This is only a preview of the September 2013 issue of Silicon Chip. You can view 41 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 "Speedo Corrector, Mk.3":
Items relevant to "LifeSaver For Lithium & SLA Batteries":
Purchase a printed copy of this issue for $10.00. |
Vintage Radio
By Ian Batty
Best Of British: the Bush
TR82C Mk.2 transistor radio
A classic 7-transistor set from the early 1960s
manium transistors in the front end,
it was a solid performer. By contrast,
the TR82C Mk.2 described here used
alloy-diffused AF117 transistors in
the front end.
A Mullard design brief in 1960 for
a 6-transistor set with three alloydiffused transistors described it as offering “outstanding performance”. So
was Bush TR82C based on this circuit?
A quick check in my Mullard “Reference Manual of Transistor Circuits”
revealed that the Bush set is almost
identical to Mullard’s design.
In addition, having previously given
Raytheon’s T-2500 (also a 7-transistor
set) a thorough going-over, I was curious to see what differences there were
between it and the TR82C.
Face-off: T2500 vs TR82C
Bush Radio began in 1932, becoming part
of the Rank empire in 1945. Along with
the iconic DAC90 and DAC10 valve radios,
they also produced the distinctive TV22 TV
set. Here we take a look at their distinctive
TR82C Mk.2 7-transistor radio.
I
F THE BUSH TR82C’s classic styling evokes the era of rock’n’roll, it’s
with good reason. But it’s not exactly
unique, the styling having been based
on an earlier valve portable designated
the MB60.
Released in 1957 and designed by
the brilliant David Ogle, the MB60
just screamed “modernity”. It set a
siliconchip.com.au
benchmark for style which was wellmatched by its performance and sound
quality.
Background to the TR82C
With the 1950s transistor revolution
well under way, Bush responded in
1959 with the TR82. Initially kitted
out with alloy-junction OC44/45 ger-
Whereas Raytheon used a craftsman-built timber cabinet, Bush settled for an elegant moulded cabinet
with clean, bold lines. A large dial
dominates the front, with its anodised
red scale set back in a well behind the
tuning wheel. As a result, unlike the
Raytheon’s “fiddly” dial, the TR82C is
easily tuned using either a single finger
or a thumb and fingertip spanning
across the dial wheel. In addition, the
TR82C uses a slow-motion dial for ease
of tuning (as does the T2500).
The volume, wave-change and on/
off-tone controls all sit in a well at
the top of the case. The volume and
on/off-tone control knobs are wellknurled and easy to operate, while
the MED(ium) and LONG waveband
switches are easy to access and respond positively. Ergonomically, this
is one of the best sets in my collection.
By contrast with other sets, the cabinet uses a variety of trims. The metal
parts are chrome-plated and the plastics are either in their original colours
or “flashed” with bright finishes. The
control legends are recess-moulded
and filled with dark paint, making
them highly wear resistant. In my set
though, the dial wheel has yellowed
and grazed with age. This badly dims
September 2013 91
Fig.1: the Bush TR82C is a 7-transistor superheterodyne set with a push-pull audio output stage. Transistor VT1 is the
converter stage, VT2 & VT3 are IF amplifier stages and VT4-VT7 form the audio amplifier with T1 acting as a phasesplitter. Switch S1 selects between the AM broadcast band and the long-wave (LW) band.
TR82C, so it was interesting to see
how it handled strong signals (see
“Performance” section below).
Circuit description
The TR82C’s volume, wave-change and on-off/tone controls sit in a recessed
well at the top of the case. A large dial “wheel” on the front of the set is used
for tuning and is easy to operate.
the appearance of bright red anodising
on the tuning scale.
In summary, in the design and usability race, the Bush TR82C is the
clear winner over Raytheon’s T-2500.
Design basics
Like the Raytheon T-2500, the Bush
TR82C uses a conventional metal chassis. It’s made from aluminium and is
fitted with insulated mounting pins for
the transistors (somewhat reminiscent
of valve sockets). The transistors are
mounted on the rear side of the chassis,
92 Silicon Chip
allowing easy access to measure pin
voltages. It also makes it easy to desolder and replace individual transistors.
