Re: electrical noise - was Re: A Question to MARISS

Guys,

You're absolutely right and I should have been more specific. It is not
necessarily the step and direction signals that are causing the
problem if they are correctly wired. It is more likely noise getting back
into the PC (or the step & direction lines) via the limit switch and
auxilliary inputs which Raul may have also connected.

The other problem I experienced was that the parallel ports I am
using (two different brands) could not reliably drive the 2531's inside
the Geckos without additional buffering.

I used a 74HCT240 on the parallel port outputs and drove a 2531 on
my pcb from that. I did double isolate the Step & Direction lines but
as you correctly described the situation with slow rise times etc., I had
to insert a schmidt trigger (74HCT14) inverter after the 2531 to square
up the signal and drive the Gecko's opto isolated input. Looking on a
scope everything now works well, level wise. As time permits, I'm still
working with Art's circuit design for jogging (joystick) each axis via a
switch matrix with germanium diodes and also learning a different
software environment to good old Flashcut.

Cheers, Peter