Page 5


The scan coils on the replacement tube will also need to be hitched up. I shall use the original interconnection lead with its British 7-pin chassis plug for this. However, now the scan coil wiring has been disturbed, it turns out that the rubber insulation has lengths that are perished.

So new rubber wiring is inserted into the original fabric shroud and connected to the plug.

Both of the scan coils attached to the tube are about twice the impedance of the 907's originals, so it will be interesting to see how they perform and what modifications to the circuitry will be required.


The round base connector for the original 3/3 tube does not fit the 6/5 tube that's now going in. Instead, the 6/5 uses two special side-fitting connectors. There will be no room for these inside the box extension to the cabinet back and in any case I haven't got any - so instead I've decided to use modified, car-type 'bullit' female connectors. These will push straight onto the tube base pins and fit within the room available. They've been sleeved in black rubber so they won't look 'new'.

Two extra wires have been added to the harness going to the tube. These feed the accelerator and the focus anodes.

The modifications are now complete. Next, to switch the set on and see what happens !



First, of course, all the various units have to be hitched up. At this point, I give myself a little warning lecture about SAFETY. There will soon be 3,500 volts DC swilling around here, and I don't want to come into contact with it, because if I do - I won't ever get another chance. At this point, the workshop also becomes out of bounds to cats - and other people.

The first thing that happens as I power up the chassis on the Variac is a 'spak' - the sound of a large spark! This is a one-off event and so I carry on advancing the Variac. The spark is coming from somewhere the main chassis. It's likely caused by arcing over of the HT somewhere, as it surges before the set warms up. I'll have to look into this later. Such sparks have the unpleasant habit of going away again when you go looking for them, only to return later...

Eventually something forms on the screen. But it isn't a picture - it's what I call a 'zizz'. Something is 'taking off' (oscillating) somewhere. I have an old note about one of the IF valves possibly causing this. The KTZ41 is duly replaced.

  ... and after doing this, and correcting some incorrect component connections in the sync separator, and some jiggery-pokery on the controls, the British Heritage Television test card appears.

However I am unsure as yet whether the instability is really cured. Only time will tell.

This non-original tube seems to be well suited to the job. By sheer chance I've even got the scan coils connected round the right way too. The picture, seen from the rear, of course shows a mirror image view.

The new tube is more sensitive to vision modulation than the original. After trying various schemes to reduce this sensitivity at the vision output, I eventually decide to put things back to original. This gives the best picture.

The picture is bright enough to be easily viewed in a well-lit room, showing that pre-war televisions didn't have a dim picture when new. A small ion burn has rapidly developed though. It's mild and not really noticeable.

Next, I run over the vision IF alignment. For a while, some suspicion falls on the condition of inductor L25, so I strip it out. Here you can see its brass slug emerging. All is well here, so I put it all back.

L25 looks after the profile of the upper vision sideband. The original alignment spec on this set only provides about 1.7 MHz of vision signal bandwidth, which will give a decidedly 'woolly' picture. Since the upper sideband points away from the sound carrier, this presents the possibility of an almost 'free lunch' in extending this bandwidth and thus also the picture sharpness. Therefore I tweak it out a bit more. This trick will only work with a double sideband signal such as that coming from the Aurora. It wouldn't give any advantage with a vestigial upper sideband signal, as might be radiated to London in future by British Heritage Television.

While carrying out the alignment I made myself a special chart, based on a digital photo, showing the positions of most of the adjustables. Here's the basic photo I started with. It gives a good general view of the main chassis.

Further testing has revealed another fault, the synchronisation becomes degraded after 5 minutes from switch on, and the test card 'cogs'. This sort of 'warm up' effect is likely to be caused by a valve. I've tried changing the MSP4 sync separator valve with no improvement however. Maybe it's time to get the 'scope out - and see what circuit conditions are actually changing... let's hope the presence of a x10 scope probe won't stop this happening!

The volume control is crackling at the start of its track and is not really acceptable. This is a shame because I had already replaced it with a new (vintage) one. You can see that the mains switch is separate from the potentiometer and actuated by a rotating peg. Another vintage-style one was duly found and fitted and the crackling is no more.

The sync separator action was improved by careful setting of the synchronising potentiometer, which controls HT on the sync separator stage. As the valve warms up, the sync separator cut-off action drifts. The trick is to set the potentiometer so it covers for this drift. The situation isn't helped in that I'm now operating the set at a reduced contrast setting, to compensate for the increased modulation sensitivity of this tube.

The whole set is quite sensitive. It seems happiest with a 6dB attenuator in the aerial lead from the Aurora.

The sparking has never returned. I like to think that this is because of some re-routing of a wire that I did. The instability too has gone. Replacement of the KTZ41 seems to have cured it.

  The alignment of the radio section is now checked. A very standard type of radio receiver is included in this set. There is no separate RF amplifier; the aerial goes straight into the mixer stage via some tuned circuits.

All goes well until I have to set up the front end on the short waves. First I check the oscillator section is calibrated correctly and the oscillator level stays adequate over the whole band. The front end alignment then involves inserting an insulated 'wand' with a nick in its side into the appropriate coil. This has a loop of wire that is manipulated with the wand to fine tune the inductance. The wand moves the loop OK, but the range of adjustment proves insufficient...

I therefore make a plunger out of ferrite and insert it into the coil from the top of the chassis. Its circumference is padded out with a boot to ensure it stays where it is put. The short wave adjustment can now be correctly made.

