theory and simulation ❌
keep cutting down the trace stub until it's roughly on frequency ✅

continuing with the copper tape shenanigans, I drilled through the board and shorted the end of that stub to have it notch out ~10 GHz instead. It came out a bit low, so I had to shorten it a couple of times. I guess the copper going through to the back of the board effectively adds some length to the stub.

Then I added another stub to notch 5 GHz again.

Now for the actual reason for these stubs: I'm trying to make a subharmonic mixer to play around in the 10 GHz amateur radio band.

I cut a gap between the stubs and added a pair of anti-parallel diodes. The stubs should notch out feed-through of the 10 GHz RF and 5 GHz LO.

Then added a third port for the IF output, just through an inductor as a basic low-pass for now.

@miek which diode is that? Surprised it still does something at 10GHz ..
@trcwm @miek it attenuates by 25 dB. 😄
@dtl @miek yes, but it still mixes.