This is only a preview of the March 2024 issue of Silicon Chip. You can view 39 of the 112 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. Articles in this series:
Items relevant to "Laser Communicator":
Items relevant to "Pico Digital Video Terminal":
Items relevant to "‘Nunchuk’ RGB Light Driver":
Items relevant to "Mains Power-Up Sequencer, Pt2":
Items relevant to "Arduino for Arduinians":
Purchase a printed copy of this issue for $12.50. |
Vintage Radio
The Bush MB60 portable radio
By Ian Batty
We have previously described two Bush transistor radios:
the early (1957) TR82C in the September 2013 issue and the
VTR103 (1961) in August 2021. The MB60, also released in 1957,
is the first valve-based Bush radio to grace these pages.
T
he Graham Amplion Company,
founded in 1894, was well-known
for loudspeakers from the early 1920s
until their closure in 1932. The Bush
radio company took over the remains
of Amplion in 1932, deriving their
name from their Shepherd’s Bush
(London) facility.
Initially trading as a subsidiary of
the Gaumont British Picture Corporation, Bush became a subsidiary of the
Rank Organisation. Bush was a major
manufacturer of radios and merged
with Murphy Radio in 1962.
While their corporate history has
been a roller-coaster, their products
were among the best from England.
Bush launched their popular DAC90A
and DAC10 radios in 1950, followed
by their distinctive TV22 television.
David Ogle (MBE DSC) was a British industrial and car designer who
founded Ogle Design in 1954. After
the war, he studied industrial design
at the Central School of Art and
Design in London. He subsequently
92
Silicon Chip
joined Murphy Radio, leaving Murphy in 1948 to join Bush Radio. While
at Bush, he was responsible for the
iconic design of the MB60 portable
radio. The MB60 set a benchmark for
style, well-matched by performance
and sound quality.
Valve lineup
The MB60 uses Dx96-series directly
heated valves. Released in 1940/41,
the RMA/RETMA 1R5, 1T4, 1S5/1U5
and 1S4/3S4/3V4 series established
the all-glass design that would continue almost until the end of receiving valve evolution, followed only
by the short-lived Nuvistor and all-
ceramic types.
The initial release featured 1.4V filaments drawing 50mA (100mA for the
1S4/3S4/3V4 output pentodes). These
appeared in the Mullard-Philips system as Dx91~93 releases. At 50mA per
valve, a four-valve portable set would
demand 250mA from the A battery.
A compact set using a single ‘A’ cell
Australia's electronics magazine
would get less than ten hours of filament battery life.
The Dx96 series halved the filament
current consumption while giving
near-identical performance, making
portables more practical.
The DK96 pentagrid converter differs from the familiar DK91/1R5. It’s
the classic pentagrid, providing a committed oscillator anode. By comparison, the 1R5 inherited the dual screengrid design from the octal 6SA7.
The DF96 pentode, DAF96
diode-pentode and DL96 power pentode use the same electrode structures as their predecessors, the 1T4,
1S5/1U5 and 1S4/3S4/3V4.
Circuit details
My redrawn (and hopefully clarified) circuit is shown in Fig.1. I am
using the Bush’s own service manual
circuit as my reference, as the Wireless and Electrical Trader 1403 version
is impractical. I have preserved the
component numbering but their strict
siliconchip.com.au
Fig.1: the circuit diagram for the Bush MB60. Note the extra IF
amplifier (DF96, V3), making this set very sensitive.
first-to-last numbering order has been
upset by my aim of making the circuit
more understandable.
The MB60’s dual-band design (long
wave and medium wave/broadcast
band) is accommodated by a ferrite
rod antenna with two windings, and
an oscillator coil with just one. This
design was reused in the follow-on
TR82 that was mentioned in the introduction.
The circuit parallels the ferrite rod’s
two tuned windings for medium-
wave reception. This gives a lower
inductance than either winding by
itself, allowing the antenna section
of the tuning gang (VC1) to tune over
526~1605kHz for the medium wave/
broadcast band.
Bush advises against adjusting the
antenna coils for low-end alignment,
so this is done by adjusting the oscillator coil for maximum sensitivity at
600kHz. Top-end alignment is performed using trimmer TC3.
Revised antenna coupling
The initial release’s antenna input/
car radio socket connects to the top of
the tuned circuit via a 5.6pF capacitor. As noted below, this is not highly
effective, and can put the antenna circuit off-resonance. The second issue
of the MB60 uses the accepted design
of a dedicated primary winding, as
shown in Fig.2.
The converter operates with zero
bias and is gain-controlled from the
AGC circuit via grid resistor R2. The
oscillator section uses a secondary-
tuned Armstrong circuit formed by
transformer L4/L3. As the DK96 is
a 6A8-style pentagrid, its oscillator
anode (pin 3) is supplied from HT via
resistor R5. Feedback is coupled to L4
via capacitor C13, while L4 couples
inductively to the local oscillator (LO)
coil’s tuned primary, L3.
515pF padder C11 ensures tracking
between the antenna and oscillator circuits for medium-wave reception. L4/
L3 is slug-tuned to allow adjustment
at the bottom end of the medium wave
band. Trimmer TC4 provides top-end
alignment, while the LO is tuned by
the oscillator section of the gang, VC2.
For long-wave tuning, the antenna
circuit uses only the L1 winding on the
ferrite rod, with L2 switched out of circuit. L1 alone, tuned by tuning capacitor VC1, now shunted by capacitor C3
(160pF) and the two trimcaps (TC1/
TC2), restricts the antenna circuit’s
siliconchip.com.au
Australia's electronics magazine
March 2024 93
Fig.2: the dedicated primary
winding of the revised MB60.
tuning range to only 158~280kHz.
A local oscillator’s tuning inductance is usually changed for different
bands by switching in a different coil
set, as changing tappings on one coil
would modify the feedback ratio and
affect the converter’s injection voltage.
This is undesirable, as pentagrids must
have a defined minimum injection
voltage for optimal conversion gain.
Instead, the MB60 switches extra
capacitances into the circuit. C9
(450pF) is connected across tuned secondary winding L3, restricting the LO
tuning range to around 630~750kHz.
As C9 has a fixed value, low-end alignment and correct tracking rely on the
adjustment of L4/L3’s ferrite core,
which was set during the medium
wave alignment.
The LO’s top-end frequency is
restricted by 33pF capacitor C10 and
adjusted by trimcaps TC5/TC6.
The converter drives the first intermediate frequency (IF) transformer
IFT1’s primary. As with the other
two IF transformers (IFTs), it has an
untapped, inductance-tuned primary
and secondary.
The first IF amplifier operates with
zero bias, with gain control via the first
IFT primary. The second IF amplifier
is similar, driving the third IF transformer, IFT3. Both stages get their
screen supply via 33kW resistor R7,
bypassed by 40nF capacitor C15.
The secondary of IFT3 drives the
DAF96’s demodulator/AGC diode.
Demodulated audio develops across
500kW volume pot VR1, with the IF
signal filtered out by 68pF capacitor
C18 and 27kW resistor R10.
The automatic gain control (AGC)
signal is picked off via 2.7MW resistor R9, with filtering and voltage division by 40nF capacitor C16 and 2.7MW
resistor R8. All controlled stages are
fed with the same AGC voltage.
The audio signal is conveyed to
the first audio pentode section of the
DAF96. This operates with low screen
and anode voltages, as is common. The
low anode current – which reduces the
valve’s transconductance and thus its
voltage gain – is compensated for by
the high value of the 1MW anode load
resistor, R13.
The valve gets contact potential bias
due to the action of 10MW grid resistor
R12, allowing the grid to ‘drift’ weakly
negative.
The amplified audio signal is fed to
the DL96 output valve’s grid via 3nF
capacitor C24 and 330kW grid stopper
R17. The DL96 gets about -5V bias via
1.8MW grid resistor R16 from the backbias developed across 560W resistor
R18, filtered by 50μF capacitor C26.
The DL96 valve drives the speaker
via output transformer T1. The output transformer’s natural resonance is
damped by 3nF capacitor C28. This is
shunted by the tone control network
of 10nF capacitor C27 and 100kW tone
potentiometer VR2.
Audio feedback is picked off from
the loudspeaker and returned to the
bottom end of 500kW volume control
potentiometer VR1 via 10kW resistor
R15, 40nF capacitor C23, 100nF capacitor C22 and 1kW resistor R11.
My set is powered by a combined
1.5V/90V battery pack or from the
mains. The later issue used two parallel D cells for the LT supply and a
separate 90V B battery for HT. Mains
transformer T2, with a multi-tapped
primary, supplies full-wave rectifier MR2a/MR2b. After filtering by
two-section pi low-pass filter C35/
R21/C34/R20/C33, it delivers about
1.35V to the filaments.
The filament voltage from the mains
supply is stabilised by shunt regulator MR2c.
As the HT supply needs to deliver a
lot less current, it is half-wave rectified
by MR1 and filtered by C32/R19/C31.
Mains/battery switching, via switch
poles S2a/b/c/d, is performed by
inserting or removing the mains plug.
A quick glance had me puzzled. Was
part of the battery HT+ wiring really
going via the mains transformer primary’s wiring?
Sure enough, it does, but only when
the power plug is removed and S2
changes over to the battery position.
This unusual connection effectively
turns the set off via the On/Off switch
in volume pot VR1: it cuts the mains
input when on AC power and the HT
supply when on battery. For battery
operation, the LT supply is switched
by S3a.
Cleaning it up
I got this set from a fellow HRSA
member, happy to close the loop on
Above: the controls for the Bush MB60 are located
on the top of the cabinet.
Right: a close-up showing the underside of the IF
transformers with the added ceramic capacitors
circled. Their values are in Table 1 shown opposite.
94
Silicon Chip
Australia's electronics magazine
siliconchip.com.au
this line of distinctively designed
English radios. It had been made to
work, then smashed in transit. My
friend and I divided the job – he would
repair the case, and I would do the
electronics.
On receipt, it was working, but I
reckoned it was a bit ‘deaf’ for a set
with two IF stages. I recalled the Astor
Aladdin (described in August 2016;
siliconchip.au/Article/10049), which
used a similar lineup. That set had
only four valves but used two in the
IF strip and employed one as a reflex
stage for the first audio amplifier.
Given the improvements in valve
and component design, the MB60
should have been at least as good. The
audio checked out OK, so it was on to
the RF/IF section.
All the IF transformer slugs were
coated with white paint. A bit of gentle
heating showed that I wouldn’t be able
to soften it and free the slugs, a trick
I had used on the Astor APN. What
to do? At the converter grid, the IF
responded best at 472kHz. So why did
I get the best performance at 467kHz
on the first IF grid and at 478kHz on
the second IF?
The bandwidth was wide enough
to drive a truck through, confirming
that, whatever the true intermediate frequency should have been, the
various IF-tuned circuits disagreed.
Also, it needed 20μV at the converter
grid for 50mW of output, much worse
than I expected.
Believing that the manufacturer’s specification of 470kHz could be
fiddled with a bit, I got a handful of
1~10pF trimmers, popped one across
each tuned winding, and adjusted
them for maximum gain. The final
intermediate frequency of about
460kHz was lower than the specification, but the gain came up pretty well
– see Table 1.
A bit too well, in fact. I had been
ready for IF oscillation with the trimcaps bodies hanging out of the circuit
wiring, but expected that the feedback would be absent once I popped
in small, fixed ceramics. It was stable
but still a bit ‘chirpy’, so I dropped a
470kW resistor across the second IF
primary.
That did reduce the sensitivity at the
converter grid from 6μV to 12μV, but
the improved stability was preferable
to instability.
I then checked the antenna/LO
alignment and found that the set working about as I expected.
