> Most likely what you have is the 'standard' IR (as in TV remote) signal
> which is either 40khz (or sometimes 38khz) on-off modulation, which
> is then further 100% (as in AND) modulated by the 'iintelligence', i.e.
> the raw RS-232 1/0 stream. If you have a scope (sounds likely) and

No. It is definitely not modulated. I used a real photodiode properly
biased so I would be sure to see any modulation. Here's an example that
shows the same character transmitted out a serial port and then the IR
port:

serial port:
---   ------   ---      ------
   ---      ---   ------      ---

IR port:
-     -  -     -        -  -
 ----- -- ----- -------- -- -----

Sorry that was an 11 bit stream but I don't want to redraw it, both are 10
bits per character(1 start, 8 data, 1 stop). You can see that when the
serial port has two adjacent positive bits, the IR port sends it as two
pulses. Chances are the IR input has at least a couple of selectable
filters depending on the mode it's used in and the simplest one just
filters out low frequencies, so the pulses come through fairly unscathed.
Raw serial port data would be mangled.

It's a late-model laptop without any IRDA drivers loaded, the IR port
shows up as com3 and the terminal program seems to have no problems
talking to it(or listening). If I take the serial out of another computer
and AND it with a 9600 HZ pulse wave, and dump that signal into an IR
LED I can talk to the laptop just fine as long as the 9600Hz signal is in
phase, but since I was using a function generator the best I could get was
working for 1 second, then bad data for a couple of seconds and so on.

If no one has heard of this I'll probably write a little program that'll
convert serial to these pulses and give it a shot. I'll let y'all know how
it turns out.

Cheers,
Bob