On the design front, the TR82C uses
a fairly conventional front-end: a selfoscillating mixer (converter) followed
by two IF stages and a diode detector
which also applies AGC to the first IF
stage. By contrast, some competing
designs (including Mullard’s 1960
proposal) employed an auxiliary AGC
diode to reduce converter gain on very
strong signals.
This refinement is absent on the
Fig.1 shows the circuit details. The
Bush TR82C is a 2-band set covering
both medium-wave (MW) and longwave (LW) frequencies. These bands
are selected using a somewhat unusual
inductor/capacitor switching method.
The tuning gang (CV1, CV2) uses
identical sections for the aerial and
local oscillator (LO) circuits. Note that
many single-band sets use a simplified
LO section without a padder capacitor.
However, this is impractical with a
multi-band set such as the TR82C, as
the LO would not track on any band
other than the MW (broadcast) band.
For MW, both tuned sections of the
ferrite rod antenna (L2, L4) are in parallel. This gives better signal coupling
but yields a total inductance lower
than either section alone.
The MW antenna trimmer (CT1) is
permanently connected across the MW
tuned winding (L2), while the tuning
gang’s antenna section (CV1) is wired
across the LW tuned winding. When
the LW band is selected, the MW
tuned winding is disconnected and
the LW winding (L4) only (still with
the tuning gang connected) is used.
In addition, the LW antenna trimmer
(CT4) and shunt capacitor C5 are also
now switched into circuit.
S1 switches the input to the converter between the two low-impedance
siliconchip.com.au
This view shows the speaker (front) side of the chassis. Note the large ferrite rod antenna and the point-to-point wiring
method used (ie, no PCB). All the parts are easy to access.
antenna windings, depending on the
band selected, ie, winding L3 for MW
and winding L5 on LW.
The LO uses a single coil assembly,
with L6 acting as the primary, L7 providing feedback and L8 used with CV2
for tuning. This means that a single
adjustment affects the low-frequency
end of both bands.
Basically, the designers have relied
on correct alignment for the MW band,
with padding and shunting capacitors
added to correct for the LW band. For
MW, a 556pF padder (C9) ensures correct LO tracking over the 995-2075kHz
range, ie, a consistent 470kHz (the IF)
above the MW band’s range of 5251605kHz. The trimmer capacitor is
CT3, with C3 in parallel.
Alternatively, when the LW band is
selected, shunt capacitor C1 (515pF)
restricts the LO range to about 528650kHz, ie, 470kHz above the 158280kHz LW band.
Self-oscillating mixer
Transistor VT1 is basically configured as a self-oscillating mixer. This is
an AF117 transistor and uses collectoremitter feedback, thereby reducing
radiation from the LO.
siliconchip.com.au
As mentioned, the intermediate
frequency (IF) section operates at
470kHz which is slightly more than
the customary 455kHz. The IF section
begins with IF transformer L9/L10
which has tapped and tuned primary
and secondary windings.
Transistor VT2 is the first IF amplifier stage. This is another AF117 and
is biased via the AGC circuit, with
R14 supplying a bias current. The
rectified output from diode detector
D1 (an OA90) “bucks” the bias (via
R12 & R13) with an opposing positive
voltage that increases with increasing
signal strength.
This AGC voltage changes the base
bias and hence the current gain of
VT2, thereby helping to maintain
near-constant volume regardless of the
signal strength.
The second IF transformer (L11/L12)
also uses tapped and tuned primary
and secondary windings. This in turn
feeds the second IF amplifier which is
based on transistor VT3, the set’s third
AF117. This stage operates with fixed
bias for maximum, constant gain.
Note that neither IF transistor uses
neutralisation/unilateralisation to combat the effects of collector-base feed-
back (Miller Effect). This is quite low
in alloy-diffused transistors and gives
no problems at low radio frequencies
(such as 455-470kHz).
The third IF transformer (L13/L14)
has a tuned, tapped primary and an
untuned, untapped secondary. This
secondary winding matches the low
impedance of the detector diode (D1).
