A Technical History of Short wave Receiver Design
A personal view by WA2MZE
Part One: Regenertive Receivers and Vernier
Dials
This is a work in progress. New chapters, corrections and
additions will be up loaded to this site as they become available.
Part One: Regenerative Receivers
The first short wave receivers were regenerative
sets using battery powered triode tubes. The accepted design used
two or three stages, a detector followed by one or two stages of audio
amplification. Commercially made broadcast sets didn't use regenerative
detectors, they had two or three stages of neutralized radio frequency
stages, a triode detector, and two stages of audio. The switch to
ac operated sets came sometime after the movement from 200 meters down
to today's present HF bands. It was the type 224 (aka 24A) tetrode
tube that really made ac operated short wave sets practical.
The regenerative detector was a simple
device, which offered a lot for very little. But the regenerative
detector was a monster needing to be tamed if it was to perform.
Antenna coupling was critical. Couple too tightly and the detector
would not oscillate, or would be uncontrollable. Couple too loosely
and signals would be weak. Antenna length and movement in the wind
would de-tune the set, or make the regeneration setting dependent on the
frequency the set was tuned to. Operation of the set with the detector
oscillating (necessary for cw or ssb reception) could couple the 'back
wave' back into the antenna causing QRM on the frequency for others.
To get the best of the regenerative set, several antenna coupling controls
were usually required. Series capacitors and inductors were used,
along with variable link coupling to the input tank circuit. In the
simplest circuits the antenna lead coupled with a gimmick capacitor to
the hot side of the input tank. A gimmick capacitor was just two
insulated wires twisted together.
Adding an RF stage ahead of the regenerative
detector solved most of the above problems all at once. The RF stage
isolated the antenna from the detector preventing radiation of the oscillating
detector. Coupling of the RF stage to the detector could be set very
loose since the RF stage would make up for the loss of signal. If
the RF stage was gain controlled, then overloading of the detector could
be avoided and just the right amount of signal strength to the detector
for good operation could be provided. The 224 tetrode was perfect
for the RF stage, it could be gain controlled with the addition of a variable
cathode resistor and required no neutralization for stable operation.
The 224 was also perfect for the detector stage .. even, stable control
of regeneration was provided by variation of the screen (grid #2) voltage.
If headphones were used (as they usually were for amateur sets) then a
single type 227 triode was used as an audio stage following the detector.
The National Radio SW3 receiver was one of the first commercially made
sets of this type. It was made using several different tube lineups.
The 6.3 volt version used type 39/44 tetrodes in the RF / Detector stages,
and a type 37 triode as the audio stage.
Control of regenerative feed back
was done in several ways. If feedback was via a 'tickler' winding,
the position of this coil could be varied. Sometimes a variable capacitor
in the tickler circuit was used. In many later receivers the position
and number of turns of the tickler coil was set to give just enough feedback
for proper regeneration in the middle of the band and the operating point
of the detector was changed, usually via varying the plate or screen voltage.
Sometimes the feed back path was in the cathode circuit, with the cathode
connected to a tap on the tank circuit. Any circuit that looked like
an oscillator could be wired as a regenerative detector circuit.
For some strange reason the best operation
always seemed to occur with a plate (or screen) voltage of 45 volts with
grid leak bias. It did not matter what type of tube was used,
the best operation was at 45 volts! Note that if a resistive plate
load was used the plate supply voltage would be higher, but the voltage
at the plate would still be 45 volts. In the case of a screen grid
tube the screen would be set at 45 volts, the plate voltage would be higher.
A screen grid tube acted as two triodes in series with corresponding higher
gain. The detector action taking place between the first and second
grids, and amplification taking place between the second grid and the plate.
This is one of the reasons that tetrode (and pentode) tubes made better
detectors than triodes.
These early receivers did not have
frequency calibrated dials. They used planetary gear reduction dials
with veriner readouts. These veriner dials could provide 3 or 4 place
accuracy reading from 0 to 100 (or 0 to 180, or even 0 to 200). The
relative dial position was applied to a graph to determine what frequency
the set was tuned to. The X axis of the graph could show the dial
position vs the Y axis showing the frequency. A function line running
diagonally down the graph plotted dial position vs frequency. A graph was
prepared for each plug in coil supplied with the receiver. Each receiver
had to be calibrated with the set of plug in coils supplied with it.
No two receivers calibrated exactly alike as each was hand assembled using
parts that differed from one sample to another. It would be some
time after the introduction of commercially built superhets before this
would change. This was not yet the age of precision components.