Hello,

I picked up an ADS lowpass filter schematic from another designer. The lowpass uses lumped element coils and is placed in a housing. The original designer placed some mutual coupling elements between the inductors to tweak the response due to the coil proximity and housing.

I am interested in calculating the power dissipated in the inductors. If I disable the mutual coupling, the calculated dissipations look reasonable (relatively light coupling--disabling minimally effect the response). However, with the mutual coupling elements enabled, the power dissipated among the inductors (four inductors in my case) are significantly different and in some cases negative. However, if you sum the power among the four inductors (with mutual enabled), the total power dissipation looks reasonable. The total loss of the network looks approximately the same whether the mutuals are enabled or not. However, the power distribution inside the filter is not right.

So, my question(s): Why does the power dissipation distribution change so radically when adding the mutual coupling. [Power is calculated as 0.5*real(V x I*) and it is calculated into the device and out of the device--the difference being Pdiss]. Does this mean that the coupling terms are physically wrong? Or, does something get messed up in the matrix when mutuals are enabled such that the power dissipated becomes confounded internal to the structure? Is there a more generalized formulation for calculating dissipated power?

I picked up an ADS lowpass filter schematic from another designer. The lowpass uses lumped element coils and is placed in a housing. The original designer placed some mutual coupling elements between the inductors to tweak the response due to the coil proximity and housing.

I am interested in calculating the power dissipated in the inductors. If I disable the mutual coupling, the calculated dissipations look reasonable (relatively light coupling--disabling minimally effect the response). However, with the mutual coupling elements enabled, the power dissipated among the inductors (four inductors in my case) are significantly different and in some cases negative. However, if you sum the power among the four inductors (with mutual enabled), the total power dissipation looks reasonable. The total loss of the network looks approximately the same whether the mutuals are enabled or not. However, the power distribution inside the filter is not right.

So, my question(s): Why does the power dissipation distribution change so radically when adding the mutual coupling. [Power is calculated as 0.5*real(V x I*) and it is calculated into the device and out of the device--the difference being Pdiss]. Does this mean that the coupling terms are physically wrong? Or, does something get messed up in the matrix when mutuals are enabled such that the power dissipated becomes confounded internal to the structure? Is there a more generalized formulation for calculating dissipated power?