Doc #131 — Radio Equipment Fabrication

---

RECOMMENDED ACTIONS (BY ACTUAL URGENCY)

Phase 1: First year (while commercial equipment still works)

1. Inventory all vacuum tube equipment and spare tubes nationally. Many older amateur operators retain valve-era equipment and tube stocks. Maritime and military surplus may include additional tubes. Every vacuum tube in NZ is a strategic asset — they are the bridge technology between commercial solid-state equipment and locally fabricated devices. Cost of delay: low in Year 1, but tubes discarded during estate clearances of deceased operators are lost permanently.

2. Identify and catalogue all ZL operators with homebrew construction experience. The amateur radio community includes operators who have designed and built their own transmitters and receivers. These individuals hold practical fabrication knowledge that is critical for this program. Many are older — knowledge capture (Doc #160) is time-sensitive. Cost of delay: skill loss through attrition is irreversible.

3. Collect and preserve technical references. The ARRL Handbook, the RSGB Radio Communication Handbook, ARRL QST magazine back issues, and NZ’s own Break In magazine contain decades of practical radio construction articles.² Ensure multiple copies survive. These are the design libraries for locally fabricated equipment.

4. Stockpile key components from electronic waste. As consumer electronics fail and are discarded, salvage: variable capacitors, fixed capacitors, resistors, potentiometers, transformers (especially power transformers and audio output transformers), coil formers, wire of all gauges, and most critically, any quartz crystals. Electronic waste streams will increase as devices fail; establish a systematic recovery program rather than allowing components to be lost to landfill.

5. Begin experimental construction. Assign 2–3 experienced homebrew constructors to build simple CW transmitters from salvaged components as proof-of-concept demonstrations. This validates that the skill base exists, identifies tooling gaps, and produces working reference designs before commercial equipment fails. Cost of delay: negligible in absolute terms, but early construction builds institutional knowledge that compounds over time.

Phase 2: Years 1–3

6. Establish a radio fabrication workshop. A dedicated workshop with: a winding lathe for coils and transformers, a vacuum system for tube fabrication (initially from salvaged laboratory vacuum pumps), basic metalworking tools, soldering equipment, and test instruments (at minimum, a salvaged multimeter, an oscilloscope if available, and a frequency counter or calibrated receiver for frequency measurement).

7. Begin vacuum tube fabrication experiments. Using salvaged glass envelopes initially, then glass from NZ production (Doc #98). Target: triode tubes suitable for RF oscillator and amplifier service at HF frequencies. This is the hardest single task in the program and should begin early to allow time for iterative improvement.

8. Develop crystal oscillator capability. Assess NZ quartz deposits for piezoelectric quality. Begin experimental crystal cutting and grinding. If NZ quartz proves unsuitable for precision oscillators, develop alternative frequency-control approaches (LC oscillators with thermal compensation, cavity resonators).

9. Produce first locally fabricated CW transmitters. Target: 10–25 W output on 3.5 MHz and 7 MHz (the domestic NVIS and trans-Tasman workhorse bands). Initially using salvaged tubes; transitioning to locally fabricated tubes as capability develops.

10. Build regenerative receivers. The simplest useful receiver architecture — one tube or transistor, a tuned circuit, and an audio amplifier stage. Regenerative receivers are selective enough for CW reception and adequate for SSB voice with practice.

Phase 3: Years 3–7

11. Develop superheterodyne receivers. More complex (requires a local oscillator, mixer, IF amplifier, and detector) but provides the selectivity and sensitivity needed for reliable weak-signal work. Uses 4–6 tubes or transistors.

12. Integrate germanium transistors from Doc #115 materials pipeline into radio designs as they become available. Transistor radios are lighter, more power-efficient, and more robust than tube designs.

13. Begin production at network-sustaining scale. Target: 50–100 complete transceiver sets per year, sufficient to replace failing commercial equipment across the national network’s approximately 150–250 regional and community stations.³

Phase 4: Years 7+

14. Refine designs for higher performance. Improved frequency stability, higher power output, multi-band capability, voice (AM or SSB) transmission.

15. Establish training program for radio fabrication. The workshop should produce not only radios but trained radio builders — a self-sustaining capability. Integrate with trade training (Doc #157).

---

ECONOMIC JUSTIFICATION

The cost of losing radio communication

The HF radio network described in Doc #128 provides two irreplaceable functions: international communication (primarily with Australia, the Pacific Islands, and eventually South America and Southern Africa) and domestic backup communication independent of the telecommunications network. If radio equipment fails and is not replaced, NZ loses both.

International communication value: NZ’s trade relationships depend on communication. Maritime trade scheduling, cargo manifesting, weather routing, and diplomatic coordination all flow through the HF network. Without it, every sailing vessel (Doc #138) operates blind — no weather information, no port coordination, no distress communication. The economic value of maritime trade to NZ’s recovery is difficult to quantify precisely but is clearly large: imported rubber (Doc #33), metals, pharmaceuticals, and other critical materials all move by sea and are coordinated by radio.

Domestic backup value: The baseline scenario assumes domestic telecommunications continue functioning for years. But this assumption depends on continued maintenance of complex electronic infrastructure (Doc #128). If regional telecommunications fail — through transformer failure, cell tower electronics degradation, or localized grid problems — HF radio is the only alternative for inter-regional communication. The cost of losing this backup is measured in lives: delayed medical evacuation requests, missed disaster warnings, inability to coordinate emergency response across regions.

The cost of fabrication

Labour: The radio fabrication workshop requires approximately 4–8 full-time skilled workers (electronics technicians, metalworkers, glassworkers for tube production). At production scale (50–100 units per year), this represents approximately 5–10 person-years of labour annually.

Materials: Copper wire (from Doc #70 pipeline or recycled wire), glass (Doc #70), iron for transformer cores (from scrap or NZ Steel — Doc #70), aluminium for chassis, and small quantities of specialised materials (tungsten or thoriated tungsten for tube filaments, barium compounds for tube cathode coatings, mica or locally produced ceramics for insulation).

Comparison with the alternative: The alternative to local fabrication is not “no radios” — it is “radios until the commercial fleet fails, then no radios.” The fabrication program is an investment that converts a declining asset (finite commercial equipment) into a renewable capability (locally producible equipment). The 5–10 person-years per year of fabrication labour is modest relative to the communication capability it sustains — the national HF network employs an estimated 200–500 operators (Doc #128), whose time is wasted if they have no functioning equipment.⁴

Breakeven: The fabrication program pays for itself the moment the first locally built radio replaces a failed commercial unit that would otherwise leave a station silent. Given commercial equipment attrition rates, this occurs within 3–5 years of program start — well within Phase 2.

---

1. RADIO FUNDAMENTALS: WHAT WE NEED TO BUILD

1.1 The minimum viable radio system

A radio communication system requires:

• A transmitter: A device that generates a radio-frequency (RF) signal at a specific frequency with sufficient power to reach the intended receiver. At minimum: an oscillator (generates the RF signal), a power amplifier (increases the signal power), and a means of modulation (impressing information onto the carrier — for CW, this is switching the transmitter on and off with a telegraph key).
• A receiver: A device that detects and converts the RF signal back into usable information. At minimum: a tuned circuit (selects the desired frequency), a detector (extracts the modulation), and an audio amplifier (makes the signal audible).
• Antennas: At both ends. Covered in Doc #128 Section 7 — antennas can be fabricated indefinitely from wire, insulators, and timber, and are not the limiting factor.
• Power supplies: DC power for the active devices and, for vacuum tubes, a high-voltage supply and a filament supply.
• Frequency control: Some means of ensuring the transmitter operates on the intended frequency and the receiver is tuned to the same frequency.

