Hi,

I've read your email, and can offer some more specs on what is needed.

- its low distance, only about 50 to 100 metres.
- its single unit <-> single unit.
- the master unit activates and communicates with its slaves one at a time.

I'll look forward to your tutorial. Could you send me the link once you have finished?

Cheers,
Shane.

> -----Original Message-----
> From: shb7@cornell.edu [mailto:shb7@cornell.edu]
> Sent: 16 August 2001 11:07
> To: Shane Tolmie
> Cc: shb7@cornell.edu; PICLIST@MITVMA.MIT.EDU
> Subject: Re: Whats the best radio for a 20kbps transceiver, with
> variable packet length (1 byte to say 64 bytes) ?
>
>
> Hi Shane,
>
> It depends on a number of things, such as how far you need to transmit
> (i.e., what kind of signal to noise ratio are we dealing with), how much
> latency you can tolerate, are there going to be multiple stations sharing
> the same channel, etc. If the ends are being controlled by something
> simple like a PIC, an XMODEM-type protocol (where each packet must be
> ACKed) is a good general-purpose protocol. However, your overall
> throughput will probably only get up to about 80% of your actual transmit
> speed (so to average 20kbps you would need hardware that could handle
> around 25 or 30kbps and had very fast transmit/receive turn-around
> times).
>
> If you need to get every last bit of throughput performance and the
> channel is pretty free of interference, you might be able to get away
> with a ZMODEM-type protocol, where the transmitter sends long blocks and
> then just listens for a short time to see if the receiver needs any of it
> resent. This system can more easily break down in the presence of
> interference, though.
>
> I just recently completed a demonstration project which uses 16F876s
> along with Linx TR-916-SC modules to do 4800bps with
> error-correction(retransmission) direct from RS-232 port to RS-232 port.
> It uses an XMODEM-type protocol and the packets have only about 25 bytes
> of data (variable length with 25 being the maximum), so the actual
> throughput is only about 2400 bps.
>
> This demonstration project is a part of my tutorial on wireless comms for
> microcontrollers which is almost completed. I will be announcing the URL
> on this list when it is ready. As usual in my projects, I have had some
> delays in getting it ready. At least the Linx module section should be up
> very soon (within a week).
>
> If you add external RAM, an external UART, and make a few simple
> modifications, you
> should be able to achieve your 20kbps spec with this system, since the
> Linx modules can go to 33.6kbps. The reasons why I shot for a much lower
> goal were that the 16f876's ram limited the packet size (since we also
> had to have serial receive and transmit buffers in the RAM) and because I
> had to implement a second UART in software, along with the rest of the
> fairly complex code.
>
> Sean
>
> On Thu, 16 Aug 2001, Shane Tolmie
> wrote:
>
> > Hi guys,
> >
> > Just imagine I had a board with a 20kbps, di-directional radio transceiver, RS232 in, RS232 out.  What sort of radio
> protocol would
> > you recommend to communicate between two devices?
> >
> > The protocol would have to have elements of the following:
> >
> > - variable packet length, 1 byte to say 64 bytes
> > - CRC for error detection
> >
> > Other that that, its an open possibility.
> >
> > Regards,
> > Shane.
> >
> > --
> > http://www.piclist.com hint: PICList Posts must start with ONE topic:
> > [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads
> >
> >
> >
>

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads