Dan Maloney
Hey guys, I'm a bit early but here's a great starter on desktop EDM:
Hi Cooper, welcome to the show!
Finishing up some stuff but I will be ready for questions noon pacific time
Hi Dan!
aaaaand I'm back
OK, then, let's get started. Dan here, I'll be moderating today along with Dusan as we welcome Cooper Zurad to the Hack Chat for a discussion about Desktop EDM. And apologies in advance for the inevitable typos -- getting used to a new keyboard here.
Hi Copper -- want to get us going with a little bit about yourself?
I remember that video, good stuff.
Thanks, Dan.
See? Copper != Cooper.
Grrr...
Hi everyone! My name is Cooper Zurad. I have years of experience in electrochemical machining (ECM) and have recently been developing electrical discharge (EDM) systems. All of this work has been on desktop processes.
I am co-founder of a company called Rack Robotics. We produce affordable EDM power supplies meant to convert your 3D printer into an EDM machine.
We are close to shipping our first product, the Powercore. We launched this project on Kickstarter
a few months ago. Link here: [https://www.kickstarter.com/projects/rackrobotics/powercore-cut-through-solid-metal-with-edm](https://www.kickstarter.com/projects/rackrobotics/powercore-cut-through-solid-metal-with-edm?ref=user_menu)
If you want a quick rundown on how the Powercore works and how it can convert your 3D printer to an EDM machine, check out this great video by my co-founder, John:
Other than that, I’m open to any and all questions regarding desktop EDM and it’s future. Excited to be here!
How do you deal with electrode wear?
Electrode wear is an interesting problem
There are several ways to deal with this
In our current implementation we move the electrode downwards in the Z axis as we cut in the XY plane
this allows us to have a consistent kerf width in the process
I'm interested in using EDM for 'drilling' deep, precise holes in metal. The problem I see when using the powercore with a 3D printer, is that chips would be stuck in the hole. How do you deal with this?
To do so we've forked LaserWeb4 and added a custom post processor
(where deep > 100 mm)
Redfast, I would recommend remixing the EDM tool holder to allow thru flushing of dielectric
The files are on our printables page. Link incoming.
This would allow the waste to be removed consistently
Files + CAD here: https://www.printables.com/@RackRoboticsO_631255
Do you need a hollow electrode for that? Or will a standard brazing rod do the trick?
Yes, Dan, you'd need a hollow electrode.
Would an injection needle work? In effect it is a stainless steel tube.
The files are now setup for solid rods as we've focused on emulating the function of a fiber laser on desktop. But the CAD is available for remixing
Does that leave a "slug" inside the electrode>
Thomas, I would not recommend it. I've used those for ECM plunging, but you really want brass or copper for EDM
How do you 'mill' out the bottom of the hole with the hollow electrode?
Dan, yes it would leave a slug if it was a blind hole.
The edm electrodes for drilling are not just plain pipes.
Let me find and image for you Nyles. There are specialized electrodes for drilling blind holes
They usually have several smaller pipes inside and you can rotate them.
What happens if stainless is used? What's the drawback?
I assume that the electrodes are only in a limited amount of dimensions, how would you then go about drilling other dimensions?
Stainless is too high resistance generally
it will get hot
These are the blind hole electrodes
Cool, thanks.
I'd imagine you could bore the hole with a hollow electrode, then change to a solid electrode and remove the slug. Just a naive guess, though
Np. They are available on Aliexpress etc
Would your power supply be usable for wire EDM?
That's not a bad idea, Dan
Andy, I believe so. We have done preliminary testing and have had good results
Would it get too hot even if cooled with the dielectric used to flush the hole?
You need to be able to draw the wire at consistent speeds, which is an engineering challenge in of itself.
(LInuxCNC was modified a few years ago to allow negative feed-override, which reverses the path, specifically for wire-EDM purposes)
Thomas, I have tried doing EDM with steel electrodes and getting a consistent "burn" is hard. It really will not remove much
In the end you want 2 different materials. One that is electrically conductive and one that is less so
Brass electrodes and tubing are abundant and work very well for EDM. Having experimented with both, it is just easier to use the brass
I'd imagine there's a relationship between power needed and electrode size, especially for big tooling like for sinker EDM. Again, a naive assumption maybe, but it seems like there has to be a limit to what the power supply can handle. True?
Hi @Andy Pugh!
I think that sinker EDM is usually graphite. This is probably partly due to it being easy to machine.
Have you tried graphite?
Andy, the LinuxCNC thing is interesting. Marlin was also modified to do this
Yes we have tried graphite. It was initially the only thing we could use to cut steel
Graphite works very well for EDM
the only issue is its relatively high internal resistance compared to brass. It will get hot enough to melt our 3D prints and Wago connectors
Hello Copper when will the first units be shipping and when can we expect the open source release for the Power Core?