1.2 Active devices: the core challenge

The active device — the component that amplifies or generates RF signals — is the component NZ cannot currently produce. Everything else (wire, coils, capacitors, resistors, antennas, power supplies) can be fabricated from NZ materials with existing workshop capability. The active device is the binding constraint.

Three options exist, in order of fabrication difficulty:

Vacuum tubes (valves). Glass envelope containing a heated cathode (electron emitter), one or more grids (control elements), and a plate (electron collector), all in a vacuum. Vacuum tubes were the foundation of all electronics from 1906 to the 1960s. They are the most fabrication-accessible active device because:

• The underlying physics is well-understood: thermionic emission (heating a metal surface releases electrons) and electrostatic control (a charged grid controls electron flow)
• The materials are available in NZ: glass (Doc #98), copper or nickel for electrodes, tungsten for filaments (limited NZ supply, supplemented by salvage from incandescent light bulbs), barium compounds for cathode coating
• The precision requirements are modest compared to semiconductors — a vacuum tube is a centimetre-scale device fabricated with metalworking and glassblowing, not a micron-scale device requiring clean rooms
• The fabrication infrastructure overlaps with laboratory glassware production (Doc #98)

Germanium transistors. Solid-state devices fabricated from purified germanium (Doc #115). More power-efficient, more robust, and smaller than vacuum tubes, but requiring a significantly more advanced materials pipeline (germanium extraction, purification, and zone refining). Germanium transistors become available when the Doc #115 program produces them, likely Phase 4–5 for quantities sufficient for radio production.

Crystal detectors. A point-contact between a fine wire (“cat’s whisker”) and a natural crystal of galena (lead sulfide) or iron pyrite acts as a simple RF detector. Crystal detectors cannot amplify — they are passive — but they can demodulate an AM signal and were the basis of all radio reception before vacuum tubes. A crystal detector receiver requires no power supply and no active components. It is the absolute minimum receiver and can be built today with NZ materials.⁵

1.3 Target specifications by stage

Stage	Transmitter	Receiver	Power Output	Frequency Range	Modulation
1	Single-tube oscillator	Crystal detector or regenerative	5–25 W	3.5–7 MHz	CW only
2	Oscillator-amplifier (2–3 tubes)	Regenerative (1–2 tubes)	25–100 W	3.5–14 MHz	CW, possibly AM
3	Multi-stage (transistor or tube)	Superheterodyne (4–6 active devices)	25–100 W	3.5–21 MHz	CW, AM, SSB
4	Refined multi-band transceiver	Integrated transceiver	50–200 W	1.8–30 MHz	CW, SSB

---

2. VACUUM TUBE FABRICATION

2.1 What a vacuum tube requires

A triode vacuum tube — the simplest type useful for radio transmission — consists of:⁶

• Glass envelope: A sealed glass container maintaining an internal vacuum. Borosilicate glass (Pyrex-type) is ideal due to its low thermal expansion, but soda-lime glass works for lower-power tubes. Glass production is addressed in Doc #98.
• Filament (heater): A wire that, when electrically heated, reaches a temperature sufficient to cause thermionic emission — the release of electrons from a metal surface. Tungsten wire is the standard filament material (melting point 3,422degC, operating temperature for thermionic emission approximately 2,000–2,500degC). Thoriated tungsten (tungsten with approximately 1–2% thorium oxide) emits electrons at lower temperatures and lasts longer.⁷
• Cathode: The electron-emitting surface. In directly heated tubes, the filament itself is the cathode. In indirectly heated tubes (more common in later designs), a separate cathode sleeve coated with barium oxide or strontium oxide is heated by the filament — this arrangement allows the cathode to operate at a uniform temperature and simplifies circuit design. For initial fabrication, directly heated designs are simpler.
• Grid: A fine wire mesh or helix positioned between cathode and plate. Applying a voltage to the grid controls the flow of electrons from cathode to plate — this is the amplification mechanism. The grid must be close to the cathode (typically 1–3 mm) and must not touch either the cathode or the plate.
• Plate (anode): A metal surface (typically nickel or copper) that collects the electrons. In power tubes, the plate dissipates significant heat and must be designed accordingly.
• Vacuum seal: The envelope must maintain a hard vacuum (below approximately 10^-3 Pa) throughout the tube’s operating life. Any gas leakage causes the tube to fail. Glass-to-metal seals where the electrode leads pass through the envelope are the most common failure point.⁸

2.2 The fabrication challenge

Vacuum tube fabrication is not conceptually difficult — it was routine industrial practice from 1910 to 1970. But it requires several capabilities that must be developed:

Glassworking: Forming the envelope, making glass-to-metal seals, and connecting components inside the envelope before sealing. NZ will need trained glassblowers working with appropriate glass (borosilicate preferred). Doc #98 addresses glass production; the glassblowing skill itself must be developed through practice. Scientific glassblowers at NZ universities (if still practising) are the starting point.⁹

Vacuum pumping: Evacuating the sealed envelope to a hard vacuum. Laboratory-grade rotary vane pumps (salvaged from NZ university and hospital vacuum systems) can achieve the required vacuum levels. Mercury diffusion pumps — constructible from glass and mercury, both available in NZ — provide the higher vacuum stages needed for quality tubes. The vacuum system is not the hardest part of tube fabrication, but it must work reliably.¹⁰

Tungsten wire: For filaments. NZ does not produce tungsten. The immediate source is salvage: incandescent light bulbs contain tungsten filaments, and NZ has millions of light bulbs in homes, warehouses, and retail stocks. Each bulb contains a very small amount of tungsten (approximately 0.5–1.0 g for a standard 60–100 W bulb), but millions of bulbs represent a meaningful stockpile. Longer-term, tungsten must be imported via trade (Australia has tungsten deposits) or substituted — thoriated tungsten from welding electrode salvage (TIG welding electrodes are 97–98% tungsten).¹¹

Electrode fabrication: Forming the grid, plate, and cathode from nickel, copper, or steel sheet and wire. This is conventional metalworking at small scale — within NZ machine shop capability (Doc #91).

Glass-to-metal seal materials: The seal wires must have thermal expansion coefficients matching the glass envelope. Kovar (a nickel-cobalt-iron alloy, approximately 29% Ni, 17% Co, balance Fe) is standard for borosilicate glass seals but is not produced in NZ and must be salvaged from existing electronic equipment or imported. For soda-lime glass envelopes, copper wire (available from Doc #70) can be sealed directly, but soda-lime glass has higher thermal expansion and is more prone to cracking — limiting tube reliability. This material constraint may determine whether borosilicate or soda-lime glass is used for tube envelopes.¹²

Getter material: After sealing, a “getter” — typically barium or magnesium — is flashed inside the tube to absorb residual gas molecules and maintain the vacuum over the tube’s lifetime. Barium is not common in NZ, but small quantities can be obtained from barium sulfate (barite), which occurs in some NZ mineral deposits, or from salvaged electronic components (CRT television tubes contain barium getters).¹³

2.3 Realistic fabrication yield

Historical vacuum tube production in the 1920s–1930s, when the technology was mature but production was less automated than later decades, achieved yields of 60–80% — meaning 20–40% of tubes produced were rejected for failing to meet specifications.¹⁴ An initial NZ fabrication effort, working from first principles with improvised equipment, should expect significantly lower yields — perhaps 20–40% initially, improving to 50–70% with practice and process refinement.

