CNC Plasma Table

Plasma Cutter + CNC Table + Auto Load/Unload = Tool to build other projects faster

Similar projects worth following
Overall Goal: CNC plasma table that is auto loaded/unloaded through network interface.

Having a plasma table to cut metal parts would be nice. Loading the plasma table with 4'x8' or 4'x10' sheets can get awkward. Having this auto-loaded with a click of a button would be ideal. Industry has loader/unloaders either in sheet form or in roll form so the concept is there. This project is to focus on developing the software side of the loading and unloading. To get to that development stage, the mechanics will have to be hammered out:

Plasma Table
Sheet rack

The process of the mechanical build will keep the software side in mind, ie the I/O's, motor controls, etc. Then the development stage of the software for the loader/unloader can be addressed:

Expandable to multiple material/sizes
Interface with Plasma Table
Interface with Computer that will "print" the cut

Links to Logs-

Phase I - Build a Plasma Table

Progress - Table is built and functioning. Hard limit randomly faults when enabled. Axis loses steps. GRBL/Universal GCode Sender appears to start/stop movement in a cut with multiple short line segments.

Controller - GRBL running on UNO

Interface - Universal GCode Sender running on Raspberry Pi

Mechanical - Many designs out there. Pick one and wing it.

Plasma Cutter - Use one on hand. PakMaster

Phase II - Build Material Rack

1-29-20 update.  This has been put on back burner.

Progress - Not Started. Main rack being built with loader/unloader in mind. Not developed, trying to keep open build concept. 

Mechanical - Stackable rack that is expandable to hold more sizes of to hold multiple sheets of material. Design for 4'x8' 4'x10' sheets. Allow to be expand to multiple sheet size as needed.

Phase III - Build Material loader/unloader

1-29-20 update.  As much as I would like this to become part of the setup, space has become an issue and this is on the back burner until room is available again.

Progress - Not Started. Building main table that will have expansion/modification ability as the project evolves.

Controller - Use the Raspberry Pi IO or capable of interfacing with the Raspberry Pi.

Mechanical - Be able to slide material out from rack, lower raise material to plasma table height, slide under plasma table. Reverse order for uncut material unload. Finished parts will fall down when trap door under plasma table hinges downward. The finished cut *should* drop down if fully cut. Then the uncut material, if completely exhausted, will be cut from the main sheet. Trap door to drop exhausted material down same path as finished part.

Phase IV - Develop/Find network capable "cutting" queue to send file from office computer to plasma table.

Progress - Not Started

  • 3 × 452 oz in steppers Bipolar bought off of ebay
  • 1 × ?? oz in stepper Z axis motor pulled from some copy machine or printer at some point
  • 99 × Miscellaneous bearings These were pulled from copy machines and printers through the years.
  • 3 × A 6A55M072NF1512 72 teeth pulley x and y axis drive
  • 6 × A 6A55M024DF1510 24 teeth pulley x and y axis drive

View all 20 components

  • Project with plasma cutter

    willbaden04/19/2020 at 13:22 0 comments

    It is always nice to see someone build a project, but I would like to see what kinds of projects come of those builds.  So to keep with that theme, here is another example of what this plasma table has been used for:

    Meshing Fusion with this plasma table is definitely worth doing.  Previously, I had been using CamBam, which is a great software, but was limiting in that it would be time consuming to put together 3D sheet metal projects.

    With Fusion, I can do things like this house and the fighter jet I had done prior at a faster pace. . . faster used loosely for someone who has other time commitments. . . as I just cleaned up a kiddos spilled milk . . . 

    Things that I am seeing would be nice:

    1. Torch height control - I haven't experienced one, but I gather from what other people have shown on the internet, one would be nice.  You can see glimpses in the above video of me reaching down to bring up the torch.  If the torch is left to drag on the sheet metal, the arc will extinguish.  The last cut in the video, before the pan across the completed cuts (time 0:41), was due to a restart required on that piece.  The torch drug across the material and went out.  Once it goes out, it doesn't re-fire until the torch fire signal is released and re triggered. 

