1946 HMV 1804.

[ppppenguin]

That's a remarkably good looking video signal.

[Geordie McBoyne]

Hi Jeffrey,
since my last post the video developed across the diode load has risen to over 15volts.
I must keep faith with this set. It might turn out to be a fantastic performer.
The vision circuits are not the problem but the sound is.

Geordie McBoyne.

[Murphyv310]

As Jeffrey says that is an impressive video signal. Someone has done their homework very well indeed with that front end.

[ppppenguin]

One of the big problems in designing a TV set is providing video to drive the CRT. You need lots of volts with wide bandwidth. Hence lots of gain * bandwidth product, something in short supply with earlier valves.

Some prewar EMI sets such as the 702 used what is effectively a small RF power amplifier to feed the detector which drove the CRT directly. They also used a high slope CRT to minimise the swing needed. As valves improved we arrived at the conventional low level detector followed by a video amplifier and, occasionally, a cathode follower. Usually driving the cathode as this was slightly more sensitive than the grid.

The problem arose again in the early days of transistor TVs where making fast high voltage transistors was a hard problem.

It's not hard to imagine that the guy who modified David's 1804 was a skilled engineer from Pye who knew exactly what he was doing.

[Geordie McBoyne]

The set seems to give the best performance on Channel B2, hopeless on channel B4. All this would suggest the set was modified for the Holme Moss transmitter. Was that TX receivable in Cambridgeshire?

Geordie McBoyne.

[ppppenguin]

That's a very long distance. Sutton Coldfield and Ally Pally were both closer.

[Murphyv310]

Don't forget about Joe who was receiving transmissions from south of the border in Anstruther!

[boater sam]

We regularly get Winter Hill in Cropredy when the Oxford signals are rubbish. Hills seem to stop Sutton Coalfield!

[Geordie McBoyne]

Readjusting the aerial input coil increases the vision gain. With the Aurora set to channel 2 the video developed across the vision demodulator load resistor is 15 volts P - P, and that's with the contrast control set to minimum.
So the performance of the vision stages is improving but it's not so with the sound. The performance of the sound channel is absolutely hopeless! The attachment shows the sound take off from the anode of the mixer valve.
The set came with a Mazda SP61 as the sound IF amplifier, this was subsequently replaced by an MO W61 vari-mu pentode. I've now replaced this valve with a Marconi Z66, a valve that resembles the SP61, so in a sense the sound IF amplifier is now back to the original design.
Because the timebase is inoperative, without any special measures the HT voltage will rise to a dangerous level, as high as 400 volts. To reduce the HT voltage to a safer level a 15Kohm resistor is inserted between the negative tag of the HT reservoir capacitor and chassis.   This was in fact done in early production 1804 and VT50 sets which had the facility to operate sound only, the resistor is bypassed when the set is switched to normal TV operation.  The vision off switch was ganged to the brightness control.   The Ekco TSC30 has the same control arrangement.

Geordie McBoyne.

[Murphyv310]

Was the Z66 not reckoned to be a pretty rubbish valve? So did you try the sound prior to using the Z66, perhaps the W61 would have worked OK.