Plasma Stress Test

Hi Guys,

I have recently upgraded ( or downgraded depending on how you look at it) my machine to use Leadshine closed loop stepper motors instead of the DC servos I was using. (Leadshine call them Easy Servos)

They are rated at 7Nm @ 100 RPM and 0.6 Nm @ 2400 RPM. I am using a 10:1 reduction so I am hoping they will have enough torque to move my 140kg gantry at the same speed as my old servos. All the calculation I have done suggest they will work.

What I was wondering, what would be a good item to cut that will really test out the limits of the motors? I dont think the rectangle plates I usually cut with four holes will place much stress on the system.

I think a series of high speed direction reversals including multi axis moves (all axis moving at once and changing directions together) is probably the best test and should stress the motors as much or more than most any cutting moves.

Next question, what is considered high speed for a large gantry machine? 12500mm/m?, 24500mm/m?

i would start at 12500 mm/m and work up until you see a failure and then back it off 15 to 20 % if it were me and compare what you get with the servo motors you are replacing and see how well these closed loop steppers do.

Stupid question, What would failure look like? Will the machine just not move any quicker past a certain point? Or will it shut down?

well basically your encoder feed back and motor won't match the controller if you loose steps and your drive should fault in that case. plus when you loose steps you have a visual on a stalled stepper as well.

one last thing that might help you out is pay close attention to how the motors sound. if for some reason your drives don't fault out stepper motors make a very distinctive sound when they start loosing steps. some times they flat out stall and don't move and some times they cut in and out. if you notice any sound changes that would most likely be lost steps. that's when you start getting into your acceleration and speed type adjustments to fine tune your system. i don't know enough about linux in that department to offer any help.

I hang a 6Kg vice from the torch mount and have it run one of the two attached gcode fro two days, checking at the end if it returns to the starting point correctly.

linuxcnc.org/docs/devel/html/plasma/plasmac-user-guide.html

thefabricator03 wrote: Stupid question, What would failure look like? Will the machine just not move any quicker past a certain point? Or will it shut down?

I can answer this now. High acceleration and lots of small parts, heaps of pierces will cause motors to overheat. When they overheat, they miss steps! I don't think closed loop steppers will help here.

So you can reduce the current but you loose torque so using a smarter driver such as those from Lam Technologies, you can reduce current when at constant velocity and only smash the amps at the motor for those few milliseconds its accellerating.

At this point, if you have not installed the drivers correctly, then the drives themselves will overheat. They can tell Linuxcnc of this via an error pin. But use the correct pin, this is treated similar to an estop. I need to do more work on how to handle it so it can recover (when they cool)

So then once you fix this, the plasma cutter can overheat. If this happens, it drops the ArcOK and Plasmac handles it very gracefully. It pauses and once the error subsides, pressing pause, the job continues.