Cooper explains that they’ve prepared a special fork of the LaserWeb4 for use with EDM that will constantly lower the Z height at the proper rate to counteract the electrode itself becoming shorter over time.
@easytarget @cprezzi have they been in touch about this?
No, not that I have noticed anyway.
That said, this isn’t the first time I’ve seen an article about people doing EDM with 3d printers, I saw an article with a similar setup recently (might have been the same one). An interesting project.
I’m guessing they modified lw-comm.server rather than the LaserWeb app itself. That’s what I’d do anyway; it’s easier to add an ever increasing Z offset there than having to play with the gcode generator(s!) in the main webapp.
Wouldn’t you want to add to Z only while actually in contact with the metal being cut?
I just skimmed the HackChat, lots of interesting discussion there and a classic ‘expert’ trying to prove it’ss a scam, somehow.
Anyway, there is some discussion of firmwares etc, and the announcement that there will be a LW4 fork for this (targeting Marlin, presumably). I’m OK with that so long as they publish their mods, and retain credits etc.
I’ll look further at this when at home and I have a desktop for easier reading. I’d like to touch base with them since, apart from anything else, I’m curious about exactly how they do it.
Part of the ‘cleverness’ of this is that you do not need to modify the Ender’s firmware, they apply the ‘wear rate’ calculation to the gcode instead, which seems pretty sensible to me, beter than requiring a firmware mod.
Yes, I’m imagining starting with mill cut path (that will do Z lift between cuts) coupled to a post processor that adds a z offset continually while doing G1 cuts but not otherwise.
This could be done as part of the path generation itself, but more likely as a post-processing operation, either in LW or in lw-comms.server (which is, imho, easier to implement. The path generators are complex, fragmented and not nice to work with).
I have implemented automatic Z height adjustment with a modified Smoothieware firmware, a PLC and a special circuitry board. It is very machine specific though as the laser cutting heads needs to support this. I guess with the EDM it may even be easier.
Hi. I’m far from being an expert. But there are many things they at least wanted people to think they can do while knowing that they in reality won’t be able to do. Let’s look at it a little:
First: The title of the kickstarter was “Cut solid metal”
Ok… Aluminum is metal. But are thin sheets real solids?
Second: Use your 3D printer with stock firmware.
Tricky. Works for aluminum sheets but steel is a different story. Sometimes you need cutting speeds at 1mm/minor below. Marlins default minimum feed rate is 0.05mm/s = 3mm/min if I’m not wrong.
Third: They mentioned on their discord that they did tests up to 50khz. Turns out the product is rated 2khz.
Fourth: They claim that “first short tests” indicate that it will work for wire EDM.
2khz produces heavy sparks. On my build even 5khz@60v with low duty cycles instantly wants to melt electrode to the work piece. I won’t even mess around with 2khz and doubt that it is suitable for wire EDM. May work with a very very thick wire diameter.
Sixth: The marlin problem:
Marlin uses a planner and segment buffer with acceleration and deacceleration. You can’t just stop and retract an axis. At least the steps remaining in the segment buffer have to finish. Even if you use a feedbackloop and detect a short circuit the axis will keep moving until the current segment buffer block is finished. After working with a custom GRBL fork this turned out to be a massive problem.
Number seven: I’ve read from multiple people how happy they are to get one and how excited they are to cut steel with it and I feel a little bad for them. Infos like “we think it is possible” or “a short test indicates that it can do X” dafuq? Didn’t they used their product before selling it and if they had success wouldn’t they share it? I gave them tips how to improve their cutting process. After some talk they tried it and announced that they made progress with titanium or whatever material it was. At some point they shared an image of a steel cut. It was visible what a pain this cut was and that they had to manually adjust the electrode like 100 times.
The new video indicates the product sold is the version 1 (V1). To me it sounds like they already want to move to version 2. So how good can it be?
But let’s wait for some reviews. The boxes are delivered now and reviews from happy customers should come soon. I personally haven’t seen any happy owners yet.
There are other things but time is precious. I just wanted to add this small opinion of mine to the thread.
Have a nice day.
At least the source code is on Github now (GitHub - Rack-Robotics/EDMWeb: EDM CAM software for generating EDM gcode).
Seems he added a new gcode generator and change the css to dark mode ( 
I like).
We probably should try to get this guy aboard the LW Dev team (as he is in the code already) 