The Past and Present of the Solid-State Circuit Breaker

In 1879, in a laboratory in Menlo Park, New Jersey, Thomas Edison sketched an idea for a device that could automatically cut power to an overloaded circuit. He had just lit the first practical incandescent bulb and needed a way to keep the wiring feeding it from burning down the building around it.

One hundred and forty-seven years later, in June 2026, an Eaton engineering team stood on stage in Nuremberg to accept a Best Paper Award at PCIM Europe for a smart circuit breaker built for low-voltage DC power grids. The applause wasn’t for a mechanical contact snapping open. It was for a silicon carbide chip that can interrupt several thousand amps of fault current in roughly 700 nanoseconds — close to a million times faster than a human eye can blink.

That’s the arc this article follows: from Edison’s sketch to a chip that reacts before a fault can do any damage at all. It’s a story with two very different chapters — the century-long reign of the mechanical circuit breaker, and the much shorter, faster-moving story of the solid-state device now displacing it in the applications that need speed the most.

Chapter One: The Age of the Mechanical Circuit Breaker

Before there were circuit breakers, there were fuses — a wire that melted under overload and had to be physically replaced after every trip. Electrification needed something reusable.

• 1900 — Granville Woods, a prolific American inventor, patented one of the earliest automatic circuit breakers.
• 1904 — Cutter Manufacturing in Philadelphia began mass-producing breaker products, though early units remained bulky and expensive.
• 1923–1924 — In Mannheim, Germany, engineer Hugo Stotz combined a thermal trip and a magnetic trip into a single, resettable device — the first modern miniature circuit breaker (MCB). His company was later absorbed into what is now ABB, which still marks 1924 as the birth year of the modern MCB. The thermal-magnetic, molded-case architecture Stotz introduced is still the basic template for the most widely used circuit protection device on Earth.

For the next several decades, mechanical breaker technology advanced along two main axes: the medium used to quench the arc (oil, then air, then sulfur hexafluoride, then vacuum) and the voltage class it could handle (from a few hundred volts to well over a million). Westinghouse patented an SF6 breaker in 1951; electronic trip units arrived in 1973.

But every one of these advances ran into the same wall: speed. AC breakers get a natural assist — current crosses zero roughly 100 times a second, giving the arc a moment where it can be extinguished with comparatively little drama. DC current never crosses zero. Once a fault starts, current climbs in a straight line, governed only by system inductance, and a mechanical breaker has to physically pull contacts apart against that climbing current. Detection to interruption typically takes 10 to 50 milliseconds — fine for a lightbulb, potentially catastrophic for equipment that can’t tolerate even a few milliseconds of uncontrolled fault current.

Chapter Two: The Solid-State Breaker Is Born

Semiconductors began creeping into power engineering in the 1960s. Engineers started asking an obvious question: what if a transistor or thyristor did the switching instead of a mechanical contact? Electrons move far faster than any physical part ever could.

The first thyristor-based “solid-state circuit breakers” appeared in the late 1970s — but early thyristors were semi-controlled devices that could only turn off at a natural current zero, which is exactly what DC systems don’t have. The real turning point came in the 1980s, with the arrival of fully-controlled devices: the Gate Turn-Off thyristor (GTO) and the Insulated-Gate Bipolar Transistor (IGBT), both of which could be switched off on command, at any point in the waveform.

• 1987 — At the University of Texas, researcher William Kernaghan built one of the first GTO-based solid-state DC breakers, rated at a modest 200V/15A. Small as it was, it proved the concept was viable.
• 1990s — ABB and Mitsubishi introduced higher-power GTO-class devices; in 1999, researcher Jeffrey A. Casey and colleagues published a systematic look at applying solid-state DC breakers to distribution networks.
• 2005 — The U.S. Power Electronics Systems Center built prototype DC breakers rated at 2.5kV/1.5kA and 4.5kV/4kA.

But a new problem emerged as these designs scaled up: an all-solid-state breaker keeps its semiconductor in the current path permanently, and even the best power devices have a real forward voltage drop. For a breaker carrying full load current around the clock, that translates into constant heat, accelerated aging, and a real electricity bill — a tradeoff that mattered a great deal once solid-state breakers moved beyond the lab.

Chapter Three: The Hybrid Compromise Takes Over HVDC

Rather than choosing between “fast but lossy” and “efficient but slow,” engineers combined the two. In a hybrid circuit breaker (HCB), steady-state current flows through a fast mechanical contact with near-zero resistance; the moment a fault occurs, a parallel semiconductor branch takes over and clears it in microseconds. This is the architecture that carried solid-state protection into high-voltage DC transmission — a market mechanical breakers alone could never have served.

• 2012 — ABB announced a prototype hybrid DC breaker for 320kV transmission systems, capable of interrupting 9kA within 5 milliseconds — widely regarded as the first genuine engineering breakthrough in HVDC breaker technology.
• 2013 — Alstom followed with a design clearing over 3kA within 2.5 milliseconds.
• 2016 — China’s NR Electric unveiled a 500kV hybrid breaker prototype rated at 20kA/3ms, around the same time a modular cascaded hybrid breaker went into service on Zhoushan’s five-terminal HVDC project — the world’s first hybrid HVDC breaker in actual operation.
• 2017 — NR Electric introduced a bridge-rectifier hybrid topology and built a 535kV breaker interrupting 25kA in under 3ms; China XD Electric completed type testing on a separate 500kV prototype the same year.
• 2020 — Multiple Chinese research institutes and grid operators commissioned 500kV-class hybrid breakers on the Zhangbei flexible DC grid — the world’s first flexible HVDC grid project, with breaking times under 3ms and 25kA maximum interruption current.

Hybrid breakers became the dominant technology for large HVDC transmission projects through the late 2010s. But they remained, by their nature, large, complex, and expensive — engineering achievements suited to grid-scale infrastructure, not to compact, high-volume applications.

Chapter Four: The SiC Era Arrives

By the mid-2020s, three trends converged to pull solid-state breaker technology out of niche transmission projects and into the mainstream.

First, AI computing’s appetite for power. As GPU rack density climbs from roughly 100kW toward the megawatt range, data centers are moving toward 800V DC distribution architectures that eliminate several stages of lossy AC-DC conversion — a shift IEEE Spectrum has covered in detail as major infrastructure vendors race to commercialize it. These low-capacitance DC buses don’t store enough energy for a traditional fuse to operate reliably; only a breaker that reacts to fault current directly, at semiconductor speed, will do.

Second, the maturity of silicon carbide. SiC devices offer breakdown field strength, thermal conductivity, and switching speed that silicon simply can’t match. SiC MOSFETs bring meaningfully lower on-resistance and an order-of-magnitude reduction in switching loss; normally-on SiC-JFETs are naturally suited to fault-current-limiting topologies. Both are following in the footsteps of earlier DC breaker research programs — Infineon’s NEST-DC consortium, which brought together Airbus, Siemens, and university researchers years earlier to lay the groundwork for fully electronic DC grid protection.

Third, sustained industrial investment. Eaton has published a four-year SSCB product roadmap: a 1500V/4000A IGCT water-cooled unit for rail transit in 2021, an 800V/600A IGBT design explored for aviation in 2022, a 1500V/1000A marine variant in 2023, and a 1500V/1250A hybrid SSCB in 2024 — the design lineage that led to its 2026 PCIM award. Infineon has stood up a dedicated team developing SSCB-specific SiC-JFET devices alongside a three-tier reference design spanning branch, feeder, and main-breaker protection.

At PCIM 2026, no single approach dominated. A European path favors normally-on SiC-JFETs pushed to their physical speed limit — Fraunhofer IISB has demonstrated a monolithic 150A/900V device with on-chip snubber integration, and Airbus has tested a 400A bidirectional SiC-SSCB reaching interruption speeds around 700 nanoseconds. A more industrialized path, championed by Eaton, favors silicon IGBTs with advanced silver-sintered packaging — trading some raw speed for cost, reliability, and manufacturability at scale.

Old Technology, New Technology: A Side-by-Side View

The Standards-Setters’ Warning: Hybrid May Not Be a “Transition”

Not everyone in the industry agrees that all-solid-state technology will simply replace hybrid designs once SiC costs fall further. Researchers working on China’s high-voltage DC circuit breaker standards — the same group behind some of the HVDC hybrid breaker milestones above — have made a pointed thermodynamic argument at recent industry conferences: falling chip prices shrink the purchase-cost gap between hybrid and all-solid-state designs, but they do nothing to close the topology-driven loss difference. A semiconductor that carries full rated current 24 hours a day for years will always dissipate more cumulative energy — and age faster — than one that only engages during the rare fault event a hybrid design is built around.

Their broader point is that the industry’s most-cited spec — microsecond breaking time — hides a metric almost nobody outside the design team checks: short-circuit withstand time (SCWT), the window a semiconductor can survive full fault current before it must be shut off.

• IGBTs typically tolerate 10–30 microseconds — a workable margin for detection and control electronics to react.
• Commercial SiC MOSFETs often tolerate only 2–6 microseconds, and once voltage ringing, sampling delay, and gate-drive propagation are subtracted, the real margin can shrink to almost nothing.

That gap is why a single successful short-circuit test proves far less than a design that has survived thousands of repeated fault events without measurable drift in device parameters — the kind of endurance validation called for under the emerging IEC 60947-10 standard for low-voltage DC switchgear.

Beyond SCWT, the same researchers point to three engineering challenges that separate a working prototype from a bankable industrial product: selective coordination (ensuring a fault trips only the nearest breaker, not the entire bus, which requires large validated fault datasets rather than simulation alone); busbar stray inductance (above roughly 3–5kA, parasitic inductance in the busbar — not the semiconductor chip — often becomes the true limiting factor, since even 1 nanohenry can produce hundreds of volts of induced overvoltage); and MOV endurance (the metal-oxide varistor absorbing each interruption’s energy must survive thousands of high-energy impulses over its service life, a far more demanding duty cycle than the occasional lightning surge MOVs were originally built for).

Their closing caution for the market: it’s a mistake to extrapolate a headline “future addressable market” figure directly onto near-term company revenue. An SSCB is one component within a much larger DC distribution build-out — if the surrounding rectifier and liquid-cooled busbar infrastructure stays expensive, halving the price of the breaker alone won’t be enough to drive adoption.

Device Choice Today: IGBT, SiC-MOSFET, or SiC-JFET

Epilogue

From Edison’s sketch to a chip that clears a fault in 700 nanoseconds, circuit protection has traveled 147 years — from safeguarding a single lightbulb to guarding AI compute clusters worth tens of millions of dollars. It has been, at every stage, a negotiation: between cost and performance, between speed and reliability, between the elegance of a simple mechanical part and the raw speed of a semiconductor that never touches anything at all.

That negotiation isn’t finished. It’s being carried forward today by engineers in Nuremberg, by standards committees weighing hybrid against all-solid-state, and by manufacturers turning laboratory breakthroughs into products that have to survive years of real service. The line Edison drew in 1879 hasn’t broken yet — it has simply gotten faster.

Frequently Asked Questions

Who invented the first circuit breaker?

The first automatic circuit breaker is generally credited to Granville Woods around 1900, though the device most people recognize today — the resettable, thermal-magnetic miniature circuit breaker — was invented by Hugo Stotz in Germany and patented in 1924. That basic architecture is still the template for the majority of circuit breakers manufactured worldwide.

Why don’t mechanical circuit breakers work well for DC systems?

AC current crosses zero roughly 100–120 times per second, giving a mechanical breaker a natural moment to extinguish the arc it creates when contacts separate. DC current has no zero-crossing, so a mechanical breaker has to physically interrupt current that just keeps climbing — a slower and more damaging process that typically takes 10 to 50 milliseconds.

What’s the difference between a hybrid and an all-solid-state circuit breaker?

A hybrid breaker routes normal operating current through a low-loss mechanical or contact-based path and only switches to a semiconductor branch when a fault occurs, trading a small amount of response time for lower everyday conduction loss. An all-solid-state breaker keeps the semiconductor in the current path at all times, achieving the fastest possible response but generating continuous conduction loss for the life of the installation.

Why is silicon carbide (SiC) central to the current generation of solid-state breakers?

SiC devices offer higher breakdown field strength, better thermal conductivity, and faster switching than silicon, which lets solid-state breakers reach microsecond-to-nanosecond clearing times with lower conduction loss. Normally-on SiC-JFETs are additionally well suited to fault-current-limiting circuit topologies, though their normally-on behavior adds complexity to gate-drive and safety design.

Why are AI data centers driving demand for solid-state circuit breakers right now?

AI computing is pushing GPU rack power from roughly 100kW toward the megawatt range, driving a shift to 800V DC distribution to avoid impractical amounts of copper. These high-voltage, low-capacitance DC buses don’t store enough energy for a traditional fuse to trip reliably, so operators need a device that reacts to fault current directly at semiconductor speed — which is exactly the role a solid-state circuit breaker fills.

Ready to Bring Solid-State Protection Into Your Design?

If your project involves 800V DC data center distribution, battery cluster protection, or another zero-arc-tolerant application, HIITIO’s HBS1 solid-state circuit breaker series spans eight current platforms from 40A to 4000A across both all-solid-state and hybrid (SSHCB) topologies, with breaking times as fast as 20μs and breaking capacity up to 250kA. The line is engineered alongside HIITIO’s own HVDC contactors, semiconductor fuses, and DC protection portfolio rather than sourced as a standalone switch. Send your voltage class, prospective fault current, and enclosure constraints, and an applications engineer will return a matched configuration.