0
\$\begingroup\$

I'm trying to figure out how eddy currents change with frequency. While learning about this its occurred to me that eddy currents might be impacted differently as frequency changes, if the voltage is a square versus a sine wave. (Assuming the peak voltage reached does not change).

A square wave with a negligible rise time is essentially a DC voltage applied to the transformer (for this discussion really just an inductor) for half a period of the operational frequency. As frequency changes, the amount of time the voltage is applied changes, but the resultant slope of current, mmf, H-field, and B-field, is the same regardless of frequency. di/dt = V/L .

Since eddy currents are a result of induced voltage across core resistance, and induced voltage is proportional to slope of flux (V = L * di/dt), then eddy currents remain constant as frequency of a square wave driving voltage changes. (The peak flux reached changes but that does not result in a higher induced voltage.)

But with a sine wave, the slope of the applied voltage actually changes with changes in frequency, so therefore the slope of the flux changes, and this would changes in induced voltages which create eddy currents. For instance higher frequencies induce higher voltages and therefore higher eddy currents.

Is this correct?

\$\endgroup\$
  • \$\begingroup\$ Are you taking into account that if you fed an xfrmr a bipolar square wave, the inductance of the xfrmr will filter your square wave into an ugly jaggy-with-harmonics sine wave? To actually produce a square wave current in the primary is unlikely. \$\endgroup\$ – K H Dec 26 '18 at 4:35
0
\$\begingroup\$

Voltage waveform has a big impact, especially the related dB/dt. In literature you will find versions of the Steinmetz equation which take this into accont. I usually use the "improved generalized Steinmetz equation" which is discussed e.g. in this paper (equations (1) - (3)).

\$\endgroup\$
  • \$\begingroup\$ I'm seeing some thing which are telling me I don't really know whats happening here. a) Quora pages on this topic have a wide variety of answers from engineers which directly conflict with each other, such as here: quora.com/… b) There are research papers like the one you linked that discuss this, including from the 1900's which means the situation is probably not as simple as I described (and I dont know how to explain that, except that maybe there are derivatives of derivatives at work?) \$\endgroup\$ – Not Really Dec 31 '18 at 16:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.