# Ethernet transformer saturation at low frequencies

I am using a 10/100BASE-T Ethernet transformer TG110-E050N5RL (Circuit A in the datasheet) in a circuit. I am applying a 60 Hz 500 Vac hipot across C9 (T1 pin 16 and T1 chassis ground) as shown here: . Pin 14 is open, to simulate a test stand failure. Nothing is connected to T1 pins 9, 10, 11.

I am seeing nothing -- flat, 0 Vdc -- on the corresponding secondary, T1 pins 1 and 2, with a scope during hipot.

My best guess is that core saturating because of the very low frequency of the hipot signal relative to what the transformer was designed for, but I am surprised that I am seeing nothing at all. I have measured the windings' resistances, and the windings do not appear to be damaged.

What, if anything, should I be seeing?

• This is a transformer for signals with some MHz, but not for very, very low frequencies like 60 Hz. The necessary core weight is inversly proportional to frequency. – Uwe Jul 12 '18 at 16:34
• Does Urms=4.44fNAB ring a bell? – winny Jul 12 '18 at 21:19

Consider that the transformer windings have inductance of 0.4 µH, from the specs you provided, and the capacitor is 1,000 pF. At 60 Hz, the capacitive reactance is ~2.7 MΩ and the inductive reactance is ~150 µΩ.

Calculate the ratio -- how much voltage is across the capacitor and how much across the inductor?

BTW, if you were to put a spark gap, perhaps 0.5 mm, in series with the HiPot supply, you probably would get measurable RF voltage out.

This on-line simulator should help you understand. I've entered the capacitor and winding inductance value in order to see what voltage you can expect to see across the driven winding: -

So Vout above is representing the voltage across pins 15 and 16 of your transformer (no need to look at the secondary): -

As you can see, below about 8 MHz it acts as a 2nd order high pass filter and the signal level a hundred times below this frequency (about 80 kHz) is down by 80 dB. At 800 Hz it will be down by 160 dB and at about 80 Hz (near enough 60 Hz) it will be down by 180 dB.

If your hipot voltage is 500 V, then the voltage seen on pins 15 and 16 will be 0.5 uV RMS. Is your measurement equipment capable of seeing this?