[Robin Hewitt]

Hi

This is not a build log, I'm doing that somewhere else, it's all sums and boring so you're not missing much. This is just a few ideas for anyone considering aluminium and wanting to talk about it. I'm building a 2.5 x 1.25 metre plasma table, that's 8x4 plus a bit for those of you who still do Imperial.

I am aiming for a tool position tolerance of 0.1mm, not because I expect to get 0.1mm from the cutter, even the lasers don't offer that, it's more because I have CNC and I can.

The picture shows work in progress, if you run the X to it's end stop you can get the X Y and Z bearings all in the same pic.

The X gantry is 2.7 meters of 4" aluminium box section supporting a 15mm Hiwin rail with slightly oversized balls to keep it running true. The box weighs 10kg A 10kg load in the middle will bend it 0.5mm. The rail weighs 4kg. I am aiming for a gantry plus torch weight of around 22kg. The gantry is the long axis because when I did the sums it seemed the better bet.

I need to keep the torch carrier extrememly light so I can accelerate Y with the torch in the middle of the gantry without bending the gantry more than my tolerance. No possibility of hefty stepper motors and fat screws for the Z. If it takes >2 kgf to accelerate it in Y I am ouside my tolerance.

My Z only moves 40mm (1.5") and is driven by a tiny DC motor. It tracks up and down on two 8mm shafts. It has a mechanical memory for the workpiece surface and dials to set pierce and cut heights.

I don't think it is possible to make a plasma cutter gantry and bed in one piece and make it accurate, so I am not going to try. If I bolt the XYZ to the bed I think it will bend, so I'm going to rest it on 4 pins, one in each corner and all adjustable. The Y axis has two 30mm round rails and I am depending on their weight to hold everything down. Being able to lift it off the bed could be handy.

Getting the Y axis bearings perfectly square to the gantry is essential. I decided to use two plates at either end, bolted together. One fixed to the gantry one to the linear bearing. That way I can machine the top plate square to the gantry before attaching it to the bearing plate if necessary.

I insisted on screw from behind Hiwin rail for the X. Don't want any indentations in the front of the rail to carry plasma dust inside it's seals. This meant making 23 spacers to go inside the box section so I could bolt it from behind without bending anything. The bolt holes are 1.5mm oversize so I can bend the rail up in the middle should I need to.

I'm using 15mm steel reinforced T5 belts and NEMA 23 motors geared 40mm travel per rev. One motor for the Y with an axle across the back.

If I ramble on much more this will get exceedingly boring...

Have fun

Robin

[asuratman]

Hello,
Do you use THC?

[Robin Hewitt]

Hello,
Do you use THC?

I don't use anything yet, I haven't finished it :banana:

I will do THC, but there is no stepper

The Z axis gets 3 command wires which it decodes to 8 possible commands, 000, 001, 010 etc. It has one reply line because 4 core screened cable is cheap and flexible.

I decode with a logic array rather than a microprocessor, so if it gets spiked by the plasma pilot arc then it can only get it wrong for the duration of the spike. It's self healing.

The 8 states are

Move up to the top
Move down to touch or bottom stop
Move to pierce height
Move to cut height
Move up (THC)
Move down (THC)
Free wheel
Brake