K-B Featherlight (VC11) restoration

[dazzlevision]

Hello,

Recently, I was given this little set (made in 1964) by a fellow Forum member. The cabinet shell and front panel have scrubbed up well and the front edges of the cabinet shell are undamaged. This example also has a UHF tuner fitted.

In essence, the original KB “Featherlight” portable TV chassis is a mechanically re-arranged version of the STC VC3 hand wired dual standard chassis electrical design, in order to fit into the cabinet shell and hug the 12” CRT’s cone. Standard Telephones & Cables (STC) were K-B’s parent company. It might look oddball, but it works (rather well, IMHO), like its big brother VC3 chassis. Component accessibility is also pretty good. Apart from some germanium, selenium and silicon diodes, the only transistors are to be found in the UHF tuner.

Evidence of previous repair work was that some of the valves had been replaced (Brimar valves were fitted at the factory, due to STC’s connection with Brimar). The line oscillator valve should be a PCF802 but someone had fitted a PCF80, which I replaced. A few carbon resistors had been replaced (none of which were the correct value) and the Dubilier “Bluecon” mains filter capacitor had been snipped out of circuit at one end. Bluecon was a Dubilier trade mark, used for their mixed dielectric paper capacitors (also sold by Radiospares, but RS branded and with a grey outer case).

After giving the set an initial dust out and clean, I applied power and it produced a raster with no discernible vision modulation. This was due to an open circuit vision detector diode (an ST&C type GD12). The diode was inside the final vision IF transformer’s screening can. These STC “GD” series diodes all seem to have expired at this stage in their lives. I guess their environmental sealing wasn’t good enough. I used an OA90 to replace it.

I then had to clean all of the turret tuner’s contacts and was rewarded with video but no line sync. This was due to an o/c OA81 phase comparator diode. As a precaution I also replaced the “Dubilier” “Bluecon” boost HT reservoir capacitor and a Bluecon heater chain decoupler.

I now had a reasonable picture (albeit a little under scanned) but the boost HT started to decline after a few minutes, which set alarm bells ringing over the state of the LOPT! The PY801 was rather tired, as the boost HT took longer to establish fully than one would normally expect.

I then replaced all the high value resistors in the boost HT circuit (as they tend to go high or o/c with the passage of time) and the Boost HT was set to 660V, as per the service manual. However, the LOPT issue remained. To avoid any unnecessary stress on the LOPT, I also replaced the two “S correction” capacitors (405 = 0.3uF and 625 = 0.15uF). The originals were very leaky and out of tolerance (TCC “Supamold” and Dubilier “Bluecon”).

I then found that the cathode bias decoupling electrolytic for the PCF80 sound output valve’s pentode section was completely s/c (it was one of those Irish made “SDS/Callins” black cased types, so popular with Thorn/Ferguson). The cathode bias resistor was also out of spec and was replaced at the same time.

The VHF fine tuner didn’t work (it is a Philips AT7650 miniature turret tuner, with printed circuit coils on the channel “biscuits”) and it looks like someone has sprayed a lot of oily gunge inside it in the past. The fine tuner operates when the fine tuner control knob is pushed in and turned. This pushes a short metal shaft with a cross point at its end into the tuner’s outer casing. It then engages with a white plastic moulding that has a recessed cross moulded into it and thus engages with the metal shaft. These white mouldings are fitted around the periphery of the channel coils rotor; one per channel biscuit. On the other side of the plastic moulding is an eccentric cam and it is this that sets the position of the fine tuner slug for each individual channel, when it is selected. Due I think to the gunge applied to the tuner’s innards, two of these white mouldings have broken apart and fallen off. I am not sure they can be repaired and refitted.

I unsoldered the external connecting wires and completely removed the tuner from the set. I then took out the rotor drum with the channel coil biscuits clipped into it, in order to obtain access to the other components. That also gave me a chance to clean the fixed channel coil wafer contacts and see if the 22kOhm PCF801 g2 feed resistor of the mixer stage was still within tolerance (it wasn’t, so it was replaced).

I thought my Aurora unit might be playing up, as I was getting its channel 1 output on the channel 3 setting of the VHF turret tuner (and when I set the Aurora to Band 3, channel 9, the TV was set to channel 7). I checked the fitting of the channel biscuits was correct, in relation to the markings on the channel selector knob, which they were.

I freed up the Nylon slug (with aluminium sleeve) that forms part of the fine tuner variable capacitor (these were a stock fault on Philips made VHF tuners of this era; the Nylon part splits in two). I also removed the goo that had been applied (liberally) inside the tuner at some point in the past (all over the turret biscuits as well). Upon reassembly, I was dismayed to see that the results were unchanged! The set is very insensitive, which must be due to the local oscillator frequency being way out compared to the other tuned circuits for each turret biscuit.

I could see that the fine tuner slug was now moving in and out of the ceramic outer sleeve correctly, so the overall tuning capacitance in the local oscillator circuit must be wrong. As the fine tuner has no effect, it must be faulty or not making correct electrical contact with the rest of the LO circuit. Holding my finger near a rod ceramic by the fine tuner capacitor, I could alter the LO frequency. So, it looks like a second dismantle will be necessary (luckily, fairly straightforward).

One bit of good news is that the LOPT problem is changing. Up to a day ago, the boost HT rose to a maximum as the set warmed up (from a cold switch on) and then started to decline around ten minutes later, in the usual way. When I tried the set the next day, the Boost HT started to increase for around 10 minutes after a raster appeared, stabilised and then started to decline after 25+ minutes. I could hear a low level “sizzle” noise that seemed to be emanating from the LOPT’s primary winding. I was slightly hopeful that by exercising the LOPT in this way, but switching off before serious overheating occurred, it might stabilise.

Well, the next day, I tried the set again and the Boost HT remained steady for two hours! It does look as if the LOPT has self-healed. I think the most likely reason would be the evaporation of moisture and (possibly) the reflow of wax in the wax impregnated paper used in the primary winding’s construction. I will run the set in its cabinet, to allow it to reach a higher operating temperature, in order to drive out as much moisture as possible. Only time will tell if the LOPT is going to last.

To be continued………

Regards,

Dazzlevision

[dazzlevision]

Some more pictures............

[Tractorfan]

Hi,
That's one thing it certainly wasn't - featherlight!
Cheers, Pete

[Heatercathodeshort]

Great write up. I had one of these receivers in my kitchen in S.W. London for many years. Very reliable. Check to see if the fine tuner core [wand] is complete. Often the end drops off with the little brass ferule. Flatten the wire end of a 2 watt resistor, heat in in a flame then jam it into the bit held in the clamp. Unscrew when cold. I did have a few of those around somewhere.
Check the anode load resistor to the anode of the oscillator, typically 6.8K. It may have gone up to around 30k. These receivers are very sensitive and produced good pictures on their 'rabbits ears'. The tuner is excellent. John.

[FERNSEH]

I remember these sets well, very popular with customers. A derivative was the RGD "Night Owl"
The Philips tuner was the same unit fitted in the Philips Style 70 models.
Many other manufactures used that tuner.

DFWB.

[dazzlevision]

Hello again,

I have dismantled the VHF tuner (again) and I now know why the fine tuner isn't working.

The fine tuner slug's metal sleeve should be held in a fixed position on its Nylon core core by a Nylon "shoulder" pressing against the metal sleeve. In my tuner, the shoulder has disntegrated (I suspect due to the liberal application of that gunge in the past), allowing the metal sleeve to move along the Nylon shaft. As a result, although I could see the slotted end of the Nylon shaft moving when the FT mechanism was operated, the metal sleeve part wasn't!

Luckily, I have a NOS spare fine tuner slug left over from when I replaced them in TVs in everyday use (a long time ago!).

Pictures to follow and I hope to report a successful conclusion to the VHF tuner repair tomorrow.

Regards,

Dazzlevision

[dazzlevision]

Hello,

The attached photos show the good and faulty fine tuner slugs and you can clearly see the missing Nylon shoulder holding the metal sleeve in position. The other picture shows the fine tuner mechanism.

