Re: Pulse Gen

Mariss Freimanis wrote:

> Hi,
>
> Jon, I respectfully disagree for two reasons:

OK, in the specific sense that you were using this on an embedded system,
where you decided how to set up the interrupt response, etc. then I agree,
there is no problem. A lot of the people here are talking about using this
on a PC, under Windows 95 or thereabouts, with little or no Real Time
support. That is a definite problem, as interrupt service is really
unpredictable in that environment.

> 1) This is based on an idustrial product I originally designed about
> 10 years ago. It was a dual axis pulse gen integrated with a Z80
> computer on the same board to do the math. It was way too expensive
> ($750) but several hundred were sold. It worked perfectly.
>
> 2) It doesn't matter when precisely you output the velocity update,
> only that you update every interrupt. The data is "clocked" in on the
> interrupt edge. You have the entire subsequent interrupt period to
> present the next update.

Jon