[H-GEN] Remote process monitoring & control

[Michael Anthon]

(Note reply-to: being general at humbug.org.au vs Michael Anthon <mca at tams.com.au>)

I used to work for an electronics company that did a lot of work
interfacing microprocessor controlled devices with other equipment. 
Always be extremely wary of things that claim to be optically isolated. 
Even using the normal little 6 pin opto-isolators that everyone uses,
the data sheets show track layouts and spacing for the boards the must
be met for the isolation requirements and in practice, these
requirements are rarely met, particularly when designed by someone who
has never actually done any formal testing of these requirements (I am
speaking from experiance here... I learnt the hard way 8^).  

It is actually quite simple to achieve the isolation that you would
require for this job using standard componenets, BUT, I would carefully
check the PCB design of any kit before using it.  The last thing you
want to do is fry your computer.  The commercial systems tend to use
plug in isolation "modules" which have the opto-isolator circuit placed
in a small plastic box and then filled with epoxy with a set of pins
protruding through the bottom for plugging into carrier boards.  This
achieves 2 purposes.  For inputs and outputs it protects the isolation
circuits from a build up of contaminants (the spacings on the boards
these modules plug in to is generally MUCH greater than required for
these isolation levels so these do not need similar protection) and for
outputs, the epoxy acts as a bit of a heatsink as well.

Cheers
Michael

Doug Young wrote:
<snip>
>reliable ways of doing it. The Dick Smith kits are supposed to have
> "opto-isolation" whatever that implies

--
This is list (humbug) general handled by majordomo at lists.humbug.org.au .
Postings only from subscribed addresses of lists general or general-post.