Re: [CAD_CAM_EDM_DRO] Re: moving Bridgeport steppers with Gecko drivers - help!

theokspindoctor wrote:

>Jon, I was confused by the comment that a sowftware-only solution
>would not work.
>

I was specifically saying that you can't get 48,000 steps/second with
software-generated step pulses.

> I am also using a stepper board motion controller
>that is sending the step and dir commands to the Gecko driver
>board. It is supposed to be effective, but the setup is apparently
>not so straight forward. From reivewing a previous post you made
>earlier in the month, the half winding mode at 7 amps is the
>preferred connection, which is what I think I have done.
>
>However, there is still the question of whether the 470 uF capacitor
>is necessary across the power connection. Mariss writes that this
>is required when the power supply is more than a 18 inches away or
>uses a fuse, both of which are true for my setup. I assume that is
>the length of the wire from the power supply to the Gecko drive. He
>says this is necessary because the drive draws supply current in
>20khz pulses and needs "flywheel" for proper function.
>
>

Yes, you may see problems if you don't use the capacitor where recommended.
I had the power supply about 6 inches of wire away, and still used some
small
caps there, just to be sure.

But, the original post didn't make it clear whether the drives were losing
sync or the software just wouldn't put out step pulses fast enough.

Given the calculations in a previous message, you have 2:1 belt drive
with a 5 TPI
leadscrew, so you need 10 turns of the motor per inch. With a 200
step/rev motor
and the Gecko 201's 10 x microstepping, you need 2000 step pulses/rev,
or20,000
step pulses per inch. To go 120 IPM, you need 2 IPS or 40,000 step
pulses/second.
Can CNC Lite produce such a pulse rate? Since it uses a Galil (presumably
DSP) board, it probably can do this, if the setup parameters are correct.

Jon