Jeff Liebermann wrote on 8/6/2017 7:20 PM:
On Sun, 6 Aug 2017 12:17:16 +0100, ldagan
wrote:

Another thing is attenuation in waveguides. I assume that it would be
the case with many tens of GHZ, for rectangular device. But, as I am
only guessing, as I have been out of the game for long.

At GHZ frequencies, the skin depth is quite shallow. Therefore, only
the inside surface plating really needs to be low resistance. I've
worked with silver plated rigid molded PLASTIC WR90 X band waveguide
which worked but had mechanical rigidity problems. We also had
problems with cracked plating on outside 90 degree corners, such as on
flange to waveguide connections.

https://www.microwaves101.com/encyclopedias/waveguide-construction
http://www.eenewsautomotive.com/news/plastic-waveguides-future-communication-networks

Silver is only some 10% better conductivity than copper and you lose about
half that back due to the skin effect. So in reality it is only around 5%
more effective (unless I'm remembering the numbers wrong and it's actually
20 and 10 per cent).

From what I've read a much larger effect is surface roughness at the really
high frequencies. If the roughness is comparable to the skin effect depth
it creates a longer, more resistive path.

--

Rick C