Re: [CAD_CAM_EDM_DRO] interpolation and algorithms of multiple axes was Re: Pulse Gen

Mariss Freimanis wrote:

> Hi,
>
> Jeff, I could not agree with you more. The "pulse multiplied" step
> drive is indeed a bandage for a system that cannot generate pulses
> fast enough.
>
> Why can't it? Because the computer is being misused by having it
> execute mindless timing loops for freq. generation. This is best done
> with very simple external hardware. That way you could have megahertz
> step rates AND a fancy Windows GUI.
>
> The output would indeed be a velocity command, but the ext. hardware
> would also generate an interrupt from the same xtal oscillator that
> runs the pulse gen. Position would then be the digital integral of
> the velocity; i.e. simply add/subtract the velocity to/from the
> position register every interrupt.
>
> I have used this pulse gen circuit with a Z80 CPU this way and it
> worked perfectly.

I really don't think this will work, if you need to keep track of absolute
position. I have thought about this, and it just seems that there will
always be an occasional sampling uncertainty that can cause lost
(or doubled) counts when the step pulse occurs just as the computer
changes the velocity. Did a step come out just before the change,
or not? The only way to know is to count them!

So, that is what I had in mind with my rate generator. It would generate
both step/dir and quadrature pulses (useful for stepper drivers that
take phase inputs). The quadrature outputs would be coupled to a
quadrature encoder counter (in systems without encoder feedback)
so that an exact position reading would be maintained at all times.

I suppose you could arrange software to forbid a rate that would put a
pulse too close to the end of the sampling period, so that the uncertainty
could be prevented.

Jon