    2. Water Table - plasma cutting gets smoky.  A home built filtration setup is used.  Built from an old blower fan from a house furnace with a filter that my work was throwing away.  It works, but the particles kicked up by the cutter doesn't allow the filter to keep up.  Along with better air quality, it would be nice to have something cool the parts being cut to prevent as much warping.  

    I have a few more projects that I would like to utilize the plasma table with and will plan to post those as they get completed.

  • Its been a while

    willbaden01/30/2020 at 01:02 0 comments

    Its been a while since I last typed up a log on this project.  The plasma table is still cutting and the latest cut can be viewed below:

    I haven't really made many updates to the machine.  The remaining two plastic ball bearing guides for the floating head have been replaced with aluminum ones.  The plastic ones were allowing the head to float left, right, forward and backwards too much.  after replacing, it had less play.

    The plasma cutter supply also had a shelf made to allow it to stay mounted on the table itself.  This helped out with some room around the table.

    Lastly, the torch cable got something to hang it from.  Cable management was an issue keeping it out of the cut and off of the material.  A 1x1" square tube was cut to length.  A simple swing hook was placed up top with the cable zip tied to it. The 1x1" tube was then welded to a round tube that seated in a larger tube on the bottom. This allows the hanger to swivel around as needed with the cable.

    To make the bend near the top, multiple cuts were made through to the other side.  The tube was then bent and welded.  

    So far so good. 

    As for the missing steps issue. . . it has not gone so well.  Depending on the wear on the tip, the plasma table can jump/miss steps when firing the arc.  Recently, I switched over to Fusion 360 jumping on the free software bandwagon.  Which turned in to me subscribing to the yearly subscription fee, but that is a side note.  The switch to Fusion was a good decision since there is more control on the post processor.  I modified the grbl post processor and added a homing sequence prior to the first cut.

      // absolute coordinates
      if (properties.useZ) {
        writeBlock(gAbsIncModal.format(90), conditional(properties.useZ, gPlaneModal.format(17)));
      } else {
      zOutput.reset() ;
      //reset any z axis offsets
      writeBlock("G92.1") ;
      writeBlock("G90G0G59" ) ;
      writeBlock("X0.Y0.Z0." ) ;
      //Add dwell to prevent the homing cycle from stopping the x0y0z0 move
      writeBlock("G4P.1") ;
      //Home Machine to start with fresh zeroes
      writeBlock("$H") ;
      //Set G59 location to remember the shortest location to home 
      writeBlock("(Set G59 X0 Y0 Z0 Location)") ;
      writeBlock("G10L20P6X0.Y0.Z0.")  ;
      //Set G54 Z height to speed up plasma cutting height detect
      writeBlock("(Set G54 Z0 Location)") ;
      writeBlock("G10L20P1", zFormat.format(properties.homeZOffset))  ;
      //Move to G54 Z zero to save time on the first cut
      writeBlock("G0G54" ) ;
      currentHomeCutCount = properties.numberCutsHome ;

    Then an option was added to home the plasma table after a set amount of cuts:

          if (properties.useHoming){
    		if (currentHomeCutCount <= 0 ){
    		  //home machine 
    		  currentHomeCutCount = properties.numberCutsHome ;
    		  isHoming = true ;
    		} else {
    		  currentHomeCutCount -= 1 ;  
          if (isHoming){
    		//home machine 
    		writeBlock("G0G59" ) ;
    		writeBlock("X0.Y0.Z0.") ;
    		//Add dwell to prevent the homing cycle from stopping the x0y0z0 move
    		writeBlock("G4P.1") ;
    		writeBlock("$H") ;		  
    	    writeBlock("G0G54" + xOutput.format(getCurrentPosition().x), yOutput.format(getCurrentPosition().y), zFormat.format(0)) ;

    The video above shows this homing between cuts on some of the pieces for the jet.

    Along with the homing additions, the torch tip touch off was added prior to every cut: 

    	  writeBlock("G38.2Z-300.F1000") ;
    	  writeBlock("G92Z-19.") ;
    	  writeBlock("G90G0Z.5") ;
    	  writeBlock("M03S10000") ;
    	  if (isHoming || isFirstTime){
    	    writeBlock("G4",pFormat.format(properties.homingDwell)) ; 
    	  }else {
    		writeBlock("G4",pFormat.format(properties.normalDwell)) ;
    	  isFirstTime=false ;
    	  writeBlock("G1 Z-0.05") ;	

     All in all, the plasma table is coming together being a good hobby tool.  I have gone down the route of production yard art, but find that without torch height control, a water table or a more powerful plasma cutter, it will...

