jes notes Index Gallery . Clavichords Cyclecar Dehumidifier Financial Data FreeCAD MSVA Philosophy Signed Distance Functions Shaft passers Snap issues

2026-10-23

Last modified: 2026-10-23 13:28:04

< 2026-10-22

FluidNC plasma cutter

I wired it up with a Leadshine servo and tested it again, again 50+ arc strikes and no apparent problems, so probably I'll use the Leadshine ones.

They are meant to take a minimum of 20v or something but I just used the 12v supply I had to hand and it seemed to work fine so I am inclined to run it this way. If it proves insufficient I can buy a 24v supply some other time.

I found a 2-relay module inside the cabinet of the CNC milling machine so will use that as the torch relay.

So next step I think is let's come up with a CAD model for the physical frame and start building it.

The electrical cabinet doesn't need a lot of thought, I have a stainless steel box I got from a charity shop years ago which has been looking for a use. I can put all the parts on a piece of plywood and then screw the plywood into the box, will just want holes for cables to pass through, and a power switch and emergency stop button on the front.

Annoyingly the X200 charger seems to have died overnight, laptop won't turn on and charger output measures 0v.

The shelf area I want to put it on is 80cm wide, 55cm deep, 60cm tall.

I'm thinking the X axis will be a piece of steel box section at the back of the machine with a carriage on it made of 6 bearings, 2 of which are adjustable for tensioning. X axis will be a 3d-printed rack-and-pinion. Both motors will go on the X axis carriage, with wires going off the back towards the controller on the left-hand side. The Y axis will be a piece of steel box section protruding towards the front of the amchine, with the Y carriage also riding on 6 bearings, and the torch clamped to a manually-adjusted Z axis which hangs off the right-hand side of the Y carriage, with the torch cable exiting the front of the machine towards the plasma machine which is off to the right.

The X axis box section can probably sit directly on the shelf. It will want "legs" protruding towards the front at the left and right edge. In between the legs I will put a water tray with levelling feet, and on top of or inside the water tray I will have slats to put the workpiece on.

I'm going to try out https://solidcraftcad.com/ instead of FreeCAD for this. It is only 2 weeks old but looks pretty good already.

OK I loaded it up and tried to mess about with a basic design composed of multiple bodies just to see if I could work out how to do what I needed to do, and I couldn't figure it out, and as it's such new software it's really hard to tell what's a bug and what's working as intended. I'm going to give this a miss for now and use FreeCAD as usual.

I can copy the Leadshine motor model out of the milling machine project.

I wonder if I should be using an assembly workbench? Typically I use "LinkGroups" to group things under 1 common coordinate system so that I can visualise driving individual axes to their 2 limits to make sure nothing interferes, but this triggers a bug where all the geometry disappears every time you open a new sketch until you exit and reopen it. But it does avoid the problem with assembly workbenches where constraints can get confused when you edit a body and things get attached in the wrong place. Hmm. What is the latest on FreeCAD assembly workbenches?

Just use LinkGroup I think. Or maybe I don't even need to be able to slide it back and forth at all.

For attaching plastic parts to 1.6mm thickness steel I am going to use the sheet metal roofing screws. They want a 6.5mm clearance hole.

Actually instead of a 3d-printed rack-and-pinion I might just use GT2 belt as a rack, with like a "omega drive" for it. That way the problem of reducing backlash on the rack is reduced to tensioning the belt, plus I already have belt & reduces time/material in 3d printing.

I have pieces of GT2 belt 65cm long and 109cm long. Will need to print pulleys to go on the Leadshine motors.

Y axis will be omega drive, just needs the 1 pulley on the motor plus a couple of idler bearings.

X axis will be a looped belt, needs 1 pulley on the motor plus an idler pulley at the other end.

Tensioning in both cases can just be a part with a bolt that pushes into the belt to stretch it, in a place where it's not going to have to run over the pulleys.

I generated a 30-tooth GT2 pulley with the TimingGear macro from the addon manager.

I have a pair of bearings labelled 6002RS to use as idlers on the omega drive. And I have 12x 6701 bearings to use for carriage wheels.

Hmm, thinking on it further, I think 6 bearings per carriage is not really enough to constrain it because it will be able to rotate around the length of the box section a little bit. That's annoying. In that case I'm going to have to order some bearings, so I may as well use the "608" bearings for both the wheels and the omega drive idlers.

So that's the 2nd purchase of this project. I ordered a 40-pack of 608 bearings on Amazon for £6.75, so total expenditure now £86.75. Costs are spiraling.

I think 7 per carriage may be enough? On the "long" face of the box section have 2 towards the bottom and 1 towards the top on one side, and 1 in the middle on the other side that is adjustable. And then on the "short" face you have 2 in the middle of one side and 1 adjustable one on the other. And then that is enough to tension it in both axes and stop rotation?

< 2026-10-22