Re: Phantom power on RS485 network

It occurred to me that I might be able to feed phantom power down
the RS485 signal lines to the remote boxes, but I don't know how. I
'Googled' on various combinations of 'RS485' and 'phantom power' but
came up with nothing useful.

X10 uses the AC-network to transmitt it's signals. PoE can use all four
wires to power devices and still keep 1Gbps data asfaik. Etc.. (google..)
It's all about seperating DC & AC.

There will be no baseband audio and the data rate will be very low. The
idea is that there will be several boxes around the stage. Each will
have two buttons and two LEDs (red & green). The master station will
monitor all boxes for button presses and flash the appropriate LED on
the box where the button was pressed. All LEDs can be cancelled from
the master station which will have clear buttons and an LCD display to

How many boxes will there be?, Led-only box seems kind of light for a scene ;)

indicate which box was activated. I am planning on using ATmega8's in
all the boxes and a serial protocol using ASCII messages. The protocol
will require two-way communication on the network and the network will
probably be around 75 metres in total length. I am thinking of RS485 as
it doesn't radiate emissions as badly as an unbalanced system and there
are microphone cables running nearby.

Two or even one io pin of the mcu (atmega8) could be used to setup a very
cheap transmission system. A capacitor in series with the signaling pin
should keep data signals seperate. A coil in series with the mcu powerpin
ought to supply power. Care would have to be taken to resonance and stability
of volt level. You have to verify this stuff on powering the mcu thoe :)
For RS-485 operation additional driver ic is needed (costfactor).

However you have given me food for thought and I could use a system
similar to 1-wire. A pull-up at the master station and all USARTs could
connect their TX pins to the bus through open-collector drivers. RX
pins could simply hang on to the wire. I don't know if this would
system would be capable of 75 metres, but I'll kludge something
together and see what happens.

Sounds like a reasonable plan, but split the pullups at both ends.
Driven with an NPN transistor to ground, or the equivelent ic output, I
would go with about 200 ohms pull ups, for a total of 100 ohms. Then a
slow data rate - say 1200 baud, because you are not sending much data

Given l=75 meters, cable velocity vf=.42 (?), m=5x propagation marginal

1/(l /(vf * c)*m)
1/(75/(.42 * c)*5) = 336 kbps possible ..?

Just a quick guess..

nor needing especially fast response times. I double that this will
induce much noise into nearby, ballanced, shielded, microphone cables.

Maybe some rc filter could fix that by slowing rise and fall times.

.