Hi,

Since I currently need such a circuit too I'll answer to
Russell's questions if may I take the liberty :)

> It's should just be a matter of designing the circuit for worst case
specs.
> Optos are notoriously bad for parameter shift over lifetime and also
often
> have a wide initial parameter spread.
> The answers to the following questions would be useful:
>
> - How fast a data rate do you wish to be able to handle?

4800bps (9600bps max)

> - What voltage and current is available on the output side?

Current available on the output side is direct drive from PIC RX/TX
pin.
But I also have a DC source of 9..35V/50mA available which can be
stepped down to e.g. 6VDC with a small smps. Cost is not a very big
issue
for this external RS232 isolator. But it's absolutely neccesary that
no power supply is needed on the PC side.

I've already an isolated RS232 application running with a MAX253
oscillator running a small ferrite pushpull transformer to obtain
isolated secondary (PC-) side power for the MAX232. The RX/TX lines
are isolated using 6N136 opto's. While working OK, this solution is
probably the most expensive one and it takes too much board space.
This
last requirement is also important since the isolator has to be housed
in a small external box.

> - What isolation voltage do you require?

500 to 1000V

> - What is the purpose of the isolation eg safety of user, protect
PC,
> protect output equipment, meet regulations ... ?

Protect PC and output equipment against ground loops etc.

Thanks!

Daniel...

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics