interpolation and algorithms of multiple axes was Re: Pulse Gen

Hi,

What I would like to do is to take a look at what the goals are and
to see if they are practical.

The problem:

It presently takes a fairly powerful computer to generate relatively
mediocre step pulse rates (<100 kHz for a 500 MHz Pentium). No one I
know complains their step frequencies are too high.

The solution:

Relieve the PC of the burden of having to generate step pulses
directly with external hardware.

What form the solution should take:

1) A common standard. For this to be effective it should be an "open
architecture". The hardware circuit and specifications must be non-
proprietary. Anyone should be able to "roll their own" if they want
to.

2) Very simple. The hardware only generates step pulses. Possibly
general purpose I/O such as limit switches and fault inputs as well.

3) Modular. One pulse gen card per axis to keep cost down.

4) Low cost. Less than $5 to $10 per axis in parts.

For my part I will post a complete circuit diagram of the pulse gen
as a starting point. It generates 32,768 evenly spaced frequencies
from 0 to 4 MHz. It has been tested and works perfectly. Anyone may
build it and do with it as they please; I will retain no rights to
the circuit.

I will build a surface mount version, the pulse gen cards will be
about 1.5" by 2" per axis, so 4 axies will fit in a 2" by 2" by 1"
envelope.

Comments and suggestions needed.

Mariss