Part three: Frequency Calibration
As soon as commercial broadcasting got
started it was necessary for listeners to be able to find their favorite
stations on the dial of their receivers. Stations began to publish
their program listings in the newspapers just as TV stations do today.
A station was identified by its' frequency on the dial, and the receiver
had be able to quickly dial up any given station. When there were
few stations on the air this was not a problem, but as the band got packed
with stations receivers had to get better. With their greater selectivity,
superhets could separate stations on the band. Now they had to be
set to a given spot on the band, and do it on demand.
The first receivers used venire dials
with relative readouts, say from 0 to 100. The user could log where
each of his favorite stations came in. Some receivers had two scales
on the dial, one blank and one with the relative scale. The user
could pencil in the station call signs on the blank part of the dial.
Eventually the dials were simply calibrated with the frequency the set
was tuned to. This required that each receiver be aligned at the
factory so the dial tracked the actual frequency the set was tuned to.
When a set was serviced, the repairman had to check the alignment (changing
a tube could throw off the calibration since there were variations between
tubes, even those of the same type.)
As mentioned in part two, there were
several types of dials: Airplane dials had fixed circular scales with a
moving pointer. Meter type dials had a moving circular dial scale
with a fixed pointer. Slide rule dials had a linear scale with a
sliding pointer. These used pulleys and string to move the pointer.
When multi-band receivers came out, the slide rule dial had the advantage
of having all scales of equal size. The circular dials would crowd
the scale on the inner radius of the dial.
The first multi-band short wave sets
used plug in coils, later designs had switches changing coil sets.
These superhets used variable frequency oscillators which covered a different
frequency range for each band. The tuning rate for each band was
different. Each band had to be calibrated separately. The tuning
was faster on each higher band, because the tuning ratio was about the
same band per band. If a variable capacitor with a maximum
capacity of 350 400 was used (as common in broadcast sets) the set would
have a 3 to 1 tuning ratio. This is about right on the broadcast
band tuning from 550 khz to 1600 khz. Switching to the first short
wave band the set might tune from 2 to 6 mhz. This is still a 3 to
1 tuning ratio, but now the khz per dial rotation is twice as much as it
was on the broadcast band. The set tunes faster. On the next
short wave band the set might now tune from 5 to 15 mhz. The set
tunes even faster.
To make it easier to tune these short
wave receivers on the higher frequency bands a second tuning control, called
a band spread tuner was added. The band spread tuner had a separate
tuning dial calibrated with several tuning ranges for various sub-bands.
Usually these were for the amateur bands, although some receivers also
had band spread scales for a few of the international broadcast short wave
bands. To use the band spread tuner the main tuning control had to
be set to a mark which corresponded to a specific scale on the band spread
scale. For example, there would be a mark at 3.5 or 4.0 mhz which
corresponded to the band spread scale for the 3.5 to 4.0 mhz amateur band.
There were two types of band spread controls. One type was purely
mechanical. It consisted of a separate slow motion gearing which
provided for the band spread action using the main tuning capacitor.
The second type was electrical. It used a separate tuning capacitor
with a smaller maximum capacity than the main tuning capacitor. This
capacitor was wired in parallel with the main tuning capacitor, and was
provided with a separate tuning dial and gear drive. The second type
of band spread control was the most common in use.
The classic multi-band short wave
receiver with a band switched variable frequency first oscillator had its'
limitations. It was difficult to make such an oscillator stable on
the higher frequencies. Receivers of this type tended to drift when
used on twenty meters and above. Temperature compensated circuits,
voltage regulators, and separate power supplies for the vfo to allow it
to remain powered up when the rest of the receiver was shut off (to eliminate
the warm up period) helped and were used on the more expensive sets.
There was a better way and soon after world war II new ground was broken
in short wave receiver design.
The Collins 51J and 75A receivers
introduced a new design concept that would change short wave receivers
and amateur radio forever. Instead of using a band switched vfo in
the front end with a fixed IF, Collins made the first IF variable and used
a crystal controlled first oscillator. Band switched crystals solved
the drift problem on the higher frequencies. The vfo now tuned a
single frequency range so the same tuning rate existed on all bands.
The receiver could be calibrated once for all bands (the first oscillator
crystals did have to be set to the exact frequency but this was a small
adjustment that could be made once per band, anywhere within the band.)
The resultant receiver was actually a single band receiver with a band
switched crystal controlled converter stage ahead of it. The result
was a double conversion receiver. The icing on the cake was the newly
invented mechanical filter. These IF filters, first made at 500 khz,
later at 455 khz, provided a nearly brick wall response curve. They
were made in several bandwidths: 16, 8, 4, 2, and .5 khz (and any others
by request). The IF could now be designed for gain, without considering
bandwidth. The bandwidth was provided by the filters.
Collins also designed the first linear
tuning vfo. This device changed frequency at a rate constant with
the rotation of the shaft. Collins design used a variable inductor,
but similar devices were built using variable capacitors. With a
linear tuning vfo a turns counting dial could be used as a direct frequency
readout. This was in fact done on the famous R390 series of military
receivers designed by Collins. National Radio also produced an SSB
transceiver using such a dial.
A variation on the above theme was
to keep the fixed first IF, but pre-mix the vfo with a crystal oscillator.
This also resulted in a receiver with low drift on the higher frequency
bands, and a fixed tuning rate. The Drake R3 and R4 series of receivers
(and corresponding transmitters / transceivers) used this scheme.
They had a first IF around 5.5 mhz using a crystal filter in the IF.
Yet another way to obtain a constant
tuning rate using one variable oscillator range was to make use of the
image response. If an IF of 1700 khz was used with a vfo of 5.2 to
5.7 mhz the receiver would tune both 3.5-4.0 mhz and 6.9-7.4 mhz.
By mealy re-tuning the input of the receiver both the 80 and 40 meter bands
would be tuned. Another such combination was to use a vfo of 5.0
to 5.5 mhz with a 9 mhz IF. This would tune 3.5-4.0 mhz and 14.0-14.5
mhz, covering the 80 and 20 meter bands. Many early SSB rigs used
this design, which explains why the 75 and 20 meter bands were the first
to have SSB phone operation replace AM.
