Ever Ready Model B2
|
|
This job is a little different
to the usual type of repair, being to sort out the IF amplifier
stages of a battery operated portable radio chassis.
The pictures below show the
chassis as it was received (together with a set of various replacements
to be used as necessary)
The Ever Ready B2 is not too
disimilar to a recent addition to
my collection, a Model N.
|
 |
 |
|
|
|
As one of the problems
was reported to be dodgy IF transformers, I measured across the
pins of both (L3/L4 & L8/L9) and found circa 10-13 ohms for
each of the four windings so I guess they may be OK.
I removed L3/L4 for testing
and found it was a replacement marked "Wearite" and
with a coverage of 456Kc/s to 475Kc/s so might need a small capacitor
across each winding to allow it to comfortably tune to 455KHz.
Opening the case I was able to note the markings on the tuning
capacitors.. these were both marked 65pF +/-3%.
This means that the coil must
tune from 1874uH to 1727uH to match the marked coverage.
Ignoring worst case analysis
and stray capacitances etc, adding only 3.3pF lowers the tuning
range to 444Kc/s/463Kc/s thus neatly meeting our 455Kc/s requirement.
With the transformer in place
the lower coil will be tuneable using a thin plastic rod with
a flat end slid between the components. The original lower tuning
slug used a metal screw and this secured a solder tag carrying
a couple of wires. This was unsecured as the Wearite transformer
doesn't have a metal fitting so I'll tidy this later. I noted
that three of the wires to the two transformers were dry jointed.
Below are copies of the pages
from Volume I of Radio & Television Servicing
|
|
 |
|
 |
 |
|
Maybe this isn't the usual
method of testing an IF transformer, but it looks like it might
work.
As with most laboratory RF equipment
there are certain things to bear in mind. In this case the output
impedance of the tracking generator (TG) and the analyser input
(IN) have an impedance of 50 ohms. To cater for this, to make
the test as close as practical to the radio circuit, I'm using
a pair of capacitors to provide a rough and ready match to the
primary and secondary windings. As a guide C1 and C2 might be
33pF which has an impedance of about 10Kohms at 455KHz. I'll
try this value and change it as necessary. The aim is to check
the quality of the transformer and ensure it covers 455KHz at
roughly mid point of the dust cores.
Apart from having trouble with
a bad coax lead the test went very well and was quite illuminating
as it revealed a weakness regarding IF transformers using adjustable
dust cores.
It's pretty obvious when you
think about it but each of the two cores will affect not just
the coil being tuned but also the second coil to a lesser extent.
This will be more pronounced if the core being adjusted has been
pushed beyond the centre of the first coil so it's approaching
the second coil.
I also discovered that it isn't
possible to properly test a transformer outside its screening
can because the tuning range was tens of KHz higher. It's also
important to ground the screening can during the test.
Below is the result of the test
on the Wearite M850 IF transformer tuned to 455KHz.
|
 |
|
 |
To get this result I used a
couple of 33pF coupling capacitors (C1 & C2) and set the
Rigol DSA815's tracking generator to 0dbm.
Although the transformer is marked 456KHz to 475KHz it easily
tuned down to 455KHz.
Each vertical division is 20KHz
wide so the peak is adjacent to the 450KHz line.
As I mentioned above both cores
shifted the tuned frequency so requiring a series of retunings
to simultaneously get the strongest peak at pecisely 455KHz.
Of course the tuning will need
to be carried out again once the transformer has been refitted
into the radio but the test proved it was a good transformer..
Next I'll test the other Wearite
IF transformer.
|
|
|
Thinking a little more
about the observed results.. first the effect of removing the
aluminium can. This had the effect of shifting the resonant frequency
of the transformer coils upwards in frequency. This can be explained,
at least partly, because when one has two coils relatively close
together each will borrow some inductance from the other so,
let's say L1 sees an extra 10% from its mate. This means that
removing the can must have reduced the coupling between the coils
so that our extra 10% with the can in place has lessened, lowering
the inductance and raising the resonant frequency.
Turning to the effect of the
dust cores. Bringing the cores into the area bridging the two
coils will increase their coupling and increase the inductance
of both. This will decrease their resonant frequency. This effect
is enhanced because the approaching dust cores of the two coils
on their mates will naturally increase their inductances and
lower their resonant frequencies even more.
This experiment incidentally
was not to tune the IF transformers to fully match the Ever Ready
radio's requirement, but to check that they're OK (they were
suspected of being faulty). Once fitted, the tuning will have
changed because of different stray capacitances. Currently those
30pF capacitors C1 & C2 will probably shift the transformer
tuning downwards and a coil or two may need the extra few pF
to lower the resonance down to 455KHz (at least with the Wearite
coil whose tuning range falls to 456KHz). The final step once
the transformers are refitted will be to set them up to provide
a decent bandwidth centred on 455KHz because we're not listening
to narrow band morse code but to AM broadcasts.
Why is it important to set the
coils to 455KHz? Well, with a typical radio the dial will be
marked with frequencies (or to be precise here, in wavelegths)
and (being a superhet) the local oscillator will have been carefully
matched to these markings. The designers would have arranged
that the overall performance of reception would be flat across
the whole tuning range, and altering the IF and changing the
local oscillator setting would make a mess of the radio's performance.
One of the clues to the careful
design of a radio such as this example is the value of the padder
condensers. Look at the circuit diagram above and you'll note
that the oscillator padders, C4/C5/C6 have odd values. C4 for
medium wave tuning is 430pF and for long waves C5 and C6 are
switched in, being 140pF and 375pF.
|
|
 |
This is the second of
the two Wearite IF transformers but to which I added 3.3pF across
each winding and again tuned to 455KHz. The top of the curve
is more flattened and about right for AM reception.
Compared with the response of
the first Wearite example I used 10dB less tracking generator
signal (-10dBm instead of 0dBm). The frequency span for these
traces is 100KHz rather than the 200KHz used for the first trace
above.
|
|
 |
Two other IF transformers
were supplied. These look identical to each other and neither
was marked. The coil DC resistances for this example were 16.5
ohms and 24.5 ohms.
One winding tuned OK to 455KHz
but the core for the second winding had no effect whatsoever.
|
|
 |
This is the response of
the second unmarked transformer. Both windings measured 16.5
ohms and tuned normally. |
|
 |
This is the response of
the same transformer. Both windings tuned normally but the input
and output were reversed with no real change in the curve. I
guess this means that the coils are identical. |
|
|
And here's the explanation
of why the radio failed to work. This is the test result for
the second IF transformer fitted to the set. Apart from the metal
adjuster at the top of the IF being seized and immovable, the
bottom adjuster managed to achieve this frequency of about 740KHz
as its minimum instead of 455KHz or less.
The problem is either the IF
transformer is designed for a much higher frequency or it should
be fitted with external tuning capacitors.
Another possibility is that
it has a faulty winding because the DC resistances of the windings
measured 9 and 16 ohms.
I'd need to detach the screening
can to see if there's an explanation.
|
 |
|
|
I was also supplied with
a transformer missing its can. This was the same type as the
last one tested above (with metal adjusters) but had a 16 ohm
winding and a 28 ohm winding. All the samples above use Litz
wire which has something like 10 strands of enamelled wire wound
to form parallel insulated connections and if several of these
went open circuit either by fusing or from poor soldering or
even a simple fracture the DC resistance of the winding would
rise. For example... if ten strands equalled 16 ohms then each
strand would contribute 160 ohms. If five strands are open circuit
the measured resistance of the remaing five would be 32 ohms.
So in our case a resistance of 28 ohms reflects three open circuit
strands. Why use Litz wire? Well, DC current uses all of the
copper in the wire but in the case of AC, nearly all the current
is conducted in the outer surface of the copper with hardly any
using the body of copper. So in order to increase the conducting
surface to AC lots of insulated copper wires are used in parallel.
The much reduced AC resistance makes a coil work better at radio
frequencies. Losses are less and tuning is sharper. |
|
 |
|
The B2 chassis fitted
with two transformers that tested OK and after a little rewiring.
Testing will need a substitute aerial coil as this is in the
lid of the radio which I don't have. The chassis is said to require
a 1.5 volt (250mA) supply for the valve filaments and an HT supply
of around 70 volts (max 8mA).
With its 455KHz IF the local
oscillator on medium waves tunes from 1055KHz (600m) to 1955KHz
(200m) and around 655KHz (1500m) for long waves. Alignment uses
trimmer capacitors TC1 & TC2 (MW) and the cores in L5 (MW)
and L2 (LW). As the layout of coils is less than ideal the alignment
is a bit vague with a final adjustment made at 1300KHz = 230m
(TC1). This would have roughly matched the aerial for 208m, Radio
Luxembourg and 247m, BBC Light Programme.
The IF transformers will be
finally adjusted using a spectrum analyser to set the best IF
bandwidth.
|
|
 |
|
This is the testbed for
the chassis. I'm using my home made universal
power supply designed for battery operated sets. The HT current
was 10mA at about 75 volts and the LT current at 1.47 volts was
250mA.
My TinySA provides the necessary
signals starting at 455KHz and I found to my surprise that without
the connections I used to test the IF transformers they tuned
to about 520KHz but did tune to the required 455KHz without a
problem other than the core at the base of the first IF transformer.
To make matters easier I removed the upper core and tuned the
lower one before refitting the top core and tuning that. It's
possible that a small capacitor might be needed to get the tuning
perfect but that can be added later. The chassis seems to work
properly as both wavebands tuned but the missing aerial coil
(L1) for medium waves needs to be emulated later. If the tuning
capacitor and trimmer is 500pF the coil needs to be 140uH to
tune to 600KHz.
As expected, the IF tuning was
quite critical and needed the cores to be well inside the coils.
Also, having secured the two parts of the chassis together (they'd
been unscrewed for coil access) it was virtually impossible to
get to the lower core in the second IF transformer. Two things
are needed.. one is to add say 10pF across each of the four coils
and the second to add grounding wires between the two parts of
the chassis to ensure good stability once the securing screws
are removed.
In the event I didn't need to
add extra tuning capacitance as the circuit strays took care
of that. Below is the final test setup. I scanned the receiver
at the frame aerial connection (that blue coil which I didn't
attempt to resonate is in place of the frame winding) with a
signal centred on 455KHz and monitored the detector anode at
the DAF91 via my high impedance test probe and a series connected
15pF capacitor and 390kohm resistor (roughly 415kohm) to reduce
damping of the circuit.
|
|
 |
 |
This is the scan before
making any adjustments. The odd shape is a little puzzling but
is due to the receiver picking up some of the scan in its medium
wave circuits. This was resolved by adding a 20dB attenuator
in the tracking generator output. Initially I'd thought one of
the IF tuned circuits was out of tune but none of the tuning
cores altered the position of the anomalous hump. |
 |
Here's a decent curve
after tuning to 455KHz and reducing the tracking generator output
by 20dB to -40dBm. I guess the 10dB change in amplitude is partly
due to the AGC action in the receiver.
After completing the IF tuning
I set the DSA815 to zero span and checked the response of the
medium waveband. This was OK, readily tuning 1.5MHz and 600KHz
at the band edges.
The lower IF tuning cores are
awkward to get at so any further tweaking, say to flatten the
response slightly, can be made at the top cores.
|
|
|
|
| |