 Next I try the television on programme materials rather than test card, to ensure the sync separator remains happy. It does! The picture is bright and the quality is lovely.

Final assembly now beckons. First I apply another thin coat of lacquer to the top of the cabinet. I've had to use thin coats to prevent the central transfer being attacked. Also, heavier coats applied after a delay can later cause the finish to craze or crack...

It now looks a lot better than how it was when I first took charge of this cabinet - with lifting veneers and the base plywood coming apart underneath. .

It's important to preserve the original 'deep brass' colour of the speaker grille, originally created in the factory by overspraying the brass with toner (lacquer).

After blowing out the debris and dust with compressed air, the surface is thoroughly cleaned, then dressed with a rub-on coating normally used for vinyl dashboards! It now 'glows' and looks great.

 The original identifier tab is remounted on the lower rear of the cabinet. It was crooked originally and is put back in its original position! The serial number of this set is 4168.

 The radio dial frame and glass is cleaned and goes back; the little central 'television' red indicator lens too...

The armour screen glass is removed and this and the rubber mask carefully cleaned. The glass turns out to be nearly 7mm thick and I see the area of the mask around the tube has been painted originally with deep brown paint.

I mount the tube into the cabinet. Only... I'm now going to have to do the same job again when I return to it on Tuesday, because I've just noticed there are a few small specks of dust on the tube face! There are also some bubbles in the glass, but I can't do anything about those...

You can see how the replacement tube juts out at the back. Adapting the back for this will be more 'fun' in store!

Everything now goes back in. By the way, the flash from the photography is here making the interior black surfaces look much more 'chalky' than they really are.

Let's hope it all still works when I switch it on!

Here are the knobs, prior to cleaning. Unfortunately three of them have chips in their edges where someone has tried to lever them off in the past and this will have to be repaired. Another half-day's work!

"Milliput" epoxy putty is used as the repair filler. This can be deftly shaped with water. When set, the result is finished with wet 'n dry paper and dark brown paint dabbed on and later merged in.

Here are all the knobs, including the repaired ones.

... and they are fitted back on to the set. Some of these knobs fix to forked ¼-inch shafts in time-honoured Marconi/HMV fashion, using special drive screws. Believe me, any old self-tapper won't work! Fortunately I have enough of these original-type screws in stock.

Now it's time to think about the cabinet back. The replacement tube is longer than the original and now juts out quite a way. After some rumination, I decide to make a new protection/extension box in the same pattern of the original - but deeper.

Here we see the original box in the top left corner, after removal from the back. The new box will be made of hardboard, which is being marked out on its smooth side. Using the rough side of the hardboard as the exterior surface of the new box and, then sanding and painting it, will give a good approximation to the original stippled finish.

The new box is then painted. Here is the new larger box alongside the original.

The mains lead had been replaced before. Only the cable used had been too fat and the live wire had become crushed where it entered the connector - a possible safety hazard.


A more slender fabric covered cable is now fitted. The connector mounts into the back, ensuring the mains is disconnected whenever the back comes off. This feature must have saved the lives of inquisitive owners, since when the back comes off there's 3,500 volts about!

However, this feature meant that re-assembling the connector became a right fiddle. The internal parts of the connector would fall out as the housing was assembled. In the end, as a temporary measure, I held the thing together using extra-long screws and tightened them in stages as assembly proceeded. Once everything was clamped up, the original screws could go back.

Here's the back with its big new protrusion. The fitting of this didn't go perfectly to plan. The back hinges at the left side and swings in (and out). In its original position, the aperture and the box would foul one side of the tube base as it swung in. So I have moved the box over to the right by half an inch and everything now fits as it should.

In essence, the set is now finished, though I haven't yet switched it on. When this set came to me, it had already had "attention". Nevertheless all the original parts that I've replaced have been retained, to assist historical reference.

Anyway, here's a peek at the picture. This is much as how it must have looked in 1939. Sharpness isn't its strong point and there is a small ion burn, but it's bright enough to be viewed under most conditions. The tube may further improve with use.

I have had to add a centring magnet behind the scan coils to centre the picture - the electron guns on these tubes are seldom perfectly aligned.


But one other thing... that pesky surge 'spark' has appeared again occasionally just after switch on. It only is present after a rest of several days - otherwise it doesn't occur.

Therefore I have taken the chassis out again, re-connected it to the rest of the set, and will switch it on every few days to try to locate the source of this brief event. The first trial will be tonight, but I'll have to wait 'till it's dark.


Of course the spark then went away! At least, until I had put everything back together in the cabinet again - then it came back!

However this time I saw roughly where it was coming from - the region of the height and width potentiometers in one corner of the chassis. Things are rather crunched in here and high tension is present on tags near to the metalwork.

I took this section apart and reassembled it, making sure there were adequate spaces between the tags. I also added rubber sheet insulation in crucial areas. No more spark!

Another thing I did was to re-route the cathode lead running to the tube separately, away from the other tube lead. The cathode lead conveys video to the tube. This has reduced stray capacitance and has made the picture sharper than seen in the photo further up and removed a slight tendency to waviness or 'cogging' on the test card.

The restoration story ends here. However the story of the set itself will continue, as now it returns to its owner - to be used and enjoyed.

A 75 year-old television set has been re-born and once again can occupy pride of place in the home.


Main Street, Sedgeberrow, WR11 7UF, United Kingdom.
Click HERE to email us.