Moreton Cheyney Receiver
Having succeded in hearing broadcasts
on medium and long waves, and having struggled to understand
the automatic gain controls operating within the receiver, I
decided to review and correct deficiencies in the circuitry.
One problem I discovered is the condition of the Westectors at
X1 and X2. So I could progress I replaced both of these with
OA91 germanium diodes, but having considered the forward voltage
characteristics of the Westector I need to further check for
a better substitute. The W6 has six copper oxide elements and
if each drops say 200mV when forward biased, this will amount
to 1.2 volts. The OA91 will have a drop of say 200mV so the careful
specification of the MC circuit will not do exactly what the
designers intended. Basically the W6 requires enough AC to develop
a rectified voltage of 1.2 and once this is passed will produce
the intended AVC control voltage. Essentially this is what is
referred to as "Delayed AVC"... or in other words,
AVC action is suspended until a broadcast signal of sufficient
strength is tuned. The OA91 may be so sensitive as to develop
AVC action by rectifying mere noise.
However, the MC circuitry has components
in place to provide a delay so the OA91 will not produce AVC
control too prematurely.
Above is one of two Westectors,
marked "6" located inside the can of IFT4. Data on
these is fairly sparse bit a WX1 has a rating of 6 volts RMS
and a forward current of 100uA and is good up to 1.5MHz. The
WX15 can handle 90 volts so a 6 element device will handle around
36 volts RMS.
There are two types of
volume control (nowadays referred to as gain control) within
the MC receiver, normal AVC which is always active plus what
they call QAVC which is intended to quieten the receiver when
tuning between broadcasts. Both AVC and QAVC use a Westector,
which I've labelled X2 and X1 respectively. As QAVC can be switched
off, I'll first consider AVC, but as the two circuits are interlinked,
what happens within the set when QAVC is grounded? V6 cathode
is grounded via R23 and R24, X1 cathode is grounded and V8 is
turned on slightly more than it was previously, but has no effect
on the receiver. VR5 is disabled and the QAVC end of R55 is grounded,
but having no measurable affect the AVC circuit.
AVC is produced by X2
which is reverse biased by R49/R50, developing a little over
3% of the HT supply voltage, or 8 volts with 250 volts of HT.
This means that AVC should not be active until enough RF is applied
to X2 anode to overcome the delay bias plus the forward drop
within X2, say 1.2 volts or a total of a little over 9 volts.
With an OA91 at X2, this would be about 8 volts.
With no signal and with QAVC
turned off, the AVC system should be inactive unless there is
a lot of noise present, say from the aerial or general valve
hiss etc. If you examine the circuit you'll find V1, V2, V5 and
V7, are used for AVC, with broadcast signal amplification using
V1, V2, V5 and V6. I can only assume this difference would allow
for a different response, perhaps a less broad response for AVC
(and QAVC). An interesting side effect would be to miss AVC action
completely if IFT4 is mis-tuned by a few KHz.
As per usual practice AVC is
applied to the RF valves, V1, V2 and V4 plus IF amplifiers V5,
V6 but not V7 (the AVC/QAVC amplifier). This will mean that all
those valves with AVC control have to be variable mu (such as
the 6K7 type and V7 non-variable mu (such as a 6J7 type).
Turning to QAVC. This is provided
by X1 whose cathode is wired through R38 and VR5. This is a low
impedance path so will need sufficient current to be generated
to have any effect on the system. V8 is the QAVC amplifier and
this can supply a decent current to drive the AVC circuit to
a level sufficient to reduce the gain of the RF and IF amplifiers
when a broadcast is not present. VR5 is available as a preset
control to set the inter-station background noise level to a
level which allows the user to hear stations and not to hear
internally generated valve noise or general annoying electrical
noise. How does the QAVC system work, other than modifying the
AVC line via R55? Within V8 are a pair of rectifier diodes which
are reverse biased through R23, R24, R38 and VR5 which together
place a small positive voltage on V8 cathode. This is primarily
governed by the cathode currents of V6 and V8 but adjustable
via preset VR5. The bias voltage is stabilised or maintained
by C13 and is added to the normal AVC voltage supplied via R82.
Once the reverse bias is overcome by a signal of sufficient level
present at the secondary winding of IFT3, demodulated audio passes
through the low pass filter comprising C77/R72/C78. If the receiver
is used for weak signal reception QAVC can be turned off at S1,
the Gram/Radio selector switch.
A quick check of some
voltages revealed much as expected. With no aerial connected
the IF amplifiers were getting around plus 0.5/1.1 volts of bias
with the QAVC turned off/on. The AVC diode was cut off at about
8 volts, the QAVC diode was reverse biased at 22 volts and with
QAVC off at -0.3 volts. Alignment should be performed therefore
with QAVC off and at a test signal input level below that required
to activate X2 (about 8 volts). Because of the high impedance
of the audio output circuits an external audio amplifier will
be necessary. Alignment of IFT4 would need to be carried out
with a test signal of 465KHz and ideally a spectrum analyser
to narrow the response of the coils compared with IFT1/2/3.