This table started out with the purchase of the famous (or infamous?) JD's Garage Budget Plasma Cutting Table plans. Once I studied the plans, I quickly realized that a 24x32 table just wasn't going to do me any good -- I own a heavy-haul trucking company, and have to cut steel for repairs WAY bigger than that.
So, following the hints in the plans about enlarging the size of the table, I built a 50x54 inch frame and used heavy-wall 2x2 steel tubing for everything but the gantry -- that was 2x2 aluminum tube. I bought the "hard parts" from JD's Garage, and it didn't take me long to have a functional 3-axis plasma cutting table with floating Z-axis using an Ardino Uno & the stepper drivers suggested in the plans (DM541, TB6600).
One of the first things to happen was that, at the high temperatures that close to the torch, the 3D printed plastic Z-axis floating head melted... so I was down for 2 weeks waiting for parts from JD's Garage. While waiting for the parts, I took the opportunity raise the Y-axis rails so I could keep the torch a bit further from the Z-axis (my torches are all PowerCutter IPTM80 or PE85C shielded).
So... parts arrive, Z-axis is rebuilt, and -- it starts melting again! Finally, sick of the melt-downs, I bought a RattmMotor 100mm actuator on Amazon. That cured the melt-down problem.
I wasn't very impressed with the speed, or to be perfectly honest, the accuracy at that point. Being one to want the best out of something I built, I began looking for something faster & more capable. I read several places where the Arduino Mega 2560 was more capable, and that GRBL for the Mega (in it's different forks) was capable of a dual Y-axis and gantry auto-squaring, so I built one -- first with gnea/grbl-mega then with Grbl-5X -- but I could never realize the speed increase or the gantry auto-squaring.
I still wasn't happy with the speed of the basic mechanical machine, or the alignment of the axis carriers. That was when I found the OpenBuilds Parts Store... my wife wishes I hadn't...
I set about ordering enough C-Beam, wheels, gantry plates, screws, etc to completely rebuild the upper part of my machine; I also decided at that time to switch to ball screws instead of the cheesy belt drive system from JD's plans. I also built my own floating Z-axis -- out of aluminum -- and all of this really improved the speed and accuracy of the basic mechanics.
Unfortunately, I had not been able to implement a torch height controller that actually worked; the closest I came was modifying Hale's "Stand Alone THC" from another build here into an interrupt-driven THC that worked with the Arduino Megs. Unfortunately, the Mega simply wasn't up to the job and decided to dive into the plate I was cutting, wrecking the shield cap & consumables on my PE85C torch (those are expensive, too!!)
All in all, I just wasn't happy with the entire Arduino controller system; too limited in power, and the G/M codes it supports are just a sub-set of the NIST code definitions.
About 2 weeks ago, I had been reading a lot about EMC2 (LinuxCNC) -- it seemed to offer everything I was looking for. So, since I had some old PC's in my "closet", I decided to give it a whirl. Man, I wish I had done it sooner!
I bought 2 SainSmart ST-v3 CNC parallel breakout boards and a PCIe parallel card for about $60, and splurged on a larger steel enclosure -- the old Arduino system's box was a bit too crowded for my taste.
While I was waiting on them, I installed Debian 12 Linux on an old Dell Optiplex 755 (2.8 gHz Core2Duo w/ 4 gb of RAM) and installed LinuxCNC with the RTAI real-time kernel and started familiarizing myself with the software. Having not even looked at code since my college days (no, you don't want to know how long ago that was...) it took me a few days to get my "logic system" in my head back on line.
Don't get me wrong -- I sing the praises of LinuxCNC for what it is capable of, not the effort required to get there. While there are utilities like StepConf and Pncconf wizards, they only help you build a barely functional machine; you then have to do a great deal of coding yourself if you want a very capable and versatile CNC machine. Luckily, since LinuxCNC has been around for so long, there is a wealth of forums, Wiki's, and Guru's around on the web who are very willing to help out a "newbie".
At this point, my machine has a dual powered Y-axis, Y-axis gantry squaring, dual proximity home & limit switches on all 4 physical axis' (X, Y, TandemY, and Z), a torch height controller that actually works (directly through LinuxCNC), and all without the use of Mesa cards.
Another thing about LinuxCNC is that the latest incarnation for plasma, QtPlasmaC, does away with those cobbled together macros for initial height sensing -- just set your float travel, how far above the work you want the torch to slow to "probe speed", your pierce height & cut height, and it is automatic whenever your program issues the M3 command.
Arduino Mega to EMC2 (LinuxCNC) Plasma Table Conversion
Build in 'Plasma Cutter Builds' published by cnc-trucker, Jul 29, 2024.
This is a story about a guy who started with a $25 set of plans, and through 4 iterations/ rebuilds, ended up with a professional quality CNC plasma table capable of cutting anything steel up to 2 inches thick.
-
-
-
- Build Progress:
-
- Build Complete!
-
Special Notes
If anyone wants to try their hand at LinuxCNC without spending a lot of time coding, I have posted my complete build at GitHub - ccwtruck/LinuxCNC-4x4-PlasmaTable: Complete working LinuxCNC 54" x 60" Plasma Table Control Software With Arduino Mega 2560 THC
CatiaDr, JustinTime and Mark Carew like this. -
-
Build Author cnc-trucker, Find all builds by cnc-trucker
-
- Loading...
-
Build Details
- Build License:
-
- GNU (GPL3+) General Public Licence
Reason for this Build
After we purchased a wrecked 60-ton 2+2+2 heavy-haul trailer, we needed to cut some heavy steel parts. Cost to have them cut commercially was prohibitive, so we decided to go the DIY route.Inspired by
Too many to list here! -
Attached Files:
-
-
Attached Files:
-