New plasma build - planning stage

[spumco]

Robert,

I'm sold on using mechanical switches. I'll figure out a switch arrangement with a roller or something like it so I don't have to worry so much about ramming the switch with overshoots. Anyone have a link to some switches they think are excellent?

So... basic design:

Gantry is 60" long for about a 52"-53" cutting width. Y-axis rails are 72" long for about a 64"-65" cut length.

Gantry is 3"x3" T-slot channel with 4"x.125" steel flats on the sides for rails and V-groove bearings. Rails will have bevels milled on them. I'm estimating the X-axis trolley to weigh about 8-10 lbs with both X and Z motors. Gantry will be likely be around 30 lbs (including X-trolley) after I mill a bunch of big holes in the steel gantry rails. Trolleys are all .250"Tm aluminum plate bolted together. Gantry is at the same elevation as the Y-axis rails - there's no 'riser'. Height above table is achieved/set by raising the Y-axis beams/rails off the table rails. Imagine a large "H" mounted with brackets above a steel table - all three 'trolleys' will be at the same elevation.

Using NEMA23 3-phase closed loop steppers from Automation Technologies at 48vdc. Fixed belt drive (HTD 5M x 25mm) using 3.5:1 reduction from the motor and a 15T pulley driving the belt. Same reduction/setup on the dual-Y trolleys. Calcs put the motors at about 400RPM at 300IPM. I tried to stay in the fat part of the motor torque curve up to about 300-400IPM.

Z-axis drives a 1204 ballscrew, with a 1.5:1 ratio to get the 400IPM max listed for the Neuron THC at 1500RPM (motor).

My target is (up to) 300IPM cutting speed, and the drive system has a mechanical resolution of 0.0029" per full step (1.2 degrees per).

Top speed? Acceleration? No clue. I know what the theoretical top speed is, but since I don't have a reasonably accurate gantry weight yet I can't figure out if it'll take all day to get up to speed. Worst case, I've left room to change the primary reduction up to a 4.5 or even 5:1 ratio to get better acceleration. If it turns out to be a total dog, motion wise, I've designed it to accept NEMA34 motors and I'll just talk him in to some 750W servos and make it walk across the floor.

Plasma is intended to be used as a hobby thing. Friend will likely cut anything he can get on the table, but primarily thinner material as I think his wife wants to make Etsy trinkets and stencils for her home business. This will not be a production situation.

Wasn't planning to use a floating head to save some weight on the X-trolley. This is my first build, but I was under the impression that ohmic was superior on thin material since (assuming it works) it doesn't deflect warped stock as much as a floating head might. Again, if this turns out to sub-optimal, my Z-backplate could be easily modified to accept a rail and switch of some sort to float the torch holder. I was also planning to use an air blast prior to initial torch height moves to clear water or crud from the torch & plate.

Helps having a CNC mill while building this. It really helps having two 6030 Fadals available at another friend's shop. Gang-drilling the steel rails for the T-slot mounts is going to to be SO nice - cycle start, shift rails 12", cycle start Op 2 - done.

Thanks for the help,
Ralph

[beefy]

Allow for axis deceleration though at limit switches from rapid motion to stop.

Just knocked this up to show what I did on my table to resolve this issue (hope the colours don't hurt anyones eyes). These are used on the home switches. My limit switches are separate and have the long metal springy arms on them because they don't need accuracy.

There's various other mechanisms that will do the same thing, whatever the mind can conceive. This gives accuracy and repeatability, and so long as nothing seizes up, it completely protects the switch from overrun. I have hard stops past the limit switches, so it's impossible for the overrun to go more than the compression distance of the spring.

Keith

[Robertspark]

Thanks Keith, you've given me an idea (reverse acting microswtch) from a simple bolt, depressed bolt breaks the miroswitch.

[spumco]

Well that's pretty clever. Totally thieving that idea for the switches.