> Date: Sat, 6 Jul 2013 22:15:26 -0500
> From: el...@pico-systems.com
> To: emc-developers@lists.sourceforge.net
> Subject: Re: [Emc-developers] "Open" Development
>
> Chris Morley wrote:
> >
> > I bet there is a following error spike right when you do reversal or maybe
> > not badly because the
> > spindle does not react with infinite jerk either.
> >
> Yes, and that''s why you want to control the rate of spindle reversal so
> the following axis has a better change of keeping up.
> > So if your tap reversal changes direction faster then the machine can
> > react, it doesn't matter
> > if you have jerk limiting or not - there will be error.
> > I bet we'd b surprised how much error rigid tapping can tolerate. machine
> > and part flex is forgiving.
> >
> > And that is what I am saying, without jerk limiting, we are asking the
> > machine to move faster
> > then it actually can, so we get a small following error spike, most
> > pronounced at reversals.
> >
> So, for rigid tapping, you want the spindle to reverse more gradually
> than it
> might be capable of doing. Then, the jerk is reduced.
So you wouldn't need to turn jerk limiting off for G33.1 then.
Just as we don't turn acceleration or speed limiting off for G33.1
The TP should never ask the machine to move faster then it can.
> > I remember the graphs I looked at were when the mill cut circles - there
> > were reversal spikes at each of the four quadrants.
> >
> A tight servo should have no spikes at the reversals. The arc interpolation
> is quite gradual at these reversals, unless the system is trying to do
> backlash
> compensation (which I consider to be a big mistake.)
>
> Jon
>
Well gradual is a relative term :)
It may have been a very fast move of a small circle on a large machine.
Also I don't remember the amount of the spike. it certainly wasn't thousands of
an inch.
This was an company trying to sell something so would give worse case.
I'm pretty sure it was hiedenhain.
I wish I could find the write up.
Chris M
------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:
Build for Windows Store.
http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers
