2026-07-11
Before power MOSFETs and thyristors dominated, engineers built magnetic amplifiers ("mag-amps") — devices that used the nonlinear B-H curve of a saturable core to gate AC power with a small DC control current. They're still around today as post-regulators in multi-output switching supplies, where they trim a secondary rail without adding an entire switcher.
The principle: a toroidal core with square-loop material (like Metglas 2714A or square-loop permalloy) has an inductance that collapses from very high to nearly zero the instant flux reaches saturation (Bsat). Wind a power winding carrying AC and a DC control winding on the same core. When the control current biases the core closer to saturation, the AC winding saturates earlier in each half-cycle, and load current begins flowing sooner. The result is a phase-controlled AC output — but instead of a triac firing at a gate pulse, the core itself decides when to "open the gate."
Post-regulator example: A multi-output flyback produces +12 V (main, regulated) and +5 V (auxiliary, sloppy). Insert a saturable inductor in series with the +5 V rectifier. Every switching cycle, the inductor blocks current until it saturates — the delay is set by how close to saturation a small control winding has biased the core. A simple error amp comparing +5 V to a reference drives the control current. You get tightly regulated +5 V with ~95% efficiency, no extra switch.
Sizing rule of thumb: The volt-second product the core must block before saturating equals N · Ae · 2Bsat. For a Metglas 2714A core (Bsat ≈ 0.57 T) with Ae = 0.1 cm² and 20 turns:
Design caveats:
Historically, mag-amps ran everything from radar transmitters to naval fire-control computers. Today they're niche, but unbeatable when you need a robust, radiation-tolerant, silicon-free regulator on an auxiliary rail.