Electrostatic discharge machining for nozzles

Hi Everyone,
I had a problem yesterday printing a really tiny part and I figured out the problem was the nozzle diameter. I have some 0.3mm drills but i'm sure theres a limit to how small you can realistically go with a twist drill. I was wondering what would happen using something like eletrostatic discharge machining. According to wikipedia it can leave a very smooth side. My idea was to use a piece of wire as an electrode and use it to erode a hole through an acorn nut using a spark generator (maybe something could be rigged together using a large inductor and some high voltage mosfets). It might be possible to get a smooth 0.15mm nozzle if I could get hold of 0.1mm wire. Obviously there'd be problems cleaning such a small nozzle and it would print really slowly so only useful for very small parts.
Has anyone tried this?
Pete

Wow that is a cool idea a lot of the electronics is there already.

How about a Wire EDM Reprap maybe a cheap way to get into a home machine shop..

Humm Reprap Prusa EDM

Quote

maybe something could be rigged together using a large inductor and some high voltage mosfets

To get decent surface quality for such tiny items - the lower the voltage, the smoother the surface - the voltage required isn't exactly high. Something like 1 V to 12 V, so chances are good you get away with the usual PWM heater output.

The biggest problem is to get rid of the produced waste, it quickly clogs your just-drilled hole. Usual mean to solve this is to press (distilled) water through a hollow drill (tube instead of wire) into the hole.

Also I don't think inner surface quality of a nozzle is that important as long as it's reasonable straight.

Here you can learn how this stuff works: [nptel.iitm.ac.in] It's part 39 of the course, "Electro - Discharge Machining"

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I was only thinking about nozzle design but I could see it working for general hole drilling. The main problem I've read about is electrode erosion as the spark erodes on both sides. Blind holes would be difficult unless you can find a good way to measure depth and replacing electrodes would be frustrating.
Circuits I've seen are based on ramp generators but they need some form of detection circuitry to detect a short circuit. That could be pretty easy to do with a MCU. The process would be to drop the z axis a notch at a time and keep driving sparks to create a hole. If you held the workpiece above the base, a noticeable increase in the voltage for a spark may indicate the electrode has gone all the way through. A hypodermic needle may make a good electrode, though the high resitance of stainless may cause issues (little current involved so might be helpful as ballast instead). I think its worth a try.

p

Quote

The main problem I've read about is electrode erosion as the spark erodes on both sides. Blind holes would be difficult unless you can find a good way to measure depth and replacing electrodes would be frustrating.

Yes, the electrode erodes, too. But in a predictable way. So, if you want to have a hole 1 mm deep, you have to forward the electrode 1.5 mm or something like that. How much exactly is the result of tests with that electrode material/workpiece material pair.

To find the distance between workpiece and and electrode one usually measures current instead of voltage. Zero current means: hole is done.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

This has completely digressed from drilling a small smooth hole in a nozzle so probably in the wrong forum, but it could be good for milling pcbs as it would only remove the copper.
If its predictable, what factors would affect it? The ideal case for me would be low erosion of a stainless steel electrode against a copper substrate as we could use a blunt syringe needle as an electrode. If we clamped a big ground plate to one end of the pcb and eroded the tracks out in a raster pattern from the other we could get a nice smooth pcb but it would be very slow. The process wouldn't damage the substrate so you'd get a very clean removal.
A flat tray with a pcb fixed to the bottom and filled with distilled water (or kerosene) would work and pumping water down the tip would help remove the debris. I might be able to chuck together a prototype with an old take away tray.
p

Quote

If its predictable, what factors would affect it?

You didn't watch the video I linked to, did you?

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Your right, I didn't watch it, hopefully tonight! From other sources, I've seen that the erosion is due to a number of potential factors. There would be

• Localised resistive heating on either side of the spark which would be proportional to the resistance of the material. High conductivity in the electrode would counter this.
• Heat erosion due to the plasma created by the spark. High melting point might help a bit, so my solder coated crimp pins are probably not the best!
• Possibly some mechanical erosion when the plasma collapses (possibly similar to cavitation). Harder electrode material should help

The wikipedia page mentions tungsten being used and also says that brass erodes more than copper.

I have some copper clad board and some candidates for electrodes (crimp pin terminals) to play with and I'll try connecting an RC circuit to the heater output. A suitably sized resistor in place of the thermistor could trick the board to keeping the extruder heater on and save reprogramming the board every time I want to test.

p