Switching on a transformer

from Wikipedia, the free encyclopedia

When switching on a transformer can in unfavorable phasing of the electrical power to a greatly increased inrush come because of the iron core into the saturation is driven. This effect is also known as the rush effect or switch-on rush . The level of the switch-on current depends on the switch-on time in relation to the time profile of the applied alternating voltage and the magnetic flux stored in the transformer core, the residual magnetism ( remanence ).

Current peak when switching on a transformer in the voltage zero crossing with a negative polarity before switching on, maximum residual magnetism in the core

Basics

Current peak when switching on a 100 VA toroidal transformer

An inductance like the primary coil of a transformer cannot be applied (switched on) to a voltage source like an electromagnet for direct voltage without problems, at which the current increases slowly. The magnetizing current reaches its peak value after less than a quarter of the period of the alternating voltage in the worst case (switching on at zero crossing) despite low remanence and is then a multiple of the peak value of the magnetizing current in the steady state at the end of the voltage half -wave. Transformers with low remanence have an air gap in the iron core, for example welded EI cores, and therefore have a shear (to the side) of the hysteresis curve . The iron cores of conventional transformers, with little or no air gaps, cause problems when switched on. For weight and cost reasons, transformers are designed with the smallest possible air gap in the core and also so that the iron core is just not saturated when the alternating voltage on the primary winding changes polarity after half the period. This is guaranteed in steady-state operation, but not under unfavorable conditions immediately after switching on the voltage:

The worst case is when the voltage is switched on when the voltage crosses zero, if the core is also premagnetized by high residual magnetism in the direction of the switch-on polarity and thus the voltage-time area of the applied voltage wants to drive the magnetic flux in the same direction for a half-wave immediately after switching on. Then the maximum flow can even be up to three times as large as in the steady state, whereby the iron core , which is then not designed for this , is driven far into saturation. This leads to a strong increase in current which is only limited by the copper resistance of the primary winding and the source impedance.

The principles described here for 50 Hz transformers (period of 20 ms) apply analogously to all frequencies. Even with switching power supply transformers ( transformers , the drive electronics company) should this into account, so that it is not overloaded.

In an efficient transformer, U induced is almost as large as U at every moment . You can equate both and transform the induction formula :

This equation must be integrated from the switch-on time to t . The point in time from when the primary winding is fed and whether the iron core was previously magnetized is of decisive importance for the result . There are four combinations in total, which are explained separately:

Switch on at the peak value of the voltage

The magnetic flux in the transformer core after switching on at U  = max. Without residual magnetism, Φ must start at zero
The magnetic flux when the iron core is unfavorably premagnetized when it is switched on. is increased by the value of the integration constant

This relationship describes the temporal flow of flux Φ as a function of the voltage U at the primary coil when it is switched on at its peak value U max .

No residual magnetization, only possible for cores with a large air gap.

The constant of integration C is zero, that is, at t  = 0, Φ is zero. This is synonymous with the fact that the iron core was demagnetized before switching on; the sine function in the picture above applies . This figure also applies to the steady state of the transformer during operation: The magnetic flux periodically changes polarity and runs symmetrically to the zero line: From 0 to 5 ms, wird increases and reaches the peak value of the magnetic flux at 5 ms. The primary voltage, which has meanwhile reversed its polarity, then demagnetizes the iron core up to 10 ms.

With residual magnetization, remanence, for cores with a small or no air gap.

The residual magnetization is influenced by the voltage time area of ​​the last voltage half-wave before switching off. If the iron core still has residual magnetism before it is switched on, the sinusoidal curve of the magnetic flux shifts up or down accordingly, because the flux starts from the remanent flux , then C  ≠ 0 (see blue area in the picture below). At the time of 5 ms, the magnetic flux Φ max is greater or less than in the steady state. If it is larger, the core is briefly saturated and  a positive current peak flows through the primary winding at time t = 5 ms. If this residual magnetism had been known, it would have been possible to switch on 2 ms later and avoid saturation. Unfortunately, it is far too time-consuming to measure the residual magnetism in the iron core before switching on.