Thus far, the design is similar to the
Mk.1 version, except that the Mk.1
used lower-performing alloy-junction
transistor types, ie, OC44/OC45. However, these did require neutralisation
networks for proper operation.
Audio stages
The audio section in the earlier
Mk.1 version was unusual, with the
first two stages being direct-coupled.
It achieved good thermal stability and
dispensed with one coupling capacitor, thereby improving low-frequency
response.
The Mk.1 set also used an unusual
“local” feedback scheme in the ClassB output stage. This involved using
two extra windings on the output
transformer, one for each emitter connection, with the resulting feedback
reducing distortion in the output stage.
September 2013 93
This is the view inside the set with the back cover removed. Note that the transistors are all mounted on this side of the
chassis, with their leads terminated on insulated solder pins, making it easy to desolder and replace them if necessary.
The method of construction used is reminiscent of that used for valve sets.
By contrast, the TR82C Mk.2 circuit ditches the direct-coupled audio
preamplifier stages and reverts to the
more common 2-stage configuration
with resistance-capacitance coupling.
As shown on Fig.1, the detected
signal from D1 is fed to volume control
(RV1) and from there to the first audio
stage (VT4, OC71) via capacitor C19
(8µF). Tone control RV2, connected
to the base of VT4, applies adjustable
“top cut” to this audio signal. This
tone control also integrates the power
switch, unusually switching both positive and negative battery leads.
VT4 in turn drives the second audio
stage (VT5, OC81D). This transistor
operates with a collector current of
about 2mA and only dissipates about
20mW, yet it is mounted in a clip
heatsink. This heatsink would appear
to be unnecessary and may be there
simply to provide a convenient way
of anchoring the transistor in place.
Transistor VT5 drives the primary
of transformer T1 which operates as
a phase splitter. Its centre-tapped secondary then drives a push-pull output
stage based on transistors VT6 & VT7
(both OC81D) and these in turn drive
the centre-tapped primary winding of
speaker transformer T2. T2’s second94 Silicon Chip
ary then drives the speaker or a set of
headphones (or an external speaker)
via a headphone socket.
Unlike the Mk.1 version, the TR82C
Mk.2 includes trimpot RV3 to allow
the output stage bias to be adjusted
to minimise crossover distortion. It
also includes a common emitter resistor (R28) to provide a small amount
of feedback. The Mk.2 version also
provides a feedback loop from the
collector of output transistor VT7 to
the base of driver stage VT5, ie, via
R25. This feedback further reduces
the audio distortion.
The output power is quoted as
325mW at 10% distortion. It’s fairly
modest but enough to provide comfortable listening levels.
Cleaning up
Despite its age, the chassis was still
in good condition and was operating
normally. All that was required was
a quick touch up of the alignment
adjustments. During this process, I
discovered that both aerial trimmers
needed extra capacitance for optimum
performance and I ended up adding a
10pF capacitor in parallel with each
one.
As obtained, the set also needed
a good clean, both inside and out. I
began this process by first dismantling
the case, separating the front and back
covers and the central section.
These two covers were in good
shape physically but were grubby and
had dirt lodged on the bottom lands of
the grilles. A little “elbow-work” with
some widow cleaner and a toothbrush
soon had them sparkling again. The
dial well was also grubby and this
responded to a careful once-over with
a cream cleanser.
In addition, the chrome bands were
covered in some form of hardened
grease/dirt composite. They were
cleaned up with some judicious scrubbing and came up looking like new.
Unfortunately, the cabinet’s blue
middle section was a different matter. It was covered with some kind of
brown, greasy film (probably from a
kitchen) that proved to be quite stubborn to shift. After using a sponge
wetted with a liquid kitchen cleaner to
little effect, I tried using to a microfibre
pad. This had proved to be effective
in the past for removing all kinds of
accumulations, including marker pens
and paint.
It did a reasonable job but the blue
colouring gradually built up on the
siliconchip.com.au
The photo at left shows the appearance of the bright red dial scale after the tuning wheel had been removed, so that the
chassis could be removed from the cabinet. The rear cover is removed by undoing a single screw.
pad, showing that part of the plastic
surface was coming off with the grease.
