The Cause Might be Magnetic Saturation of a Transformer Core
Power Supply Fluctuation and Loads Containing Transformers
Power supply disturbance tests are generally performed to determine whether devices that run on the AC mains supply will malfunction when exposed to disturbance. These tests take several different forms, including the short interruption test, the voltage sag test, and the high harmonic induction test. When testing devices that use transformer-based power supplies or are connected via step-up transformers, depending on the testing parameters, current spikes may be arise in the transformer, causing the AC power supply to become overloaded and compromise testing. In this column, I’ll explain what causes transformers to produce current spikes, and how to configure testing equipment to prevent such spikes from arising.
Current Spikes and Their Causes
Figure 1 shows a setup used to perform a power supply disturbance test.
Figure 2 shows waveforms that were yielded when a PCR-LE series power supply was used to create a momentary interruption in the supply current. As you can see, when the current switches back on, the next negative portion of the sine wave causes a spike in current at point (b).
This phenomenon is caused by the transformer becoming magnetically saturated at point (b). That is, if the input current had not been interrupted at point (a), the transformer’s flux would have been reset by the positive portion of the sine wave that would have arrived. Because of the interruption, however, the flux was not reset, causing a negative magnetic flux to remain in the transformer core. When the following negative portion of the sine wave arrives, it causes the flux density to become even more highly negative, causing saturation magnetization to be exceeded. This ultimately causes magnetizing inductance to fall to more or less zero, causing the spike in current at (b). (See Figure 3.)
Later, both current and voltage become erratic as the AC power supply’s overcurrent protection cuts in and out.
How to Interrupt Current Without Causing Spikes
Figure 4 shows the voltage and current waveforms elicited by a current interruption that did not result in a spike. The short interruption begins when the 50 Hz wave is at a phase angle of 90° and lasts for 10 milliseconds.
We can see that the voltage peaks at point (c) and (d) in Figure 4 are of the same magnitude. In other words, the positive and negative magnetic flux are of equal magnitude and cancel each other out, causing net flux to be reset to zero. Therefore, if the positive voltage waveform and the negative voltage waveform within a single cycle are symmetrical, you will not get current spikes.
Actual standards for voltage dips and short interruptions such as IEC/EN 61000-4-11 come with notices such as that below, which urge the standards committee for the device in question to exercise caution:
When, stipulating performance standards for interruptions of 0.5 stages (1/2 a cycle) in duration on devices fitted with power supply transformers, the device committee should pay particular attention to effects of input current. In such devices, flux saturation in the transformer core after a voltage dip can cause current to reach 10 to 40 times nominal levels.
In cases where a current overload is unavoidable, it is, I believe, preferable to switch to a higher capacity AC power supply. If you are using a power supply from Kikusui’s PCR-LE series, you can simply connect in parallel a second supply from the same series with a high enough rating to compensate for the shortfall. For example, Kikusui’s 2000VA model (the PCR2000LE) connected in parallel to the 4000VA model (the PCR4000LE) using the optional parallel driver (sold separately) can be used in lieu of a 6000VA AC power supply. Note that both supplies need to be run on the same firmware, which should be updated to the latest version. Firmware can be downloaded from Kikusui’s website.