The pulse width at the detector varies because of the interaction
between the bandpass filter and the AGC.  The rise and fall delays
are not equal.  This shouldn't be a problem in practice, if you
do the learning function with a strong signal.  The stronger signal
will tend to equalise the rise and fall delays.

There is no standard format; each manufacturer seems to use a different
protocol.  I have found that just timing each transition of the
sequence (with reference to the first pulse, and a time resolution of
about 50-100 microseconds) gives adequate results.  Note that some mfrs
use alternate codes for the same button, presumably to allow
distinction between 'auto repeat' and multiple presses of the same
button.

All remotes that I have tested start with a relatively long pulse.
This is probably done to allow the AGC of the receiver to stabilise,
and possibly to grab the attention of the processor.

Regards,
SJH
Canberra, Australia

> From: Robert Zeff <rzeff@AINET.COM>
>
> Hi,
> I'm trying to "learn" the codes of various IR remotes.  I've
> found that the pulse widths from the remotes vary by as much as
> 30% depending on distance.  I'd appreciate at suggestions,
> specifications, etc.  that anyone can give me regarding
> the remote formats.