I will reassemble the tuner tomorrow and fingers crossed, that should do the trick.

Regards,

Dazzlevision

[FERNSEH]

Originally Posted by dazzlevision

The other picture shows the fine tuner mechanism.

A beautiful piece of mechanical engineering.

DFWB.

[dazzlevision]

Hello again,

Whilst I had the VHF tuner dismantled, I replaced two HT feed resistors in the mixer/oscillator stage, as per Heatercathodeshort’s advice.

After fitting a replacement VHF tuner fine tuner slug, I populated the rotor with its channel coil biscuits. I then applied suitable lubricants to the detent roller (which holds the rotor in the correct “stop” position for each channel biscuit) and to the rotor earthing contacts (that form a screen between the RF and mixer/oscillator stages).

I refitted the tuner into the set and soldered its connections in place. I then switched on and was rewarded with a good Test Card C picture with the tuner (and the Aurora!) both set to channel 1. It was clear that the set’s gain was much higher now the local oscillator tracked correctly with the other tuned stages in the tuner (so much so that I had to fit an 18dB attenuator). The free end of the fine tuner slug has a slot in it, to allow for adjustment. I used a trim tool to centralise the fine tuner range of adjustment on a Band 3 position, where it had the greatest effect.

Now the set was working correctly on VHF/405 lines, I turned my attention to UHF/625 operation. Moving the system switch to “625” and applying a UHF test pattern signal to the UHF aerial socket, all I could obtain was a faint 1kHz tone whilst tuning through the UHF bands.

I measured the LT supply to the tuner and it was only 3V rather than the 13V specified in the service manual. Uniquely (IIRC), in the VC11, the 12V supply is derived from the cathode of the second vision IF stage (an EF184 pentode). In hybrid sets from the VC11 era, such as the Sobell 1012 and Bush TV135, the dc supply for the transistor stages was derived from the cathode of the PCL85’s pentode section (frame output). In most other chassis, a simple resistive dropper was used to feed the UHF transistor tuner from the set’s HT line.

In order to get 13V at the EF184’s cathode, a positive bias is applied to the control grid from the HT line via a 2.7M 5% resistor and it was open circuit. Replacing this brought the UHF tuner supply up to around 11V and it now produced a fairly good picture, albeit with some moiré patterning. I found that I could greatly reduce this by moving the VHF turret setting to a Band 3 position.

Now in this set’s VHF tuner (made by Philips) as well as in the VC3 chassis (which uses an AB Metal Products VHF tuner), the UHF tuner’s IF output is applied to the control grid of the mixer stage through a “bridge” circuit, which doesn’t require the VHF tuner to be set to a “UHF” position, where a special biscuit would be brought into circuit (as is used in, for example, the Thorn dual standard sets from the 850 to the 1400 chassis).

However, I have noticed on designs using the “bridge” approach, that when the VHF tuner is set to Channel 1 and the set is switched to UHF/625 operation, there is a slight change in the UHF picture (e.g. when looking at a test card). I believe this is because channel 1 is so close to the IF pass band and it does have some influence on the overall IF response on UHF/625.

On the VC11’s Philips VHF tuner, there is an adjustable coil in the bridge circuit and I will try a small “tweak” to this and see if there is any change.

Of course, getting stuck into a VHF tuner's innards as I have done runs the risk of upsetting the circuitry. I was very careful not to move any components. I also replaced any faulty resistors with carbon composition types, rather than carbon film, which may have a different self inductance/capacitance at radio frequencies (and I ensured they were fitted exactly as originals were; lead lengths, lead bends, etc).

To be continued………

[Hybrid tellies]

Nice write up. Great little sets these and I think the only all valved portable dual standard set to be made. I repaired one for a friend many years ago. It was very reliable and sensitive on both systems. When my friend got promotion and had to travel around the country he would carry this set around in the boot of his car. The reliability of the set was much reduced so he scrapped it without telling me.

