Wide bandgap characterization

Measuring switching losses in ultra-fast Wide Bandgap (WBG) semiconductors like SiC and GaN requires DPT. Because standard operating sweeps happen too quickly to isolate individual transitions cleanly, DPT isolates a single turn-on and turn-off event under controlled current and voltage baselines to calculate precise energy dissipation.

Switching loss is an energy measurement (expressed in Joules) derived by multiplying the instantaneous voltage across the transistor by the current passing through it, and then integrating that power waveform over the duration of the switching transition.

To measure the ultra-fast nanosecond switching edges of GaN and SiC semiconductors accurately, you must eliminate probe-induced noise and match your hardware bandwidth to the speed of the transition edge.

Standard oscilloscope probes with long alligator ground clips introduce parasitic inductance that distorts the signal, creating artificial voltage overshoots and ringing that can mask true device behavior. To prevent this, engineers must bypass standard grounding wires entirely, opting instead for dedicated PCB-mounted test fixtures, such as MMCX coaxial connectors or short metal ground springs pressed directly against the local ground plane.

Instrumentation must be selected based on edge rise time rather than switching frequency, utilizing the high-speed knee frequency formula to ensure the oscilloscope and probe combination provides enough bandwidth headroom (typically 1 GHz or higher for a 1 ns edge). For critical high-side measurements in a half-bridge layout, standard differential probes lack the necessary Common-Mode Rejection Ratio (CMRR) to withstand extreme voltage steps; therefore, optically isolated fiber-optic probes (like R&S®RT-ZISO) must be deployed to completely isolate the instrument and block common-mode leakage.

Finally, because power calculations multiply voltage and current waveforms simultaneously, probes must be physically aligned using a calibration deskew fixture beforehand, as a timing mismatch of even a single nanosecond can alter calculated switching metrics by more than twenty percent.

RDS(ON) represents the total internal resistance between the Drain and Source terminals of a MOSFET when it is fully turned on and operating in its linear conduction state. This parameter serves as the primary bottleneck for efficiency while a transistor is actively passing current, as it directly dictates the system's conduction loss.

Conduction loss is the continuous thermal energy dissipated as heat during the entire duration of the transistor's "ON" state. From Joules’ Law, it scales exponentially with the root-mean-square (RMS) drain current passing through the device. Minimizing RDS(ON) is critical because excessive conduction loss degrades overall power converter efficiency and generates significant thermal stress that can threaten system reliability.

WBG semiconductors like SiC and GaN provide massive advantages here because their superior material properties allow for a much higher electron density, enabling them to achieve an ultra-low RDS(ON) in a significantly smaller physical die size compared to legacy silicon.

Measuring the dynamic on-resistance ,RDS(ON), of a power MOSFET while it is actively switching inside a circuit is one of the most difficult challenges in power electronics. This difficulty arises because the oscilloscope must measure a tiny millivolt drop across the transistor when it is ON, but survive hundreds of volts across the exact same terminals a microsecond later when the transistor turns OFF.

1. Clamping Circuit Method
The clamping circuit method overcomes the extreme voltage differences between a transistor's ON and OFF states to protect the oscilloscope from input amplifier saturation. When a probe is connected directly to the drain, zooming in to resolve a tiny millivolt drop during the ON-state will cause the high voltage of the OFF-state to severely blind and overdrive the scope, rendering subsequent data meaningless. To prevent this, engineers insert a specialized fixture utilizing ultra-fast diodes to cut off the high OFF-state voltage spikes, allowing the oscilloscope to safely measure the true, un-saturated ON-state voltage drop RDS(ON) at a highly sensitive vertical scale.

2. Direct Mathematical Formula
The direct mathematical formula allows an oscilloscope to calculate and map RDS(ON) continuously across the time domain using its real-time math function. Once the clamping circuit isolates the true ON-state voltage drop, the scope divides this time-varying voltage waveform by the concurrent drain current I_D captured via a high-bandwidth current sensor, such as a current viewing resistor (CVR). The resulting formula, [RDS(ON)(t) = V_DS(ON)(t) / I_D(t)], lets engineers visually observe exactly how the transistor's internal resistance settles immediately following a fast-switching transition.

Static and dynamic testing represent the two fundamentally different environments used to evaluate a transistor's internal resistance. Static testing uses a curve tracer or Source Measure Unit (SMU) to apply steady DC signals to a non-switching device, which is ideal for baseline datasheet verification but fails to replicate actual system operation.

Conversely, dynamic testing measures resistance while the device is actively switching under load, such as during a DPT, which is critical for capturing real-world operational anomalies like the temporary dynamic RDS(ON) spikes caused by high-voltage charge trapping in GaN and SiC devices.

Switching emission refers to the unwanted electromagnetic interference (EMI) and high-frequency noise generated by a transistor during its ultra-fast turn-on and turn-off transitions. When a power semiconductor switches states, the rapid change in voltage over time (dv/dt) and current over time (di/dt) forces high-frequency energy to radiate through the air or conduct through circuit traces, causing signal distortion or component malfunction.

High common-mode measurements are necessary because key components in a power converter, specifically the high-side gate driver and high-side transistor in a standard half-bridge configuration, do not connect to a stable system ground. Instead, their reference node (switch node) rides on a massive voltage waveform that violently swings hundreds or thousands of volts up and down between the DC bus rails during every single switching event. To capture the delicate millivolt signals controlling these high-side switches without destroying test equipment or corrupting data, instrumentation must be capable of floating atop and ignoring these immense common-mode voltage swings.

Consequently, conventional high-voltage differential probes are no longer sufficient for evaluating modern wide bandgap (WBG) technologies. Characterizing ultra-fast switching edges requires high-bandwidth performance (up to 1 GHz or more) paired with exceptionally high Common-Mode Rejection Ratio (CMRR) across that entire frequency spectrum. Because legacy probe designs exhibit rapidly degrading CMRR at high frequencies, specialized optically isolated measurement systems must be deployed in these demanding testing scenarios.