This means producing 100 working tubes may require fabricating 200–500 tubes. The labour and material cost is not negligible. But vacuum tubes, once working, have long operational lifespans — thousands to tens of thousands of hours for receiving tubes, and hundreds to thousands of hours for transmitting tubes operating at moderate power levels.¹⁵

Estimate: A trained team of 2–3 workers (glassblower, metalworker, vacuum technician) could produce approximately 5–15 working tubes per week once processes are established — enough to supply the radio fabrication program and build a tube stockpile.

2.4 Tube types for radio service

Not all tube types are equally important. The fabrication program should focus on:

Triode (3 elements: cathode, grid, plate). The simplest amplifying tube. Suitable for oscillators, RF amplifiers (at HF frequencies), and audio amplifiers. A single triode type, designed for medium-power RF service (10–25 W plate dissipation), could serve as the universal building block for both transmitters and receivers in Stages 1–2.

Tetrode or pentode (4 or 5 elements). More complex but provides higher gain and better stability at RF frequencies. Pentodes are particularly useful for IF amplifiers in superheterodyne receivers. These are a Stage 2–3 development target — not needed for initial CW transmitters.

The strategy is to design radio circuits around a single tube type initially, even if this is not optimal from a circuit design perspective. Manufacturing one tube type well is far more practical than attempting to produce several types simultaneously.

---

3. PASSIVE COMPONENTS FROM NZ MATERIALS

3.1 Inductors (coils)

Inductors are wound from insulated copper wire on a form (cylinder or toroid). NZ can produce these immediately:

• Wire: Copper wire from Doc #70 pipeline, recycled motor windings, or salvaged electrical cable. Wire must be insulated — enamel-coated magnet wire is ideal (salvaged from motors and transformers), but cotton-covered or varnished wire works. Wire gauges from 0.5 mm to 2.0 mm diameter cover most radio applications.
• Coil forms: Ceramic tubes (from NZ clay — Doc #70/pottery), PVC pipe sections (from existing stocks), wooden dowels (sealed with varnish), or cardboard tubes coated with shellac or varnish for moisture resistance.
• Design: Coil inductance is determined by the number of turns, diameter, length, and core material. Air-core coils are standard for RF applications. Iron-core coils (using laminated transformer iron) are used for audio frequencies and power supply filtering. All design calculations are available in the ARRL Handbook and similar references.¹⁶

3.2 Capacitors

Capacitors store electrical energy and, with inductors, form the tuned circuits that select specific radio frequencies. Several types can be fabricated locally:

Variable capacitors: Interleaved metal plates (aluminium or brass) with an air dielectric, rotated to vary capacitance. These are precision mechanical devices used for receiver tuning. Salvaged variable capacitors from old radios are the best short-term source. New fabrication requires machine shop work (Doc #91) to produce smooth-operating plate assemblies — achievable but requiring careful attention to plate spacing and alignment.

Fixed mica capacitors: Thin sheets of mica (a natural mineral) interleaved with metal foil. Mica capacitors are stable, precise, and handle high voltages — ideal for RF circuits. NZ has mica deposits, though their quality and accessibility require assessment. Salvaged mica capacitors are available from electronic waste.¹⁷

Paper-and-foil capacitors: Aluminium foil interleaved with waxed paper, rolled into a cylinder. Suitable for lower-frequency applications (audio, power supply filtering). All materials are available in NZ. These were the standard capacitor type from the 1920s through the 1960s. Performance is adequate for audio and power supply filtering but inferior to modern film capacitors: higher leakage current (typically 5-50x higher than polypropylene film), lower temperature stability, and shorter operating life (5–15 years versus 20–30+ years for film types). They are also physically larger for the same capacitance rating and are unsuitable for precision RF tuning applications.¹⁸

Ceramic capacitors: Produced from NZ clay (kaolin), formed into discs or tubes, coated with silver paste electrodes. Ceramic capacitors are the most demanding to produce — they require high-temperature firing and precise dielectric formulation — but they offer excellent RF performance and are worth developing for Stage 3+ designs.

3.3 Resistors

Resistors limit current flow and set bias voltages. Several types are locally producible:

Wire-wound resistors: Resistance wire (nichrome from salvaged heating elements, or constantan/manganin from instrument salvage) wound on ceramic or composition forms. Precise and high-power capable, though fabrication requires careful winding technique to achieve target resistance values within useful tolerances (typically +/- 5–10% for hand-wound units, versus +/- 1% for commercial equivalents). The main limitation is that wire-wound resistors are inductive, which can cause problems at RF — this is acceptable for DC bias and audio circuits but not for RF signal paths.

Carbon resistors: Carbon rods or carbon-composition elements provide non-inductive resistance suitable for RF circuits. Pencil lead (graphite) is a crude but functional source of carbon resistance — however, graphite pencil-lead resistors are fragile, temperature-sensitive, and difficult to produce at consistent values, making them suitable only for prototyping. More refined carbon-composition resistors require mixing carbon powder with a ceramic binder (kaolin from NZ clay deposits), forming into rods, and firing at 800–1,000degC in a kiln (Doc #102 for charcoal-based carbon source; Doc #98 for kiln capability). Achieving consistent resistance values requires careful control of the carbon-to-binder ratio and firing temperature — expect +/- 20% tolerance initially versus +/- 5–10% for commercial carbon-composition resistors.¹⁹

3.4 Transformers

Power transformers (mains voltage to tube supply voltages) and audio transformers (for coupling between amplifier stages) are wound from insulated copper wire on laminated iron cores. NZ can produce these:

• Cores: Laminated silicon steel from transformer scrap (recycled from failed distribution transformers — Doc #69) or cut from silicon steel sheet. Grain-oriented silicon steel is preferred for efficiency but standard mild steel laminations function, albeit with higher losses.
• Wire: Magnet wire from Doc #70 or recycled motor windings.
• Insulation: Varnish, shellac, paper, or cotton between layers.
• Design: Transformer design for radio power supplies is well documented in the ARRL Handbook and electrical engineering references. The calculations are tractable for trained technicians; the construction requires patience and care, particularly in achieving adequate inter-layer insulation and tight, uniform winding to avoid excessive leakage inductance and voltage regulation problems.²⁰

---

4. FREQUENCY CONTROL

4.1 Why frequency control matters

A radio transmitter must operate on a specific frequency to be received by stations listening on that frequency. If the transmitter drifts off frequency, contact is lost. For CW (Morse code) communication, frequency stability of approximately +/- 500 Hz is adequate — the receiving operator can track small drifts by retuning. For voice communication (AM or SSB), tighter stability is needed — approximately +/- 100 Hz for SSB.²¹

4.2 Crystal oscillators

The gold standard for frequency control is the quartz crystal oscillator. A thin slice of piezoelectric quartz crystal, cut to precise dimensions, vibrates at a specific frequency when an electric field is applied. This frequency is determined by the physical dimensions of the crystal and is extremely stable — typically within a few parts per million.²²

NZ quartz deposits: NZ has quartz deposits, but the critical question is whether any NZ quartz is of the optical-grade purity required for piezoelectric oscillator crystals. Most natural quartz is polycrystalline or contains inclusions that prevent it from functioning as an oscillator element. The Coromandel Peninsula (particularly the Thames-Coromandel district with its extensive epithermal quartz vein systems), the Reefton area on the West Coast, and the Otago goldfields region are the most promising sources for large single crystals, but suitability must be assessed by cutting and testing sample crystals — a process requiring lapidary equipment (grinding and polishing) and RF test equipment to verify oscillation.²³

