This is only a preview of the October 2017 issue of Silicon Chip. You can view 34 of the 104 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Items relevant to "0.01Hz - 6+GHz touchscreen frequency meter, Part 1":
Items relevant to "El Cheapo Modules Part 10: GPS receivers":
Items relevant to "One hour project: Kelvin – the very clever cricket":
Items relevant to "3-way Active Crossover for speakers, Part 2":
Items relevant to "Deluxe eFuse, Part 3: using it!":
Purchase a printed copy of this issue for $10.00. |
Vintage Radio
By Associate Professor Graham Parslow
HMV 1955 Portable
Model 12-11
If you think the HMV set
featured this month looks
very similar to the model
B61D featured in the June
2017 issue, you are quite
right. But even though
both sets use the same battery valves, the same case
and even the same chassis,
there are significant differences in their circuits.
How can that be? Partly this is explained by the fact that the later set
has a 4-valve superhet instead of five
valves but offsetting this is fact that it
can be powered from batteries or from
its inbuilt 240VAC mains supply.
Externally, there are few differences
between them since the same case was
used for a number of HMV portable
radio models between 1951 and 1956.
One subtle difference between the
12-11 and B61D is in the brass Little
Nipper badge on the front. There is
a line across the bottom of the 1955
badge, while the 1951 badge had the
words “HIS MASTER’S VOICE” instead. The badge on the model 12-11
does contain those words but they are
written in a smaller font, below the image of Little Nipper (the dog listening
to His Master’s Voice from the gramo88
Silicon Chip
phone) and above the horizontal bar.
When I received this radio, the exterior was quite grubby but internally it
was quite clean. Luckily, the exterior
cleaned up well and now matches the
clean sound that it produces, which
is about as good as a portable of this
type can get.
The circuit
The speaker and some other components on my set are stamped February
1955, so this one is reliably dated. Its
circuit appears in the 1955 compilation of the Australian Official Radio
Service Manual (AORSM) and is reproduced in Fig.1.
Both these sets use the same chassis and the same loop antenna with
external aerial coupling. However,
there was a welcome change in the
Celebrating 30 Years
later 12-11 set with the use of a plug
and socket connection of the aerial to
the chassis so that the back can be easily removed. The loop antenna is part
of the first tuned LC circuit.
And that is where the first major
change to the circuit becomes apparent
in that there is no tuned RF amplifier
stage and the top of the chassis reveals
an unused hole for the missing valve.
At the same time, the tuning condenser is 2-ganged rather than 3-gang
and with no RF preamplifier, the tuned
signal feeds directly into the control
grid (pin 6) of the 1R5 pentagrid frequency changer, V1.
From that point on, the arrangement of the four remaining valves in
this largely conventional superhet
circuit is quite similar to the B61D
model. It has an almost identical 1R5
siliconchip.com.au
frequency changer circuit and the
intermediate frequency is the same at
457.5kHz.
Neutralisation
siliconchip.com.au
Fig.1: two aspects of this circuit are unusual. The seriesconnected directly heated cathodes of the four valves are at
different potentials by virtue of their position in the series
heat string. That necessitated a separate voltage divider
(R1, R2, R3 etc) to correctly bias the grid of each valve.
And the two diodes in the 6V4 rectifier are used as halfwave rectifiers to provide the HT and LT rails. Note also the
charging (“reactivation”) facility for the dry cell batteries.
This radio has a neutralisation capacitor, shown on the circuit diagram
connecting the two grids of the 1R5
via the local oscillator; its value is
not specified.
Neutralisation in valve circuits refers to cancelling the effect of internal
inter-electrode capacitance in order to
reduce its tendency to oscillate and
this also usually improves the stage’s
bandwidth.
Typically the neutralisation capacitor is connected between a point which
is 180° out of phase with the anode of
the mixer stage and its control grid.
