The Same Test, Three Completely Different Systems
At PCIM Europe in Nuremberg, double pulse testing was everywhere.
That is not a coincidence. The double pulse test (DPT) is the standard method for characterizing switching losses and dynamic behavior in SiC and GaN power devices. As SiC-based traction inverters and power converters move from niche high-end applications to mainstream production, the pressure to perform this test — accurately, efficiently, and at volume — has never been higher.
Walking the show floor, I encountered three distinctly different approaches to the same fundamental measurement. Each one reflects a different stage in the evolution of how power electronics engineers actually do this work: a precision bench instrument running a live demo, a fully automated production-ready test cell, and a compact fiber-optic isolated system designed for the development lab. Together they show what the DPT ecosystem looks like right now.
Setup 1: R&S MXO 5 + Wolfspeed — Precision Bench Characterization
The R&S and Wolfspeed joint booth carried the banner DOUBLE PULSE TESTING, and the centerpiece was a live switching transient capture on a Rohde & Schwarz MXO 5 Series oscilloscope (MXO58 · 2 GHz · 12-bit ADC).
The waveforms on the screen during the demo told the complete story of a SiC turn-off event:
→ White trace — Vds ringing: The drain-source voltage oscillation after turn-off, reaching a peak of approximately 720A equivalent (reading from the left-axis current scale, indicating this was cross-calculated or overlaid). The damped sinusoidal ringing is the characteristic signature of parasitic inductance in the power loop resonating with device output capacitance. The frequency of this ringing — visible as approximately 8–10 full cycles in the 1.8 µs capture window — gives the engineer the bus loop inductance value directly via the formula L = 1/(ω²·C).
→ Blue trace — drain current turn-off: The current falling from the on-state value (~300A) with the characteristic turn-off spike caused by the reverse recovery of the freewheeling diode and the load inductance di/dt. The speed and shape of this falling edge defines the turn-off switching loss.
→ Purple trace — Vgs: The gate-source voltage falling from on-state to off-state, with the Miller plateau visible as the brief flat region during which the switching node voltage swings.
→ Green trace — current tail: The small residual current after the main turn-off event — in SiC MOSFETs this is typically very short, confirming the unipolar device behavior that makes SiC superior to IGBT for high-frequency switching.
The probe configuration visible on the front panel confirms a professional multi-channel measurement chain:
→ CWT Rogowski current sensor (black, left BNC input) — a current transformer wound around the power loop conductor, providing electrically isolated current measurement without inserting resistance into the circuit
→ RT-ZHD16 high-voltage differential probes (blue, × 2) — Rohde & Schwarz's high-voltage differential probes rated to 480V, used for Vds and Vgs measurements in the switching node environment
→ RT-ZHD07 (further right) — a lower-voltage differential probe for gate signal characterization
The oscilloscope settings — 12-bit ADC, 2.5 GS/s, 200 ns/div, 2 GHz bandwidth — represent the measurement architecture discussed in Article 034: sufficient bandwidth to capture the full frequency content of SiC switching transitions, 12-bit resolution to resolve the millivolt-level gate detail on top of the kilovolt-range switching event.
Setup 2: The iPE Automated Test Cell — From Bench to Production
The right half of the R&S + Wolfspeed joint booth was occupied by something that immediately distinguished itself from every other oscilloscope demo at the show: a complete automated power electronics test cell from iPE (·PE·).
The system is a floor-standing instrument cabinet with a transparent safety glass enclosure. Inside the enclosure, a precision aluminum linear rail mechanism — the kind used in industrial automation and precision metrology — provides a motorized platform for DUT mounting. The mechanism's threaded leadscrew and guide rails are visible through the glass, indicating a system designed for repeatable, reproducible contact with power module test points.
The operational workflow this system enables:
→ DUT insertion: The power module (a Wolfspeed SiC half-bridge module, in this case) is placed on the mounting fixture
→ Automated contact: The linear rail mechanism lowers the probe assembly to make precise, controlled electrical contact with all measurement points simultaneously — Kelvin contacts for Rds(on) measurement, current sensors for Is, voltage probes for Vgs and Vds
→ Automated test sequence: The R&S test software (visible on the touchscreen GUI at the control console) drives the MXO 5 oscilloscope through the complete DPT sequence — applying the double pulse stimulus, capturing the switching transients at both turn-on and turn-off, extracting switching energies and device parameters
→ Pass/fail evaluation: The software compares extracted parameters (Eon, Eoff, Vth, Rds(on), Qg) against specification limits
→ DUT release: The mechanism retracts, the DUT is extracted and the next module loaded
The yellow safety sensor at the door base confirms that the system implements safety interlocks: the high-voltage DPT stimulus cannot be applied while the door is open, protecting the operator from the energy stored in the bus capacitance during testing.
The significance of this system at a power electronics exhibition is the transition it represents in the industry. Double pulse testing was historically a manual, lab-based characterization activity performed by expert engineers on individual prototypes. As SiC power modules move into automotive-volume production — where a single EV traction inverter uses six half-bridge modules and production volumes reach tens of thousands per month — the characterization requirement scales proportionally.
The iPE system is the answer to that scaling problem: a DPT cell that can be operated by production technicians rather than application engineers, with standardized measurement protocols, automated data logging, and pass/fail evaluation that doesn't require interpretation of oscilloscope waveforms by hand.
