Re: [CAD_CAM_EDM_DRO] Re: solution for optoisolator 3.3 or 5 volt

Hi Tom,
I do need to filter out some of the HF noise on the limit lines and that's one
reason for the opto isolators. I think you're saying that if the computer
case makes contact with the stepper driver case then I might have problems.
PP pins 18 to 25 are connected to the chassis of the computer chassis/case
and that connection is carried through to the D shell at the end of the
cable. The equipment ground from the computer chassis (round plug pin /
green wire ) is bonded to the the electrical panel which is grounded.
Can I float both the negative side of the stepper driver supply rails and
the 5 volt logic supply rails off the equipment ground so neither will be
connected to the signal ground - is that a solution? Is this problem caused
by ground loops?
cheers,
Denis

On Wednesday 09 January 2008 11:25:34 caudlet wrote:
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, denis <denisc1@...> wrote:
> > Hi Tom,
> > thanks for the info. I knew there was a bandwidth limitation but
>
> didn't know
>
> > it was that low. The stepper drives are not opto isolated. I'll try
>
> buffering
>
> > the PP output lines to drive the 6N136 cathodes. The separate
>
> supplies are
>
> > there already - totally isolated supply rails but common grounds.
>
> Those aren't isolated supplies if they share the same ground. It's
> the common ground that is the conductor for the common noise. Motor
> DC has motor noise and PWM stuff on it. With the same ground on both
> sides the only use of the Opto's is to do some level translation and
> since the response of the opto is so poor it will kill high freq noise
> ON THE SIGNAL LINE ONLY. If noise is on the signal ground (since it's
> tied to the "dirty" outside ground) it just skipd right on through and
> you can still have noise problems.
>
> If you have a scope do this: Ground your scope probe over on the PC
> logic side then go out to the Motor ground at the driver card and see
> wha the signal looks like. Should be a straight line.....You will be
> amazed at the spikes you see.
>
>
>
>
>
> I need
>
> > isolation on the limit switches for noise immunity, and was thinking
>
> of using
>
> > TIL117 - speed is probably not an issue for this application, do you
>
> think
>
> > that's a good choice for the limit switches?
> > cheers,
> > Denis
>
> Virtually any opto will work for switch isolation. No speed required
> there. Once again though if you run the same ground most of the
> advantages of isolation are defeated.
>
> In a lot of driver cards you have no choice because they use a common
> ground with the logic. Most of those cards are for smaller motors and
> the noise issue is not as critical. I do know that the Xylotex card
> (which uses a common ground) had noise problems early on and they had
> to make some circuit mods to get rid of it.
>
> Designing a motor drive is easy (stepper or servo). Making it work
> reliably and survive driving big inductive loads is not!
>
> I admire anyone that wants to build and design their own stuff if the
> objective is to learn. If you are doing it to try and save money it's
> marginal over buying a ready to run BOB. Even more so for a motor
> drive (unless you you have a free source of parts and your time has no
> value).
>
> > On Monday 07 January 2008 22:39:33 caudlet wrote:
> > > Depending on the opto you choose and parallel port drive capability it
> > > may or may not work.
> > >
> > > First problem is that some ports can't turn on an opto that takes 15
> >
> > /snip
> >
> > > TOM CAUDLE
> > > www.CandCNC.com