SiC Power Device|Application

Points to Note When Measuring SiC MOSFET Gate-Source Voltages: General Measurement Methods

2024.07.24

Points of this article

・When an extension cable is soldered onto a DUT terminal and a voltage probe is connected to perform measurements, if switching speeds are fast, the observed waveform changes considerably.

・Due to the effect of an extension cable that has been installed to perform measurements, a waveform that is completely different from the original waveform is observed.

・When performing measurements, one must always keep in mind that a waveform being observed may actually be different from the original waveform.

SiC MOSFETs have excellent switching characteristics, but because of the extremely large changes in voltages and currents that occur during switching, it is necessary to accurately measure the gate-source surges that occur, as explained in the “SiC MOSFETs: Behavior of Gate-Source Voltages in a Bridge Configuration” article of the Tech Web Basic Knowledge section on SiC Power Devices. In this article, points to be noted when measuring gate-source voltages are explained. The explanation uses SiC MOSFETs in examples, but it can be referenced as knowledge that is common to power devices in general, such as ordinary MOSFETs and IGBTs.

Points to Note When Measuring SiC MOSFET Gate-Source Voltages: General Measurement Methods

Power switching devices that are used in such products as power supply units often have a heat sink mounted for cooling, and when measuring voltages across device terminals, a voltage probe or the like cannot usually be mounted directly on the terminals. For this reason, extension cables may be soldered onto device terminals, and a voltage probe is then connected outside the product housing to perform measurements.

Fig. 1 is an example in which a heat sink is mounted on a ROHM evaluation board (P02SCT3040KR-EVK-001), and extension cables are connected to a voltage probe. Extension cables (approx. 12 cm long), used for connection to a voltage probe, are soldered to terminals of the devices under test (DUTs); the extension cables are twisted to suppress the effect of radiated noise. In this example, the measurement method was employed to observe waveforms when double pulse tests were performed using the bridge configuration shown in Fig. 2.

Fig. 1. Measuring gate-source voltages using
extension cables connected to a voltage probe

Fig. 2. Double pulse test circuit

ROHM MOSFETs (SCT3040KR) were mounted on the high side (HS) and low side (LS) of the double pulse test circuit, and the HS was switched while the LS was kept in the always-off state (gate voltage 0 V). The extension cables shown in Fig. 1 were soldered directly to the HS gate and source terminals.

Fig. 3 shows measured gate-source voltage waveforms. When the external gate resistance RG_EXT is 10 Ω, the effect of the extension cables is not very large, but when RG_EXT is set to 3.3 Ω and the switching speed is raised, noise due to voltage and current changes as well as circuit-based high frequency action are induced, and the measured waveforms change considerably. This is an example in which the effect of extension cables installed for the purpose of measurement cause changes in the frequency band of the measurement instrument, and addition of a superfluous impedance causes a waveform that is completely different from the original waveform to be observed.

Turn-on waveforms

Turn-off waveforms

Fig. 3. Gate-source voltage waveforms measured with extension cables mounted. The waveforms are completely different from the original waveforms.

As a point to be noted here, observations must be performed while always keeping in mind the question of whether a waveform that is observed is actually the original waveform, or is instead different from the original waveform due to an effect of some kind. In light of this, in addition to using a method that enables accurate observations, it is also necessary to understand the factors that can affect observations.

Fig. 4 shows the equivalent circuit of the differential voltage probe used in the measurements (see references *1, *2). Ordinarily, the frequency characteristic of a voltage probe is set including the probe head. However, when extension cables are mounted on measurement terminals of a DUT, if observing fast switching waveforms on the order of tens of nanoseconds, resonance phenomena are induced between a floating inductance LEXT and the input capacitance C of the voltage probe bodies. As a result, high-frequency voltage ringing occurs superimposed on the voltage waveform, and a surge larger than the actual surge may be observed.


Fig. 4. Equivalent circuit of the differential voltage probe

  • *1. Reference documents: “ABCs of Probes”, Application Note (No. EA 60W-6053-14), Tektronix, January 2016, and “WaveLink Medium Bandwidth(8-13 GHz) Differential Probe” Operator’s Manual (924243-00), TELEDYNE LECROY, May 2014
  • *2. Reference Document: “WaveLink Medium Bandwidth (8-13 GHz) Differential Probe” Operator’s Manual (924243-00), TELEDYNE LECROY, May 2014

    SiC Power Device

    Basic

    Application

    Product Information

    FAQ