Cooper ** Sorry
Hi darkomenz! I remember you from the Discord ;)
Indeed hope your day has been going well.
I did some very rudimentary attempts to EDM copper-clad boards. Managed to get nice clean burn just through the copper on single spark. Can this be used as an alternative to milling circuitboards?
We expect to be shipping Kickstarter units before the end of July. Source will release when we begin shipping.
Just dropped off 700lbs of aluminum to the powder coater today
I do not believe EDM is a good choice for milling PCBs. Tool wear and inconsistent burn completeness may cause issues
also if you mill anything into isolation, it is no longer on the EDM circuit.
Something we didn't promise backers, but will be delivering, is a custom software for doing EDM with your Powercore. A fork of LaserWeb4 that will make it very easy to set up complex, multi path cuts.
I thought Gridfinity was released under a non-com license.
We are not selling Gridfinity
You print your own parts
oh
It is a great standard that we find very useful for organizing tools though. Love Zach
It will. But tool wear can be compensated (we do not have depth to go into, and a single spark of proper energy removes a nice circle of copper (tested with 0.1mm wire electrode and capacitor discharge) with both getting nicely through and not making a crater with melted-out walls. The suspicion I have is that it will be less troublesome than adjusting a mill. The toolpaths would of course have to be designed to not make insulated islands. (Side-to-side line-by-line raster is a bruteforce approach, could be done smarter.)
I am disapointed to find that the Kickstarter is US-only. I was actually about to press the button.
Sup boys. Hope you have a good day.
Will there be a schematic for the power supply?
Thanks Andy. We are expecting to be able to serve Europe and Canada starting around October if all goes well.
Happy to finally found some really knowledgeable guys to talk to about EDM.
Thomas, yep! We're going to be releasing everything as we get the KS units out the doors. We wanted to make sure our Backers got theirs before anyone else.
I built the XY platform for a wire-edm project, but stalled at the PSU stage.
How would you go about drilling deep holes with non-standard diameters that don't have an electrode for them?
That's definitely the hardest part of everything. The rest of the build can be very similar to 3D printers
Red, that's a good question. I suppose it comes down to your specific needs. You might turn a part to the correct OD.
If you ask me the hardest part is the software. Marlin, Klipper and LinuxCNC won't work very well out of the box.
@redfast00 Buy the next size up and machine down?
Otherwise you could probably drill a pilot and then orbit around to make the size correct
Yes, software is very difficult. We've been fortunate enough to be able to pay a professional to do our software.
Can a nonstandard electrode be made by electroforming copper into 3d-printed mould?
Marlin is something I don't think I will ever work with again if I can help it lol
I would not electroform it. You will not get enough current through that coating and it will quickly be worn away by the EDM process.
I agree I think the two large challenges in this space is a working pulsed power supply and software that can handle the specific needs of the edm process.
So what motion control do you want to use?
@gedm-dev LinuxCNC would need a HAL component to monitor the arc and tweak the feed-override. But that shouldn't be too hard.
(In fact, I have done it already)
Solid metal or graphite is the way to go
Electroforming, as massive millimeter-class thickness (and then peeling off the printed part that formed the outside constraints). Not mere electroplating.
gedm-dev, we are using a custom fork of LaserWeb4 for planning. Otherwise you can run our EDM implementation on a stock printer
@Andy Pugh That's only half the truth. Marlin and GRBL are inspired by LinuxCNC. They all use some kind of ringbuffer. A planner buffer to prepare instructions and a segment buffer for the execution.
no closed loop yet, but that hasn't been an issue for us very much.
If you need to retract you need to wait until the last block in the segment buffer is executed.
@gedm-dev LinuxCNC does not use any buffering. Trust me on that, I am the lead developer.
Dominik Meffert tried to make Marlin closed loop for his EDM work. He has had some success.
https://hackaday.io/project/181551-sinker-edm-machine
I have worked with him on getting his fork working, but it took a long time and was pretty unreliable for what we were doing
I putzed with Marlin's guts some time ago. The stepper control could have a retract-return mode that issues fixed rate of pulses to z-axis, fixed number, should be possible to shoehorn into the segment processing in the stepper.cpp part.
Maybe I'm outdated but wasn't there something on the linuxcnc page and why it will be hard to use for EDM?
I've seen people running automated retracts on LinuxCNC mills. This makes me thing it'd work pretty well for it
Userspace possibility. Break the paths to very short segments. Lots of overhead on the serial port. Much faster reactions.
I would be interested in trying to make it work on Klipper just because more people are familiar with it though.
@gedm-dev Maybe, we added the negative adaptive feed only a few years ago.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.