Help with homing on dual axis PlasmaC

phillc54 wrote:

You cannot rely on microsteps to increase resolution. There were some good points made on a recent thread on the users mailing list, I'll see if I can find it.

EDIT: There is a lot of stuff to wade through but there are a couple of interesting tidbits in there:
www.mail-archive.com/emc-users@lists.sou...ge.net/msg81808.html

Phill, I did not say anything about increasing accuracy with increased microstepping but it will smooth motion as per this quote from your link..

> > > If you can hear it, it isn't smooth. I'd recommend using the
> > > finest micro stepping setting your driver offers, but simply not
> > > counting on more than 1/4 of a full step in your resolution
> > > accuracy. But just because the motor and drive may not be able to
> > > physically achieve positioning accuracy much better than 1/4 step,
> > > doesn't mean that having the finer micro step resolution won't
> > > significantly contribute to smoother quieter motion.

This is where I think the modern drivers from Lam and Trimatic have a significant advantage over older designs like the Geckos. I suspect Gecko had their eye on accuracy when they settled on 10:1 not smooth motion.

snowgoer540 wrote:

Back to the topic... I guess micro stepping is really just a way of toning down vibration and chunky movements? I legitimately always thought it was a means of increasing resolution. Maybe it's different for ballscrew machines?

I found the thread I was thinking of! It was Tommy who offered up the advice:

Tommy's post in thread

Greg, you should know that there is no need to go searching through the forum for the good oil! The whole intent of the plasma primer was to put it all in one spot
linuxcnc.org/docs/2.8/html/plasma/plasma...per_motor_revolution

That way you don't need to repeat the brutal learnings I had.
I was actually thinking about that just recently and wondered if I would get better results from a 5:1 reduction instead of the 3:1 I ended up with. That would get that motor down to about 18mm/rev

The other advice I have read many times is from Tom Caudle from CandCNC, He maintains that 1" (say 25mm) is the ideal distance per rev.

so Phill at 60mm per rev, you have diagnosed your own problem. I can share a design for that reduction drive as it just bolts on in place of a NEMA23.

phillc54 wrote:

I guess one should really bin it all and go to servos.

I don't think thats necessary becasue the forces on plasma machines are totally predictable becasue there is no cutting forces.
In any case, you'd need to revisit gearing to get the torque you need as servos are high speed low torque vs steppers low speed high torque....

Just fess up. You broke the rules you pay the penalty (of more quality shed time revising your never ending build!)

phillc54 wrote:

But I always thought rules were made to be broken...

You are only allowed to break the rules when you know what they are
(Rod's Law)

Learning the rules can be painful!
(Rod's Corollary)

Man that bit hurt!

There is such a massive skills set in building a plasma machine.

phillc54 wrote:

rodw wrote:

There is such a massive skills set in building a plasma machine.

HaHa, now I know why mine is broken...

So thats why Plasmac is so successful!

Seriously though, you look at where Plasmac is now, you can see different features and see all the pieces different people have contributed. So the skillset you borrowed has been pretty broad! I still have so many ideas on how to improve Linuxcnc's plasma cutting support. The other day, I found out to my dismay that with one of my ideas I've chased for about 2 years, I was rediscovering what Hypertherm already knew for years. But I can guarantee that no other entry level motion system can give a way forward to knock on Hypertherm's door!