As a corollary to this, can an off-line switching power supply's transformer, say the one from a switching wall-wart or laptop supply, be used as a substitute for the linear mains transformer in the linked question when making isolated waveform measurements? Or would feeding it with a voltage that has low-frequency components in it damage the transformer somehow? It seems like the bandwidth (tens to hundreds of kHz) and isolation (several kV) on such a unit would be more than adequate for the application...

  • 1
    \$\begingroup\$ Depends if you can handle the voltage-time area before saturation and some mean to reset the core. \$\endgroup\$
    – winny
    Jan 8, 2017 at 18:03
  • \$\begingroup\$ @winny -- if a bipolar waveform is applied to the core, is a reset means needed? \$\endgroup\$ Jan 8, 2017 at 18:04
  • \$\begingroup\$ Staircase saturation comes to mind unless you have perfect balance. Or if you can just AC-couple it with a capacitor without ruining your signal. Also, how long duration for your kV signal? Your SMPS lasts 1(+) second during production and hi-pot. If you keep it at that level you will break down the isolation. \$\endgroup\$
    – winny
    Jan 8, 2017 at 18:06
  • \$\begingroup\$ Hrm, sounds like a 100nF or 1uF X2 cap in series might be needed... \$\endgroup\$ Jan 8, 2017 at 18:07

2 Answers 2


As it's designed for high frequency switching, its primary inductance will be hopelessly low for 50Hz or 60Hz use. So it'll appear as a low impedance shunted across the AC supply or system under test, drawing a high current.

That current, in turn, will saturate the transformer core, further decreasing the inductance - unless the measurement is across a low voltage.

Hopefully, if you are measuring across the AC mains supply, it'll just trip a breaker or blow a fuse before anything worse happens.


A ferrite core designed for high f uses a much lower permeability e.g (600 vs 6000) than Electrical Steel laminate and will not work at line frequencies. The energy is not stored in the magnetic material but in the gaps. Low f Ferrite uses hard magnetic iron manganese particles in a ceramic binder which forms the distributed gaps to store B field energy.

Using >10x as many turns which causes other problems with conduction losses and space limits for the copper.

As I recall, you may still find large line 50W transformers using ~ 200x40mm Dia. torroidal cores in some Stereo amplifiers to reduce weight. I believe they more A/m of primary H field and have poor power factor due to the lower inductance.

But Silicon Cold Rolled Grain Oriented Steel (CRGOS) laminate is far more efficient at line f.

e.g. below enter image description here

  • \$\begingroup\$ Since you bring it up... worth also looking up metglas.com and AMT/AMDT transformer technology. Also, a nifty video on pole transformer manufacturing using CRGO: youtu.be/tUO3o5JTGhQ \$\endgroup\$
    – jonk
    Jan 8, 2017 at 22:26
  • \$\begingroup\$ Yes AMT = 15 to 30% of no load losses at 3x cost but full load losses only reduced ~20% of full load loss not used in cheap pole XFMRs \$\endgroup\$ Jan 8, 2017 at 22:40

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.