    Read more »

  • Clean Cuts Come Hard

    willbaden02/25/2017 at 02:30 0 comments

    As there have been a few cuts under the plasma tables belts, it is time to get the cutting experiences documented.

    Speed changes (I think, it was a while ago that these pictures were taken)

    Go too slow and overcut happens:

    Go a faster speed and things look better:

    Dragging a torch along flat steel causes the arc to go out. The loss of arc path is highlighted in yellow:

    Having a non perpendicular head can cause splatter/incomplete cut in certain directions:

    To be continued. . .

  • Crashing Comes Easy

    willbaden02/25/2017 at 02:14 0 comments

    Using a floating head on corrugated tin seemed it was the ticket. Especially when torch height control is still not implemented. Touch off of the material, back off the material to allow an arc to fire, plunge the head down on top of the material so the plasma cutter drags along the height changes of the tin. Then the head drops into a hole and gets ripped off the first time. No big deal, the v'd plastic guides that support the floating head deform enough to allow the head to pop off. The second time wasn't as fortunate:

    So new pulleys were made out of more rigid bearings and the v'd guide itself made out of aluminum. In the picture below, the old plastic piece is off to the left with the bearings that popped out laying beside it. The left side of the plastic pulley has a black line where it was sheared:

    In the process of machining the aluminum pieces, the bearing bore was too large on one. The installed bearing would slide in and out smoothly. Instead of making another one, the pulley was dinged with a punch to help trap the bearing:

    With two new v'd guides created it was assembled and installed. Back in business:

    To help with this kind of problem in the distant future, a magnetic break away floating head should probably be built.

  • Exhausted Material Cut

    willbaden08/30/2016 at 02:18 0 comments

    With the table capable to cut across the width of the material (48 inches), a program was created to act as the final cut to remove the completely cut material from the main sheet.

    G21 G54
    This code was ran after homing the machine ($H) and resetting G54 location5 mm from the edge of the homing resting location. It made a clean cut, except the beginning.

    I had forgotten about the delay to fire and will have to implement a primary cut that is near the cut line, but not on it.

  • Material Cut Table

    willbaden08/14/2016 at 01:12 0 comments

    The table that holds the material being cut is going through updating. The test table was four tubes clamped to a welded assembly that was re-purposed for a material table.

    It has served its purpose and is time for an upgrade. 2x2x1/8" material was ordered for the material table and for the first revision of the material rack. Using a chop saw (would have been nice to use a cold saw) the material cut table was cut up.

    There will be four lengths on the top and bottom with 4 short risers to separate the top and bottom. Below is an in process view of the material cut table acting as jigs for the first revision of the material rack.

    The plasma cut table is coming together, here is a mockup of the new table. At this point it is only tacked together for a trial fit under the plasma axis. The four holes in the bottom will be used for qty. 4 - 1/2" bolts that will be used as leveling feet. The bed has a slight skew to it after being welded together, but should be able to level with the cutting frame.

    On the top horizontal bars are three holes on both sides. These will be used to bolt the plasma cut table to the stock loader/unloader. It will also be used to bolt the plasma cut table frame work that holds the material and eventually will hinge on the closest side to allow the cut or spent material to drop out the back side.

    The table needed adjustable feet to allow for leveling. The 1/2" holes weren't enough. So 3/8" steel cut 2x2 were drilled and tapped for 1/2-13.

    These were then mocked up and welded to the table.

    The hinged table will be considered consumable. Current design is made by 1x1x1/8 steel tubing. This is intended to be a quick setup to cut some parts that I would like to see finished in the near future.

    This was set in inside the larger frame for test fitting.

    Outside of leveling the CNC table and cut table, there did not appear to be much for concern. The CNC was taken to the limits in the X axis to see where the consumable table would sit. Then the hinges were made up.

