Technical Specifications

Main Features and Advantages

Snubberless IGCT topology = smaller, cooler, simpler:The ABB 3BHB012961R0001​ is the poster child for why ABB pushed IGCT over GTO in the late 1990s–2000s. A conventional GTO at 4.5 kV / 2 kA needs a bulky RC snubber (du/dt limiting) and a -Ve gate supply (-15 V or deeper) to force turn-off; the ABB 3BHB012961R0001​ IGCT turns off with a transistor-like gate pulse (fiber in, GVC gate unit drives the disc) and the device’s internal structure holds the tail current so short that the external snubber can be eliminated or drastically reduced. For the ACS6000 drive designer, that means: smaller busbar, less snubber-loss (snubber resistors dissipate ~1–2 % of drive rating on GTO; on IGCT it’s near-zero), and the ABB 3BHB012961R0001​ runs cooler for the same kW through the valve. The offshore-wind case [web 8]: GTO→ABB 3BHB012961R0001​ retrofit, cooling power -30 %, inverter efficiency +0.6 % — that 0.6 % on a 33 kV / 50 MW offshore inverter = ~300 kW saved per turbine-hour, which pays the retrofit in <2 years.Thyristor-level conduction loss, transistor-level switching:On-state drop of the ABB 3BHB012961R0001​ is ~1.7–2.0 V at 2645 A — that’s 4.5–5.3 kW dissipation per disc at full load, which sounds high until you compare to a 4.5 kV IGBT half-bridge module where two IGBTs + two FWDs might drop 3–4 V total (but the IGBT switches faster, so the tradeoff flips at higher frequency). At the ABB 3BHB012961R0001‘s 0.2–2 kHz sweet spot (MV drive, grid converter — not 20 kHz VFD), the low Vce(sat)-equivalent of the IGCT wins: the ABB 3BHB012961R0001​ dissipates less in conduction than an IGBT pair would at 2 kHz / 4.5 kV, and the IGCT’s shorted-failure mode (press-pack shorts -> fuse the stack, protection trips) is safer for series strings than an IGBT’s open-failure risk. That’s why 6 kV / 10 MVA+ drives stayed IGCT (ACS6000) while <1 MVA went IGBT (ACS880). The ABB 3BHB012961R0001​ is the disc that makes that economics work.Press-pack uniformity and predictable failure:The ABB 3BHB012961R0001​ disc is a sandwich: two molybdenum/copper buffer plates, the GTO+diode chips in the middle, ceramic insulator rim, stainless steel pressure frame. Current flows axially (face→face), so the clamp force must be uniform — ABB specifies the torque/clamp-force per valve design, and the ABB 3BHB012961R0001​ is designed so that if a chip fails short (the dominant failure mode), the disc as a whole shorts, the valve’s DC link fuse or crowbar clears, and the stack is swap-able. Compare to an IGBT module where a bond-wire lift = open, which can cause over-voltage on the other leg — the IGCT’s shorted-failure is “fail-safe” for the inverter topology (NPC/H-bridge just trips, no dc-link overvoltage event). For metallurgy plants where a drive trip costs 50k–500k/hr, the ABB 3BHB012961R0001‘s failure mode is a feature, not a bug.Fiber-optic gate = EMI-immune in high-dv/dt:The ABB 3BHB012961R0001​ doesn’t have a wire-bond gate pin in the traditional sense — the gate electrode is a small stud on the disc periphery, and the gate unit (GVC7xx, mounted nearby on the valve PCB) talks to it via a short fiber pigtail + local gate driver PCB on the disc edge. The fiber isolates the 2–10 kV/µs dv/dt on the main terminals (ACS6000 switches 6 kV in <1 µs) from the gate control — a wire-gate device at 4.5 kV / 2 kHz would pick up enough displacement current to false-trigger. The ABB 3BHB012961R0001​ + GVC gate solves that. When troubleshooting a “valve arm fault” on an ACS6000, the fault usually traces to the GVC gate unit (fiber dirty, desat threshold drift) not the ABB 3BHB012961R0001​ disc — but if the disc has punched (short across anode-cathode, DC link fuse blown), the ABB 3BHB012961R0001​ is the swap item.2 kHz switching for MV / grid, not for VFD:The ABB 3BHB012961R0001​ is not trying to be a 20 kHz device — at 4.5 kV / 2.6 kA, switching loss scales with V×I×freq, and 2 kHz is already ~17 MW × 2 k × switching-loss-coeff — the disc would melt if pushed to 10 kHz. But 2 kHz is perfect for: 6-pulse NPC MV drive (output fundamental 0–75 Hz, PWM carrier 1.2–2 kHz), HVDC Light VSC (PWM 1–2 kHz at 4.5 kV DC link), STATCOM (same). The steel-mill case [web 8]: 0.5–200 MVA drive, response time -40 % vs old GTO drive, noise -25 % — the IGCT’s faster turn-off (≤30 µs vs GTO’s 10–20 µs plus tail) lets the current regulator loop带宽 wider, so the rolling-mill main drive responds faster to screw-down load steps. That’s the ABB 3BHB012961R0001‘s real plant-floor value: not the disc itself, but the faster current loop the drive can run because the ABB 3BHB012961R0001​ switches cleaner than the GTO it replaced.Water-cooled, double-sided:The ABB 3BHB012961R0001​ clamps between two water-cooled sinks — cold plate → thermal interface (no TIM paste on the disc faces, the clamp pressure makes metal-metal contact) → disc anode face / cathode face → thermal interface → cold plate. Double-sided means the heat from the GTO chip (center of the disc) conducts both directions to the sinks, halving the thermal resistance vs single-sided. For a 5 MW valve arm (say 3× ABB 3BHB012961R0001​ in parallel per NPC phase-leg position), the water flow temp Δ is manageable at 2 kHz / 4.5 kV. The disc itself has no built-in temp sensor (some ABB IGCT valves embed a PT100 on the sink near the disc edge; the ABB 3BHB012961R0001​ datasheet may note case T sensing point). When swapping a ABB 3BHB012961R0001, the sinks must be cleaned (no old thermal-paste residue — there shouldn’t be paste, but debris/dust yes) and the clamp torque re-set to ABB’s valve doc — uneven torque = hot spot on one side of the disc = premature failure.