Hallicrafters S-20 Sky
I first saw this
old receiver at an auction house in Christchurch in April 2000.
It was painted in a sort of cream coloured emulsion over its
original finish with plenty of rust showing through. The chassis,
visible under its lift-up lid was red with rust and altogether
it was in a very sorry state. I can't exactly remember what I
bid for it but it probably wasn't more than a couple of pounds
as old radios, only a few years ago, rarely fetched more than
£15. After getting it home I decided to restore it to what
I imagined was its original condition, being considerably helped
by finding a picture of it in my 1938 copy of The ARRL Radio
Amateur's Handbook, where it was listed at $49.50.
I dismantled the metalwork,
removed the old paint and rust from the panels, and repainted
them with a dark grey crackle finish Hammerite. After derusting
the chassis with a proprietary gel, which I think, turns the
rust into a harmless phosphorus compound, I reassembled everything.
The dial mechanism and gears needed cleaning and oiling and worked
OK (see further down this page) although the dial surface shows
a lot of wear. Under the chassis the majority of the components
were serviceable but to my dismay I found that someone had added
a long wave band at the top short wave position, discarding the
original coils, padders and trimming condensers in the process.
This must have been when this particular band had been dead due
to the vagaries of sunspot activity. Being keen on restoring
the receiver to its original spec I decided to remove the long
wave band and refit the original band which covered the unusual
range of 17 to 44Mc/s. After some experimentation this turned
out to be no mean feat! To carry out the reconversion I had to
turn to my computer and Lotus 123 (remember this was the last
year of the last millenium!), which, apart from its usual role
of producing spreadsheets, allows complex equations to be processed.
I managed by experiment, using a test coil and a grid-dip meter,
to measure the minimum and maximum values of each section of
the tuning condenser gangs. The position of the wavechange switch
in relation to the tuning condenser and the grids of the valves
wasn't too bad at low frequencies but very awkward at low VHF.
I had to measure stray capacitances of leads and valve grids
and the wavechange switch wafers etc. By developing an algorithm
based on all the factors for each of the three coils (2xRF and
oscillator) I was able to determine the required inductance of
each coil. This included calculating values for trimmer condensers
for tuning and a padder condenser for the oscillator coil. The
latter is necessary because the oscillator tracks the RF coils
by the IF frequency which I also fed into the equations. I also
had to determine the proper side of the RF to run the oscillator
although I recall, that given the known parameters, my equations
told me this was only possible one way.
Eventually after much
trial and error, and by careful selection of components of the
right vintage at the bottom of my junk box, I wound the coils,
got the oscillator to oscillate and the RF stages to amplify
and NOT oscillate, everything turned out OK and the desired short-wave
coverage was achieved. I have a little suspicion, although I
am writing this after some years have passed, and looking back
without too much clarity, I might have cheated when it came to
overall performance and added a little extra amplification in
front of the audio output valve. Who's going to look for a transistor
under the chassis of a pre-war radio! Since then I've taken additional
pictures and can confirm there is a transistor lurking under
the chassis of this 1938 receiver.
The cone of the old loudspeaker
had perished and I remember fitting a new unit designed for a
car radio. Apart from this and a new mains lead and 13amp plug
(and I suppose a transistor) the restoration is fairly original
although I never got round to finding a couple of replacement
aluminium extrusions missing from the sides.
For the reader's information
the receiver has the following valve line-up: -
6K7 RF stage; 6L7 frequency
changer; 6J5 RF oscillator; 6K7 IF amplifier; 6Q7 detector, AVC
and AF amplifier; 6F6 power output; 6J5 BFO and a type 80 rectifier.
And the frequency coverage is
545Kc/s to 1800Kc/s
1.7Mc/s to 5.8Mc/s
5.8Mc/s to 18.8Mc/s
17Mc/s to 44Mc/s
Above the metal tuning
dial on which the previous owner has marked the various amateur
Left, the fine-tuning logging scale
which rotates in line with the main logging scale markings (0-24)
on the metal dial.
A neat layout of trimmers
for the RF coils.
I recall using the cathode
bias voltage from the output valve to power the transistor. I
used to listen to Radio New Zealand and added extra audio amplification
to give me better results.
An advertisement for the Sky
Champion from 1939.