This is only a preview of the June 2005 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 "The Coolmaster Fridge/Freezer Temperature Controller":
Items relevant to "PICAXE Colour Recognition System":
Items relevant to "PICAXE In Schools, Pt.2":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
Vintage Radio
By RODNEY CHAMPNESS, VK3UG
Signal generators – what they
are and how to fix them
An RF signal generator is a vital when it
comes to servicing and accurately aligning
vintage radio receivers. However, it’s no
good having a generator if it isn’t working
correctly or isn’t calibrated.
A
RADIO FREQUENCY (RF) signal
generator (or modulated oscillator) is an instrument that can act as a
substitute for a radio station. It can be
set to generate any frequency over its
range and the resulting output signal
level can also usually be varied before
being fed to the radio under test.
This allows several things to be
checked and/or adjusted in the receiver, as follows:
(1) the accuracy of the dial calibrations;
(2) the receiver’s sensitivity, along
with its ability to handle both weak
and strong signals;
(4) the amount of frequency drift in
the local oscillator; and
The Advance 62 signal generator is capable of generating frequencies from
150kHz to 300MHz over six ranges. It also features CW and tone modulation
and is just the shot for aligning vintage radio receivers.
96 Silicon Chip
(5) the receiver’s alignment, especially the IF (intermediate frequency)
stages.
In addition, a signal generator can
feed a variable or fixed-level audio
modulating signal into the RF oscillator for checking the performance of the
audio section of the receiver. And of
course, it can provide an audio signal
for directly testing audio amplifier
stages.
Top of the range signal generators
can also be used to perform a number
of other tests on high-performance
receivers. We’ll consider some of these
tests later on.
Typical units
The term “signal generator” should
really only apply to units that have
a high order of frequency accuracy
with rigidly controlled and calibrated
output levels. In addition, it should
be possible to control the output level
down to a fraction of a microvolt at
radio frequencies.
One such unit is the Hewlett Packard 606B signal generator. As with
similar units, its output level is accurately controlled and its frequency
accuracy is set by either an in-built
crystal-controlled marker oscillator
or by an external source. In addition,
its internal filtering and shielding is
such that the only signal likely to be
detected from the generator will be at
the RF output terminal (and this will
only be at a controlled level).
On the other hand, “signal generators” like the Leader LSG10 and
LSG11 should more correctly be called
“modulated oscillators”. A modulated
oscillator has little or no filtering or
shielding to prevent an uncontrolled
level of RF signal from “escaping”
siliconchip.com.au
Above: inside the Advance 62 signal generator with the RF
shield in place. The view at right shows the unit with the
shield removed (ie, by unscrewing it).
from the unit. This signal may completely bypass the attenuator, which
means that the output level is quite
arbitrary.
These and many similar older units
are still useful for testing restored
vintage radios, although they cannot
be used for some precision tests. The
output frequency from these modulated oscillators is usually accurately
calibrated on the dial (making them
useful for checking alignment and dial
calibration, etc) but the output level
cannot be relied upon.
Before passing judgement on any of
these instruments, it must be considered what work each was intended to
do. For example, the Hewlett Packard
606B was designed for testing and
aligning high-performance 2-way radios and was an expensive instrument
when new. It was able to test many
parameters other than RF sensitivity
and frequency stability and an indication of its quality can be gauged by
the fact it was used during the 1960s
and 1970s by the Department of Communications (now ACA) in their Type
Approval Laboratories.
By contrast, signal generators such
as the Leader LSG10 and LSG11 are
much more modest but will still do a
reasonable job for most vintage radio
restorers.
Between these two extremes are
other instruments that will not only
do the job for vintage radio restorers
but will also meet the needs of radio
amateurs. The latter typically require
more accuracy than an LSG11 can
give. Units such as the Advance P1
and 62 signal generators, for example,
have quite good shielding and filtering to minimise signals escaping from
siliconchip.com.au
the case or being radiated down the
power lead.
Fixing an Advance P1
A friend who suffers from Parkinson’s disease is largely confined to a
wheelchair and is unable to hold a
soldering iron with a steady hand. He
had been trying to restore the Advance
P1 signal generator but due to his disability, was not making much progress.
As a result, I agreed to do the job for him
– he would get the unit back in working
order and I could use the P1 as part of
an article on test equipment.
When he obtained the unit, it appeared as though it had been worked
on by its previous owner, the power
supply and audio oscillator section
being the obvious “casualties” of this
attention.
My friend told me that the P1 came
with a 5Y3GT rectifier for the power
supply. However, this didn’t seem
right to me as the power transformer
was too small to provide enough power
for even the 5Y3GT’s filament. Also,
the valve socket was right alongside
what I took to be the audio oscillator
transformer.
I traced the circuit out around the
octal socket and found that it was
connected to the audio transformer.
Lacking a circuit, I had no way of
knowing what valve had originally
been used in the audio oscillator but
experience told me that a 6J5 triode
may suit. I looked up the pin-outs
and found that it would indeed make
a Hartley audio oscillator circuit if a
6J5 was plugged in.
Having worked out what the octal
socket and its adjacent circuitry did, it
was time to look at the power supply.
The 240V AC mains supply came in via
a switch to the primary of a small transformer. I checked all the windings and
found there was a tapped primary and
two untapped secondary windings,
one for the valve heaters and the other
for the high tension (HT) supply.
The mains was supplied by a
3-core lead and someone in the past
had wired the Neutral (black wire) to
the Active terminal of the plug. Due
to this inaccurate wiring, the Active
wire did not have the switch in series
with it – 240V AC was applied to the
power transformer whether the unit
was switched on or off. This was corrected, so that the switch is now in
the Active line.
As part of the RF filtering, two 600V
1nF capacitors were wired between
the Active and Earth and Neutral and
Earth. However, 600V DC-rated capacitors are not at all suitable this job and
may puncture due to voltage spikes on
the mains. They were replaced with a
purpose-made suppression block rated
to work at 250V AC.
Next, I removed the shield from the
June 2005 97
The Leader LSG11 is a low-cost generator but is still useful for aligning vintage
radio receivers. It produces frequencies ranging from 120kHz to 130MHz.
the unearthed end of the “HT” winding
was attached to an adjacent lug on the
terminal strip. I fitted a 1N4004 diode
between these two lugs and then tried
the power supply again. A DC voltage
of about 77V was the result, which
seemed to me to be in the ballpark.
I turned it off, refitted the 6J6 and
the dial lamp, and then turned it on
again. The generator warmed up and
its signal could be heard in a portable
receiver close by. I then decided to fit
a 6J5 into the audio section and this
also proved successful, with a tonemodulated signal now being heard in
the receiver.
This job turned out to be easier than
I had expected, especially in view of
the previous owner’s modifications. It
was just luck that his “improvements”
hadn’t caused damage to other sections
of the circuit.
With the unit now working, the
next step was to spruce up the unit’s
appearance. The cabinet was the
obvious place to start, since its paint
was in poor condition. It was sanded
down and given several coats of matt
black spray paint (although I suspect
that the original cabinet colour was a
hammertone grey).
The front panel was more of a
problem and I achieved only partial
success by filling in the spots where
the black paint was missing. However,
some bare spots were too close to the
lettering and I couldn’t afford to damage that.
The accompanying photo shows the
final result. It looks a lot better than
before and it’s now a fully-functioing
unit.
Advance 62 signal generator
This is the view inside the Leader LSG11 signal generator. Note the switched
coils and the tuning capacitor.
RF oscillator section and this revealed
a 6J6 valve. This valve was removed,
along with the dial lamp, and power
applied to the unit. The filament voltage was up around 7.5V and the winding which I thought may have been for
the HT produced just 55V AC.
There was absolutely no way that
a valve rectifier could have ever been
98 Silicon Chip
fitted to this unit. Instead, a solid-state
rectifier had obviously been fitted
originally – perhaps a small selenium
unit.
Closer inspection showed that a
wire went from a terminal strip located on the top of the transformer to
an electrolytic capacitor in the power
supply filter network. Furthermore,
Another friend loaned me his Advance 62 signal generator (a later
model) so that I could use it to work
out what had been altered in the P1. I
had expected them to use similar audio
oscillators but found that there was a
significant difference between the two
circuits. The power supply sections
are almost the same, though.
Although I did work out most of
the altered circuit section of the P1,
the 62 provided a handy way to check
whether my basic ideas were accurate
or not. And the differences? – the
P1 uses a 6J6 (both sections) as the
RF oscillator and a 6J5 as the audio
oscillator/modulator. By contrast, the
62 uses one half of a 12AT7 as the RF
oscillator and the other half as the
siliconchip.com.au
audio oscillator/modulator.
