PCB RF behaviour

This is an excellent talk on RF behaviour on a PCB. Before you dismiss RF behaviour as unimportant for micro controllers, watch the video. This is not beginner stuff though.

On the subject, this was the book that opened my eyes to this sort of thing back in the 1980's, and
its probably still pretty relevant today (even though the references to TTL are dated!):

http://www.ivorcatt.org/digital-hardware-design.htm

And it rightly refers to the pioneering work of Oliver Heaviside, a much under-estimated figure in
the history of electronics/telegraphy.

PerryBebbington:
This is not beginner stuff though.

That's a serious understatement :slight_smile: An interesting talk for sure.

Since watching this lecture I no longer think of decoupling capacitors as decoupling capacitors. I think of them as the local power supply to whatever device they are close to because they are the immediate source of any variation in current to that device. Of course, that doesn't mean they don't decouple, but it seems to me their immediate purpose is short term power supply with decoupling as a secondary effect.

I've always seen decoupling as just that: taking care of immediate needs (typically ps to ms scales) allowing the main power supply to react. But then I don't think I've ever seen a proper definition of "decoupling" in this context.

If the capacitors were intended as reservoir capacitors, they would be far larger. The typical values of decouplers (bypass capacitors) are chosen to short out any HF and reservoir functions would only be a tiny side effect.

This TI paper discusses bypass caps (which we refer to as decoupling caps more often) and why they are needed at each VCC pin with a short trace.

I think of them as the local power supply to whatever device they are close to because they are the immediate source of any variation in current to that device.

which is what I have been saying in the forum for years.