The Transmitter Chain

A receiver that misbehaves annoys you. A transmitter that misbehaves annoys the whole neighbourhood, and it is your licence that is at risk. That is why the syllabus puts "principles and operation of low power transmitter" in Section A and why the station-conduct conditions in Section B come back to the same faults — harmonics, key clicks, hum, spurious emissions and over-modulation. Learn the chain of stages and what each one is for, and every one of those faults has an obvious address.

What a transmitter must do

Four things, in this order: generate a stable signal on exactly the frequency you intend; impress the information on it; amplify it to the licensed power; and deliver it to the aerial clean — with everything except the wanted output suppressed.

The low-power transmitter chain

The master oscillator, and why stability is a legal requirement

Whatever the oscillator does, the rest of the transmitter faithfully amplifies. Drift of 100 Hz at a 3.5 MHz oscillator becomes 600 Hz on air after a multiplication by six. A crystal oscillator ties the frequency to the mechanical resonance of a quartz slice, which is far more stable against temperature, supply voltage and ageing than any LC tuned circuit — see Oscillators and Frequency Stability for the mechanism and for what makes a VFO drift.

The buffer — the stage that earns its keep by doing nothing

Connect an oscillator straight to an amplifier and the oscillator's frequency shifts whenever that amplifier's loading changes — and the loading changes every time you press the key, because a keyed stage draws current on key-down and none on key-up. The oscillator is pulled. On the air this is heard as chirp: the pitch of every dot and dash sliding as it starts. It is instantly recognisable and instantly unpopular.

The buffer sits between them. It has a high input impedance so it barely loads the oscillator, and a gain of about one — it is not there to amplify. Its whole job is isolation, so that the oscillator sees a constant load no matter what the stages after it are doing. A well regulated supply for the oscillator cures the other half of chirp, the supply-voltage dip on key-down.

Frequency multipliers

A multiplier takes an input and delivers a whole-number multiple of it: a doubler, a tripler, occasionally a quadrupler.

How? By deliberate distortion. A class C stage conducts for considerably less than half of each cycle, so its output current is not a sine wave but a train of narrow pulses — and a pulse train is rich in harmonics of the input. Put a tuned circuit in the output, tune it to the second harmonic and you have a doubler; tune it to the third and you have a tripler. The very non-linearity that makes class C useless for SSB is exactly what makes it the natural choice here. See Amplifiers and Classes of Operation for the conduction angles.

Worked example — 3.5 MHz crystal to the 15 m band

A 3.5 MHz crystal oscillator feeds a doubler, and the doubler feeds a tripler. What comes out?

3.5 MHz × 2 = 7 MHz at the doubler output.

7 MHz × 3 = 21 MHz at the tripler output.

The multipliers multiply in cascade, so the overall factor is 2 × 3 = 6, and 3.5 × 6 = 21 MHz — the 15 m band. Every paper that asks this offers 17.5 MHz as a distractor, which is what you get by adding the factors (2 + 3 = 5). Multiply them.

MOPA

MOPA is Master Oscillator Power Amplifier: an architecture, not a component. The frequency-determining oscillator is kept small, low powered and lightly loaded, and a separate amplifier supplies the power. The alternative — a single power oscillator generating the frequency and the power together — is simpler and always worse, because the load, the supply and the keying all pull the frequency directly.

MOPA has a second virtue for CW: you key the amplifier, not the oscillator, so the oscillator runs continuously and never has to restart on frequency. That is the structural cure for chirp.

The power amplifier

The PA supplies the output power, and it is the stage the Indian licence measures — the conditions require meters of standard accuracy to read the DC power input to the final RF stage. Its class depends on what it is amplifying:

• Class C for CW and FM. Both are constant-envelope modes carrying nothing in their amplitude, so the distortion class C introduces does no harm and its efficiency — 70% or better — is worth having.
• Class AB, or class A for SSB and AM. These carry the information in the envelope, so the amplifier must be linear. Efficiency is sacrificed to keep the envelope faithful.

Neutralisation

Every amplifying device has some capacitance between its output and its input. That capacitance feeds a fraction of the output back to the input, and when the stage has enough gain, the feedback is enough to sustain oscillation — the amplifier becomes an oscillator, usually at some frequency of its own choosing, and radiates rubbish. Neutralisation deliberately feeds back an equal and opposite signal through a small adjustable capacitor, cancelling the internal feedback. It is adjusted with no drive applied and the stage tuned for minimum output.

Parasitic oscillation

A parasitic oscillation is an unwanted oscillation at a frequency unrelated to the operating frequency — not a harmonic. The stray inductance of the leads and the stray capacitance of the device form a resonant circuit of their own, often up at VHF, and the amplifier oscillates there while apparently doing its proper job at 14 MHz. The symptoms are unexplained heating, erratic tuning and interference nowhere near your band. The cure is a parasitic suppressor — typically a small resistor with a few turns of wire in parallel with it, fitted in the lead, which damps the VHF resonance while passing HF unharmed.

The output tank and the low-pass filter

The pi-network tank at the output of the PA does two jobs with the same handful of components, and the exam asks about the second one:

1. Impedance matching. The final stage wants to work into a load of some hundreds or thousands of ohms; the feeder is 50 Ω. The pi-network transforms one into the other, so the PA delivers its rated power. See Feeders, SWR and Matching.
2. Harmonic suppression. A pi-network is a low-pass shape, so on the way through it attenuates everything above the operating frequency.

A separate low-pass filter between transmitter and feeder finishes the job. It passes the operating frequency and rejects everything above it, and the reason it matters is arithmetic: the second harmonic of 7 MHz is 14 MHz, but the second harmonic of 21 MHz is 42 MHz and the third is 63 MHz — squarely in the VHF television and broadcast bands. Note that a low-pass filter does nothing whatever to the SWR of a mismatched aerial; it is a filter, not a matching device.

Spurious emissions

"Spurious" means anything leaving your aerial that is not the wanted signal. They have distinct names, distinct causes and distinct cures, and the examiner tests whether you can tell them apart.

TVI and BCI

TVI is television interference; BCI is broadcast interference. Harmonics from an HF transmitter are the classic cause of TVI, and the cure is at your end — a low-pass filter and good screening and earthing. But the fault is often at the other end: a neighbour's cheap set overloaded by your perfectly clean fundamental, which is fixed by a high-pass filter at their aerial socket, not by you reducing power. Diagnose before you apologise.

The dummy load

A dummy load, or artificial aerial, is a non-inductive resistor — normally 50 Ω — in a housing that can dissipate the power. It presents the transmitter with a perfect load and turns the output into heat instead of radiation, so you can tune up, measure power and fault-find without putting a signal on the air.

Semiconductor transmitters

The syllabus asks for "basic knowledge about construction of semiconductor based transmitters", and the practical difference from a valve set is ruggedness under mismatch.

A valve PA is forgiving. It runs at high voltage and low current, it tolerates a bad match, and if you over-drive it briefly it survives. A transistor PA runs at low voltage and very high current, and its junctions have almost no thermal mass. Feed it a high SWR and the reflected power raises the voltage and current at the wrong points in the cycle; the device can be destroyed in a fraction of a second — long before you notice the needle.

Hence SWR foldback protection: the transceiver monitors reflected power continuously and automatically reduces drive to the PA as the SWR rises, so power output falls away instead of the transistors failing. If your 100 W radio delivers 20 W into a badly matched aerial, nothing has broken — the protection circuit is doing exactly what it was built to do, and the answer is to fix the aerial, not to look for a fault in the radio.