>> I presume that it would work because of the low currents involved and the
>> fact that RS-232 "0" can be represented by 5V.  Does this mean that a 1 is
>> anything less than 3V (which is what i believe the EIA Spec says).
>
> The very simple, answer is "YES." This is because any non valid input
> voltage is defined as a logic 1 on the output.  Don't become confused,
> this is the essential point.  BUT, as Mike Fahrion pointed out, there
> are a some cases the negative swing is required.
>
> RS232 input to a pic can be through a single resistor [...]

If the PIC is receiving data via RS232 from another device, then this is
true.  The sender will generate both positive and negative voltages.  You
should have a resistor to limit pin current and a diode to ground to keep
the PIC's input pin from being driven to far below 0V.  Be aware that you
are giving up some noise margin, but it should work fine.

If the PIC is trying to send data via RS232 output to another device, it
won't work to significant numbers of devices.  It may work fine to a PC
or a Mac, but...  Lots of equipment wants to see the negative voltage.

I have a terminal.  It's -12V power supply failed (rest are OK).  So the
RS232 TxD output is 0V or +12V (no negative).  It will work hooked to
itself in loopback.  It will work with  _some_  equipment in the office.
But other pieces of equipment simply refuse.


> Of course, this sort of interface belongs to the M2L (Mickey-Mouse logic)
> family

M2L is perfectly adequate for prototyping and development work (if you
know the limitations).  I just wouldn't ship a product built on it.

                                                Lee