Re: [CAD_CAM_EDM_DRO] USB options ...

On 06/20/2014 04:11 AM, Hannu Venermo gcode.fi@...
[CAD_CAM_EDM_DRO] wrote:

> The subspindle is opposite the main spindle (and of lesser power).
> This allows machining the back of the part, without touching it.
>
> For many manufacturers, live tooling, ie y-axis, is today more common
> than without it.
> Subspindles are less common, and more expensive.
>
>
> They can syncronise the 2 spindles, while running, so that the
> subspindle picks up the part while running without stopping.
>
>

OK, so there are a couple areas that would need some
development.
Normally, the spindle is run as a velocity device, you tell
it how fast
to spin, and leave it at that. It IS possible to to run the
spindle as
a positioning axis, and then have some HAL code that would sync
another spindle with the first. LinuxCNC would be unaware that
this is happening, it would just be sending speed to the primary
spindle, and then tripping an auxiliary code to tell the
subspindle
to sync to the primary. This probably is not very hard to
do, given
that the machine had appropriate spindles, drives and encoders.
The primary spindle would have a PID component to convert
speed command to what the spindle drive needed and close the
speed loop. The subspindle would then have a PID to sync
position
to the main. There would be some servo tuning involved.

Another issue is live tooling, which requires switching LinuxCNC
from spindle mode to C axis mode. Most of what was used to
set up the main spindle PID would be re-used, but the command
from LinuxCNC would be switched over from the spindle speed
command to a C axis position command. Again, this could be
set up with suitable HAL components, like mux and scale. I
think
there may be some live tooling sample codes around, but I'm
pretty sure I've never heard of a subspindle setup.

Jon