Surely it's a case of the output RF stage amplifier passing through a Co-ax switch to route RF via a CU to whichever antenna is in use for that particular TP.
Ohhhh... thats tricky stuff. Switching full power signals makes it very vulnerable to
failing coax relays. Having switches before the final amplification makes things easier.
Combiners are not necessary on every TP then. This assumes the TWT's are wideband and
linear which they hopefully are. Pure guesswork but narrowband combiners are often larger
than ones with wider bandwidth. I've opened some scrapped fancy coax relays and
the autopsy was "interesting".
Anyone got information on the RF system "layout" in these birds?
Edit - answer ended up inside qoute.