Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!uakari.primate.wisc.edu!caesar.cs.montana.edu!ogccse!orstcs!guille!daver
From: daver@guille.ece.orst.edu (Dave Rabinowitz)
Newsgroups: comp.sys.handhelds
Subject: Re: Format of the hp28s IR led output.
Message-ID: <14017@orstcs.CS.ORST.EDU>
Date: 29 Nov 89 00:03:21 GMT
References: <2371@gmu90x.gmu.edu>
Sender: useneta@orstcs.CS.ORST.EDU
Reply-To: daver@guille.ECE.ORST.EDU (Dave Rabinowitz)
Organization: Oregon State University, E&CE, Corvallis
Lines: 31

In article <2371@gmu90x.gmu.edu> peraino@gmu90x.gmu.edu (peraino) writes:
>
>Since the infrared led is also used by companies other than Hewlett Packard, I
>want to make its code (as far as me and some friends of mine have cracked it)
>public.

The details of the IR interface code were published in the Hewlett-Packard
Journal about three years ago.  Your local HP sales office should be able to
locate a copy for you.

>The data and checksum are frequence modulated, and every mark is modulated with
>6 pulses.

The spec allows 6 or 7 pulses modulated at 32768Hz.

The coding splits each bit time into a 1 and 0 slot.  A burst in the 1 time
slot indicates a 1 bit, a burst in the 0 slot indicates a 0 bit.  A burst in
both slots or neither slot is recorded as an error, and the error correction
bits allow correction of up to two errors of this type.

A frame starts with three bursts spaced 1/2 bit time apart, a sequence which
cannot legally occur in the middle of a frame.

Your interpretation appears to be equivalent, but I think this is a simpler way
to look at it.

>	- Gaps between S and D are maximum 1 msec; it is up to the hardware,
>	  to resynchronise with every byte.

A bit time is 27 cycles of 32768Hz, or 824 usec.  The printer resynchronizes
its timing at each burst.