Service manual

Now, the original idea was to use a simple transistor mixer to combine the V.C.O. output

with the output from IC2.

The difference frequency would then be filtered out by an L-C

'tank' circuit to be passed to the LC7137. This idea failed to work in practice and was

abandoned in

favour of the digital approach (Ref. 1) pioneered by Bill Sparks G8FBX and

Colin

Horrabin G3SBI.

Instead of mixing and filtering, we simply blank

out a certain number of Hz from the

V.C.O. output during each second. The LC7137 does not really care about the waveform

shape since it incorporates digital circuitry itself.

The occasional missing pulse, therefore, causes no problems. Since the count is less than

that produced by the V.C.O. the LC7137 increases the V.C.O. frequency until it receives

the correct average rate of pulses. The object, then, is to remove one V.C.O. pulse for

every square

wave which is produced by the 10.24 MHz divider circuit.

You might be tempted to think that simply to differentiate the square wave to form a

spike

which drives a

monostable would be a satisfactory method of generating a blanking pulse

for each leading edge of the square wave. Well, it would be, were it not for the fact that the

V.C.O. and the 10.24 MHz crystal oscillator are not

synchronised.

Also, the length of the blanking pulse needs to be varied between

transmit and receive

because the V.C.O. frequency changes. Finally, the length of the blanking pulse would

also have to be adjusted according to the channel

frequency which also alters the V.C.O.

frequency. The problem is solved by the use of two "JK

Flip-Flops" in IC3. IC3a is

programmed with J high and K low so that each time the clock input CP1 sees the falling

edge of a square wave, output Q1 goes high. This action sets J2 high. (The output of IC3b

has been reset low by the combination of R4/C3 at switch-on and is unaffected at this time

by the state of J2).

Now, thanks to the efforts of Messrs Sparks and

Horrabin who developed a way to use a

74COO as an amplifier, we have a nice, sharp V.C.O. pulse waveform appearing from

IC4b and feeding the clock input CP2. (This explanation is not difficult, just 'wordy'

;

follow the circuit in Fig. 1.)

When CP2 sees the falling edge of a pulse, the contents of input K2 are transferred to /Q2

so this output goes low. IC4d is now disabled so the next V.C.O. pulse from IC4b can not

pass. IC3a input RE1 goes low and resets Q1 low so input K2 goes high. When the next

clock pulse falling edge appears from IC4b it transfers the contents of K2 to the output

which goes high, thus enabling IC4d which can then pass more clock pulses until the next

square wave falling edge hits Cp1.