Often, a tap on the IF transformer, or
the IF transformer secondary is the
connection point and so the IF transformer provides the necessary phase
inversion. This provides positive feedback at lower frequencies, improving
bandwidth.
But at higher frequencies, inherent
phase shifts, including those due to
the reactance of the neutralisation capacitor, cause this feedback to become
negative and this is why it reduces the
tendency of the amplifier to oscillate
at an unwanted frequency.
In this circuit, the connection of
the neutralisation capacitor is a little
unusual. V1 drives the local oscillator
at 457.5kHz above the tuned station’s
frequency.
Now the input and output sides of
the oscillator are normally 180° out of
phase at the oscillator’s operating frequency. In this case, they are the anode (pin 2) and grid (pin 4).
So the designers have taken advantage of this existing phase inversion
from the anode of V1 and are simply
connecting the neutralisation capacitor
between the local oscillator and main
control grid.
The signal path is slightly different
for neutralisation (via C2 rather than
C3) but the phase shift of both paths
will be similar and hence the neutralisation is effective.
There has been some correspondence to the Editor recently about the
subject of neutralising, with much
disagreement over exactly how it
works. To look into the topic a little
more deeply you might like to start
with the Wikipedia entry at https://
en.wikipedia.org/wiki/Neutrodyne
Celebrating 30 Years
October 2017 89
Reproduced from a label stuck to the underside of the chassis, this diagram
shows the dial cord stringing arrangement, chassis arrangement, battery
replacement instructions and the alignment frequencies.
IF stage and biasing
Moving on now, IF transformer IFT1
feeds the 475.5kHz signal to the 1T4 IF
amplifier, V2. This stage is stabilised
by shunt capacitor C8.
The amplified signal is demodulated by the diode in the 1S5 valve (V3)
and the audio appears across R8 in
series with the volume control VR1.
The junction of these two resistors
becomes more negative under strong
signals and this provides feedback for
automatic volume control (AVC, otherwise known as AGC).
At this point, it’s worth mentioning the somewhat unusual biasing arrangement in this set.
Both mixer/oscillator V1 (1R5) and
IF amplifier V2 (1T4) have different
negative AVC bias voltages applied
to their grids via resistors R1, R2 and
R3. V2’s screen grid is connected to
HT via a decoupling network comprising R7 and C9, while V3’s screen
is similarly connected to HT via R15,
filtered by C16.
Series-connected filaments
All the filaments of the five valves
in the earlier B61D model ran from a
1.5V cell but in this set, all the filaments are connected in series to run
from a common 9V B supply which
can be a battery or the in-built 240VAC
mains supply.
Note that these are directly heated
cathodes and that means for V1-V4, the
cathode connection at pin 1 is shared
with one side of the filament (heater).
And that means that the cathodes of
V1-V4 are all at different potentials.
V3’s cathode is at ground potential
while V1 is higher, V2 higher again
and V4 the highest.
This meant that the designers had to
go to special lengths to correctly bias
the grid of each valve and this was arranged in two ways. First, while the
grid of V4 is connected to chassis via
a 1MW resistor (R8), the grids of the
other three valves connect to a voltage
divider comprising three high value
resistors (R1, R2 & R3) together with
the volume control VR1.
At the same time, three of the four
heaters (V1, V3 and half of V4’s tapped
heater) are shunted with resistors and
these have been chosen to fine-tune
the grid bias voltages of the various
valves. Note the two RC filters in the
filament network, to reduce the noise
and ripple coupling into the most sensitive stages, V1 and V2.
Audio amplification
Audio from volume control pot
VR1 is AC-coupled to the pin 6 control grid of V3 (1S5) which is the first
audio amplification stage. The signal
is then coupled by C17 to pin 6 of
V4, the control grid of the 3V4 output pentode.
V4’s screen is connected directly to
the HT rail and capacitor C20 is con-
While this is the same chassis as used for the HMV B61D described in the June 2017 issue, the layout is quite different
with four valves rather than five, a 2-gang tuning condenser rather than a 3-gang unit and three extra capacitors.