Crystal fabrication process: Assuming suitable raw quartz is available:

1. Select clear, inclusion-free quartz crystal specimens
2. Orient the crystal using X-ray diffraction (if equipment survives) or, less precisely, by optical methods
3. Cut thin blanks (approximately 0.5–3 mm thick depending on target frequency) using a diamond saw or abrasive slurry
4. Grind and lap the blanks to precise thickness — the resonant frequency is inversely proportional to thickness, so the target thickness must be controlled to within approximately 0.01 mm for HF crystals
5. Apply electrodes (silver paint or evaporated metal) to both faces
6. Mount in a holder with light spring pressure on the electrodes
7. Test for oscillation at the target frequency and adjust by further grinding if necessary

Feasibility assessment: Crystal fabrication is achievable if NZ quartz is suitable, but it requires precision grinding equipment and patience. Each crystal must be individually tuned by grinding. Production rate would be slow — perhaps 2–5 crystals per day by a skilled worker. This is adequate for equipping a radio network (each transmitter needs 1–3 crystals) but not for mass production. Rated [B-C] — feasible if raw material is suitable, difficult if not.

Alternative: cultured quartz. Industrial quartz crystals are normally produced from cultured (hydrothermally grown) quartz. This process requires an autoclave operating at approximately 350–400degC and 100–170 MPa (1,000–1,700 atmospheres) with a seed crystal and a nutrient solution of sodium carbonate or sodium hydroxide. This is within the capability of NZ pressure vessel fabrication (Doc #112, #94) but is a significant engineering project in itself. It is a Phase 4+ development.²⁴

4.3 Alternative frequency control

If crystal oscillators are unavailable or in limited supply:

LC VFO (variable-frequency oscillator). An oscillator using an inductor-capacitor tuned circuit. Frequency is set by adjusting the capacitor or inductor. LC oscillators drift with temperature, vibration, and supply voltage changes. Drift can be reduced by:

• Using high-quality components (silver-plated coil wire, air-spaced variable capacitors, stable fixed capacitors)
• Mechanical rigidity (rigid mounting, vibration isolation)
• Thermal compensation (using components with opposing temperature coefficients)
• Voltage regulation of the power supply

A well-built LC VFO can achieve stability of approximately +/- 200–500 Hz at 7 MHz over a period of hours — adequate for CW, marginal for SSB voice. Many amateur transmitters of the 1940s–1960s operated successfully with LC VFOs.²⁵

Calibration approach: Even without crystal-controlled transmitters, crystal-calibrated receivers allow accurate frequency measurement. A single crystal (say, 1 MHz) produces harmonics at every integer multiple of its frequency (2 MHz, 3 MHz, 4 MHz…) that serve as frequency markers across the HF spectrum. A transmitter can be set to a known frequency by tuning it to zero beat with the nearest crystal harmonic and counting the offset.

---

5. TRANSMITTER DESIGNS

5.1 Stage 1: Single-tube CW transmitter

The simplest possible transmitter: a single triode tube operating as a self-excited oscillator, keyed for CW.

Circuit description: The tube’s grid and plate circuits are both connected to a single tuned circuit (coil and capacitor), with feedback from plate to grid sustaining oscillation. The antenna is coupled to the oscillator either directly (for simple designs) or through a separate coupling coil. A telegraph key in the cathode circuit switches the oscillator on and off to produce Morse code.²⁶

Components: - 1 triode tube (salvaged or locally fabricated) - 1 variable capacitor (100–365 pF range) - 1 inductor (hand-wound, approximately 10–30 turns of 1 mm copper wire on a 40 mm diameter form for the 7 MHz band) - 1 telegraph key (fabricated from brass strip and a wooden base) - 1 power supply (300–500 V DC for the plate, 6.3 V AC for the filament) - 1 antenna coupling coil - Miscellaneous: resistors for grid leak, bypass capacitors, chassis, connectors

Performance: Output power depends on the tube type and plate voltage. With a medium-power triode at 400 V plate supply, 10–25 W RF output is achievable. This is enough for reliable NVIS communication within NZ on 3.5 MHz (80m) and for trans-Tasman communication on 7 MHz (40m) under favourable propagation conditions.²⁷

Limitations: Self-excited oscillators have poor frequency stability (the antenna load affects the oscillator frequency — “pulling”). They also produce relatively broad signals with significant harmonic content, which can interfere with other stations. These limitations are acceptable for initial CW communication but should be addressed in Stage 2 designs.

5.2 Stage 2: Crystal-controlled oscillator with power amplifier

Separating the oscillator from the power amplifier solves the frequency stability problem. The oscillator runs at low power with a crystal or stable LC circuit; a separate amplifier tube boosts the signal to transmission power.

Circuit description: A crystal oscillator (one tube) generates a stable RF signal at the crystal frequency. This feeds a power amplifier (one or two tubes) that amplifies the signal to 25–100 W. The amplifier’s antenna coupling does not affect the oscillator frequency because they are isolated.²⁸

Components: - 1 crystal or stable VFO - 2–3 tubes (1 oscillator, 1–2 amplifier) - Tuned circuits for each stage - Power supply (higher current capacity than Stage 1) - Keying circuit (with shaping to reduce key clicks — a resistor-capacitor network across the key)

Performance: 25–100 W output with stable frequency. If crystal-controlled, frequency stability is excellent (parts per million). This is a professional-quality CW transmitter, comparable to commercial amateur equipment of the 1940s–1950s.

5.3 Voice transmission

Voice (AM or SSB) transmission requires modulating the RF carrier with audio:

AM (amplitude modulation): The simplest voice modulation method. Requires an audio amplifier chain (microphone, voltage amplifier, power amplifier) with output power equal to approximately 40–60% of the RF stage power for full modulation depth.²⁹ AM transmitters require fewer precision components than SSB but use spectrum and power less efficiently — an AM signal occupies twice the bandwidth and uses roughly three times the total power of an equivalent SSB signal for the same voice intelligibility.

SSB (single sideband): The standard voice mode for HF communication. SSB suppresses the carrier and one sideband, concentrating all power in a single sideband containing the voice information. SSB transmitters are significantly more complex to build — they require a balanced modulator, crystal filter (for sideband selection), and linear amplifier chain. SSB is a Stage 3–4 target, not a priority for initial fabrication.³⁰

---

6. RECEIVER DESIGNS

6.1 Crystal detector receiver

The simplest possible receiver: a tuned circuit (coil and variable capacitor), a crystal detector (galena or germanium diode point-contact), and headphones. No power supply required. Detects AM signals and, with practice, CW signals (CW produces a clicking sound in a crystal detector; adding a BFO (beat-frequency oscillator) — a separate local oscillator that produces an audible beat note with the CW signal — makes CW reception practical).

NZ construction: Galena (lead sulfide, PbS) occurs in NZ, particularly associated with gold-bearing quartz veins in the South Island. A small piece of galena crystal with a fine wire point contact (the “cat’s whisker”) is the detector element. The entire receiver can be assembled in a few hours by a worker with soldering experience, though finding a functional galena specimen and adjusting the cat’s whisker contact point requires patience and some trial and error.³¹

Limitations: No amplification. Useful only for strong signals — a local station within a few hundred kilometres, or a high-power broadcast station. Not suitable as the primary receiver for the HF network, but valuable as a no-power backup, as a training tool, and as a proof of concept.

