What the transmitter has to do
The transmitter (Components of the radar system) generates the radio-frequency pulses. There are two fundamentally different designs for this, and the difference determines what the radar can do:
- A power oscillator generates the pulse itself, directly at high power. Simple and cheap, but every pulse starts with a random phase.
- An amplifier chain (master oscillator – power amplifier) first makes a weak, extremely stable signal and then amplifies it. Every pulse has a known phase relative to a reference.
That known phase is called coherence, and without coherence there is no Doppler processing: to measure a small phase shift from pulse to pulse (Pulse-Doppler radar and Moving Target Indicator), the radar must know with which phase each pulse left.
Magnetron: the classic oscillator
The magnetron is the tube that made radar practical in the Second World War. Electrons spin in a strong magnetic field past a ring of resonant cavities and so directly generate a powerful microwave. The magnetron is compact, robust and cheap — it is still in almost every marine navigation radar (and in the microwave oven in your kitchen).
Its drawbacks follow from the fact that it is an oscillator: every pulse starts with a random phase, and the frequency drifts slightly with temperature and wear. A magnetron radar is therefore non-coherent. With a trick it can still do MTI: the radar measures the phase of every transmitted pulse and locks a local reference oscillator to it. This is called pseudo-coherent or coherent-on-receive — good enough for simple MTI, but too coarse for high-grade pulse-Doppler.
Coherent amplifier chain
A coherent radar builds its transmit signal from two stable oscillators. The STALO (stable local oscillator) provides a high, stable frequency; the COHO (coherent oscillator) a lower reference. Their sum is amplified into the transmit pulse. On reception the radar mixes the echo with the same STALO back down to the intermediate frequency (The receiver: from echo to plot) and compares it with the COHO: any phase shift is then Doppler, not chance.
For the final amplifier there are four families:
| Amplifier | Characteristic | Typical use |
|---|---|---|
| Klystron | very high power and high gain, but narrowband (a few per cent) | large ground radars, illuminators |
| Travelling-wave tube (TWT) | very wideband (up to an octave and more), less peak power | airborne radars, PESAs (PESA — Passive Electronically Scanned Array) and above all jammers |
| Crossed-field amplifier (CFA) | compact and efficient, moderate gain | naval radars, including the final stage of AN/SPY-1 |
| Solid-state (GaAs, GaN) | low peak power per transistor, very reliable; long pulses with pulse compression | T/R modules of AESAs (AESA — Active Electronically Scanned Array), modern surveillance radars |
Solid-state transmitters do not deliver megawatt pulses. They compensate with long pulses and a high duty cycle (Peak power and average power of the pulse radar): the same average power, spread out in time, and still fine range resolution through pulse compression.
Every transmitter leaves a fingerprint. A magnetron gives itself away through small frequency jumps from pulse to pulse; a tube transmitter through the characteristic rise and fall time of its pulse and through unintentional modulation on pulse (UMOP). At that level ESM can recognise not just a radar type but sometimes an individual set: specific emitter identification (SEI). And because jammers have to be wideband, the TWT is also the workhorse tube of the EW world.
Radartutorial.eu describes every tube with cross-sections and principle of operation: radar transmitters.