SECTION 7
There are other designs of radios that do not use the
superheterodyne principle.
CRYSTAL SETS
These radios only have about 4 parts in them, so they are easy
to fix, but also hard to fix, as each part must perform to a high standard.
The most important item is the tuning device, which can be a coil and variable
capacitor, or just a tapped coil (using self capacitance). Some designs
may use a variometer, which is one rotating coil inside another. The 2 coils
assist or oppose and vary the inductance. The Q of these coil must be high
to get loud stations and sharp tuning. There was a lot of thought put into
coil design, and the basket wound coils with Litz wire (multi-stranded) using
air core with no former, performed the best. The detector could be
a galena crystal (cat's whisker) or germanium diode. Silicon diodes work fine
in the metropolitan area. The headphones must be high impedance, so as not
to load the circuit. If the headphones are low impedance, use a speaker transformer
as a matching transformer. Any speaker transformer will do. A large aerial
and a good earth are essential.
The easiest way to check the coil is with a grid dip meter.
It can tell you the resonant frequency, and (by the suddenness of the
dip) a rough indication of the Q (quality). Common faults are corrosion
of the wire. If all the stations are cramped up one end of the dial, then
the self capacity is too high. If the coil is wound with cotton covered
(CC) wire or double cotton covered (DCC) wire, check for moisture
in the covering. Gently drying it in the sun is the best way. Insulating
the coil with glue or spray lacquer may reduce the Q. Check the variometer
axles and connecting wires.
Common headphone faults are an open circuit coil due to
corrosion. Rejoin or rewind the coil. Check for faulty leads. Tinsel
wire is hard to rejoin. To fix it place some fine strands of normal copper
wire in the join, then bind it with cotton, to match the covering. It works
sometimes. The ends can be terminated, by winding fine tinned copper wire
around them, and then soldering. Check and clean
the metal diaphragm. A special hardened steel is used. Transformer laminations
can be used, but they are usually not wide enough.
HORNS
A radio horn is just the same as a single headphone, except it is
heavier duty and it has a large metal tube of increasing cross section.
The horn couples the sound from the headphone like driver, to the air. The
horn can be metal or wood. Be careful as the horns that are usually used
on radios, are connected directly in the amplifier anode circuit, and so
have HT on them. The driver suffers the same faults
as headphones. Check the horn for blockages, or air leaks as the sound can
escape and reduce the volume. This is important for wooden horns.
T.R.F. RADIOS (Tuned Radio Frequency)
These sets are very simple and can be regarded as an amplified
crystal set. They may have one or two RF stages and one or two audio
amplifiers. The TRF, due to their time of manufacture, used directly
heated valves, so there was no cathode, and therefore no way to make
cathode bias. The grid bias had to be generated from a C battery (in battery
TRF sets), or a C supply, or back bias (in a mains powered set). Another
method, was to ground the positive side of the filament, and use the negative
side as the grid return. This was the age before ganged tuning capacitors,
so each stage had a separate single gang tuning capacitor. The detector
is commonly an anode bend detector, or a grid leak detector. The anode
bend detector is merely an audio amplifier with no bias. The audio stages
usually have a negative or C bias supply. The grid leak detector is a capacitor
and high value resistor in parallel, the combination providing the detection.
The RF coils should be at right angles (or well shielded) to prevent coupling
between stages. There was sometimes a resistor in the grid lead as a "stopper"
to prevent oscillation. Another method was "neutralisation" which required
some feedback from plate to grid, usually by a small capacitor. The gain
of these radios was set by a filament rheostat, which adjusted the filament
voltage, and the gain. On some sets, there was a filament rheostat and a
tuning condenser for each stage, resulting in lots of knobs. The sets often
used headphones, or a horn loudspeaker.
The most common fault in a TRF battery set is burnt out
valve filaments, as it is very easy to connect the A and B leads wrongly.
A common fault in the detector is a high or open circuit grid leak resistor.
Faults in the RF stages are often open circuit windings and dirty wipers
on the tuning capacitors, caused by the age of the sets. Audio faults
are often open circuit transformers. Also look for corrosion on the valve
sockets and wiring. If the construction method is bare copper wiring, to
bolts or screws on the valve sockets, clean these. If oscillating, check
the grid stopper resistors. Check that the coils are connected the correct
way. Check that it is not a regenerative circuit and is supposed to oscillate!
To neutralise a stage, tune to a station, remove the filament voltage of
that stage, and adjust any neutralising control for MINIMUM volume, because
the neutralising control works by producing negative feedback to counter
any capacitive coupling in the valve.
REGENERATIVE RADIOS
These detectors were common in the 1920s and also for home made
radios. The radios used an almost oscillating detector followed by audio
amplifier stages. They relied on feedback to increase the gain, but needed
careful adjustment, so that they did not break into oscillation, and turn
into transmitters, using the radios aerial. There was fear in the war,
that ships with these types of radios, would broadcast the location of
a convoy and attract enemy attack. Some care was taken to avoid this by
adding a filter or RF stage between the aerial and the radio. Screening
of the local oscillator in a superheterodyne was elaborate in some maritime
receivers. These sets were easy to use. Just turn the reaction up full,
then tune across the band, and each station appears as a whistle. Stop on
the one you want, then back off the reaction control. The design problem
was to make the drop into oscillation gentle, as if it was too quick, it
was difficult to stay at the maximum loudness point, and out of oscillation.
Common faults with these, are no oscillation, due to a
valve loosing emission, or components aging. If the drop into oscillation
is too quick, then reduce the screen or plate voltage. If it won't oscillate,
and someone has been working on it, check that the reaction winding
has not been connected backwards. Faults in the audio section
were usually open circuit audio transformers.