See also transformer switching relays : based on the effect of the voltage-time area, it is explained graphically why inrush currents can occur and how these can be avoided entirely with targeted premagnetization.

Switch on when the voltage crosses zero

Magnetic flux in the transformer core after switching on at U = 0. Without residual magnetism, Φ starts at zero, which is why the magnetic flux curve shifts upwards

These relationships describe the temporal flow of flux Φ at the primary coil when it is switched on at the zero crossing of the alternating voltage U.

Without residual magnetization

The constant of integration C is zero at t = 0, Φ likewise, i.e. H. the iron core was demagnetized before switching on. Due to the integration over the entire first half-wave, however, the magnetic flux increases up to 2U max / ω. The function drawn in the picture has been shifted upwards so that it comes into the saturation area marked in red. With conventional transformers, the result has the effect that the iron core reaches the saturation limit, where the high relative permeability µ r of the iron quickly decreases. The core cannot carry the double magnetic flux doppelt. As a result, too little counter voltage is induced and huge current peaks flow through the primary winding. These last a maximum of 10 ms each and repeat themselves slowly, decaying during each alternating voltage period. This is equivalent to a decaying pulsating direct current in the supply line. This is followed by equalization processes that gradually shift the magnetic flux curve Φ into the negative range. Their speed is determined by the copper resistance and the impedance of the supply.

With residual magnetization

If the iron core was magnetized before switching on, the flux curve is shifted further up or down. In this case, the curve must be shifted up or down by the value of the residual magnetism. It then no longer starts at zero. The constant of integration C is not equal to zero at t = 0. As a result, the maximum of the flow can even be up to almost three times higher than the value for which the iron core was designed. From a practical point of view, the core cross-sectional area would therefore have to be designed for 3 times the flux, especially with the core without air gap, as is the case with the toroidal core transformer, so that the core saturation and thus the inrush current is avoided in every switch-on case. As this would make transformers much too large, inrush current limiters or transformer switching relays are used.

Random power on

Manufacturers of semiconductor relays also sell so-called instantaneous semiconductor relays for switching inductances. (Random switching). It can be switched on at a good point in time, but also at a bad point in time, because the remanence position and height are not taken into account. The effect is that the protection no longer triggers as often as, for example, when switching on at zero crossing. However, this does not reliably limit or avoid inrush currents.

Stress-time area

The integration with respect to time means to determine the area under the passage of time of a quantity. The voltage-time area, here the integration of the voltage over time, transports the magnetization along the hysteresis curve, i.e. builds up or reduces the magnetic flux Φ or reverses its polarity. (Without a change in the magnetic flux, there would be no induction and without induction there would be no voltage transmission by the transformer.) In the following, the effect of the voltage-time area is shown using the example of a cosine curve, i.e. when the transformer is switched on at the peak value of the alternating voltage and switched off at different times . Which, however, is rarely used in electrical engineering. Usually, the phase angle is switched on at a certain point in time and switched off when the current crosses zero. See function of the thyristor .

The green marked area between the cosine curve U ( t ) and the time axis from t  = 0 to a selectable point in time t is determined; it is called the voltage-time area (unit Vs). It is a measure of the magnetic flux reached up to this point in time ichten in the core of the transformer; the area is calculated in the adjacent figure for four arbitrarily selected switch-off times (1.5 ms, 3 ms, 5 ms and 7.5 ms) and plotted as a blue bar for the flow Φ.

Voltage-time area when switching on the voltage in the apex with remanence equal to zero, with different switch-off points

In the figure  , the voltage time areas are shown in green for four points in time from the switch-on time t = 0, in each case at the peak of the voltage up to the variable switch-off point. It can be seen that the green area, i.e. Φ, increases from t  = 0 ms to t  = 5 ms. (The magnetic flux, the remanence , is always = 0.) Then the magnetic flux Φ has its maximum value. From this point in time, the areas that continue to grow are below the time axis, are shown in light green and count as negative. As a result, the magnetic flux Φ (the sum of the two areas with the correct sign) becomes smaller again and reaches  the value zero at time t = 10 ms.

