Re: Geckodrive

You have serious problems if your mechanism is so frozen up that even
holding torque can't move it.:-)

No action is planned yet. The focus right now is to perfect the servo
algorithm, then the focus moves to how it interacts and affects other
coordinated axis followed last by handling exceptions such as you
address.

This project is in the "lab" phase. The servo algorithm work is
perhaps 10% of all that remains to be done. That means months of work
yet to be done.

Mariss


--- In CAD_CAM_EDM_DRO@yahoogroups.com, "lcdpublishing"
<lcdpublishing@...> wrote:

>
> So, then if the motor does stop rotating, call it what you will if
> that isn't what a stall is, what is planned at that point, a fault
> condition like on a servo system?
>
>
>
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Mariss Freimanis"
> <mariss92705@> wrote:
> >
> > To "keep a motor turning no matter what" is clearly impossible;

it

> > requires infinite power from the motor. Nice mental picture
> though.:-)
> >
> > Slowing down under load is a better option than the alternative;
> > having the motor skip steps, desynchronize or fault-out. That is
> the
> > practical definition of "stalling".
> >
> > Mariss
>
> > --- In CAD_CAM_EDM_DRO@yahoogroups.com, "lcdpublishing"
> > <lcdpublishing@> wrote:
> > >
> > > Mariss, this isn't making any sense, but I think it has to do
> with
> > > what you call a stall and what I call a stall.
> > >
> > > To me, a motor is stalled when it no longer rotates. Are you
> > saying
> > > that your GREX device will keep that motor moving no matter
> what?
> > >
> > > Now unless you have magic powers, this isn't possible. Which
> makes
> > > me think that your interpretation of stall is different than
> mine.
> > >
> > > Based on your description below, it appears as though the drive
> > will
> > > slow down the velocity in hopes that the greater torque at low
> > speed
> > > will overcome the load and allow the motor to rotate again.
> > > However, at some point, there will be a load in excess of what
> the
> > > drive/motor can overcome - resulting in a stalled (Stopped)
> motor.
> > > Does the GREX then fault or what is planned for this condition?
> > >
> > > Thanks
> > >
> > > Chris
> > >
> > >
> > >
> > >
> > >
> > > --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Mariss Freimanis"
> > > <mariss92705@> wrote:
> > > >
> > > > No, the motor does not stall. It is prevented from doing that
> by
> > > > monitoring the lag angle and decelerating the rotating
> magnetic
> > > field
> > > > to prevent it from getting more than 1.8 degrees ahead of the
> > > rotor.
> > > >
> > > > The net effect is the motor slows down under heavy load. As

it

> > > does,
> > > > more torque is available as speed drops until a "balance
> point"
> > is
> > > > found where available torque equals load torque.
> > > >
> > > > Relieve the load torque and the motor accelerates until a new
> > > balance
> > > > point is found.
> > > >
> > > > The lead/lag angle register is analogous to the position

error

> > > summing
> > > > node in a classic servo. A step pulse increments this
> register,
> > > the
> > > > encoder count decrements it. The contents of this register is
> > > input to
> > > > a PID algorithm; the PID algorithm output then modifies the
> step
> > > pulse
> > > > rate to keep the summing node at a desired constant value.
> > > >
> > > > The lead/lag register also is used to increase the
> > > motor's "stiffness".
> > > >
> > > > An open-loop step motor has +/- 1.8 degrees of stiffness. In
> > fact,
> > > a
> > > > step motor only delivers its rated holding torque when its
> shaft
> > > is
> > > > displaced 1.8 degrees from its no-load position. The servoed
> > > response
> > > > is to "step into" the load, canceling this error. A 500-line
> > > encoder
> > > > gives a 10-fold increase in stiffness over an open-loop step
> > motor.
> > > >
> > > >
> > > > Mariss
> > > > --- In CAD_CAM_EDM_DRO@yahoogroups.com, "lcdpublishing"
> > > > <lcdpublishing@> wrote:
> > > > >
> > > > > Okay, so the motor still stalls like any other stepper
> system.
> > > > > However, it tolerates the stall and provides feedback to

the

> > > > > controlling software which would allow for a recovery if in
> > fact
> > > the
> > > > > problem causing the stall is self-correcting?
> > > > >
> > > > > It sounds neat, yet impossible at the same time. Either
> way,
> > it
> > > > > sounds like it has the potential of allowing higher speed
> > > performance
> > > > > from stepper systems. Perhaps I should look at this as if
> it
> > > were in
> > > > > an adaptive control system and it would make more sense.
> > > > >
> > > > > Sounds cool guys.
> > > > >
> > > >
> > >
> >
>