I have been concerned about an issue I have seen with an LLC resonant transformer I have been designing in a pulsed power application. The core is designed to avoid saturation at the maximum applied H-bridge voltage and the core area/number of turns. Not a problem.

However, since this is a multi-resonant transformer, the resonance of which changes with load, there is a low frequency induced within the resonant tank when the H-bridge is turned off or the load is discharged. I have attached a photo of this phenomenon.

My question is, does one expect the transformer to run into saturation if the applied voltage is say 10x less than that which it was designed for, even if no significant magnetizing current exists? I know that the applied H-field induces the B-field which causes the saturation, but the equation for transformer saturation seems to only be a function of applied voltage irrespective of current:

Bs = (VT * DUTY/FREQ)/(N*AE).

But I do not see how a transformer can saturate without any magnetizing current as a H-field would not be produced.





1 Answer 1


If you observe no magnetising current flowing, then the transformer is not saturating. Presumably your voltage is low enough that the volt-second product at your odd low frequency is within the flux handling limit of the core. Saturation is caused by voltage being applied for too long, and can be diagnosed by an excessive current flowing.

If the voltage is 10x less than the design voltage, then the transformer should work, drawing its normal magnetisation current, at 1/10th of the design frequency.

  • \$\begingroup\$ This is in LTSpice though, which does not model transformer saturation. I have modelled the transformer saturation in an alternative simulator and the flux on the scope flies off to the saturation flux despite no or very little current flowing in the H-bridge as the switches are turned off when the voltage is reached before capacitor discharge. \$\endgroup\$
    – jvnlendm
    Commented Aug 26, 2023 at 17:33
  • \$\begingroup\$ @jvnlendm LTspice does support a non-linear inductor. From LTspice help find that there exists "a hysteretic core model based on a model first proposed in by John Chan et al. in IEEE Transactions On Computer-Aided Design, Vol. 10. No. 4,April 1991 but extended with the methods in United States Patent 7,502,723." There's a question on this site here that asks whether or not this model works with linked/coupled inductances. That may still be an open question. \$\endgroup\$ Commented Aug 27, 2023 at 3:47
  • \$\begingroup\$ @jvnlendm Just to add a few links to consider (no idea if they help, but trying anyway): EEVBLOG on coupled saturated inductors, LTwiki on the Chan model, and LTspice.IO group subfolder where find B-H.gif and Xfrmr.zip. \$\endgroup\$ Commented Aug 27, 2023 at 4:08
  • \$\begingroup\$ @jvnlendm If you read German, then this link on dimensioning may also be of some indirect help. Not sure. but I figured I'd add it, just in case. \$\endgroup\$ Commented Aug 27, 2023 at 4:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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