDM_machine_34e1e043.html

The project is almost finished and so far everything works acceptable.

The new concept was able to cut 30mm steel and also 2mm stainless steel.

PCBway helped a lot developing the PCBs and I think they deserve a little fan art for their great work.

The total cut took three hours. This is very slow and based on the feedback shown on the UI scope it should be possible to adjust the settings to make it in 1 hour.

The settings file is the same that made the 30mm cut and seems bulletproof. Not a single issue on this cut except one related to the mechanics.

The X axis belt walked off the idler bearings and finally fully popped off the pulley. That happened at the bottom of the W and from there on the X position is a little off. Shit happens.
I'm doing the recordings with a smartphone and had to swap the clips to the PC multiple times. That is why there are cuts.

Next cut will hunt the max speed.

And a look at the G-EDM EVOII boards. In total there are three PCBs. One to switch the sparkpower, one with isolated dual channel Feedback and one controlboard that runs with an ESP32..

The EVOII cube (pulseboard and sensorboard)

The Controlboard:

The final boss is tamed.

Cutting a 30mm steel block with a cheap DIY wire edm machine is a quest that almost never ends. Almost.

Fully redesigned the PCBs. The boards now have JST connectors to easy connect them. Wiring the display was the biggest pain in the past. Now it is plug and play.

The ESP board has a few more pins exposed and provides an optoisolated RX/TX UArt interface with a 3.3v to 5v level shifter to control the DPM. Enable/Disable it, change voltage and current etc.

The pulseboard provides bus voltage and high side current sensing. Also fully isolated with linear optocouplers.

The first wire extension was very specific for the G-EDM tower. It worked very well but with the new router it is time for a new module and add all the improvements possible.

Access to the wire will be easier. The bearings on the spring loaded tensioners are fully exposed to the operator.

Power feed contact is now a real EDM contact that can be shifted to the side to get a new position and it is also spring loaded. This should increase the lifetime a lot.

The internal routing of the power wire and flush hose will end in connectors. A banana jack on top of the module and a push in connector for the water hose next to it.

The flushing nozzle are threaded now and the upper guide has a grip to easily adjust it.

Not finished yet but it starts to get somewhere: