A Company You Haven't Heard of, Solving a Problem You Know Too Well
At PCIM Europe 2025 in Nuremberg, most booths belonged to the names everyone in power electronics recognizes. Then I walked past a stand with a green logo I hadn't seen before: Cleverscope.
The table skirt said it plainly:
CS548 Isolated Oscilloscope — 4 Isolated Channels · 2kV between Channels · Isolated Input Pod · Isolated Output Pod · 100V/ns CM
And the rollup banner behind it went further: "The world's only integrated fibre optic isolated oscilloscope. Measure the High Side."
Cleverscope is a New Zealand-based instrumentation company. The product lineup at PCIM was unfamiliar to most visitors. The customer list on the far panel was not: Tesla, Siemens, ASML, Infineon, Keysight, Dyson, NASA, Williams Racing, and a number of European universities. Companies that are not known for buying instruments from unknown vendors.
The engineering problem Cleverscope is solving is one that every power electronics developer eventually confronts, usually at the worst possible moment: how do you measure the gate voltage of the high-side switch in a half-bridge converter — a signal that is referenced to a node swinging at 100 V/ns — without that common-mode transient destroying your measurement or your oscilloscope?
The High-Side Measurement Problem
In any half-bridge power converter — the topology that forms the basis of motor drives, inverters, DC-DC converters, and power factor correction stages — there are two transistors: a low-side device and a high-side device. The low-side device's source is referenced to the negative DC bus. The high-side device's source is the switching node — the midpoint of the bridge.
Measuring the low-side gate voltage (Vgs) is straightforward: the reference is ground, the signal is a few volts. Connect a standard probe, read the waveform.
Measuring the high-side gate voltage is fundamentally different. The high-side source — which is also the local reference for the gate drive circuit — is the switching node. That node swings from near zero to the full bus voltage at every switching transition. In a 400V bus with GaN switching at 10 ns transition times, that's 40 V/ns. In 800V SiC systems, it can be 100 V/ns or more.
A standard differential probe, connected between the high-side gate and the switching node, sees this common-mode transient on both inputs simultaneously. Even with 80–100 dB CMRR, a 400V common-mode swing at 100 V/ns will inject millivolts of error into the differential measurement — on a signal that spans only about 20V. The error is not negligible. It can make a clean gate waveform look like it has oscillations, or make a real oscillation look clean. Either way, the measurement is unreliable.
The conventional workaround — a battery-powered floating oscilloscope attached to the switching node reference — works at lower switching speeds but becomes dangerous or impractical at SiC and GaN switching rates. The floating instrument becomes an antenna for common-mode energy, the battery ground becomes a source of injected current, and the measurement validity is questionable.
The CS548's approach: optical isolation. Fibre optic cable between the measurement point and the oscilloscope. No galvanic path for common-mode energy to travel.
CS548: The Architecture of Optical Isolation
The CS548 is not an oscilloscope with isolated inputs — it is an oscilloscope whose architecture is defined by isolation as a first principle. The specifications from the booth confirm the key parameters:
→ 4 channels, each isolated to 2 kV from every other channel — not 2kV to ground, but 2kV between any two channels. In a full three-phase inverter with six switches, this allows simultaneous measurement of multiple high-side gate signals without any channel sharing a reference
→ 14-bit ADC, 200 MHz bandwidth — the resolution to see the millivolt-level gate voltage detail that determines switching behavior, with the bandwidth to capture sub-nanosecond transients
→ 100 V/ns common-mode rejection rate — the specification that is the direct answer to the GaN and SiC switching environment. This is not a passive probe CMRR figure — it is the rate at which common-mode voltage can change without affecting the differential measurement
→ Isolated Input Pod and Isolated Output Pod — the remote sensing nodes that attach to the circuit under test, with fibre optic cables carrying the signal back to the main oscilloscope unit. The measurement point and the display unit are galvanically disconnected
→ 0–65 MHz isolated signal generator — built-in, for frequency response analysis (FRA) without requiring an external signal source
→ 4× remote fibre sockets — supporting up to four remote digitizer units in addition to the four built-in channels
The panel describing application configurations shows what this architecture enables in practice:
→ Four units: 4× 2kV isolated channels, capturing 16× 50ns pulses with better than 700 ps time dispersion — the resolution needed for GaN switching characterization
→ Two units: 8× 2kV isolated channels, capturing both high sides of a full bridge, including the voltage across the load — simultaneously, time-correlated, with no shared ground reference
The Remote Digitizer Family: Measurement Where the Signal Lives
The Remote Digitizer product line extends the CS548's optical isolation philosophy to standalone sensing nodes that can be placed directly at the circuit under test, with fibre optic cables running to the main oscilloscope:
CS1200 Voltage Digitizer — the primary voltage measurement node:
→ Fiber isolated, 1–30 m cable run
→ ±800 mV, ±8V ranges
→ 14-bit resolution, 300 µV rms noise
→ >100 dB CMRR @ 50 MHz
→ 200 MHz bandwidth
→ Compatible with standard probes or probe tip
CS1201 Current Digitizer — the current measurement companion:
→ Fiber isolated, 1–30 m cable run
→ ±63A, ±630A ranges
→ 14-bit resolution, 15 mA rms noise
→ >140 dB CMRR @ 50 MHz — the figure that matters most for current measurement in a switching environment
→ 200 MHz bandwidth
CS1501 Current Sensor — 1 mΩ shunt, 163 µH insertion inductance. The shunt that pairs with the CS1201 for precision current measurement in high-frequency switching circuits.
