Extrapolation Algorithm
This algorithm is a piece of a complex work made for my MS thesys, and it's another example of ECAD's application applied to a RF measurement's bench design.
The problem solved from this software-module is that one of the specifications of the bench is the possibility of polarize an oscillator (DUT) in automatic and customizable way.
If the oscillators/ILO is designed in one transistor configuration (for example Colpitts's oscillator presented in article 1MHz to 10MHz Electronic Controlled ILO: ), the problem became set the required voltage in the correct sequence for prevent damages of DUT and of instrumentation (SMU or other remote power supply).
For clearness the below flow-diagram:
How you can see in the diagram, the user's desired polarization's values are passed to the algorithm.
The model of transistor and its configuration and a data's base (as tollerances,max nr. of steps allowed, transistor's paramethers...) are required for the problem's formulation.
The most important thing is that this function works with Newton-Raphson's algorithm, in this case a 3x3 system.
The first thing is the initial's conditions (ICs) calcolus, then the function controls the intrumentation's bus and forces the supply to operate at ICs values.
Therefore a DC measurements of intersting node is required for check the polarization status.
The algorithm can calculate the functions (macro-model) and check the error (tollerances).
The termination's criteria give the opportunity to end the loop or calculate the Jacobian and the new values required for the next step
Thank you for reading.