Presumably only signal currents flow through that resistor, and not even much at that, because RS-485 is differential. RS-485 still requires a ground reference; it is not an isolated interface, the receiver can only read signals within 7V or so of local ground.
The point then, would be to provide a weak kind of isolation, allowing ground loop voltages to drop across the resistor, rather than draw larger currents through a lower resistance path, and without leaving it fully open which would all but guarantee the receivers see trash.
It should be bypassed with a capacitor, so that shielded cable can be used if available, preventing RF interference from getting into the cable. As shown, the resistor simply drops some voltage (at some current), regardless of frequency, which in turn allows RFI into the cable. It does however act to terminate the cable/shield, so it won't be as prone to resonances as it would be otherwise. (If bypassed or direct-grounded, a ferrite bead might be desirable on the cable, to reduce possible resonances.)
Enterprise-grade improvements would include:
- Differential and common mode filtering
- ESD/surge protection (the RS-485 transceiver has excellent ESD ratings, but other interference sources may be a concern, up to and including induced lightning surge on long enough runs)
- A TVS on VCC would be alright, too
- Isolation (very easy to do, a couple optoisolators and a DC-DC converter, adds maybe five bucks to the BOM cost)
Tim
Home
Help
Search
Login
Register