The P1 and the 62 are also quite
different in the amount of filtering
and shielding that they use. The P1
has a bolt-on cover over the oscillator
which lets some signal leak out along
the edges, while the 62 uses a large
aluminium can with a screw type lid.
This forms a cheap but very effective
shield to prevent radiation.
Filtering of the AC and DC leads
is achieved using series RF chokes,
feed-through capacitors and parallel
polyester capacitors. The RF coils
(chokes) are spread throughout both
units on reddish-brown coil formers.
The switched signal attenuator in
the P1 uses several resistors within
“view” of each other. This allows the
VHF signals to partially bypass the
attenuator. By contrast, the 62 has a
much better shielded attenuator that
isolates each section of the attenuator
from the next. Both of these attenuators
can be seen with their rear shields
removed in the photographs.
In operation, the P1 has a frequency
range of 100kHz to 100MHz while
the 62 covers 150kHz to 300MHz.
That’s one reason for the better filtering and shielding in the 62; it has a
much higher maximum frequency of
operation.
Both the P1 and the 62 proved to be
quite stable in frequency when warming up. The case radiation from the P1
is quite noticeable, although nothing
like the LSG11’s case radiation. The
62 is better still – on the broadcast
band, its signal is barely audible on
a sensitive AWA transistor portable
placed one metre away.
The Leader LSG11
The LSG11 is one of the cheaper
devices but it still generates signals
that can be used for aligning domestic
vintage radios. Fig.1 shows the circuit
details.
The RF oscillator consists of one half
of a 12BH7 (V2) which then feeds the
second half of the valve. The audio oscillator is a 6AR5 in a Colpitts oscillator which also feeds into the grid of the
second section of the 12BH7, where
the audio and RF signals are mixed to
give a modulated RF signal. The 6AR5
can also be used as a crystal oscillator which can be useful for aligning
equipment on spot frequencies or for
testing FT243 crystals.
The RF range is from 120kHz to
130MHz, although harmonics can
siliconchip.com.au
Fig.1: the Leader LSG11 circuit uses just two valves. V1 (6AR5) functions
as a Colpitts oscillator and this feeds the grid of one section of V2 (12BH7)
where it is mixed with RF signals produced by the second section.
extend this to 390MHz. However,
the stability of the oscillator is insufficient to make it worthwhile using it
on harmonics.
The RF output level isn’t controlled
in any way and it varies significantly
across each band and between the
various bands. An automatic gain
control (AGC) system of some sort
would maintain the output at a constant level but that would have added
to the cost – and this is a cheap unit.
It is different to the Advance units in
that it can put out 400Hz or 1000Hz
audio signals.
The oscillator frequency stability
June 2005 99
unit a lot more usable although it was
still not of a high standard.
As shown in the photos, the oscillator sections in the Advance units
are more compact than in the LSG11.
This means shorter lead lengths in
the Advance units and they are also
better shielded.
Hewlett Packard 606B
This is my friend’s Advance P1 signal generator before restoration. It proved
easy to get going again, despite a previous owner’s strange “modifications”.
and the frequency calibration of the
dial are remarkably good for such a
low-cost unit. However, signal radiation out of the case and across the
attenuator is quite high. This makes
comparative sensitivity checks on different receivers virtually impossible.
That said, if you don’t want to spend
much money and you can put up with
the inadequacies of the LSG11 and its
smaller brother the LSG10, they are
well-worth having.
A look inside the LSG11 reveals
just how little filtering there is, which
explains why its signal leakage levels
are so high. Back in the 1970s, a couple
of radio amateurs decided to “transistorise” their LSG11s. This reduced
the level of radiation from the device,
as there was no longer a power lead
extending from the cabinet to carry
RF signals (the modified units were
fitted a small battery pack inside the
cabinet). This modification made the
This is the view inside the Advance P1 with its RF shield removed. Be sure to
follow the existing lead dress when replacing parts in tuned circuits, otherwise
the calibration will no longer be accurate.
100 Silicon Chip
The 606B well and truly deserves
to be called a signal generator. Indeed,
this was the “Rolls Royce” of signal
generators back in the 1960s. It is big
(510 L x 350 D x 315mm H) and heavy,
with a large handle at either end, and
in its hey-day was very expensive. Today, they sell for as little as $50. That’s
cheap but keep in mind that some of
the parts are now hard to come by if
something goes wrong.