The magnetization of the core is built up in the first quarter oscillation (0 to 5 ms) and is reduced in the following quarter oscillation (5 to 10 ms). The dark green area above the time axis and the light green area below the time axis have compensated, the iron core is again non-magnetic. This applies in the steady state and also immediately after switching on, if the residual magnetism, the remanence, was previously zero.

If the iron core had residual magnetism at the beginning of the integration because the transformer was switched off at a moment that was unfavorable for switching on at the apex, this residual magnetic flux in the transformer core is added to or subtracted from the magnetic flux generated by the green area. As a result, the magnetic flux Φ can exceed its value to over saturation, whereby the primary coil loses its inductive resistance.

Possible solutions

Limitation of the inrush current

Switch-on delay with NTC thermistor and relay

There are several ways to avoid tripping the overcurrent protection.

  • In the simplest case, a slow-release or oversized fuse with increased load capacity is used and a certain loss of security is accepted.
  • First, a high-load resistor of a few ohms is connected in series with the primary winding, which is short-circuited after about 100 ms.
  • An electronic circuit determines the zero crossings of the primary voltage and switches on after a quarter of the period when the voltage is at its maximum. It does the opposite of what a zero-crossing switch does and is called a peak voltage switch. However, this is only suitable for transformers that have an extra air gap in the core and therefore only have low residual magnetism.
  • With small transformers up to around 200 W, an NTC thermistor is often connected in series with the primary winding to limit the inrush current . After each switch-off, you must wait for the cool-down time. The thermistor can be short-circuited with a relay so that it can cool down during operation. Then its service life increases considerably. The waiting time is then only required for brief switching off and on again. It does not make sense to connect several NTC thermistors in parallel because both never get hot at the same time and therefore only one carries the current and is thus overloaded.
  • When power transformers , such as those used in power supply networks at high voltage level , are switched on, the supplying generator voltage is reduced as far as possible before the transformer is switched on. Another possibility for limiting is similar to limiting the short-circuit current by means of compensation coils.

Avoidance of the inrush current

Another possibility is to completely avoid the inrush current with a transformer switching relay or a soft start switchgear. This means that a transformer can be switched on frequently in succession without increased current load and without waiting times. Because of the increased circuit complexity and thus the price, this is not used for all devices.

literature

  • Gerd Fehmel, Horst Flachmann, Otto Mai: The master's examination in electrical machines . 12th edition. Vogel Buchverlag, Oldenburg and Würzburg 2000, ISBN 3-8023-1795-5 .
  • Gregor D. Häberle, Heinz O. Häberle: Transformers and electrical machines in power engineering systems . 2nd Edition. Verlag Europa-Lehrmittel, Haan-Gruiten 1990, ISBN 3-8085-5002-3 .
  • Günter Springer: Expertise in electrical engineering . 18th edition. Verlag Europa-Lehrmittel, Wuppertal 1989, ISBN 3-8085-3018-9 .

Individual evidence

  1. Rush effect - inrush current ( memento of the original from June 11, 2010 in the Internet Archive ) Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice. , accessed on October 9, 2010  @1@ 2Template: Webachiv / IABot / www.hgaechter.ch
  2. R. Gudat, U. Schulz, B. Weidmann, M. Kurth, E. Welfonder: Instructions for action for rebuilding the partial network without external voltage specifications . , sixth GMA / ETG conference, 2003, Munich

Web links

The link above and the individual verification 1 above describe the peak voltage switching on of an idealized transformer which, for example, has no remanence, but this does not apply to conventional transformers. In addition to transformers for special applications, transformers have a more or less large remanence, depending on the type of iron core, which strongly influences the switch-on behavior when, as described in the link, the transformer is switched on at the peak of the alternating voltage. The smaller the air gap, the greater the remanence. As the remanence increases, the inrush current also increases significantly when the peak is switched on.