Physics going on in a transformer

The switch-on transient is all about the time in the cycle the volts come on, and the magentization state - or rather the lack of it, which is not a good match to the state it should be at that point in the cycle.

It is made vastly worse if you switch near a voltage zero crossing, which would be ideal for adding a flat capacitive load, it is about the worst you can do for adding a non-magnetised  inductor. Also the majority of transformers are sized in terms of turns per volt  and core size, such that the core is within a factor of two of saturation in normal use. If you switch on at the wrong point, the peak tries to be double, but the core saturates, and in effect, once all the little magnetic domains that swing around to give the magnetic response have been jammed at one end, it is as if the core then vanishes, and the windings may as well be air-cored . The result is a spectacular current peak, and several cycles of the transformer going thud-ring to recover.

The correct solution is to pre-magnetise the core via a moderate resistance, and then to short out the resistor after a few cycles,  or to apply power via a thermistor, whose resistance drops to near zero as it self heats.

Modern electronic soft starts, can put the mains on just after the top dead centre of the voltage waveform, which is the point when the core would normally not be very magnetised, so it starts on the falling voltage.


Further to the above, and concerns about how transformers relay the load from one side to the other, to be rigorous, I will clarify that  the B field in the core is only almost constant, and to first order is only a function of applied voltage,frequency, and core geometry.

Drawing current from the secondary causes the current in the primary to rise to keep the field constant, and the change in amp turns is equal and opposite on primary an secondary  sides, (well almost, losses ignored here and good magnetic coupling assumer), so the B field in the core remains (almost) constant.

Looking in at the primary side, the impedance presented is the effective inductance of the primary, then shunted by the transformed version of the secondary load.

B field in core does vary with load for non constant voltage transformers, such as current transformers .

some text book explanations are not always clear on this.

Mike

The switch-on transient is all about the time in the cycle the volts come on, and the magentization state - or rather the lack of it, which is not a good match to the state it should be at that point in the cycle.

It is made vastly worse if you switch near a voltage zero crossing, which would be ideal for adding a flat capacitive load, it is about the worst you can do for adding a non-magnetised  inductor. Also the majority of transformers are sized in terms of turns per volt  and core size, such that the core is within a factor of two of saturation in normal use. If you switch on at the wrong point, the peak tries to be double, but the core saturates, and in effect, once all the little magnetic domains that swing around to give the magnetic response have been jammed at one end, it is as if the core then vanishes, and the windings may as well be air-cored . The result is a spectacular current peak, and several cycles of the transformer going thud-ring to recover.

The correct solution is to pre-magnetise the core via a moderate resistance, and then to short out the resistor after a few cycles,  or to apply power via a thermistor, whose resistance drops to near zero as it self heats.

Modern electronic soft starts, can put the mains on just after the top dead centre of the voltage waveform, which is the point when the core would normally not be very magnetised, so it starts on the falling voltage.


Further to the above, and concerns about how transformers relay the load from one side to the other, to be rigorous, I will clarify that  the B field in the core is only almost constant, and to first order is only a function of applied voltage,frequency, and core geometry.

Drawing current from the secondary causes the current in the primary to rise to keep the field constant, and the change in amp turns is equal and opposite on primary an secondary  sides, (well almost, losses ignored here and good magnetic coupling assumer), so the B field in the core remains (almost) constant.

Looking in at the primary side, the impedance presented is the effective inductance of the primary, then shunted by the transformed version of the secondary load.

B field in core does vary with load for non constant voltage transformers, such as current transformers .

some text book explanations are not always clear on this.

Mike