Power over LVDS

Isn't "12V LVDS" an oxymoron?

Link to the PowerPoint? ...Jim Thompson

AC couple the data, dummy. Like PoE.

There was also a mention about it in a Intersil presser for their ISL76321.

I would expect something a little more exotic than a Ethernet style POE given the signaling rates.

Could be reasonable. Seems like it would be a variant of Power over Ethernet (PoE).

?-0)

Sounds like phantom feed for condenser microphones or Power over Ethernet. However, these use transformers to couple the signals.

To use transformers, the signal should not contain any data dependent DC offsets, so you would have to use Manchester coding etc.

Sorta. The schemes I've seen use capacitors at each end to couple the LVDS lines and inductors to isolate the DC. At the frequencies usually encountered, the parts are small and cheap.

I haven't seen transformers used for power over LVDS. Often there's only one LVDS line, so the common center-tapped PoE thing doesn't work.

The transmitters and receivers that support AC coupling are using

I'll look through this and see if there is anything interesting

AC-Coupling Between Differential LVPECL, LVDS, HSTL,and CML

The Intersil part ISL76321 shows ac coupling in the data sheet with

When I have some time I'll sit down and work it out....

isn't it +12 on the signal lines and ground return?

I'm not clear what the question is. The LVDS schemes I've seen are power on one and ground on the other, similar to the phantom microphone power scheme (above). I suppose you could put 12V on both and rely on another common ground for the return. Sounds like an EMI problem waiting to happen, though.

VDS

much.

he

xlr phantom power is gnd to gnd, ~48V via resistors to both signal wires

think you could use similar scheme for lvds, use inductors and caps to isolate power and signal so you can try to get proper differential termination

should few volts of common mode voltage to play with

-Lasse