Inductor in external ground path?

[Rerouter]

In light of telstra planning to switch off the 2G network, I've been tasked to clone / improve an RS232 breakout board for our M2M device so we can use an external 3/4G data modem

the original board has a 0.1 ohm smd inductor of unknown value breaking the power ground from the serial modem ground, and i am struggling to see why?,

The device will be mounted in bus / truck electrical bins, near UHF radio's and similar, so i am hoping to  remove any possibilities of noise getting in via the adapter,

Between the M2M device and the external modem are 7 connections, The Serial lines are coming out of a standard MAX232 chip.

Power +
Power Ground
Serial Ground
RX
TX
CTS
RTS

[nctnico]

I'd add some ferrite beads in the signal/power lines and leave the ground inductor out. An inductor in the ground may screw the logic levels up and reduce the noise margin.

[T3sl4co1l]

In just the ground lead, yeah, not a great idea.  Better: ferrite beads on each of the signals; also, what type TVS are those?

Also, what the...fuck...?  Who drew those symbols?  'Cuz antiparallel zeners don't do zener things at all, it's the same as two regular diodes shorting the signal to ground!  You can't buy such a device anyway, bidirectional TVSs are antiseries!

Anyway... if they're supposed to be, like, SMAJ15CA or thereabouts, they'll have enough capacitance (couple nF) to be worthwhile in combination with a ferrite bead.  So you get some LC filter action going.  If not, you may want some explicit capacitance (same magnitude).

Bidirectional TVS are appropriate for RS-232, which is a bipolar signaling standard.  Don't use clamp diodes (well.. you could if you had the spare supplies in the system, but who uses +/-15V anymore?).

A ferrite bead in just ground wouldn't be advisable, but running the whole cable through a ferrite bead will help take the edge off common mode currents, for all signals equally.  Shielded cable helps even more (but the shield has to be grounded into the circuit ground plane or grounded chassis, on both ends, to work).

Tim

[Rerouter]

Thank you nico, and t3sl for the constructive criticism

The symbols where me late at night, not thinking, now fixed,

The cable i do not have control over and is in almost all cases unsheilded, so I've added a ferrite bead to each signal, and the updated schematic included, and yes t3sl, thank you for pointing that out, the tvs's each have about 4nF of capacitance,

Now there is only 1 remaining thought, should i nudge up the (0603 Package) ferrite beads towards the connection pads and make a hole on the ground plane opposite the same shape of the trace and pad of the bead, (thinking to remove AC coupling before the bead) or just not worry?

[T3sl4co1l]

Don't worry about it.  The ferrite bead is a few hundred ohms at its peak; the capacitance to ground from that pad is probably higher at most frequencies!

More important will be keeping the chip components at enough distance for easy rework, and for avoiding strain from the connectors.

I've been known to cut away ground plane from "dirty" outside connections (like around signals before the ferrite beads, or power lines before common mode chokes), but I have no idea if it actually does anything at all.

Tim