1946 HMV 1804.

[Geordie McBoyne]

"Was the Z66 not reckoned to be a pretty rubbish valve"
Hi Trevor, that's right, I heard it from other engineers that the Z66 is not a very good valve, very variable in terms of quality. One idea that could be considered, we know the Mazda SP61 is electrically similar, it might be possible to change the Mazda octal base to international octal.

Geordie McBoyne.

[Geordie McBoyne]

A much safer source of EHT to test the CRT. This unit will after adjustment supply 4000 volts for the final anode of the 10" Emiscope type 3/4 CRT.
The ex-Pilot TV87 heater transformer has been found so it will be possible to supply the timebase valve heaters without having to employ the receivers' EHT transformer.

Geordie McBoyne.

[ppppenguin]

Very wise using safe EHT for trials. I have an old 0-5000V PSU with very limited current output which is ideal for older sets. Originally designed for school labs - I used them myself when at school.

If I needed more than 5kV I'd knock up something with an old diode split LOPT. I've got a couple of old PC monitor chassis which would be ideal for that.

[Geordie McBoyne]

Testing the sound circuits of the HMV 1804. With 100millivolts P - P of 10.5Mc/s modulated signal supplied to the input IF transformer an audio signal of 2.5 volts is developed across the demodulator load resistor. The signal across the load resistor rises slightly when the AVC is shorted out.
When you consider the input voltage is 100mV P-P and half of the modulated signal waveform is removed by diode action of the demodulator the gain of the IF amplifier is actually 50 times. I'd reckon that's a very satisfactory figure. The IF amplifier valve is a W61, a valve that resembles the 6K7G.
But this still doesn't answer the question, why isn't the sound coming through from signals supplied to the aerial socket?

Geordie McBoyne.

[ppppenguin]

Is 100mV reasonable at that point? Considering the set should work with not too many microvolts of signal that's asking for a lot of gain from the RF amp and frequency changer.

[Geordie McBoyne]

Hi Jeffrey,
this receiver in it's modified form has two RF amplifier stages plus a possible gain contribution from the frequency changer valve. I'd imagine that it's possible that 100mV could be present at the first sound IF transformer even from a weak transmitted signal.
I will in due course replace the W61 vari-mu with a high slope RF pentode.
The EHT supply unit was made over twenty years ago for a similar TV project which I can't remember. It was later borrowed by another firm which had a requirement for a high voltage module. I've tried Googling "Start Electronics Limited" but nothing has come up.

Geordie McBoyne.

[Geordie McBoyne]

News Flash!    I reckon I've found the reason why the sound was so weak.   The inner of the co-ax cable link between the mixer anode coil and the first sound IF transformer has been found to be open circuit.   I'll replace the cable tomorrow.

Geordie McBoyne.

[Geordie McBoyne]

Corrosion of the copper inner conductor of the co-ax cable was the problem. Now sorted and the sound is coming through, although sound vision spacing of the IF frequencies is way out. Nevertheless, things are starting to happen.
The audio developed across the 33Kohm demodulator load resistor is positive going, one could question why this was done because if the diode was reversed to give a negative going audio and DC component a convenient source of AVC voltage would be available and it would be unnecessary to employ a separate AVC rectifier diode
My take of this circuit is because a high slope straight pentode was employed as the sound IF amplifier it's possible that high signal levels will result in a high negative AVC voltage which in turn will cause some distortion of the negative going part of the modulated IF signals. The positive side of the modulated signal waveform will be unaffected. Once again this attention to circuit design confirms that the modifications were the work of a very competent engineer.

Geordie McBoyne.

[Geordie McBoyne]

We're coming up to the restoration of the timebases. The oscillators are known to be working and are supplying sawtooth drive waveforms to the grids of the output valves. The only modification the timebase circuits received was the addition of the double-diode interlace filter which took the form of a separate module which employed an EB91 double-diode. The circuit of this unit is similar to the Pye "positive interlace filter". In fact the circuit was not unique to Pye and was also used in the Pilot CV34.
The interlace unit was removed during the initial stages of the assessment of this receiver, it can be reintroduced later but instead of using the EB91 double-diode the existing D63/6H6 valve can be used instead.
As mentioned earlier an external 6.3 volt transformer will supply the timebase valve heaters, much safer than employing the transformer in the receiver which also supplies the EHT. Don't want 4KV cables swishing about while working on the set.

Geordie McBoyne.

[Geordie McBoyne]

The attachment shows the vision demodulator and amplifier stages of the HMV 1804 and Marconi VT50.
Negative grid bias for the video amplifier valve is derived from the voltage drop across potentiometer VR2 which is connected in series with the HT negative return.  

Geordie McBoyne.