I want to measure the isolation of the following directional coupler (for more details of that specific component, go here).

enter image description here

Isolation can be evaluated through the scattering parameter S21, but how can I measure it? Obviously I can measure S14, S41, S13, S43, S31, S33... but what about S21? I cannot connect port 2 to a vector analyzer because it is internally terminated on a 50Ohm load.


You do not need S21 because, as you point out, you can't measure it, and you're not going to use it.

What's relevant to you is S34, the unwanted, ideally 0 reverse coupled signal.

Your directivity, the difference between the forward and the reverse coupling, is the ratio of S31 to S34.

This answer assumes you're the user of this packaged coupler. If instead you're the designer of the bare metal coupler component that's inside the package, then you will want to measure S21, but then you'll have access to all four ports of the coupler.

| improve this answer | |
  • \$\begingroup\$ And what about the isolation, which should be equal to: 1/|S21|^2? \$\endgroup\$ – Kinka-Byo Dec 14 '19 at 12:30
  • 1
    \$\begingroup\$ @Kinka-Byo sorry, I thought you meant directivity. How do you define isolation? What does it mean in the design that this component is used in? 1/S21 etc sounds like reflection off the termination, which merely modifies the S11 of the coupler. if so, it's already included in your S11 measurement. You'll notice in your linked data sheet there's no specification for 'isolation', which might be a clue that it's not important, or well defined. \$\endgroup\$ – Neil_UK Dec 14 '19 at 12:35
  • \$\begingroup\$ With Isolation I mean the ratio between incident Power and the back power. It is written here (tutorialspoint.com/microwave_engineering/…). But It is also written that it is the sum of Coupling and directivity, so I think I can evaluate it by adding them after using your formula for directivity \$\endgroup\$ – Kinka-Byo Dec 14 '19 at 12:52
  • 1
    \$\begingroup\$ @Kinka-Byo That tutorialspoint link uses some wierd notation. In the industry, we use 'through power', 'coupled power', and 'isolated power'. I assumed by 'back power' you meant power reflected back to the source. Yes, by all means use coupling+directivity to estimate the isolated or back power. However, in a 3 port coupler with one port internally terminated, it's not possible to measure directly. Why do you need it, estimating temperature rise of the termination resistor? \$\endgroup\$ – Neil_UK Dec 14 '19 at 15:04
  • \$\begingroup\$ Thank you very much. Not a precise reason to evaluate it, it has just been asked to me to find it \$\endgroup\$ – Kinka-Byo Dec 14 '19 at 19:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.