6.2 Regenerative receiver

The most important receiver type for the early stages of local fabrication. A regenerative receiver uses positive feedback (feeding some of the output signal back to the input) to increase sensitivity and selectivity dramatically compared to a crystal detector.³²

Circuit description: A single amplifying device (tube or transistor) amplifies the received signal. A portion of the output is fed back to the input tuned circuit through a feedback control. As feedback is increased, the receiver becomes increasingly sensitive and selective — approaching oscillation. Just below the point of oscillation, the receiver is at maximum sensitivity. Beyond the oscillation point, the receiver becomes a simple CW/SSB detector (the BFO effect is automatic).

Components: - 1 tube or transistor - 1 tuned circuit (variable capacitor + coil, as in Stage 1 transmitter) - 1 feedback control (variable capacitor or variable coupling coil) - 1 audio amplifier (1 additional tube/transistor driving headphones or speaker) - Power supply (for tube: approximately 100–250 V plate supply; for transistor: 6–12 V battery)

Performance: A well-built regenerative receiver using a single tube or transistor provides sensitivity and selectivity comparable to a 3–4 tube non-regenerative design. It can receive CW and SSB signals from thousands of kilometres away. Many successful trans-Tasman and transpacific amateur contacts were made in the 1920s–1940s using regenerative receivers.³³

Limitations: Regenerative receivers require operator skill — the feedback control must be continuously adjusted as the receiver is tuned across the band. If the feedback is advanced too far, the receiver oscillates and radiates a signal from its antenna, interfering with other receivers. Compared to a superheterodyne receiver, the regenerative design has significantly poorer selectivity (typical 3 dB bandwidth of 2–5 kHz versus 0.3–2.4 kHz for a crystal-filtered superhet), making it more susceptible to adjacent-channel interference on crowded bands. These limitations are manageable with trained operators but constrain the number of stations that can operate simultaneously on a given frequency band.

6.3 Superheterodyne receiver

The superheterodyne (superhet) is the standard receiver architecture for all serious radio communication from the 1930s onward. It converts all incoming signals to a fixed intermediate frequency (IF) — typically 455 kHz for AM or 9 MHz for SSB — where high selectivity and gain can be achieved in fixed-tuned amplifier stages.³⁴

Architecture: 1. RF amplifier (optional, 1 tube/transistor) — amplifies the incoming signal 2. Local oscillator (1 tube/transistor) — generates a signal offset from the received frequency by the IF 3. Mixer (1 tube/transistor) — combines the RF signal and local oscillator to produce the IF signal 4. IF amplifier (1–3 tubes/transistors) — amplifies and filters the IF signal 5. Detector (diode) — extracts the audio from the IF signal 6. Audio amplifier (1–2 tubes/transistors) — drives headphones or speaker 7. BFO (1 tube/transistor, for CW/SSB) — generates a beat note for CW reception

Total active devices: 5–8 tubes or transistors. This is significantly more complex than a regenerative receiver and represents a Stage 3 development target.

Performance: Excellent selectivity and sensitivity. Stable tuning. No radiation from the receiver. Capable of receiving all HF modes (CW, AM, SSB) with good performance.

Key component: The IF filter that determines receiver selectivity. Crystal filters (using quartz crystals ground to the IF frequency) provide excellent selectivity (bandwidths of 0.3–2.4 kHz achievable) but require crystal fabrication capability. LC filters are the locally producible alternative but offer significantly reduced performance — typical LC IF filter bandwidth is 5–10 kHz at best, which is adequate for AM reception but too wide for crowded CW or SSB operation where adjacent-signal rejection is critical.

---

7. POWER SUPPLIES

7.1 High-voltage supply for vacuum tubes

Vacuum tube transmitters and receivers require DC plate voltages of 100–500 V (receivers) to 500–2,000 V (high-power transmitters). These are produced by:

1. Transformer: Steps up mains voltage (230 V AC in NZ) or a lower AC source to the required voltage. Locally wound (Section 3.4).
2. Rectifier: Converts AC to pulsating DC. Options include vacuum tube rectifiers (a diode tube — among the easiest tube types to fabricate, requiring only a cathode and plate with no grid), selenium rectifiers (salvaged), or silicon diodes (salvaged from electronic waste — silicon diodes are extremely abundant in modern electronics and will be available from salvage for decades).
3. Filter: Smooths the pulsating DC. Electrolytic capacitors (salvaged) or choke-input filters (an iron-core inductor followed by a capacitor) — the latter is more robust and uses locally producible components.

7.2 Low-voltage supply for transistor circuits

Transistor radios operate from 6–12 V DC — directly from batteries (Doc #35) or from a mains step-down transformer and rectifier. Power consumption is typically 10–100x lower than for tube equivalents (a transistor receiver draws 10–50 mA versus 500 mA–2 A for a comparable tube receiver), which is a significant advantage for off-grid and portable operation.

7.3 Filament supply

Tube filaments require 6.3 V AC (the most common standard) or other voltages depending on the tube type. Locally designed tubes should standardise on a single filament voltage — 6.3 V is recommended for compatibility with NZ’s stock of salvaged power transformers, which were universally wound for this standard. A 6.3 V filament winding on the main power transformer is standard practice. NZ’s 230 V / 50 Hz mains standard (AS/NZS 3112) must be accounted for in all locally wound transformer designs.³⁵

---

8. FREQUENCY ALLOCATION FOR RECOVERY COMMUNICATIONS

8.1 Coordination with Doc #128

The frequency plan established in Doc #128 (Section 4) defines operating frequencies for the national HF network. Locally fabricated radio equipment should be designed to operate on these frequencies. The priority bands are:

• 3.5 MHz (80m): Domestic NVIS communication. Relatively easy to design for — wavelength is long, component tolerances are forgiving.
• 7 MHz (40m): Domestic NVIS and trans-Tasman. The workhorse band. All locally fabricated equipment should cover this band.
• 14 MHz (20m): International communication. Higher frequency is more demanding on tube and circuit design — a Stage 2–3 target.

8.2 Frequency management for locally fabricated transmitters

Locally fabricated transmitters, particularly Stage 1 designs with LC oscillators, may have broader signals and greater harmonic content than commercial equipment. To minimise interference:

• Crystal-controlled transmitters should be used where crystals are available
• LC oscillator transmitters should be operated on designated frequencies separated from crystal-controlled stations
• Operators should monitor their own transmission on a separate receiver to verify frequency and signal quality
• Harmonic filters (low-pass filters using coils and capacitors, designed to attenuate frequencies above the operating band) should be included in all transmitter designs to suppress harmonics that could interfere with other services

---

9. THE NZ AMATEUR RADIO COMMUNITY: KNOWLEDGE BASE

9.1 ZL homebrewers

NZ’s amateur radio community — callsign prefix ZL — numbered approximately 3,000–4,000 licensed operators as of the early 2020s, organised through the New Zealand Association of Radio Transmitters (NZART) with approximately 70 branches nationwide.³⁶ The community has a tradition of homebrew construction that predates commercial equipment. Early ZL operators (1920s–1960s) built their own transmitters and receivers because commercial equipment was either unavailable or unaffordable in NZ due to geographic isolation and import costs. This tradition has continued in a smaller community of constructors who build equipment for technical interest rather than necessity.