90
Silicon Chip
Celebrating 30 Years
siliconchip.com.au
nected across the speaker transformer
to limit the audio bandwidth.
Negative feedback from the speaker output is provided by a centre tap
on the output transformer secondary,
which is fed back to the bottom of the
volume control pot. The volume control is earthed via the output transformer so the signal to the 3V4 valve
is diminished by subtracting an outof-phase waveform.
Resistor R11 is connected between
a tap on the volume control pot and
ground and presumably helps to ensure
that there is no output with the volume
control wound fully down and may
also serve to linearise the operation
of VR1.
The power supply
module, with the mains
transformer and 6V4
rectifier, was designed
to be shoe-horned into
the case of the radio (see
photo below).
Power supply
The separate 240VAC power supply
might look conventional, being based
on a 6V4 rectifier valve (V5). However,
the 6V4’s two diodes are cleverly used
separately, to produce both the HT and
LT rails, providing half-wave rectification for each.
A limit on maximum current and
the relatively high internal resistance
of the 6V4 rectifier (around 160 ohms)
makes a 1.5V supply providing 300mA
impractical.
Instead, the LT unit in this radio produces 20V without load which reduces to 10V under load (close enough to
the nominal 9V of the battery). Using
a bench supply, this radio drew 55mA
at 9V which is close to the AORSM
specified value of 47mA.
The HT rail was measured as 79V
from the on-board supply, a bit lower
than the nominal 90V but this made
The HMV 12-11 has
a Bakelite case and
is shown without
either of the two
batteries, which
would attach
to the sheet of
cardboard at
the bottom of
the case.
siliconchip.com.au
Celebrating 30 Years
October 2017 91
little difference to performance as assessed by using a bench supply varied
between 80V and 90V.
The power supply simply incorporates series ballast resistors to reduce
the voltage to the nominal 9V and 90V
rails based on expected current drain.
By long-standing convention, the 9V
“A” and 90V “B” batteries are physically separate.
The model information glued to
the top of the chassis, behind the tuning dial, shows a user how to install
Eveready battery types 765 (9V) and
490P (90V).
However, an intriguing extra came
with this radio. The two connectors
for separately plugging into the “A”
and “B” batteries were plugged into
an adaptor built on strong cardboard.
It served to combine the two plugs
into a single plug for a battery pack
offering the “A” and “B” batteries in
one package.
This seems to have been an innovation for HMV in 1955 because neither
the packaged information with the radio nor the AORSM data mention the
adapter.
Other manufacturers had used single battery packs from at least 1951.
The Eveready type 753 combination
battery incorporates a dummy-pin
hole, set off-centre to promote correct
insertion of the connector.
Battery reactivation
The side of the radio has a knob
marked OFF/AC/BAT/RE-ACT. The
circuit diagram shows how two Oak
wafer switches in the mains power
unit control these functions.
In RE-ACT mode, the set is off but
the mains power supply is connected
across both batteries for trickle-charging, with extra series resistors to limit the charge current to trickle levels.
HMV provide the following instructions for battery reactivation:
“After the receiver has been operated on its internal batteries the power
switch should be set to the RE-ACT
position and the mains supply to the
instrument turned on. The period of
reactivation should be approximately
six hours for each hour of use on dry
batteries. As an example a receiver operated for two hours on dry batteries
would require twelve hours reactivation and this could conveniently be
done overnight.”
“Although the time of reactivation
is not critical within an hour or so, it
is important not to exceed the recommended period by any considerable
margin. The ratio of reactivation to
battery usage time applies only to the
last daily period used.”
“For example should the receiver
be used on batteries for a total of two
hours daily for three days without reactivating, then the reactivating period
would be twelve hours, based on the
last period of two hours usage.”
“The cost of power taken from the
electric supply mains for reactivation
is very low. On the basis of power costing 3d [three pennies] per unit, the
cost of a reactivating charge of twelve
hours would be approximately one
third of a penny.”