CS1202 Transistor Digitizer — listed as "in development" at PCIM — a dedicated measurement node for transistor characterization combining Vds, Vgs, and Id measurement in a single isolated unit.
The 140 dB CMRR figure for the CS1201 is worth dwelling on. In a GaN half-bridge running at 400V with 10 ns switching transitions, the common-mode voltage rate is 40 V/ns. A current sensor with 140 dB CMRR reduces that common-mode signal by a factor of 10 million before it reaches the measurement output. The current measurement that reaches the oscilloscope is the actual inductor current — not a mixture of actual current and common-mode leakage.
The Double Pulse Test Results: Real Numbers From a GaN System
The leftmost panel at the booth — the double pulse test results for a 190A high-side GaN device — is the most technically specific content at the booth, and the most directly useful for power electronics developers.
The measurement summary visible on the panel:
→ Time alignment: Is lags Vgs by 5.1 ns — the propagation delay between gate signal and current response, a critical parameter for dead-time optimization
→ QH output capacitance: 908 pF — extracted from the switching transient, a parameter that doesn't appear cleanly in datasheets for all operating conditions
→ Bus loop inductance: 914 pH — the parasitic inductance of the power loop, calculated from the voltage overshoot during switching
→ Is fall time: 3.2 ns — current fall time during turn-off
→ VDS rise time: 2.0 ns — drain-source voltage rise time during turn-off (complementary to current fall)
→ QH RDSon: 5 mΩ — on-state resistance, extracted from in-circuit measurement
→ QH conduction energy: 77.31 µJ
→ QH turn-off energy: 6.58 µJ
→ QH turn-on energy: 77.31 µJ
These numbers are the output of a complete double pulse test characterization — not just switching waveforms, but extracted device parameters, energy calculations, and timing relationships. All computed from the same CS548 + CS1200 + CS1201 measurement chain, with the symbolic maths engine built into the Cleverscope software performing the extraction automatically from the captured transients.
For a GaN device at 190A with 2 ns rise times, achieving these measurements without optical isolation is extremely difficult. The >140 dB CMRR of the CS1201 is what makes the current measurement at these switching speeds meaningful rather than dominated by common-mode artifacts.
Additional Capabilities: Beyond Switching Transients
The third panel on the booth wall describes CS548 capabilities that go beyond switching transient measurement — addressing the full characterization workflow of a power electronics development lab:
Frequency Response Analysis (FRA) — using the CS548's built-in isolated signal generator:
→ RMS or Power response
→ Gain/Phase, Impedance/Phase, Inductance/Phase Q Resr, Capacitance/Phase DF Resr
→ Passive component measurement: inductors, capacitors, transformers — directly, in-circuit, at operating voltage
Powered Op Amp or Power Supply analysis:
→ Gain/Phase, Input Impedance/Phase, Output Impedance/Phase
→ Power Supply Rejection Ratio — the parameter that determines whether a power supply's noise will corrupt sensitive analog circuitry downstream
Streaming — long-duration captures (seconds to months) to disk, for capturing intermittent events that only appear after extended operation. The panel example shows an unexpected glitch in a buffered clock system that appeared after 1668 seconds — invisible in a standard triggered capture, found through streaming.
Maths / Maths Equation Builder — symbolic maths with freeform equations, Matlab live linkage via .m files, automation of Autosave and Signal Generator output, and full help on functions and equations. The power extraction equations used in the double pulse test results are examples of this capability.
The Customer List as a Technical Validation
The fourth panel at the booth was a customer logo grid that is unusual for a company of Cleverscope's size: Tesla, ASML, Siemens, Infineon, NASA, Dyson, Williams Racing Formula 1 team, Keysight, VW, Jaguar, Toshiba, ABB, and a number of European and Australian universities.
This list is not a marketing artifact. It is a technical validation. Each of these organizations has engineering teams who evaluate instrumentation rigorously before purchasing. The presence of ASML — whose lithography systems operate at the absolute frontier of precision electronics — and NASA alongside automotive and power electronics companies signals that the CS548's isolation performance has been validated in environments where measurement quality is not negotiable.
For a New Zealand instrumentation company, this customer base is the most credible possible signal that the optical isolation architecture delivers what it claims in production engineering environments, not just in controlled lab demonstrations.
Why This Is Worth Knowing About
The CS548 occupies a measurement space where there are very few alternatives. Traditional differential probes run out of CMRR at the dv/dt rates of SiC and GaN switching. Battery-powered floating oscilloscopes are a workaround that introduces its own problems. Optically isolated voltage probes (like the PMK FireFly covered in Article 014) address the voltage measurement but not the complete four-channel synchronized system that the CS548 provides.
The complete system — CS548 oscilloscope + CS1200 voltage digitizers + CS1201 current digitizers + fibre optic interconnects + built-in FRA + symbolic maths engine — is an integrated power electronics measurement platform that addresses the complete characterization workflow from switching transient capture through impedance measurement through long-duration streaming.
At PCIM 2025, in a hall full of power device vendors and inverter system integrators, Cleverscope was showing the measurement infrastructure that makes characterizing those devices and systems possible at the speeds and voltages that define the current state of the art in power electronics.
Products observed: Cleverscope CS548 Isolated Oscilloscope · CS1200 Voltage Digitizer · CS1201 Current Digitizer · CS1501 Current Sensor · CS1202 Transistor Digitizer (development) · CS1133 Vsat Probe · CS1070 Power Amp · CS1301/CS1302 I/O Pods · Frequency Response Analysis · PCIM Europe 2025, Nuremberg, May 2025
All photos: Thomas · @SignalByThomas