These operators hold practical knowledge of radio construction that cannot be learned from textbooks alone: how to wind coils that do not exhibit parasitic oscillation, how to lay out a chassis for stable operation, how to tune a transmitter for maximum output without damaging the tube, how to recognise and correct common faults. This is apprenticeship knowledge, and it resides in a small and aging population.

Recommended action: The skills census (Doc #8) should specifically identify amateur operators with homebrew construction experience. These individuals should be recruited into the radio fabrication program — initially as consultants and teachers, and as their health permits, as active constructors. Their knowledge is the seed from which the fabrication program grows.

9.2 NZART technical resources

NZART’s technical library, accumulated over nearly a century of operation, contains construction articles, circuit diagrams, and practical advice for radio building at all levels of complexity. The organization’s magazine, Break In, has published construction articles throughout its history. These resources, combined with the ARRL Handbook and RSGB Handbook, constitute a comprehensive design library for locally fabricated equipment.³⁷

9.3 Integration with trade training

Radio fabrication requires skills that overlap with other recovery-critical trades: electrical wiring (Doc #69), metalworking (Doc #91, #95), glassblowing (Doc #98), and basic chemistry (Doc #113). The radio fabrication training program should be integrated with the broader trade training framework (Doc #157) to develop multi-skilled technicians who can contribute across these overlapping domains.

---

10. ANTENNA CONSTRUCTION

Antenna systems are covered in detail in Doc #128 Section 7. From a fabrication perspective, the key point is that antennas are the one part of the radio system that NZ can produce indefinitely without significant technical barriers — though copper wire supply depends on the Doc #70 pipeline or recycling. Wire antennas (dipoles, long wires, inverted-Vs) require only:

• Copper or aluminium wire (from Doc #70 or recycling)
• Insulators (glass, ceramic, or hardwood — all NZ-produced)
• Feedline (open-wire ladder line is fabricated from wire and timber spacers)
• Support masts (timber poles, abundant in NZ)

Even if all electronic components fail, antennas remain producible. The fabrication program should ensure that antenna construction knowledge is widely distributed — every community radio operator should be able to build and maintain an effective antenna from locally available materials.

---

11. STAGED DEVELOPMENT TIMELINE

Period	Milestone	Dependencies	Output
Year 0–1	Component salvage, tube inventory, skill identification	Doc #8 census	Stockpile of salvaged components; identified skill base
Year 1–2	First locally assembled CW transmitters (salvaged components)	Salvaged tubes and parts	5–10 working transmitters for proof of concept
Year 2–3	First locally fabricated vacuum tubes	Glass production (Doc #70), tungsten salvage	Working triode tubes for RF service
Year 2–4	First fully locally fabricated CW transmitter/receiver pairs	Tube fabrication, copper wire (Doc #70)	10–20 working stations
Year 3–5	Crystal oscillator production (if NZ quartz suitable)	Quartz assessment, lapidary equipment	Crystal-controlled transmitters
Year 3–5	Regenerative receivers in production	Tube or transistor supply	Receivers for network stations
Year 5–7	Superheterodyne receivers	More complex tube/transistor designs	High-performance receivers
Year 5–10	Germanium transistor integration	Doc #115 program output	Transistor-based portable equipment
Year 7–10	Full transceiver production at network scale	All preceding milestones	50–100 units/year
Year 10+	Refined, multi-band transceiver designs	Accumulated experience	Self-sustaining equipment production

---

12. CRITICAL UNCERTAINTIES

Uncertainty	Why It Matters	How to Resolve	Impact if Adverse
NZ quartz suitability for oscillator crystals	Frequency control quality determines transmitter precision	Cut and test samples from Coromandel and other NZ quartz deposits	If unsuitable, rely on LC VFOs (adequate for CW, marginal for SSB) and salvaged crystals until cultured quartz is feasible
Vacuum tube fabrication yield	Determines how many tubes must be made to equip the network	Begin experimental fabrication in Year 1–2; track yields over hundreds of attempts	If yields are very low (<10%), the program consumes more labour and glass than projected; increase production team size
Tungsten supply for tube filaments	Every tube needs a filament; tungsten is not produced in NZ	Inventory incandescent bulbs and welding electrodes nationally; assess trade sources (Australia)	If tungsten is exhausted before trade provides supply, develop oxide-coated directly heated cathodes using lower-temperature metals (nickel alloys)
Germanium transistor availability timeline	Transistors enable lighter, more efficient, more reliable radios	Depends on Doc #115 program progress	If delayed beyond Phase 5, the program remains vacuum-tube based — functional but less efficient
Number of surviving ZL homebrewers with construction knowledge	These individuals are the seed knowledge for the entire program	Census (Doc #8) and NZART activation (Doc #8)	If very few survive or are available, knowledge must be reconstructed from published references alone — slower but feasible
Barium availability for tube getters	Getter material is essential for tube longevity	Assess NZ barite deposits; inventory CRT tubes and other barium sources	If unavailable, tube vacuum life is shorter; tubes must be replaced more frequently
Achievable frequency stability without crystals	Determines whether CW-only or CW+voice communication is feasible with local equipment	Build and test LC VFO designs; measure drift empirically	If stability is poor, communication is limited to CW until crystals are available — CW works but is slower for message traffic

---

13. CROSS-REFERENCES

Document	Relevance
Doc #135 — Computer Construction	Germanium transistor fabrication pipeline; shares prerequisite chemistry and materials
Doc #35 — Battery Management and Lead-Acid Production	Power supply for portable and off-grid radio stations
Doc #70 — Wire and Cable Production	Copper wire for coils, transformers, antennas, and internal wiring
Doc #89 — NZ Steel: Glenbrook Operations	Steel and iron for transformer cores, chassis, mounting hardware
Doc #91 — Machine Shop Operations	Metalworking for tube electrode fabrication, variable capacitor manufacture, chassis construction
Doc #98 — Glass Production	Glass envelopes for vacuum tubes; insulators for antennas and capacitors
Doc #102 — Charcoal Production	Carbon source for resistor fabrication
Doc #102 — Wire Drawing and Fabrication	Fine wire production for tube grids and coil winding
Doc #113 — Sulfuric Acid and Basic Chemistry	Chemical processing for germanium extraction (Doc #113), electrode preparation
Doc #128 — HF Radio Network	The communication network that this equipment serves; antenna systems; frequency plans; operator training
Doc #138 — Sailing Vessel Design	Maritime HF communication requirements drive equipment specifications
Doc #157 — Trade Training and Apprenticeship	Training framework for radio fabrication technicians
Doc #160 — Heritage Skills Preservation	Urgency of capturing knowledge from aging ZL homebrew constructors

---

APPENDIX A: BILL OF MATERIALS FOR A STAGE 1 CW TRANSCEIVER

Transmitter (single-tube oscillator, 7 MHz, approximately 15 W output):

Component	Specification	NZ Source
Tube	Medium-power triode (e.g., 6L6, 807, or locally fabricated equivalent)	Salvage or local fabrication
Variable capacitor	15–150 pF, air-spaced	Salvage or machine shop fabrication
Coil	12 turns, 1 mm copper wire, 35 mm diameter, air-core	Hand-wound from NZ copper wire
Grid leak resistor	10–47 kohm, 1 W	Salvaged or wire-wound
RF bypass capacitors	100–1000 pF (2–3 required)	Salvaged mica or locally made
Power transformer	230 V primary, 400 V + 6.3 V secondary	Locally wound
Rectifier	Tube diode, silicon diode (salvaged), or selenium rectifier	Salvage or tube fabrication
Filter capacitors	8–16 uF, 450 V (2 required)	Salvaged electrolytic
Filter choke	5–10 H, 100 mA	Locally wound on scrap iron core
Telegraph key	Single-pole switch	Fabricated from brass strip
Chassis	Aluminium or steel sheet	Recycled or NZ Steel
Antenna connector, wire, hardware	Various	Salvage and local fabrication

