Transformer as 2 coupled nonlinear inductors

[Jan10]

I am trying to model a transformer and after much testing I have modeled the primary and the secondary separately as nonlinear inductances as a function of current. I then added coupling coefficients. When I tried to run the simulation I received the following errors:

SPICE Netlist Error in schematic RefDes 'k2', Element 'LL1' referenced in this component does not exist in scope, the inductor(s) will not be coupled ( I had 6 of this type of error)

SPICE instantiation of component 'kk2' failed. Component will be ignored by simulation. ( I had 3 of this type of error).

Am I to conclude that the "inductor coupling" only works for linear inductors? Do I have to try a different approach or is there another way to couple them.

[David22JRA]

Hello Jan10

Thank you very much for posting your concern. Would you like to share with us some captures about the error and also share with us a capture about the circuit?

What Multisim Version are you using?

Best Regards
David J Ramirez

[Jan10]

Hello David22JRA,

     I am using NI Multisim 14.0 Power Pro Edition. I have attached a very simplified version of the simulation. As It is a pretty weird circuit I am going to include a brief summary as to what the circuit does. The circuit is part of protection circuitry associated with three phase power generation in an aircraft. In the upper left corner you see the stepped down phases, A phase leads B phase by 120 degrees which then leads C phase by 120 degrees. A reference signal is produced by R405, R407 and C403 which is tied to the center tap of the transformer. The current source on the left is the output of a current transformer which is monitoring the current in the C phase bus. When the current in the C phase leg is zero or if it is in phase with the C phase voltage, the output (the DC voltage at pin 18 of the board) is at a nominal level about 4 volts DC. When the current in the C phase load is either leading or lagging the C phase voltage, ie when the load is either capacitive or inductive respectively, the output is driven less than or greater than 4 volts. If the output voltage deviates enough from 4 volts, a sequence of events occurs to first isolate the various busses and then remove power from the offending bus and then reconnect busses without the offending bus.

     The reference voltage at the center tap is a relatively large AC signal = about 40 Vpk. When the reference signal is positive and the input current lags the C phase voltage by 90 degrees, the reference signal is enhanced by the action of the transformer. Similarly when the reference voltage goes negative it will be diminished. this leads to a net positive and an average value above the 4 volt nominal. Similarly, when the input current leads the C phase voltage by 90 degrees, effect on the reference signal is the opposite leading to an output that is less than the nominal 4 volts DC.

     I have modelled this circuit fairly successfully with linear, coupled inductors. The bottom line is that, the simulation is off about 60 degrees in phase and about 10% in magnitude when compared to the actual system performance. We have shown that the size of the primary and secondary currents are large enough so that probably due to core effects, the inductances of the primary and secondary cannot be described by one value.

Tests show that for the range of primary currents expected the primary inductance can range from 4.8 to 136 mH. Similarly, the secondary inductance can range from  .17H to 2.92H.

     It makes no sense to me that multisim can handle linear inductors with coupling coefficients but can't handle nonlinear inductors with coupling coefficients!

     At any rate, thanks for your consideration.

Jan10