When analyzing bridge circuits composed of power devices, some engineers have a misconception about signal frequency: "There's no high frequency; the gate signal frequency is only a few tens of kHz!"

　　This understanding lacks the concepts of fundamental and harmonic frequencies, confusing switching frequency with signal frequency. When referring to high frequency here, it pertains to high-frequency components within the signal, not the switching frequency.

　　Power semiconductor devices have advanced towards high voltage endurance, low on-state resistance, and low switching losses. This advancement is directly reflected in the large dv/dt of the device's Vce (or Vds) voltage, especially in SiC or GaN devices, where the switching time can be reduced to just a few nanoseconds. This results in high-frequency energy at the midpoint of bridge circuits composed of these power devices. Consequently, when testing the Vgs or Vge signals of wide bandgap semiconductors, there exists high-frequency high-voltage common-mode interference.

　　Now we understand that the high frequency in this signal refers to high-frequency components within the signal itself, and is unrelated to the switching frequency. Even if the switching frequency is 1Hz, there can still be high-frequency components exist. Since these high-frequency components exist in the signal, it requires a high-bandwidth probe to handle them effectively. Otherwise, high-frequency oscillations or spikes in the signal might not be visible due to limitations in the probe's bandwidth.

　　One of our customers who involves in power semiconductor equipment business, they were use a differential probe (with 5MHz bandwidth limit function) to measure the gate-emitter signal Vge on IGBT. They did this without considering the high-frequency components in the signal, resulting in very clean and smooth waveforms displayed on the oscilloscope. It’s a satisfying result for engineers, however, this type of testing equipment may pose hidden risks for IGBT production lines, as some inherent high-frequency spikes could be missed.

　　In a power semiconductor testing laboratory in Dongguan, engineers were conducting tests and validation on GaN devices. Unlike the previous case, they suspected the presence of high-frequency abnormal signals but couldn't see them. Using a high-voltage differential probe was ineffective for testing the High-side. When they switched to a 500MHz bandwidth optical isolated probe to test the Vgs of the High-side voltage, they discovered oscillation signals as high as 580MHz. These oscillations originated from board parasitics and interactions within the devices. For more accurate observation and measurement, they determined that an optical isolated probe with a 1GHz bandwidth would be necessary.

　　In conclusion, switching frequency refers to the frequency of power device switching actions, while signal frequency refers to the high-frequency components within the signal. In power semiconductor devices, with advancing technology, devices switch faster, resulting in the presence of high-frequency energy in bridge circuits composed of power devices. Therefore, when testing signals on wide-bandgap semiconductor upper transistors, it's necessary to use probes with high bandwidth to accurately observe high-frequency oscillations or spikes within the signal. Otherwise, due to the limitations of probe bandwidth, these high-frequency components may not be observable.

　　Therefore, to correctly understand the signal frequency in bridge circuits composed of power devices, it's essential to differentiate between switching frequency and high-frequency components within the signal. Switching frequency refers to the frequency of power device switching actions, while signal frequency pertains to the high-frequency components within the signal, influenced by factors such as device speed, PCB layout, and component parameters.