Rather than wreck it, I didn’t push my
luck too far and left it before I was really happy with it.
Removing The
Tuning Knob
Performance
The audio frequency response from
the volume control to the loudspeaker
is 170Hz to 12kHz but the RF/IF section crops the upper frequency limit
to just 2.2kHz. A measured selectivity
figure of -60dB at ±13kHz confirms this
narrow RF/IF bandwidth, a result of
the two double-tuned IF transformers
The audio performance is quite
good: at 10mW output, the total harmonic distortion (THD) is 1.6% at
1kHz. This falls to 1.1% at 50mW,
implying some crossover distortion or
transistor mismatch at very low levels.
At 250mW, the THD is just 1.7% and
the set easily bettered its 10% quoted
distortion at full output, giving just
under 5% at 325mW.
The sensitivity is outstanding, the
set delivering 50mW output (for a
20dB signal-to-noise ratio) for an RF
input of just 7.5µV at the aerial terminal. Note that this is with the volume
control “backed off” to give the 20dB
noise figure. At full volume, this set
needs just 1.5µV at 600kHz and 2.5µV
at 1400kHz for 50mW.
It’s pretty noisy at full gain though,
with a signal-to-noise ratio of just 3dB
at 600kHz and 8dB at 1400kHz. Its
long-wave performance is quite impressive – a sensitivity of 2.5µV for a
20dB signal-to-noise ratio approaches
the 1µV “gold standard” for valve sets
using RF stages.
The signal strength received by the
ferrite rod antenna needs to be around
100µV/m at 600kHz and 80µV/m at
siliconchip.com.au
amplifier stage (VT2) falling to near
zero at this point. The Mullard design’s additional damping diode and
two resistors (shown in dotted lines
on Fig.1) would reduce (or prevent)
this problem.
Would I buy another?
The TR82C’s tuning knob is a press
fit onto the tuning gang shaft and
Bush’s servicing manual recommends
using a suction cup to pull the knob
off. It also clearly advises against
attempting to apply pressure using
“screwdrivers or other levers’.
Another method is to wrap three
or four lengths of string around the
centre boss to form a simple “puller”
arrangement. The above photo shows
how this is done.
Given the opportunity, would I buy
another one? The answer is “yes”.
In fact, there’s also an FM/MW/LW
version, the Bush VTR103, which I’ve
ordered. It uses the same stylish case
as the TR82C Mk.2 and an equally
impressive circuit design, so I expect
it to also be a good performer. There’s
also a quite rare version that uses a
miniature valve as the converter and
transistors for the IF and audio stages.
Yet another version, the TR82D,
is identical to the TR82C except that
it has a different cabinet colour. My
TR82C has a light olive-green front
and back, with a blue intermediate
band and a cream handle. By contrast,
the TR82D is cream with a tan band.
Resurrection
1400kHz for an output of 50µW and a
20dB signal-to-nose ratio. At full volume, the sensitivity is around 20µV/m.
These measured values closely agree
with the figure in the original 1960
Mullard design paper.
The “noisiness” at full gain justifies the claim that converter noise is
a limiting factor in weak-signal performance for superhets of all kinds.
The AGC control is excellent for
moderate signals, with a 30dB increase
in signal giving just a 6dB increase in
output power. The set does, however,
go into RF/IF overload for signals over
about 20mV/m. This is confirmed by
the collector current for the first IF
Finally, the Bush TR82C’s classic look has just been revived with
the release of the Bush TR82DAB, a
modern DAB+/FM/AM/LW radio. The
TR82DAB’s cabinet looks virtually
identical to the earlier 1960s design
and a full review appears elsewhere
in this issue.
Further Reading
(1) There are several discussion threads
on: http://www.vintage-radio.com/
(2) For a company history and listing
of sets: http://www.bushradio.co.uk/
(3) The original service data is available (along with many other UK sets)
at: http://www.service-data.com/ SC
September 2013 95
|