Well, you really should tell us a bit more, ie, how are they being triggered?
If they are buttons being pressed by humans (or sensors being triggered by
creatures) then do a 10 byte ID for each transmitter at 300bps, and you'll end
up having to press two buttons within 33.3mS of each other to cause contention.
To prevent that, generate a random number, say it's based on how long the button
was pressed, or where an internal free running counter is when the button was
pressed, and retranmsit  a short time later based on that random number.  Given
that the chances of two or more buttons being pressed AND the retransmit time
being the same are EXTREMELY low (since the retransmit time is based on random
environmental factors, and is truly random, *right*?) then you can probably
stand to lose the one or two transmissions you miss each millenium.  Improve it
by sending faster (9600bps) and including the time you waited for restransmit so
the receiving clock can be more accurate when it receives a delayed signal.

If you need a system that is more reliable than this, then you should look at a
better contention system where the transmitters also receive and negotiate.

I hope this helps!

-Adam

> Anthony Clay wrote:
>
>
>
> I am using a PIC microcontroller in a relatively low-complexity application,
> but still valid enough to use a PIC microprocessor.  I want to create a small
> "intelligent transmitter" that interfaces to only one switch.  The main idea
> is to know how long the button was pressed, and on what transmitter.  I also
> plan to include a small serial communications port as well to allow the
> transmitter to interface with a "handheld receiver".
>
> Here is the general scenario:
>
> Transmitters are scattered throughout an open field,  the terrain is fair, (
> flat, some hills, thick trees ).  The end result should transmit a signal
> "blindly" and hope the receiver catches it. However, the transmitter must be
> able to take extremes in temperature abuse, as they will be outside.  A
> respectable range for this application would allow a receiver to intercept
> signals in a .3+ mile radius.  This is not a major factor, repeaters can be
> emplaced to time-stamp and rebrodcast signals.
>
> The end result will transmitter will transmit a very simple message that tells
> the receiver (which monitors all transmitters) how long the switch was
> pressed.  I though about using a RTC in the transmitter, but I have decided
> against it.  The receiver (which is a PIC, EEPROM, and a parallel port
> interface) will simply log the reception of the signal.  The signal must
> consist of an ID unique to each PIC and a data packet that contains the time.
> The transmitter must stay small, so any error correction must be one way.
>
> Size and battery power are priority design considerations but so is range.
> Speed may ease some issues, but there is so little data being transmitted, it
> makes little difference.
>
> The problem that is kicking my butt is contention.  If two transmitters
> are triggered simultaneously,  no good can come.  I though of using a "listen"
> feature, but this kills size and battery power.  I also thought of using a
> form of real time clock( so that a transmitter would act only during its
> "allotted time slot"), but an external clock would use too much power, and an
> internal routine does also.  I also thought of having the PIC send the signal
> multiple times at random intervals, to overcome the odds.  But the
> practicality of this approach diminishes quickly as the number of transmitters
> increases.  Any suggestions?
>
> I also need more information on the "free bands" that such an application
> would be allowed on.  Transmitter power is obvoiusly in the mW range and is
> suppossed to have very short range.  I do not want the signals to bother
> someone a 1.5 miles away.  I am think of somewhere around 300MHz, but I have
> no real basis to choose on,  I only require a speed of ~ 300 bit/s, less would
> not hurt.
>
> Another possibility is sound.  Can a modulated, high frequency audio tone be
> heard at that range?  It could prove to be very interesting.  Also does anyone
> know of any "matched" transmitter and receiver ICs?  They need to operate on
> an "open" frequency, which has not yet been determined.  Any idea helps.
>
> Anthony