7. Microwave Tubes (O-type) // Microwave Engineering // bhaswanth

For decades, every microwave watt above a few hundred MHz came from a vacuum tube. Solid-state has taken over below ~30 GHz at moderate power, but tubes still rule at high power and at extreme frequency. Even today, your microwave oven has a magnetron in it, your air-traffic-control radar likely has a klystron, and many satellite transponders run TWTs (traveling-wave tubes).

7.1 Why not regular tubes or transistors?

Conventional triode/pentode vacuum tubes work fine up to a hundred MHz. Above that, the transit time of the electron from cathode to plate becomes an appreciable fraction of the RF period, and the grid loses control of the beam. The triode degenerates into a randomizing chamber rather than an amplifier.

Conventional transistors face the analogous problem: above their $f_T$ (the frequency at which current gain falls to unity), gain is too low for useful amplification. Modern silicon-germanium HBTs and GaN HEMTs have pushed $f_T$ above 100 GHz, so solid-state has eaten most of the microwave-power market below 30 GHz. But for kilowatt-class continuous-wave power above 10 GHz, or megawatt-class pulsed power at any frequency, vacuum tubes still win on a power-per-dollar basis.

7.2 O-type and M-type classification

Microwave tubes split into two structural families:

O-type (linear-beam): electrons travel parallel to a longitudinal magnetic field that focuses the beam, interacting with cavity or slow-wave structures along the way. Klystrons, TWTs.
M-type (crossed-field): electrons travel in crossed electric and magnetic fields, taking curved paths. Magnetrons, crossed-field amplifiers.

Sections 7.3–7.5 cover the O-types; Section 8 covers the M-types.

7.3 The two-cavity klystron

Invented in 1937 by the Varian brothers at Stanford. The principle is velocity modulation followed by drift-space bunching.

plaintext

           DC accel V0          drift space
   cathode ──────────[buncher]─────────────[catcher] ── collector
                        ↑                       ↓
                    RF input               RF output (amplified)
                       (small)              (large)

Walk through it:

Cathode and DC accelerator. Electrons are emitted and accelerated through DC voltage $V_0$ (kV-range), forming a steady DC beam.
Buncher cavity. A small RF input excites this resonant cavity. Its gap presents an alternating field across the beam path: electrons crossing during one half-cycle accelerate, those crossing during the opposite half decelerate. The beam is now velocity modulated but still uniform in density.
Drift space. Fast electrons catch slow ones. Where a fast electron emitted later catches a slow one emitted earlier, density rises; elsewhere it drops. After an optimal drift length, velocity modulation has converted into density modulation: clumps and gaps.
Catcher cavity. A second cavity tuned to the same frequency sees a beam current with strong AC component. The bunches induce a much larger RF voltage in the catcher than the buncher needed. The beam delivered DC power, the cavity extracts RF power. Gain.
Collector. Spent electrons dump their remaining DC energy as heat into a water-cooled slab.

The insight: you cannot directly amplify a few-mW RF signal with a vacuum tube at GHz. But you can take a kilowatt DC beam, modulate it at low RF level, let drift physics bunch it, and extract kilowatts of RF. The DC power is the reservoir; RF input is a shaping control.

Bunching parameter $X = \beta_b V_1 \theta_0 / (2 V_0)$ where $V_1$ is RF amplitude, $V_0$ is DC accelerator voltage, $\theta_0$ is the buncher transit angle, and $\beta_b \approx 1$ is a beam-coupling factor. Maximum bunching occurs at $X \approx 1.841$ , giving theoretical max efficiency around 58%. Real klystrons run 40–60%, multi-cavity designs hitting 70%.

Modern uses: TV broadcast transmitters (declining), radar drivers, particle-accelerator drives, deep-space-network amplifiers. The Goldstone 70-m DSN transmitter is a multi-cavity klystron pumping 400 kW at X-band.

7.4 Reflex klystron

A simpler, lower-power oscillator variant. One cavity, no second tube, no drift space in the conventional sense. The beam passes through the cavity, hits a negatively biased "repeller" electrode, reverses direction, and passes back through the same cavity. If the round-trip time matches an integer number of RF periods plus the right phase, the returning electrons deliver more energy than the initial pass took, and the cavity oscillates.

Used historically as low-power microwave local oscillators in radar receivers. Largely obsolete now, replaced by Gunn diodes (Section 9) and synthesizers, but you will see them in older equipment.

7.5 Traveling-wave tubes (TWTs)

The klystron's bandwidth is narrow (a few percent) because high-Q cavities are tuned to one frequency. For wideband amplification (octaves), the TWT swaps discrete cavities for a continuous slow-wave structure, typically a helix.

plaintext

                 RF input            attenuator       RF output
                    ↓                    ↓                ↑
   cathode ── ╔═══╦╤═══╤╦═══╤╦═══╤╦═══╤╦═══╤╦══╗ ── collector
              ║   ╪╧═══╧╪═══╧╪═══╧╪═══╧╪═══╧╪══║
              ║   electron beam (focused by magnet)║
              ╚═════════════════════════════════╝
              helix wound around vacuum envelope

The RF wave travels along the helix wire. Because the wire is wound, the wave's axial velocity is much less than $c$ (only the axial projection counts). Set helix pitch so axial wave velocity equals beam velocity (0.1 to 0.3 of $c$ ). Wave and beam travel together. The wave's longitudinal field continuously pushes/pulls beam electrons; velocity modulation converts to density modulation, and bunches reinforce the wave they ride. The wave grows exponentially: $V(z) = V_0 e^{\Gamma z}$ .

Bandwidth is octaves rather than percent. Gain per tube is 30–50 dB, output power watts to kilowatts. Used in: satellite transponders (every comsat has TWTs), EW jammers, military radar drivers. A typical comsat TWT runs 50–200 W CW at C, Ku, or Ka band.

If the tube can sustain a backward-traveling wave, it becomes a backward-wave oscillator. Designers add an attenuator and "severs" (helix gaps) in the middle to break unwanted feedback.

7.6 Backward-wave oscillators (BWOs)

Same physics as TWT but designed deliberately to support a backward wave. Frequency-tunable by changing accelerator voltage, which changes the beam velocity and the resonance condition with a backward spatial harmonic of the helix. Used as broadband sweepers in microwave instrumentation before synthesizers became affordable.