    These were then mocked up and welded to the main frame. There are two pieces of steel angle that bolt to the consumable table.

    Below is the left hinge tacked in and ready for hinge testing.

    The table hinges down and is meant to allow the material to slide down. The 1x1 hinged table is not the final design but will have the same hinging effect.

    We will see if this will be enough angle. May have to raise the table with some risers. Cross that road later. The above picture also shows the mounting plate on the right side of the table. This will bolt to the CNC tables legs.


    The spacing of the bars going across has caused some issues with parts warping and being unable to maintain a decent height while cutting. IE the torch would initially touch off of a fairly level piece, but as the cut progressed the cut material drooping/warping would allow for the torch height to change too much. This would cause an intermittent/failed cut. To help out with this until torch height control can be implemented, more cross bars were added.

    The gap between the end piece and last support bar was intentional as the plasma cutter can't reach this far.

  • Axis Movement

    willbaden08/11/2016 at 02:05 0 comments

  • Delay to Start of Plasma Torch Fire

    willbaden08/11/2016 at 00:51 0 comments

    The plasma torch needs to fire and start an arc before moving or a detail can be missed or a drop out doesn't drop out. That was the case that I ran into. Thinking that it was a GRBL issue, I posted an issue on the github grbl page here. The thought was the G4P1.2 was not dwelling the proper about of time the first time it was called, but would work fine in successive calls. The program below was made to help with diagnosing the problem.

    G21 G90 G64
    G0 Z1.0
    G1 Z1.0 F1000.0
    G1 F1200.0 X10 Y0
    G0 Z10.0
    G0 X0 Y10
    G1F1000.0 Z1.0
    G1 Z1.0 F1000.0
    G1 F1200.0 X10 Y10
    G0 Z10.0
    G0 X0 Y20
    G1F1000.0 Z1.0
    G1 Z1.0 F1000.0
    G1 F1200.0 X10 Y20
    G0 Z10.0

    This code produced the following cuts:

    The two left columns were ran once, while the right column was ran twice in a row. It showed the same cuts the first time through, but worked the second. So it might not be a grbl issue. So a phone was setup to capture the sequence of events as the program ran. The lights on the Plasma cutter are shown below:

    Its upside down as the video shows it. Pretty straight forward what they stand for. The unit was on so the AC light was on. The gas represents when the air kicks in, the DC light, when there is juice to the plasma torch and the over temp is just that. This is located on the top of the plasma unit in the video. The relay that kicks on the plasma cutter is inside the plasma unit (lower left) and has a red light turns on.

    For the first cut, the relay kicks in to turn on the plasma torch, but there is hesitation before the DC light turns on along with the plasma torch fire. After the relay shuts off, the DC light stays on for a short time.

    The way that would be nice is if M66 was implemented in grbl (may have to dive into this, if its priority goes up). For now a cut elsewhere on the material will charge the DC side and allow the next cut to come out ok.

  • Chassis Grounding

    willbaden08/05/2016 at 22:19 0 comments

    Trying to adhere to Hypotherms grounding suggestions the following grounding methods were loosely followed. Ground bar on the machine chassis that individually connects to the Y, X and Z axis.

    The Chassis ground bar also grounds out the work ground lead from the plasma cutter.

    Which is also attached to a 4' ground rod right beside the machine.

  • Raspberry Pi Addition

    willbaden08/05/2016 at 22:13 0 comments

    Seeing that there was some info on using a Raspberry Pi and Universal GCode Sender, AND it was small enough to fit inside the controller AND I could purchase a touchscreen monitor that could interface with it at a decent price. . .the combo was installed.

    The pi was easy to find a location for.

    The monitor and driver board, not so much. But now that I have a plasma table that can cut, as long as the hard limits are shut off, why not make a monitor housing? The unbent flat cutout was drawn up in CamBam. The plasma table did a good job of cutting it out. But the first design wasn't going to be easy to bend since the lip was integral to the housing. The image below shows the flaps having little ears coming off the corners. These were intended to act as the lip. In thinking about bending this, it was scrapped.

    So one without the lip was cutout and bent to form the majority of the housing.

