From: <idc-dragon_at_gmx.de>
Date: Sun, 13 Apr 2003 11:14:44 +0200 (MEST)

Hi,

Diava (Dirk-Andreas) did some coding recently to get an RC5-signal from an
IR receiver decoded using the remote input pin, great job! However, I did not
like the need for an external supply to power it. I noticed that the remote
input pin has quite a strong pullup (1k towards 3.3 Volt, means you waste 3.3
mA with a regular headphone plug shorting it) and did some experiments to
power a receiver from that.

There are 3Volt receivers from Vishay, type TSOP1836xx3V, they need about
0.75 mA current. Here's the one I used:
http://www.vishay.com/docs/82052/82052.pdf

Connecting this makes the line drop to 2.55 volt just by the supply current.
This it out of spec for the receiver, but the samples I got work down to 2.4
Volt. The problem is that a diode is needed to decouple the receiver supply
for the case that we want to drive the line low. The voltage drop of a diode
is 0.3V for a schottky, but we can't afford that. So I use a MOSFET
transistor to short the diode for the normal case where we don't pull low. This gives
the full voltage back. The diode can probably be a regular like 1N4148 again
(0.7V drop), but with a schottky the buffer capacitor charges a bit faster.

My circuit is attached, or here:
http://joerg.hohensohn.bei.t-online.de/archos/archor_ir.gif

The transistor Q1 is needed to invert the signal, Q3 for pulling low only
_after_ we're not shorting the diode any more. Can't use the receiver signal
directly.

This kind of circuit can be useful as well for a little PIC microcontroller
used as a button remote, or doing the RC5 decoding outside of the box (no
change on Rockbox then). If a microcontroller is used, two of its pins can be
used to first open the diode short and then pulling the line low. Q1 and Q3 are
not necessary then.

Have fun,
Jörg

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
Bitte lächeln! Fotogalerie online mit GMX ohne eigene Homepage!

Received on 2003-04-13