Receiver (regenerative, 7 MHz, headphone output):

Component	Specification	NZ Source
Tube or transistor	Small-signal triode or germanium transistor	Salvage or fabrication
Variable capacitor	15–365 pF, air-spaced (tuning)	Salvage
Variable capacitor	5–50 pF (regeneration control)	Salvage or fabricated
Coil	12 turns, 0.8 mm copper wire, 35 mm diameter	Hand-wound
Audio transformer	Interstage or output	Salvaged or locally wound
Headphones	High-impedance (2000 ohm) preferred	Salvage; or low-impedance with output transformer
Power supply	100–250 V plate (tube) or 6–12 V (transistor)	Transformer/rectifier or battery
Miscellaneous	Resistors, bypass capacitors, knobs, panel, wire	Salvage and fabrication

Estimated construction time (experienced builder): 20–40 hours for the pair, excluding power supply construction. An inexperienced builder with guidance: 60–100 hours.

---

---

1. Equipment lifespan estimates based on observed longevity of solid-state amateur radio transceivers (see Doc #115, Section 10.2). Electrolytic capacitor aging and semiconductor degradation are the primary failure modes. The 10–20 year range assumes careful use and competent component-level maintenance.↩︎

2. ARRL (American Radio Relay League) Handbook for Radio Communications, published annually — the standard comprehensive reference for amateur radio construction. RSGB (Radio Society of Great Britain) Radio Communication Handbook — the British equivalent. NZART’s Break In magazine, published since 1930, contains NZ-specific construction articles. Multiple physical copies of these references exist in NZ amateur radio stations and public libraries and should be preserved.↩︎

3. Network station count estimate: based on the network design described in Doc #128, which proposes regional hub stations in each of NZ’s approximately 16 regions plus community-level stations serving rural districts. The 150–250 range depends on the density of community stations deployed; this figure should be refined as the network plan matures.↩︎

4. Operator count estimate: the 200–500 operator range assumes 1–3 operators per station across the national network, accounting for shift coverage and relief. See Doc #128 for the network staffing model.↩︎

5. Crystal detector technology is described in any history of early radio. See Aitken, H.G.J. (1985), The Continuous Wave: Technology and American Radio, 1900–1932, Princeton University Press. The crystal detector was the standard radio receiver from approximately 1906 to 1920 and remained in use as a simple receiver well into the 1940s.↩︎

6. Vacuum tube theory and construction: Terman, F.E. (1943), Radio Engineers’ Handbook, McGraw-Hill — the standard reference for vacuum tube circuit design. Also: RCA Receiving Tube Manual and equivalent publications by other manufacturers. These are widely available in NZ amateur radio libraries.↩︎

7. Tungsten filament technology: thermionic emission from tungsten requires temperatures above approximately 2,000degC. Thoriated tungsten (1–2% ThO2) reduces the required temperature to approximately 1,700–1,900degC and increases electron emission, extending filament life. See Reimann, A.L. (1934), Thermionic Emission, Chapman and Hall.↩︎

8. Glass-to-metal seals: the thermal expansion coefficients of the glass and the metal lead-through wire must be closely matched to prevent cracking during cooling. Kovar (a nickel-cobalt-iron alloy) is the standard material for seals with borosilicate glass. Copper wire can be sealed through soda-lime glass. See Partridge, J.H. (1949), Glass-to-Metal Seals, Society of Glass Technology.↩︎

9. NZ has scientific glassblowers at universities (University of Auckland, Victoria University of Wellington, University of Canterbury, and University of Otago have historically maintained scientific glassblowing workshops). The current staffing of these workshops requires verification through the skills census (Doc #8).↩︎

10. Vacuum pump technology for tube fabrication: rotary vane pumps (widely available in NZ laboratories) achieve approximately 1 Pa. Mercury diffusion pumps (constructible from glass tubing and mercury) achieve approximately 10^-4 Pa — sufficient for high-quality vacuum tubes. See Yarwood, J. (1967), High Vacuum Technique, Chapman and Hall.↩︎

11. Tungsten in NZ: NZ has no significant tungsten mining. Sources for tube fabrication include: incandescent light bulb filaments (millions of bulbs in NZ, each containing approximately 0.5–1.0 g tungsten), TIG welding electrodes (2% thoriated or ceriated tungsten, approximately 2.4 mm diameter, 150 mm long — approximately 10 g each), and tungsten carbide tool inserts (in machine shops). Australia has tungsten deposits (King Island, Tasmania) accessible via trade. Tungsten quantity estimates based on standard component specifications.↩︎

12. Glass-to-metal seal materials: Kovar (ASTM F15 alloy) has a thermal expansion coefficient of approximately 5.1 x 10^-6 /degC, closely matching borosilicate glass (approximately 3.3 x 10^-6 /degC with graded seal techniques). Copper can be sealed directly into soda-lime glass (expansion approximately 9.0 x 10^-6 /degC) using a Housekeeper seal technique, where the copper is thinned to a feather edge to accommodate expansion mismatch. See Partridge, J.H. (1949), Glass-to-Metal Seals, Society of Glass Technology.↩︎

13. Getter materials: barium is the most common getter in vacuum tubes. Barium sulfate (barite) occurs in NZ in association with some mineral deposits. CRT television tubes contain barium getters — NZ has millions of CRT television sets that became electronic waste during the LCD transition. Barium from CRT salvage is a practical short-term source. See Espe, W. (1966), Materials of High Vacuum Technology, Pergamon Press.↩︎

14. Historical vacuum tube production yields: manufacturing data from the 1920s–1930s is fragmentary but industry sources indicate rejection rates of 20–40% in mature production. See Tyne, G.F.J. (1977), Saga of the Vacuum Tube, Howard W. Sams. Initial NZ fabrication would face higher rejection rates due to less controlled processes.↩︎

15. Vacuum tube operating life: receiving tubes (low-power) typically last 5,000–20,000 hours. Transmitting tubes at moderate power levels: 1,000–5,000 hours depending on operating conditions. Higher-power operation and inadequate cooling shorten tube life. See RCA Transmitting Tube Manual for rated life specifications.↩︎

16. Coil and transformer design: detailed design procedures are available in the ARRL Handbook (Chapter on Circuit Design and Component Fabrication), Terman (1943) Radio Engineers’ Handbook, and similar references. The calculations use algebra and empirical correction factors; a trained technician with these references can design functional coils and transformers, though achieving optimal performance requires iterative testing and adjustment.↩︎

17. Mica deposits in NZ: NZ has limited mica resources. Some mica occurs in association with pegmatite deposits in Westland and other regions, but NZ has not historically been a significant mica producer. The quality and sheet size of NZ mica for capacitor use requires assessment. Salvaged mica capacitors from electronic waste are the primary short-term source. See GNS Science mineral occurrence data for NZ mica locations.↩︎

18. Paper-and-foil capacitor performance: leakage current in paper capacitors is typically 1–10 microamps per microfarad at rated voltage, compared to 0.01–0.1 microamps for polypropylene film types. Operating life depends on humidity and temperature exposure. See Dummer, G.W.A. (1970), Fixed Capacitors, Pitman — comprehensive historical reference covering paper capacitor design and performance data.↩︎