    This didn't come out without its own flaws. Notice the holes around the cutout? These are to mount a backplate and the monitor. They are oblong due to a failure in the cutout process, and having to restart. Looks to be like there is a missed steps situation becoming apparent. The bending of the flaps did not come out as well as planned, but a little bit of welding can fill that gap.

    One of the flaps has holes cut in it to allow access to the monitor driver boards setup keys.

    The flange was then cut out and welded to the base.

    The corners were also welded at this time. The monitor housing was mocked up and appears to look decent in design. Would have desired more room between the buttons up top, but it'll do.

    The housing was removed to finish designing/fabricating the monitor housing. To mount the monitor in place, some delrin round stock had a hole drilled in it and a step milled on one side to capture the thickness of the monitor.

    The above picture shows the aluminum backplate that holds the touch screen interface board and monitor driver board. With the monitor board driver attached, it was mocked up on the controller box cover. The HDMI and power cables were ran also.

    The assembly was installed and looks like clearances worked out.

    It was then attached to the controller and the unit powered up successfully.

    Now to tear it apart and paint the cover and housing.

    And the back side of the monitor housing showing the backplate.

    Looking closer at the blue touch screen driver board, the ribbon cable is broke. This was broken in the mockup of the backplate. It is up in the air if I will be purchasing another touch screen (not that expensive at $15).

    The backplate of the main controller box also had some wiring added to it. A LinkSYS 5V wall wart power supply was added to feed power to the pi, monitor driver and UNO. This will probably be updated in the future, but for now appears to be doing a good job.

    And the assembled view.

    Currently I am using a Dell wireless keyboard and mouse hoping to limit the amount of exposure the USB gets from the plasma torch. After all the work of getting this enclosed, the hard limit symptom has considerably diminished. I have had it happen recently that warranted to shut off hard limits in the settings though. Along with this, it appears that missed steps are still showing up. The separated O in the image below shows that there is a shift in the Y Axis. In prior cutout attempts, the X also showed shifting

    Breaking it down, two problems persist:

    1. Hard Limit triggering (albeit not as bad as what it started out as)
      1. Appears to be solved by pulling anything USB inside of the controller box. At this point, all wiring is either shielded cable or inside of a grounded enclosure.
    2. Stepper loss of steps
      1. Appears to be solved by adding ferrite beads around all four axis drive cables that lead from the controller box to the driver box. So much for the theory of grounding the shielding immediately inside of the box to prevent this. Next time I might try to ground the shielding...
    Read more »

View all 18 project logs

Enjoy this project?



Bharbour wrote 07/28/2016 at 14:41 point

Nice Build! Seeing this project reminds me of the CNC plasma cutter that I started on. I got the carriages designed and built before realizing that I had no space in my shop to put the finished unit. 

  Are you sure? yes | no

willbaden wrote 07/28/2016 at 15:53 point

Thanks!  I did think about a smaller portable unit that would magnetically attach to a larger piece of material, but opted for this style to be able to automate sourcing from a rack.

  Are you sure? yes | no

fredkono wrote 07/28/2016 at 01:12 point

I could add some g-code if I knew what to add LOL I guess I need to read up on gcode.

  Are you sure? yes | no

willbaden wrote 07/28/2016 at 10:23 point

No worries.  Searching for grbl and probe should show some useful information.  This setup uses the following:



Searching the above gcode at should get you pointed in the right direction.

  Are you sure? yes | no

fredkono wrote 07/27/2016 at 19:13 point

What are you using to make your Gcode? I use vcarve and pcb-gcode with eagle for my small cnc, but I don't think that would work good for a plasma?

  Are you sure? yes | no

willbaden wrote 07/28/2016 at 01:08 point

Currently using evaluation period of cambam.  If you can post g1 segments and either manually add or have the post processor automate a tool height touch off, then you should be good.

  Are you sure? yes | no

fredkono wrote 07/27/2016 at 04:17 point

Nice build.,  May I ask what software you are using?

  Are you sure? yes | no

willbaden wrote 07/27/2016 at 10:39 point

Thanks!  GRBL on UNO.  Universal GCode sender as software interface.  More logs to come.  Starting with mechanical and plan on moving into electronics.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates