Isolation barrier in the digital signal isolator.

[imrankhanPNU]

Hello,

Could anyone comment on why a capacitor is better than an inductor according to the statement below:

"In general, an isolation barrier implemented with capacitors is better for passing signals to low impedance nodes than a barrier formed from inductors."
source: https://patentimages.storage.googleapis.com/8a/f0/44/e4b751f0ae6805/US7923710.pdf
Thanks.

 

[dick_freebird]

This is in large part, opinion. Likely biased by what the
opinion-holder has on the table. Just because somebody
puts text in a patent application does not mean it's true
or comprehensively so.

There are examples of both. I've seen bond wire inductor /
xfmr / "stripline" coupling, I've built capacitive coupled
half-bridge GaN driver (worked fine), I've seen papers with
integrated coupled inductors.

Seems like they make an argument related to "receiver"
node impedance. But the "barrier formed from inductors"
seems to miss the point, that an inductive coupler is trying
to be a transformer (or Marchand (tline) xfmr, which sort
of combines capacitive and inductive in the lousy tline
formed by close spaced bond wires). The transformer
qualities will affect how well the low-Z node can be driven.
But a low value C coupling isn't goping to move a low-Z
node a whole lot, for long, either.

To me it looks like a Marketing weasel wanted to get in
on the patent pile-on, and contributed a talking point.