19. Carbon composition resistor fabrication: mixing finely ground carbon (graphite) with a ceramic binder (kaolin or other NZ clay), forming into small rods, and firing at approximately 800–1,000degC. The carbon-to-ceramic ratio determines resistance. This process is documented in early electronics manufacturing literature. See Dummer, G.W.A. (1970), Fixed Resistors, Pitman.↩︎

20. Coil and transformer design: detailed design procedures are available in the ARRL Handbook (Chapter on Circuit Design and Component Fabrication), Terman (1943) Radio Engineers’ Handbook, and similar references. The calculations use algebra and empirical correction factors; a trained technician with these references can design functional coils and transformers, though achieving optimal performance requires iterative testing and adjustment.↩︎

21. Frequency stability requirements for HF communication: CW bandwidth is approximately 100–500 Hz, so transmitter drift of a few hundred Hz is tolerable — the receiving operator can compensate by retuning. SSB bandwidth is approximately 2.4 kHz, and the suppressed carrier frequency must be reproduced accurately at the receiver — frequency errors above approximately 100 Hz produce noticeably distorted voice. See ARRL Handbook, chapter on transmitter design.↩︎

22. Piezoelectric quartz crystal oscillators: quartz crystals vibrate at a frequency determined by their physical dimensions and the crystallographic orientation of the cut. The AT-cut (used for most HF crystals) has a frequency-temperature coefficient near zero around 25degC. See Bottom, V.E. (1982), Introduction to Quartz Crystal Unit Design, Van Nostrand Reinhold.↩︎

23. NZ quartz deposits: quartz is abundant in NZ but piezoelectric-grade quartz (large, clear, inclusion-free single crystals) is relatively rare in any location. The Coromandel Peninsula, with its extensive hydrothermal vein systems, is the most promising area. Assessment requires collecting samples and testing for piezoelectric activity — a straightforward test requiring only a multimeter and an RF oscillator circuit. See GNS Science geological maps and mineral occurrence databases for NZ quartz locations.↩︎

24. Hydrothermal quartz crystal growth: the industrial process for growing synthetic quartz crystals. See Laudise, R.A. (1970), The Growth of Single Crystals, Prentice-Hall. The process requires temperatures of 350–400degC and pressures of 100–170 MPa in an autoclave — demanding but within the capability of NZ pressure vessel engineering once other industrial priorities are met.↩︎

25. LC VFO stability: a well-designed VFO using quality components can achieve drift rates of approximately 100–500 Hz per hour at 7 MHz after initial warm-up. This is documented extensively in amateur radio construction literature. See Hayward, W. (1982), Introduction to Radio Frequency Design, Prentice-Hall. Permeability-tuned oscillators (using a movable ferrite core rather than a variable capacitor) offer improved mechanical stability.↩︎

26. Single-tube oscillator transmitter designs: this is the most basic transmitter topology, used universally by amateur radio operators in the 1920s–1940s. Circuit descriptions and construction details are available in any edition of the ARRL Handbook, the RSGB Handbook, and countless construction articles in amateur radio magazines.↩︎

27. Power output and range estimates: a 15 W CW transmitter on 7 MHz with a half-wave dipole antenna produces an effective radiated power sufficient for reliable communication over approximately 500–2,000 km under average propagation conditions. Trans-Tasman contacts (approximately 2,000 km) at this power level are routine for experienced CW operators. See amateur radio propagation references and QSL (confirmed contact) records from the early amateur radio era, when comparable power levels were standard.↩︎

28. Crystal-controlled oscillator-amplifier transmitter: the MOPA (master oscillator, power amplifier) configuration. This became the standard amateur transmitter architecture in the 1930s–1950s. See ARRL Handbook, Terman (1943), and similar references for detailed circuit designs and construction guidance.↩︎

29. AM modulation power requirements: for 100% modulation depth (the maximum undistorted modulation), the audio amplifier must supply power equal to 50% of the carrier power. In practice, average modulation runs at 40–70% of full depth, and the audio amplifier is typically designed for 40–60% of the carrier power to provide headroom without excessive cost. See Terman, F.E. (1943), Radio Engineers’ Handbook, McGraw-Hill, Chapter 13.↩︎

30. SSB transmitter design: the filter method (using a crystal lattice filter for sideband selection) or the phasing method (using audio phase-shift networks). Both require significantly more components and tighter tolerances than AM or CW transmitters. See Hayward, W. and DeMaw, D. (1986), Solid State Design for the Radio Amateur, ARRL. SSB is important for efficient spectrum use but is not essential for initial CW-based communication.↩︎

31. Galena crystal detectors: galena (PbS) is a natural semiconductor. The rectifying junction formed at the point contact between a fine wire and the crystal surface detects AM radio signals. Galena occurs in NZ in association with gold-bearing quartz veins (e.g., Reefton, West Coast; Macraes, Otago). See Park, J. (1906), The Geology of the Mining Districts of the North Island, NZ Geological Survey, for historical NZ mineral occurrence data. Crystal detector receiver construction is described in any history of early radio.↩︎

32. Regenerative receiver: invented by Edwin Armstrong in 1912. The regenerative principle — using positive feedback to increase gain and selectivity — was the most important receiver development before the superheterodyne. See Armstrong, E.H. (1915), “Some Recent Developments in the Audion Receiver,” Proceedings of the IRE. Construction details in ARRL Handbook and similar references.↩︎

33. Historical trans-Tasman amateur radio contacts: the first trans-Tasman amateur radio contacts were achieved in the 1920s using homebrew equipment comparable in performance to what is described here. NZ amateur radio history is documented in NZART publications and Frank Bell’s pioneering work. See NZART historical records and Break In magazine archives.↩︎

34. Superheterodyne receiver: invented by Edwin Armstrong in 1918, it became the dominant receiver architecture by the 1930s and remains so today. The principle — frequency conversion to a fixed IF — provides selectivity, sensitivity, and stability that simpler architectures cannot match. See Terman (1943), ARRL Handbook, and any modern communications receiver text.↩︎

35. Vacuum tube power supply design: standardized filament voltages (6.3 V for most receiving tubes, 5.0 V for some rectifier tubes) emerged in the 1930s and were maintained through the end of the tube era. Locally designed tubes should adopt one of these standards to maintain compatibility with salvaged power transformers. See RCA Receiving Tube Manual for standard tube operating specifications.↩︎

36. NZ amateur radio homebrew tradition: NZART (New Zealand Association of Radio Transmitters) reported approximately 3,000–4,000 licensed amateur operators in the early 2020s, organised through roughly 70 branches. The ZL amateur community, particularly in the period 1920–1970, was noted for homebuilt equipment due to NZ’s geographic isolation from commercial equipment sources. Notable NZ amateur radio constructors and their techniques are documented in Break In magazine and NZART branch records. The current generation of homebrew constructors is smaller but still active, particularly in the QRP (low-power) community. See NZART (nzart.org.nz) for current membership data; these figures require verification through the skills census (Doc #8).↩︎

37. ARRL (American Radio Relay League) Handbook for Radio Communications, published annually — the standard comprehensive reference for amateur radio construction. RSGB (Radio Society of Great Britain) Radio Communication Handbook — the British equivalent. NZART’s Break In magazine, published since 1930, contains NZ-specific construction articles. Multiple physical copies of these references exist in NZ amateur radio stations and public libraries and should be preserved.↩︎