From: Andrew Jamieson <ajamiesn_at_optusnet.com.au>
Date: Wed, 9 Oct 2002 22:47:44 +1000

If we are going to have a circuit that is powered / re-charged at all by the
Archos (and I think that we should not dismiss this yet), then the Archos
will have to be the master, and do the polling, not the remote. This is
because the Archos will need to have the its port set to OP high at most
times to provide the power, therefore not permitting the remote to send
data.

If the Archos set the line low every 100ms for 1ms (or whatever is required
to provide a valid low given the capacitance of the remote circuit) and then
set the port to input, the remote can disconnect the power from the line,
and either send data or send a null packet. The Archos can then send any
delta info required (or the remote can request it if you like that scheme),
and re-apply power on the line. This way, the remote is only powered by its
battery during comms, hopefully dramatically extending the life of the
battery.

A


----- Original Message -----
From: "Andreas Stemmer" <Andreas.Stemmer_at_web.de>
To: <rockbox_at_cool.haxx.se>
Sent: Wednesday, October 09, 2002 10:12 PM
Subject: Re: new thread about remote control with display


> > Hmmm, yes. Otherwise we would need collision detection/arbitration
etc -
> > nasty....
>
> That's it!
>
> > Maybe the remote could periodically send some sort of "I'm here"
> > packet (which would be needed anyway I guess), after which there could
be
> a
> > specified "window of opportunity" for the archos to tx any updates.
Would
> > that work?
>
> Polling doesn't mean that all data has to be sent each time the remote
asks
> for it.
> We could of course include a "anything changed?"-command or a "not
necessary
> to send data because there was no change"-answer and that's what you
> proposed
> now!
>
> Andreas Stemmer
>
Received on 2002-10-09