Segment 1 — The Cold Open

In 1947 Raytheon put a radar magnetron in a refrigerator-sized cabinet, piped water through it to keep the tube from melting, and offered the result as a commercial oven for about five thousand dollars. The machine stood taller than a person and weighed several hundred pounds; restaurants were the only customers who could justify it. The physics inside — dielectric heating of water at microwave frequencies — was already cheap in principle. What was expensive was treating a kitchen heater as if it were still a piece of wartime radar.

Segment 2 — The Old Way (Reasoning By Analogy)

The cavity magnetron was a wartime invention: John Randall and Harry Boot at the University of Birmingham built the first high-power cavity magnetron in 1940, and it became the heart of Allied centimetric radar. After the war, the firms that already made those tubes — Raytheon above all — looked at leftover magnetrons the way a foundry looks at leftover heat. Percy Spencer at Raytheon noticed a candy bar softening in his pocket near a live tube, and the company did the analogical thing. They aimed a radar transmitter at food and wrapped a metal box around the meal so the energy would not leak. That is reasoning by analogy: the product is a radar set that happens to cook. The 1947 Radarange still needed a water connection because the magnetron was a high-power radar tube with a water jacket, and it wanted far more power than a kitchen wall outlet likes to give. Through the 1950s the home attempts stayed in that groove. Tappan's 1955 domestic model was still priced around thirteen hundred dollars and still asked for a two-hundred-twenty-volt circuit, because the industry had not dropped the military-tube assumptions. Magnetrons were built in small lots, to radar tolerances, with expensive permanent magnets and hand-finished copper anode blocks. The cooking cavity was a fabricated metal cabinet, not a consumer shell. Everyone accepted that a microwave oven was industrial equipment that might, someday, shrink a little, and the finished cost sat in the thousands of dollars because the factory still thought of the vacuum tube as a weapon component.

Segment 3 — The First-Principles Move

First principles starts by throwing out the radar set and keeping only the physics: polar water molecules in food try to follow a rapidly reversing electric field, and the lag shows up as heat. You need a frequency where that loss is useful in a kitchen-sized portion, a metal box that confines the field, and a generator that can deliver something like a kilowatt of it. Two-point-four-five gigahertz, already reserved as an industrial-scientific-medical band, gives a wavelength of about twelve centimeters and a penetration depth of a few centimeters in moist food — enough to heat a plate, not a side of beef. That is the entire physical requirement. Now the magic-wand number: if the atoms of a household oven could be arranged for free, what would they cost? The cavity and chassis are thin mild steel, because a Faraday cage at this wavelength does not need armor plate; a rough guess for a compact unit is several kilograms of sheet, not a hundredweight. Price that steel at even a full dollar a kilogram and the box is still a handful of dollars. The magnetron adds a small copper anode block, sintered ferrite magnets, ceramic insulators, and a cathode — commodity masses, not a treasure chest. The high-voltage transformer is the other lump of copper wire and iron laminations, heavier than the tube and still just metal. Add a glass tray, some plastic, a door screen, and a fan, and a back-of-the-envelope materials pile lands around ten dollars, or the mid-teens if you are generous with the copper. That is the theoretical floor. Against a 1947 price of about five thousand dollars, even if you grant that first water-cooled cabinet twenty or thirty dollars of metal and magnet — which is generous — the Idiot Index sits in the low hundreds; the atoms were not the bill. The first redesign that mattered was the magnetron itself: drop the water jacket and cool the anode with a fan, because a kitchen tube running at several hundred watts of output does not have a radar transmitter's duty cycle or power density. Losing the plumber's hookup turned the machine from a built-in into a countertop that could use an ordinary outlet — a cost cut that never appears on a bill of materials and yet decides whether a household will buy one. The magnets changed next — ferrite, pressed and sintered like any other ceramic, in place of the expensive alloy magnets that radar tubes had used — because an oven magnetron needs a magnetic field across the interaction space, not a military specification for drift. Japanese tube makers, among them Toshiba, Hitachi, Panasonic, and New Japan Radio, then treated the copper anode and the vacuum process as a consumer component: cavities sized for 2.45 gigahertz, brazed and exhausted on automated lines, rather than hand-finished to radar drawings. The resonant cavities have to be the right size; they do not have to be laboratory jewelry, and once process control replaces craft the tube's Idiot Index collapses toward its copper and ferrite. Around the tube, the cooking cavity stopped being a fabricated cabinet and became a stamped and spot-welded steel box, because the only electromagnetic job that box has is to be a conductor with holes much smaller than a wavelength. Skin depth in steel at this frequency is tiny, so thickness is set by stiffness and rust, not by radio engineering, and a press can make that shape in a stroke. The waveguide that launches energy from the magnetron antenna into the box followed the same logic — a formed sheet-metal channel instead of machined radio plumbing. Door leakage, the obvious objection, was solved with a quarter-wave choke around the frame rather than a massive gasket: a clever bit of geometry that lets a cheap stamped door meet safety limits without precision machining. Standing waves inside a cheap rectangular box still leave cold spots, so Sharp's turntable — introduced in the mid-1960s — moved the food through the pattern instead of demanding a perfectly stirred mode, which is another case of refusing to overbuild the cavity. Amana's 1967 countertop Radarange, sold for about five hundred dollars, proved the form factor in the United States, and the Japanese industry spent the next decade putting stamped cavities and volume magnetrons under that shape. Each of those moves attacks a different slice of the Idiot Index: cooling and installation, magnet cost, tube labor, cavity labor, radio plumbing, door hardware, and field uniformity — and each one is a direct consequence of naming the physics instead of copying the radar.

