Coolidge X-Ray Tube

X-rays are ionising radiation and cause real, cumulative harm. Never apply high voltage to any vacuum tube to try to produce them, and never modify this model toward that. Everything here runs at 24 V or less and emits nothing.

Coolidge separates quantity (set by cathode temperature) from penetration (set by voltage). One knob became two, and that is the invention.

Wire a lamp to a single variable supply. Brightness and current move together — you cannot ask for more light at the same current. That coupling is the old tube's flaw.

Use the same small vacuum diode as the Fleming valve build: heater on its own supply, anode on a second. Two supplies, two controls — the Coolidge arrangement in miniature.

Hold the anode at 12 V. Step the heater across its safe range and record current. Current changes while the accelerating voltage never moves.

Now hold the heater steady and sweep the anode. Current rises, then plateaus. Past the plateau, voltage adds energy per electron but no extra electrons.

Put step 5 and step 6 on one sheet. In a real tube these are the two radiographic settings: filament current sets dose, kilovoltage sets penetration.

Cut a clear tube or bottle and mount a filament stand-in at one end, a target plate at the other. Build it as a cutaway so the geometry stays visible.

Form a shallow metal cup behind the filament. Coolidge focuses the stream onto a small area — without it the beam sprays and the image blurs.

Shine a torch through a card aperture onto paper, then narrow the aperture. A smaller source casts a sharper shadow — the same reason the focal spot must be small.

Set the target plate at about 20° to the beam axis. Viewed from the side, the hot area is large; viewed along the beam, it looks tiny. Heat spreads over a big spot, the image sees a small one.

Nearly all the arriving energy becomes heat, not X-rays. Tungsten melts near 3,422 °C — the highest of any metal — which is why the target and the filament are both made of it.

Run your diode's heater a little high and watch the current wander. Coolidge needed a vacuum far harder than earlier tubes precisely so leftover gas could not take part.

Mark cathode, focusing cup, anode target, target angle and envelope. Add the two control paths. This diagram is the deliverable, not a working tube.

The patent. US 1,203,495, "Vacuum-tube", filed 9 May 1913 and granted 31 October 1916 to William D. Coolidge at General Electric. The granted title says nothing about X-rays, which is a good reminder that patent titles are often the broadest defensible term rather than the useful one.

The same man solved the materials problem first. Coolidge's earlier and arguably larger achievement was ductile tungsten — working the most refractory metal known into drawable wire, which is what made long-lived incandescent lamp filaments possible. That work is the precondition for this tube: it needs tungsten in both the hottest places, as a filament that can run at emission temperature for thousands of hours and as a target that can absorb a beam whose energy is over 99 % waste heat. An invention is often a materials advance wearing a device's clothes.

Why separating the two controls mattered clinically. In a gas tube, asking for a brighter image also changed how penetrating the beam was, and the gas was being consumed as you worked, so the tube's behaviour drifted mid-session and differed from every other tube in the hospital. Exposures were guesswork, repeat exposures were routine, and repeat exposures mean repeated dose to the patient and the operator. The Coolidge tube made an exposure reproducible: set the filament current for the quantity you want and the kilovoltage for the penetration you want, and get the same result tomorrow. Reproducibility is what turned radiography from a demonstration into a diagnostic discipline — and, because it cut wasted exposures, it was also a safety advance.

The anode angle is a genuinely elegant trick. Two requirements conflict: the beam's apparent source must be tiny for a sharp image, and the heated area must be large or the target melts. Tilting the target resolves them geometrically. The electron beam lands on a long ellipse — plenty of area to absorb heat — but seen from the direction the X-rays leave, that ellipse foreshortens to a small square. One angle, both requirements, no moving parts. Rotating anodes later took the same idea further by moving fresh metal through the beam.

Why this build stops where it does. Everything above is the physics and the geometry, and all of it can be understood at 24 V with a small diode, a torch and a piece of card. Producing X-rays requires tens of kilovolts, and an unshielded source is dangerous in a way that gives no warning at the time — the harm from ionising radiation is cumulative and largely invisible until much later. Early radiology workers, including many of the field's pioneers, were injured and killed learning this. The model is the honest place to stop.