Цитата
It has a serious design flaw: the Port 2 return loss is horrible, so the DUT is mismatched and huge ripple can be observed when measuring a coaxial cable loss (S21). How are you supposed to be able to tune a cavity filter with one port so badly terminated?
You can find there are two rf paths in parallel, so 25ohm at low frequency and whatever above 1GHz where transmission lines matters. Those RFSW6024 SPDT switches are non-reflective, so ports allways present 50ohm. Look at C128, C118, U2 and U36.
Mismatch could be alleviated placing an attenuator at port 2, but it will reduce the measurement dynamic range.
https://www.eevblog.com/forum/rf-microwave/...143/#msg1220143
The truth is out there...