Re: EMC and stepper driver
Posted by
Mariss Freimanis
on 2000-10-31 15:59:43 UTC
> Wally replys:signals
>
> I do not think the stepper drivers could have this feature. If you
> are running 2 or 3 axis and each drive is buffering the step
> you would probally loose sync of the stepper signals in relation toHi,
> each axis. This would be a disaster in circular interpolation.
>
> Wally K.
Let me clarify two points:
1) The pulse multiplier does not buffer steps in the conventional
sense. What it does is output 10 evenly spaced microsteps for every
full step input. It anticipates what the next full step period will
be based on previous periods in order to evenly space the microsteps
for every new full step input. This has the benefit of making the
motor more tolerant of poorly spaced (ragged) full steps at high
speed. Because there are 10 microsteps for every input step, the
resolution is the same as a full step drive except the motor moves
with microstep smoothness at low speeds.
2) The servodrive has a +/- 128 step "buffer"(This buffer could
overflow/underflow except for the FAULT detector. It shuts the drive
off when the error reaches +/- 120 counts).
Every step sent to the drive adds to the buffer; every step taken by
the motor subtracts from it. When both match (step issued, step
taken), the contents are zero regardless of speed.
The contents of this buffer is "0" steady-state because the PID
compensation does not permit a steady-state error. As long as what is
asked of the motor is within its ability (accel/decel), the contents
of the buffer stays zero. The only time it is other than zero is for
transient events.
One example is asking the motor to instantly change speed, say from
500 RPM to 1,000 RPM. The buffer initially goes negative (falls
behind) because it would take infinite torque to instantly change
speed. A negative-going buffer pours coals to the motor and it
accelerates to make up the difference between where it is and where
it ought to be.
As it catches up with the moving target, the buffer goes to zero,
acceleration ceases and the motor's speed and position match the
target.
Mariss
Discussion Thread
Wally K
2000-10-31 02:05:40 UTC
EMC and stepper driver
Darrell
2000-10-31 09:29:16 UTC
Re: [CAD_CAM_EDM_DRO] EMC and stepper driver
Mariss Freimanis
2000-10-31 09:56:30 UTC
Re: EMC and stepper driver
Tim Goldstein
2000-10-31 13:39:11 UTC
RE: [CAD_CAM_EDM_DRO] EMC and stepper driver
Wally K
2000-10-31 13:55:57 UTC
Re: EMC and stepper driver
Jon Elson
2000-10-31 14:04:42 UTC
Re: [CAD_CAM_EDM_DRO] EMC and stepper driver
Tim Goldstein
2000-10-31 14:11:42 UTC
RE: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Wally K
2000-10-31 14:30:31 UTC
Re: EMC and stepper driver
Wally K
2000-10-31 14:40:16 UTC
Re: EMC and stepper driver
Mariss Freimanis
2000-10-31 15:59:43 UTC
Re: EMC and stepper driver
Smoke
2000-10-31 16:22:12 UTC
Re: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Darrell
2000-10-31 16:31:27 UTC
Re: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Tim Goldstein
2000-10-31 16:48:25 UTC
RE: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Tim Goldstein
2000-10-31 16:55:46 UTC
RE: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Wally K
2000-10-31 20:17:15 UTC
Re: EMC and stepper driver
Tim Goldstein
2000-10-31 20:28:29 UTC
RE: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Wally K
2000-10-31 20:52:02 UTC
Re: EMC and stepper driver
Mariss Freimanis
2000-10-31 22:19:03 UTC
Re: EMC and stepper driver
Darrell
2000-10-31 22:58:20 UTC
Re: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver
Jon Elson
2000-11-01 21:39:53 UTC
Re: [CAD_CAM_EDM_DRO] Re: EMC and stepper driver