Re: Single Servo circuit

[Jon Elson <jmelson-at-artsci.wustl.edu>]

Mike Gann wrote:

> >From what I have read on CAD_CAM_EDM_DRO list, Jon Elson
> is very close to having this done.

Yes, the hardware IS done!  I'm just waiting for the PC boards to come
back from the fabricator.  I also need to design a motherboard for the
individual boards to plug into and get that made.

We are also waiting for the volunteer who offered to do the EMC-RTLinux
driver to finish that.

Now, this only goes from encoders to a +/- 10 V servo velocity command.
I think what the originator of this thread was interested in was a
complete
digital PWM servo drive, which would get rid of all that nasty analog
circuitry.  I'm not sure that is practical at low cost.  You will need
to
have an ADC reading motor current at a pretty high rate, and something
to do servo loop calculations and set the desired pulse width.
I do this 100,000 times a second (in other words, on every pulse).
It obviously isn't practical to have the main CPU handle it at this
rate.  I think analog is cheaper for this, although it doesn't allow for

a program to auto-tune the servo.

Jon