Segment 4 — The Result & The Limits

By the late 1970s a basic household microwave could be had for under a hundred dollars, and through the 1980s the machine stopped being a novelty and became ordinary kitchen equipment. The magnetron had completed the journey from a radar spare that cost hundreds of dollars to a volume component whose price was a small fraction of that early tube. Take that hundred-dollar oven against the rough ten-to-fifteen-dollar materials pile and the Idiot Index is no longer in the hundreds; a listener who does the division gets something like high single digits, maybe ten, depending on how much copper they put in the transformer. That is close enough to the floor that retail margin, freight, safety certification, and a few screws now matter as much as metal. Assembly labor, the bulky transformer, door interlocks, and the simple fact that the unit ships in a box through a retail chain still sit on top of the atoms, and cheap ovens are generally not designed to be repaired. The physics extracted its own trade-offs: standing waves still heat unevenly, and because the energy couples to water rather than to a hot dry surface, a microwave will not brown a roast unless you add a separate crisp plate or a convection element. The magnetron remains a vacuum tube with a cathode that wears out, which is why a twenty-year-old oven often dies by going dark rather than by rusting through. Solid-state radio-frequency transistors can already generate cooking power in some premium and industrial machines, but they have not yet been driven down the same volume curve the cavity magnetron rode in the 1970s. The cheapest kilowatt of 2.45 gigahertz in a stamped box is still that air-cooled tube. The hard unsolved piece is not confinement or stamping — those are done — it is replacing the last vacuum tube in the kitchen without climbing back up the Idiot Index.

Segment 5 — The Lesson

The water jacket on the first Radarange was a confession that the product had inherited a radar's duty cycle instead of a casserole's. Inherit the field and the frequency; leave the parent machine's plumbing and tolerances behind — that is the first portable move, and it applies anywhere a technology is slumming in a new job still dressed for the old one. The second is that a Faraday cage at twelve centimeters is a stamping problem, not a cabinetmaking problem, and the cost of the box will not fall until someone says so out loud and puts it on a press. Japan did not invent dielectric heating; it invented the magnetron as a consumer part and the cavity as a stamped commodity, and the price followed those two decisions down. Tomorrow, another example or another opening. The live question is who will do to solid-state cooking what the Japanese tube plants did to the cavity magnetron — and whether the first signal will be a transistor price, or another stamped box that no longer needs a tube at all.

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.