> > It IS possible to design a resistive input mains interface that doesn't
> > violate IC input specs and which handles reasonably probable transients
> > well.
>
> With all this huff and puffing on this subject, has anybody actually
> tried it?

That question could mean several things, but yes in several cases.

I have used a properly designed resistive input from mains to processor pin
with no problems.

I have had an improper resistive input from RS232 cause severe problems
(using a circuit recommended by a commercial source (not uChip)).

Re "huffing and puffing".
Some while ago I decided to resolutely (read bloody mindedly) address the
use-of-protection-diodes-during-normal-operation issue every time it came up
as it comes up repeatedly and there are ALWAYS people who mislead the
uninitiated by telling them you can violate datasheet specs and be sure to
have no problems. I am well aware of the exceptions, the what if's, the
maybe you can's and the characterise it yourself for your application
arguments. I am aware that you can break the rules and live (or have your
design live) on many occasions. I've just decided to go in to bat on behalf
of the beginners and the less initiated (and that's a large proportion of
those on this list in this area) so that they hear some other voice than the
"just do it, all will be well" brigade :-).

Sometime, when time allows, (probably within the next 50 years) I intend to
do some real world tests injecting current into protection diodes on various
pins and report when observed strange affects do start to arise. I have seen
it happen in real world situations using digital only pins.

Re isolation and protection. I see no problem with using "non isolated"
interfaces; as long as the system level implications are understood and
properly addressed.


        Russell McMahon

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu