Re: 3 axis 9x49 retrofit recomendations.

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Dan Sergison"
<dsergison@y...> wrote:

> Sorry to ask such a general question, but there is SO much
> information, I'm having trouble processing it all.
>
> I have a plain old 9x49 enco mill. I'd like to retrofit it to 3

axis

> cnc. (maybe 4 w/rotary table)
> I'm perfectly capable of fabricating all the brackets and hardware
> but I'm over my head on the controll system. I believe I'd like to
> have linear positioning feedback from some dro, not rely just on my
> crappy leadscrews. also would need to be able to use the mill
> manually most of the time, and have a dro function.
>
> I'm a solid modeler by trade and fairly handy with computers, I

have

> a few pc's and a hpux workstation, or could set up linux.
>
> what would you recomend for a somewhat economical approach?
>
> thank you very much,
> Dan

Dan: This may not be what you want to hear, but the first order if
you want accuracy, is that you need to replace the sloppy leadscrews
with ballscrews (zero backlash if you can justify the expense). The
problem with trying to use some form of loop feedback to compensate
for the backlash will result in serious problems. You can use an
open loop stepper design (no feedback) and program in the backlash
compensation using the function in the controller software on the
PC. If you use a servo and closed loop module with some form of
encoders (like the Gecko 320/340's) then if the encoder is mounted on
the motor shaft than the motor will be in the correct position but
your table will still be off by the backlash. The final senerio is
the worst. If you take your position feedback from the table itself
using table mounted encoders then the servo will "hunt" between two
encoder transitions. The servos will have to spin and jerk out the
backlash each "hunt" (several times per second) reversing violently
and overshooting each time; your machine will proceed to vibrate
itself to pieces.