Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucsd!ogccse!orstcs!guille!daver
From: daver@guille.ece.orst.edu (Dave Rabinowitz)
Newsgroups: comp.sys.handhelds
Subject: Re: HP28 IR link
Keywords: hp28 IR LED
Message-ID: <13672@orstcs.CS.ORST.EDU>
Date: 9 Nov 89 21:00:47 GMT
References: <2444@gmu90x.UUCP> <8538@microsoft.UUCP> <13512@orstcs.CS.ORST.EDU> <8615@microsoft.UUCP>
Sender: usenet@orstcs.CS.ORST.EDU
Reply-To: daver@guille.ECE.ORST.EDU (Dave Rabinowitz)
Organization: Oregon State University, E&CE, Corvallis
Lines: 21

In article <8615@microsoft.UUCP> alonzo@microsoft.UUCP (Alonzo Gariepy) writes:
>Thanks!  Can someone at HP dig this up for us?  Since the carrier is much
>slower than the clock speed, I would not be surprised if its period is
>determined by one of the timers in the slave display controller.  In that
>case, the 40KHz signal might be achievable.

The 32768 Hz carrier comes from a crystal which drives the real time clock.
The CPU clock is also derived from this crystal; the output of a VCO (voltage-
controlled oscillator) is divided down by a programmable divider to about
32768Hz and this frequency is compared to the crystal frequency and the
oscillator frequency is adjusted until the divided down frequency matches the
crystal frequency.  The CPU clock comes from the VCO.  The speed-up programs
change the CPU clock speed by reprogramming the frequency divider.

There are two bits which control the LED.  One turns on the LED modulated by
the carrier, producing the 32768Hz signal, but the other turns on the LED
independently of the oscillator.  A program which used this bit and toggled it
fast enough (12uS on, 13uS off) could generate a 40KHz signal.  Be very
careful!  The LED drive circuit assumes the LED will be on with well under a
50% duty cycle, so if you turn on the LED and leave it on accidently you could
potentially damage the electronics.