Re: [femm] transformer capacitance and inductance

[dcm3c@xxxxxxx]

In a message dated 9/21/01 8:12:52 AM Eastern Daylight Time, erdal@xxxxxxxx writes:


> Hi everybody,
> I'm trying to calculate transformer leakage inductance and stray capacitances. Thanks to Cornelius H. Paul's notes on 'Calculating Transformer Inductances with Femm' and Jim Bloor's valuable favour, Femm gave some results for leakage inductance. Now I have two questions:
> 1) What is the effect of frequency on the leakage inductance? When I change the frequency, flux lines and magnetic energy integral changes. Is this logical? 


Frequency typically doesn't have much of an effect on leakage, because the reluctance of leakage paths are usually dominated by air.  Changes in the apparent core permeability as a function of frequency only change the leakage inductance slightly because of all that air.

The thing to look at is how you'd defined your currents.  If you define a specific source current density in a coil region with a nonzero conductivity, the current gets attenuated due to the coil's inductance.  If you aren't trying to get that effect, you can get rid of it by setting the conductivity of the coil region to zero.

2
> ) What about stray capacitances? I think making the secondary current zero (open circuit the secondary winding) and applying the nominal voltage to primary winding and calculating the stored energy in the electric field may help us. After obtaining the stored energy, by using W=0.5*C*V^2 (half of the product of the capacitance and the square of applied voltage is equal to the energy stored in the electric field. ) we can obtain capacitance value. But how can we apply voltage to primary and how can we integrate the electric field over the volume (of course an area in 2D) by using FEMM? And if I manage to calculate this, what is the dependency of capacitance to frequency? Is there any relation between the frequency and the stored energy in the electric field ? ( Like the relation in the leakage inductance calculations) ? Thanks.


FEMM doesn't solve electrostatic problems.  In addition, the low-frequency formulation used in femm specifically assumes that capacitive effects are negligible.  To do what you're interested in, you can't neglect displacement currents any more, and you end up having to use a high-frequency formulation.  An example from the literature would be: 

Charpentier, J.F.; Lefevre, Y.; Lajoie-Mazenc, M., "A 2D finite element formulation for the study of the high frequency behaviour of wound components," IEEE Transactions on Magnetics, 32(3):1098-1101, May 1996.

Dave.
--
http://members.aol.com/_ht_a/dcm3c