AVR-Chat

At 12:40 PM 10/31/04 -0600, you wrote:

> > From: rajesh parwani
> >
> > i would suggest u for i2c bus or rs232 intrerface using rxd
> > and txd pins.
> > although with suggested scheme ur some of the wires will be
> > left out , u can use them for future development
>
>I know RS-232 won't go the required 400 feet and I haven't seen where I2C will
>go that far - or can one use RS-485 line drivers/receivers with I2C?

Maybe, but your transmit speed would be limited by light speed 
considerations.  CAN has some similar limitations and a 400 foot bus would 
be limited to a max of 250kbits/sec (probably less).  I haven't studied IIC 
that closely but I suspect that the clock stretching requirement places the 
limit rather more severely.  The basic limit is the time for a signal to 
propagate from one end of the cable to the other and back along with the 
sum of the transceiver delays.  This must be less than the minimum pulse 
width + the time for the receiver to recognize the low and respond with the 
appropriate stretching. If you know those numbers then you can determine a 
maximum length for the bus.

On the other hand RS-485  serial type (UART) doesn't much care about signal 
speed (as long as reflections are minimized), except at the packet 
level.  As long as two transmitters don't attempt to communicate at the 
same time than the receivers simply receive the whole message.  You get a 
higher baud rate at the expense of having the protocol needing to either 
ensure that no two transmitters will be active at once or that some kind of 
collision detection is used to retry transmission.

Hmm, probably overkill but arcnet would be good for this with appropriate 
chip support to offload communications (speed up to 5mbits per sec, token 
passing).

Robert

" 'Freedom' has no meaning of itself.  There are always restrictions,
be they legal, genetic, or physical.  If you don't believe me, try to
chew a radio signal. "

                         Kelvin Throop, III