The Wolfspeed Module Family: What's Being Tested
The product literature on the demo desk described the Wolfspeed Overmolded Power Module family — the devices that both the iPE automated cell and the MXO 5 bench demo were characterizing:
EM Series — Industry-Standard Automotive SiC Power Module
The flagship: 750A at 1200V, in a 55×55mm industry-standard footprint, automotive qualified. The key claim: "highest current capability without paralleling" — a single module achieves 750A where competitors require parallel devices, enabling simpler inverter designs. Gen4 SiC MOSFETs inside.
TM Series — Versatile Platform
The range-extender: 650V and 1200V options, same single-switch footprint. The competitive differentiation visible in the bar chart: TMs Up to 18% More Arms/mm² of SiC than competitor — better silicon utilization per unit area, translating to either smaller module size for the same current rating or higher current for the same footprint. The 230A typeTM reference condition: 800V, 5kHz, Tj = 65°C, standard cooler.
WM Series — SiC-Optimized High Power Density
The compact solution: 40×40mm, multi-die SiC half-bridge, "Clean Switching = Up to 38% Lower Switching Losses." The switching waveform comparison shown in the product sheet — non-optimal vs. optimal die/package — is precisely the measurement the double pulse test validates. The "clean switching" claim is not marketing language; it is a measurable waveform characteristic that the iPE test cell extracts from every module.
Setup 3: Cleverscope — Compact Fiber Optic DPT for the Development Lab
The third DPT system I encountered at PCIM was not a live running demo — it was a printed marketing sheet from Cleverscope (covered in Article 033), distributed at the show to describe their double pulse test solution at booth 7-600.
The sheet showed the complete Cleverscope DPT measurement chain:
→ CS1097 DPT Board: the pulse generator that creates the double pulse stimulus
→ CS1302 Isolated Digital I/O: 4 outputs + 1 input, isolated, 1kV — trigger and control interface
→ CS1133 Vsat Probe Head: saturation voltage probe for Rds(on) measurement
→ CS1200 Voltage Digitizer: fiber isolated voltage measurement, ±800mV/±8V, >100 dB CMRR
→ CS1201 Current Digitizer: fiber isolated current measurement, ±63A/±630A, >140 dB CMRR
→ CS548 Fiber Isolated Channel Oscilloscope: the central acquisition unit with 2kV channel-to-channel isolation
The measurement results printed on the sheet provide a direct benchmark for the system's capability, using the same 150A GaN device characterized at the Cleverscope booth (Article 033):
→ RDS ON = 2.54 mΩ
→ VON = 367 mV (on-state voltage, confirming the Rds(on) extraction)
→ Is = 150A (switch current)
→ Bus Loop Inductance = 876 pH (extracted from Vds overshoot)
→ QH output capacitance = 950 pF
→ Turn Off switch Energy = 6.58 µJ
→ Turn On switch Energy = 77.31 µJ
→ VDS rise time = 2.04 ns — the parameter that requires >140 dB CMRR current measurement to capture without common-mode distortion
→ Is fall time = 3.24 ns
→ GVS Threshold voltage = 1.55V / VGS plateau = 2.5V to 3.69V
The consistency between these numbers and those visible on the Cleverscope booth display (Article 033 — same device, same setup, 190A variant) confirms this is real measurement data from the system, not simulated results.
Three Systems, Three Distinct Problems
Seeing all three approaches at the same exhibition makes the market segmentation clear:
→ R&S MXO 5 Series: The precision characterization instrument for application engineers who need to understand device behavior in depth — extract Eon, Eoff, Qg, Rds(on), and investigate the waveform detail that determines whether a gate drive design is optimal. The 12-bit, 2 GHz, multi-channel setup enables the simultaneous capture of all relevant signals with the resolution to extract meaningful parameters. This is the tool for understanding why a device behaves as it does.
→ iPE Automated Test Cell: The production validation system for quality teams who need to verify that every module shipped meets specification. No expert oscilloscope interpretation required — the system applies the test, extracts the parameters, and issues a pass/fail verdict. This is the tool for verifying that a device meets its specification at scale.
→ Cleverscope CS548 System: The development lab instrument for engineers who need to characterize devices in their own circuit environment, with the galvanic isolation necessary to measure high-side signals safely and accurately. Portable, flexible, covering both the measurement and the analysis in one integrated software package. This is the tool for characterizing devices in context — in the application circuit rather than a standardized test fixture.
The double pulse test itself has not changed. The stimulus sequence, the measured signals, and the extracted parameters are standardized. What PCIM was showing is that the infrastructure around that test has matured into distinct product categories serving distinct use cases — a sign of an application area that has moved from early adopter to mainstream engineering practice.
Equipment observed: Rohde & Schwarz MXO58 MXO 5 Series Oscilloscope · 2 GHz · 12-bit ADC · 2.5 GS/s · with RT-ZHD16 high-voltage differential probes and CWT Rogowski current sensor · iPE automated power electronics test cell · Wolfspeed Overmolded Power Modules (EM/TM/WM Series, Gen3/Gen4 SiC, 750A/1200V) · Cleverscope CS548 Fiber Isolated Oscilloscope DPT system · PCIM Europe, Nuremberg
All photos: Thomas · @SignalByThomas