While HMV referred to it as reactivation, this shows that charging of
carbon-zinc batteries has been around
for more than 60 years, even though
battery manufacturers normally do
not recommend charging of any primary batteries.
Reversing the chemical reaction that
creates battery current is a simple matter of chemistry, but the advisability
of doing so is another matter. During
reactivation, there would also be elec-
The under-chassis layout of
this set is much less cluttered
than the B661D set described
earlier, mainly due to the
omission of the RF amplifier
stage valve.
92
Silicon Chip
Celebrating 30 Years
siliconchip.com.au
Just for reference, here is what the set looked like pre-restoration. You can
see the dial is slightly cracked along the Queensland section.
trolysis of the aqueous electrolyte releasing hydrogen gas.
Reactivation does not create a magic
pudding of inexhaustible power because the chemistry is not completely reversed.
Modern alkaline batteries can likewise be regenerated, through perhaps
ten cycles, and there are many commercial products to do this. See the
discussion at https://en.wikipedia.org/
wiki/Recharging_alkaline_batteries
The speaker
The 1951 model previously described had a round 5-inch speaker
that was labelled HMV. This 1955
model has a larger 5x7-inch elliptical
speaker branded EMI and this would
have been manufactured at the Homebush plant in Sydney.
The HMV brand was first used by the
Gramophone Company UK in London
in 1921 for gramophones and records.
In 1931, The Gramophone Company
and The Columbia Company merged
to form Electric and Musical Industries (EMI) and began manufacturing
radios. HMV radios were made in Australia from 1936 at Homebush.
From the mid-1950s onward, all
HMV radios, valve and transistor,
carried an EMI logo on the speakers.
The HMV radios of the time were also
badge-engineered as Kelvinator with
some modified case work. Using EMI
as the speaker brand disguised its origin at HMV.
AWA did the same thing when it
branded speakers MSP (Manufacturers
Special Products) so that other manufacturers would not be overtly conflicted when they used MSP speakers.
A view of the case from the back
shows the elliptical space for mounting the speaker. A picture of the rear
Restoration
This radio was a relatively easy restoration project. However, at first power-up, it remained absolutely silent.
The solution was meticulous cleaning
of all valve pins and sockets to ensure
reliable contact.
During handling, the celluloid dial
sadly cracked and disintegrated into
fragments. Happily, a reproduction
dial was at hand, printed as described
in the article on the B61D, June 2017.
This radio was one of nine HMV
portables restored as a batch. Some
were more challenging than this radio and their story may be told later.
This radio is a reasonable performer on local stations in my area of good
signal strength. The case polished up
well so this restoration had a pleasing
conclusion.
SC
This set could be powered by a
battery pack containing one 90V
and 9V battery, using a multipin connector. The
disadvantage of
this is that the
pack would
need to be
discarded
as soon as
one of the
two batteries
became flat.
This could be
mitigated by
using reactivation.
The HMV 12-11 uses
a 5x7-inch elliptical
speaker. This speaker
sports the EMI label,
which was not present on
the speaker in the B61D
four years earlier.
siliconchip.com.au
of the case also shows the slots that
guide the chassis to precisely register the knobs with their access ports.
A bonus with this radio is the internally pencilled signature (“ER”), presumably of the person who checked
this radio for dispatch.
The dial background is red, a change
from the dark brown of earlier models.
The previously described 1951
model B61D had a cluttered, tightlypacked arrangement of components
under the chassis. This radio is much
less cramped, partly because it lacks an
RF amplifier section. Also, the bulky
power filter capacitors are mounted
on the chassis, not below.
This model also incorporates more
modern compact components, notably
the resistors that have the now-familiar colour bands for indicating values.
Although this radio lacks an RF amplifier section, other HMV models such
as the 22-11 of 1956 offered both an
RF amplifier and a mains power unit.
Celebrating 30 Years
October 2017 93
|