The 606B covers the RF range from
50kHz to 65MHz and the output signal level can be accurately set from a
fraction of a microvolt to several volts.
The modulation level is variable from
0-100% using either a 400Hz or 1kHz
audio frequency from its internal audio oscillator or up to around 20kHz
from an external source.
One interesting feature is that its
accuracy can be checked and adjusted
using an internal crystal reference
oscillator.
Servicing a signal generator
In many ways, a signal generator
is much easier to service than a radio
receiver. This particularly applies to
the smaller, more modest units as the
circuitry is usually fairly simple. A
quick look at the circuit of the Leader
LSG11 (Fig.1), for example, shows that
there are only two valves and a modest
number of passive components.
Because signal generators are usually housed in a metal case, little dust
gets inside to cause problems. And it’s
easy to replace the valves or any other
component – such as a capacitor or
resistor – that’s not involved with the
RF tuned circuit.
Basically, a generator can be broken
down into two sections: (1) the RF
oscillator; and (2) the audio oscillator/
modulator. To check if the audio oscillator is working, just feed its output to
an audio amplifier – a tone should be
heard in the loudspeaker. You should
be able to vary the level by adjusting
the output level control. If nothing is
heard, check the audio oscillator section for faults.
siliconchip.com.au
The RF section can be tested using a
radio receiver. The first step is to tune
the receiver to a vacant spot on the
broadcast band. That done, you simply
connect the generator’s output to the
aerial and earth terminals and tune the
generator across the broadcast band
until a decrease in the background
noise is observed. This occurs when
the generator is set to the frequency
tuned by the radio receiver.
If the generator is now set for a
modulated output, you should hear the
tone from the receiver’s loudspeaker.
If nothing is observed, check the
circuit around the RF oscillator. As
with the local oscillator in a receiver,
a properly working stage will generate
grid current. This can be checked by
lifting the earth of the grid resistor (R11
in the case of the LSG11) and installing a multimeter (set to milliamps) in
series between the resistor and earth.
If the stage is working, you should
get a reading of a few hundred microamps. If not check around this stage.
If neither the audio or RF sections
appear to be working, check that the
heater and high tension (HT) voltages
are present and are reaching all relevant parts of the circuit.
It is quite OK to check the continuity
of the RF coils (with the unit off) but
do not try adjusting them as this will
throw the calibration out. The LSG11
has no adjustments but the Advance
units have adjustable iron dust cores
in the coils and wire type trimmers
across each coil. Unless you have a
very accurate receiver or a frequency
counter to check the tuning of the coils,
leave them alone.
When replacing parts around the
tuned circuits, make sure they are
dressed exactly the same way as the
original parts as this can affect calibration. On the other hand, the layout is
not quite so important in the audio
and power sections.
Finally, if there is any corrosion on
the case or metal shields, carefully
clean all mating surfaces to ensure that
the shielding functions correctly.
Summary
With signal generators, it’s very
much a case of “you get what you
pay for”. The Leader LSG11 is a basic
signal generator which will meet the
requirements of most vintage radio
restorers. It does have some inadequacies (such as variable output levels
across each band) but for simple alignsiliconchip.com.au
The Hewlett Packard 606B was considered to be the “Rolls Royce” of
signal generators during the 1960s. Grab one of these if you can.
ment work, it’s shortcomings aren’t
really a problem.
The Advance P1 and more so the
Advance 62 provide better performance and will be appreciated by
restorers who want to get the best out
of multi-band receivers. If you can
get one of these at a reasonable price,
grab it.
The Hewlett Packard 606B is the
unit for those who want the best at a
low price (secondhand, that is). I use
mine almost exclusively for vintage
SC
radio work.
Photo Gallery: Astor Aladdin Model FH
Manufactured by Radio Corporation, Melbourne, in 1938, the Astor Aladdin
FH was designed for use in areas without mains power. The set used 2V
battery valves and was fitted with a vibrator power supply running from a 6V
accumulator. A distinctive feature was its “Presto” tuning (with telephonestyle dial), which allowed for quick selection of preset stations. The valve
line-up was as follows: IC7-G frequency changer; IF7-G first IF amplifier/
AVC rectifier; IF7-G second IF amplifier/audio amplifier/detector; and IF5-G
audio output. Photo: Historical Radio Society of Australia, Inc.
June 2005 101
|