Having lived on a farm for around
fifteen years, I reckon I know ‘agricultural’ when I see it. The LO coil looks
like the designers forgot it, then just
Table 1 – added capacitors
IFT #
Primary
1
Secondary
10pF
2
4.7pF
5.6pF
3
12pF
8.2pF
threw it down, bolted it in place and
told the assemblers to finish the set.
Performance
It is very good; more than just a
standout example of 1960s design.
For the standard 50mW output,
it needed 60μV/m at 600kHz and
32μV/m at 1400kHz with signal+noise
to noise (S+N:N) figures of 10dB and
11dB. For the standard 20dB S+N:N,
the field strengths were 200μV/m and
150μV/m. Bandwidth for -3dB was
under ±1kHz, implying some residual regeneration in the IF section. For
-60dB, it was ±22kHz.
Audio bandwidth, from volume
control to speaker for -3dB was 140Hz
to 10kHz, antenna to speaker about
130~1200Hz. Turning the Tone control
to full cut brought the top end down
to around 1kHz.
The AGC was effective, needing a
+40dB rise of input to give a +6dB
increase in output. It would not overload even at 200mV/m field strength.
The set went into clipping at 80mW,
The front view of the Bush MB60 chassis which shows the ferrite rod antenna, permanent-magnet loudspeaker and
controls. Nearly all the discrete components are mounted on this side.
siliconchip.com.au
Australia's electronics magazine
March 2024 95
Tone
Volume
Bandchange
A labelled photograph
of the rear side of
the chassis. In the
service manual, they
recommend an Ever
Battery/Mains
Ready type B147
Switch
battery.
Antenna socket
1st IF
1st IFT
Oscillator coil
Converter
Output
Transformer
HT Rectifier
2nd IFT
LT Rectifier
2nd IF
3rd IFT
HT Filtering
LT Filtering
Demod/1st Audio
Audio Out
with 10% total harmonic distortion
(THD). At 50mW, the THD was 7%,
and 3% at 10mW out.
Versions
As noted above, the first release used
capacitive coupling from the external
antenna socket, while the follow-on
used the conventional primary winding on the ferrite rod.
The MB60 seems to have been
released in just one colour scheme: a
Mains Transformer
grey case with a red perimeter band.
There’s a moulded depression at
the lower right of the rear cover in all
three models.
The VTR103 used it for the Tape
Recorder output connector, but it was
blank in the TR82. It was originally
placed for the MB60’s mains connector plug. Mystery solved!
Special handling
Like the follow-on TR82 and
This is a portable set running from 90/1.5V; you can see the battery plug and
lead lying in the bottom of the cabinet. The cabinet was designed by David Ogle,
who also designed the Ogle SX1000 car.
96
Silicon Chip
Australia's electronics magazine
Mains Socket
VTR103, the tuning knob is a push fit.
See the TR82 article (September 2013;
siliconchip.au/Article/4404) for
advice on safe removal. That said, I
found finger pressure was adequate
to withdraw the knob.
Radiomuseum offers two online
schematics (siliconchip.au/link/abrc).
The Wireless & Electrical Trader 1403
version (like for the TR82 and VTR103)
is difficult to understand: all switches
are broken out into individual make/
break contacts. That demands that
you get out a pencil and try to work
out what is on (or off) for each band
according to the description near the
end of the article.
It also takes some work to realise
that mains voltage cannot connect
through to the HT+ line. Pity the poor
service technician.
The other schematic, titled “Radio
Servicing” is an extract from the Bush
Radio Service Instructions MB60. This
circuit is an improvement, except for
the confusing power supply wiring.
The manual contains extensive details
and modification notes.
I recommend the complete Bush
original, which you can download
from ElektroTanya (siliconchip.au/
link/abrb).
They provide free original manuals,
many from European equipment not
hosted elsewhere. If you do visit them,
consider uploading material they don’t
have, or maybe just a donation.
Radios like these come up on eBay,
but you’ll also find them at auctions
run by the Historical Radio Society of
Australia (HRSA).
SC
siliconchip.com.au
|