Impedance matching update

[andresmmera]

This is an update for the impedance matching tool.

A few months ago I looked back again at the matching tool code (#470 ) and added some stuff to fit my needs. While doing that, I realized that some stuff could be improved and did a few changes.

What’s new?

• Improved the circuit topology selection logic (or at least that's what I think):
  • The if-else structure was replaced by a switch-case condition. The code looks cleaner this way and that's certainly a more scalable approach. In my opinion, this makes easier to add new synthesis methods.
  • The number of arguments passed to the synthesis functions was reduced by using a struct.

• Added a pop-up window to select the L-section type. L-section matching has always two possible solutions depending on the "path" you choose in the Smith Chart. So far, the program was taking just one of them. IMHO, the user should be allowed to select which type of L-section he/she wants. This is specially important because more often than not the designer needs to have an specific network response (highpass/lowpass)

• Allow different topologies for the input and the output matching networks. So far, the tool implemented the same matching network topology at both sides of the device to be matched.

• The implementation settings of the matching network (as well as the substrate parameters) are put in a separate QDialog. This reduces the logic of the main window because no slot functions are needed.

• For those synthesis techniques using capacitors and/or inductors, the user may specify a finite quality factor for the capacitors and inductors. Ideal elements are used if Q = 1000 (default)

• Added the highpass response in cascaded L-section matching. (It's simply the dual of the one already implemented).

• It was implemented a function for calculated Tee/Pi-type lumped equivalent circuits of a transmission lines. Here you'll find a Sage spreadsheet with the details

  • It were added several design techniques:
    • Pi/Tee matching networks.
    • Transformer-based matching techniques:
    • Pi/Tee matching networks.
    • λ/4 matching: Trivial, but not included so far.
    • Transformer-based matching techniques:
      • Single tuned transformer matching
      • Parallel double-tuned transformer matching
      • Series double-tuned transformer matching
      • It is possible to use either coupled inductors or the π/tee uncoupled equivalents.

• Docs updated. After adding a number of new images, I noticed that qucs-doc/technical was becoming overcrowded. IMHO, this may become a problem as the docs grow, so I put the images related to the impedance matching tool in a separate directory (qucs-doc/technical/images/Synthesis/Impedance_Matching). To my mind, we should do the same for the other tools or stuff in general (in a separate PR). <= I failed to do this (I'll come here later... need to learn more about the autotools build first)

• I prepared the following "diagram". It illustrates roughly the program flow: Matching_Network_Design_Tool.pdf

References: [1] The Design of CMOS RFIC. Thomas H. Lee. Cambridge University Press. 1998 [2] Design of RF and microwave amplifiers and oscillators. Pieter L. D. Abrie. John Wiley & Sons. 1999

Checklist

• [x] It builds with cmake and autotools.
• [x] No newer warnings.
• [x] Update docs and check they build without errors! New literature references are linked.
• [x] Clang-format.