OT: low current RS485 and wake-up

[SteveS]

Off topic I know, but I think of general interest, and there are some clever minds on this list....

I need to set up a network of battery-powered devices that will be idle most of the time so I need to minimize the 'sleep' current. RS485 is a natural for the communications, as it's well supported chip-wise, multi-master, and can handle long cable lengths and noisy environments. However they have high (to my needs) current (300uA or more). Some chips have low standby currents but the receivers are inactive then (I think), unlike some RS232 chips that have a low current rx function.

So, anyone done something like this?

Solutions I can think of are:

1. Find an RS485 chip that has a low current rx (haven't yet)
2. Build a 'wake-up' circuit that sits on the RS485 line and catches a wake-up signal (possible interference with the RS485)
3. Add a separate wake-up line to the cable (not elegent and makes cable bigger, etc)
4. Use something else off the shelf or custom (not as well supported, probably more $)
5. ?

TIA,

Steve

[SherpaDoug]

I did a simillar system years ago that used a 455kHz ceramic resonator to detect a ring tone imposed on the RS485 lines and wake up the processor. The resonator allowed me to get a CMOS logic level signal without any active devices, so there was zero power consumption. We could wake up a dozen receivers on a kilometer of cable. Each receiver would look for its address and go back to sleep if it was not found within a few seconds. The processor had to clamp the wake up interrupt when it was awake so the RS485 data would not excite the resonator and cause extranious interrupts.
_________________
The search for better is endless. Instead simply find very good and get the job done.

[Neutone]

Does your entire network have to be battery powered? If you can run dual twisted pair that gets you an extra set of wires for power. A single solar cell and gel cell batter might power your entire network if no other power source is available. There are other solutions but there is always a drawback.

[SteveS]

See - I said there were some clever minds on this board!

Really interesting concept SherpaDoug! I'l have to check that out.

On the battery power - the system has no access to any power but battery and we need to keep battery size and cost to a minimum. A method we use now is to have the master cycle power to the slaves to control wake-up, but we are looking at a new configuration that will make that tough to do.