I'm designing the input/output impedance matching circuit for a RF power amplifier, based on the vendor supplied s2p file. I designed separately the input/output matching circuit at the given working frequency. But when I connect the input/output matching networks and the model based on the s2p file and do simulation in ADS, the matching frequency point went away from the designed frequency.

enter image description here

In the diagram, SnP is the model based on the s2p file, the left is the input network, the right is the output network. It seems the output network will change the S11. So the input matching network designed based no load no longer matching at the given frequency.

So, what's the correct way to do a matching based on the s2p file? Do I need matching them in order, that is, first output, then input?

  • \$\begingroup\$ Can you explain more what is your procedure for designing the matching networks? Are you just finding an impedance to put in series with the ports that gives maximum return loss? Or an admittance to put in shunt with the ports? Something more complicated? \$\endgroup\$ – The Photon Dec 16 '18 at 16:58
  • \$\begingroup\$ The one thing I can say based on what you wrote, is you normally would assume a matched load on port 2 when analyzing something about port 1, not "no load". For example, assume a matched load on port 2 while designing the matching network for port 1. (I'm also not clear if you designed your matching network on port 2 to minimize S22, or to minimize the reflection seen by port 2 looking into its load, and I'm not awake enough yet to remember if those would require the same network) \$\endgroup\$ – The Photon Dec 16 '18 at 17:00
  • \$\begingroup\$ The matching is done in ADS, the matching element is LC sections. For input matching, the source termination is assumed 50 ohms, the load impedance is extract from the s2p file, that is, S11. \$\endgroup\$ – diverger Dec 16 '18 at 18:07
  • \$\begingroup\$ Similarly, for output matching, the load impedance is assumed 50, and the source impedance is extract from the s2p file, but this time, using S22. \$\endgroup\$ – diverger Dec 16 '18 at 18:10

Your Answer

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

Browse other questions tagged or ask your own question.