anti-ENIG rant

The surface resistance loss on the rough bottom side of the trace is going to make much more difference than the top-side plating.

Besides, copper is a better conductor than gold.

Cheers

Phil Hobbs

onsdag den 12. maj 2021 kl. 23.46.32 UTC+2 skrev John Larkin:

since it is there to protect the copper so it easy to solder it would be pointless and a waste of gold to put it under the solder mask

Selective plating for ENIG seems to be the standard. If you want "full body" ENIG (plating before solder masking) you can ask for it at many fabs.

I was about to say, use GCPW instead of microstrip and you can pay less attention to a lot of things including surface roughness. But there's a decent-looking white paper from Rogers that says different:

-- john, KE5FX

Right. Most of the e-field is in fiberglas, not air, and the bottom side of most PCB traces are horribly treated to improve adhesion. The roughness makes big skin losses. We are researching using low-loss Isola laminates with relatively smooth copper on the epoxy side.

Mirror-finish traces are available but adhesion suffers.

Om my upcoming proto board, I'll include a couple of 3" long traces, microstrip and CPW, and compare risetimes. On FR4, I'd expect maybe

It would be nice to speed up runs on FR4 and avoid exotic laminates.

There's not a lot of nickel or gold on the places that do get plated. You might have scrapped it off with solder mask.

And even nicer if you could rewrite the laws if physics.

The nice thing about buried stripline is that you don't case as much about the adhesion of the buried copper to the board, so that you could use copper strip that was flat on both faces.