[dazzlevision]

Hello again,

Unfortunately, although I have the correct Philips trimming tool for the VHF tuner’s UHF IF injection adjustable coil, it refused to budge, so I have left this issue in abeyance. As long as I do not select channel 1, no patterning is present.

The width wasn’t quite sufficient on 625, even with the 625 set boost pot set to maximum. I therefore checked the values of the remaining (untested) high value resistors in the Boost HT regulation circuitry. I found three more that were out of tolerance (gone high). Replacement did the trick and the two “set boost” pots were now readjusted to give correct horizontal scan on both line systems.

The next problem was noticeable keystone distortion of the raster (the top and bottom of the raster was not parallel but noticeably closer together on the left hand side than the right. On the scan coils used in this chassis, made by Plessey, there are only two scan correction magnets and these are fitted on either side of the coils (i.e. in the East and West positions). They are disc rather than rod magnets, with a square central hole to allow the use of a suitable non-magnetic adjusting tool to rotate them. Sometimes, a bar magnet is embedded in the plastic moulding at the top and bottom of the coils, but not in this case.

Now I doubt adjusting these will cure the keystone distortion; they will mainly affect the shape of the left and right hand sides of the raster (I suspect the scan coils themselves have become in some way distorted or misaligned). I know that Plessey coils of this type have a tendency for the transluscent plastic moulding that supports the coils, wiring tagstrip, pincushion magnets, etc to yellow, become brittle and disintegrate!

Access to adjust these magnets is only possible with the rear part of the chassis lowered and even then, it is not good (your hand is over the two line output valves and their top caps!). I am going to file down a suitable trim tool from a knitting needle, or similar.

To be continued……


Dazzlevision

[AlanC]

I was given one of these sets in the 80's; it was an easy fix (snip the mains filter cap and fit a new fuse!) and it was my bedroom set through my early 20's. When the demise of Band I/III came about I tried to fit a UHF tuner to it without success- timebases ran okay on 625, but I could not get a signal through the IF stages. I tried everything, without success, and concluded that there was a fault on one of the IF coils that was only switched in on 625. Anyway, I wish I'd kept it- suspect it got slung when I moved out of my parents place together with a lot of other stuff I wish I still had! It was totally reliable and the 405 picture quality was excellent.

[jjl]

I had two examples of this type of set back in 1980, one was badged RGD, the other KB. The RGD one was the first telly that I owned.

They both developed faults where there appeared to be very little vision gain and poor horizontal sync. I never managed to fix them and managed to blow the RGD's CRT heater during one of my attempts

Having seen dazzlevision's first post, I'd be prepared to wager that the same two diodes were duff in my sets.

John

[dazzlevision]

Hello Again,

I have now tried adjusting (by rotation) the circular pincushion correction magnets. As expected, they mainly affect the left and right hand sides. However, I have only just noticed that there is a square metal plate fitted behind each magnet and that this plate can be rotated separately from the magnet, so I shall try this and report back. I have never seen these metal plates before (or possibly never noticed them!).

Pictures of the scan coils and magents/plates attached.

Regards,

Dazzlevision

[dazzlevision]

Hello,

Well, rotating the small, square metal plates had no perceptible effect. These plates are capable of being moved within their plastic housing and they are set in position with a small blob of locking paint. I presume they alter the magnetic field produced by the circular magnets next to them. However, I think that the keystone distortion is only really visible on a Test card, so I shall not experiment further.

The only other thing I had to do was apply some new adhesive tape to hold the CRT's rimband earthing braid connection in place, to eliminate some "brushing" interference spots on the picture. The later type of CRT used in the VC11 sets has a rimband, whereas the original type did not.

I think that this set's restoration has been taken as far as is practicable and the LOPT is still OK!

My final work will be to clean up the cabinet some more and try to fit some replacement telescopic aerials (both of the originals are broken).

Regards,

Dazzlevision

[Tazman1966]

A very interesting write up Dave. I look forward to seeing the pictures of the cleaned up and finished set.

As an aside, I wonder how it compares, performance wise with its contemporary set from the Thorn stable, the 980?

[dazzlevision]

Originally Posted by Tazman1966

A very interesting write up Dave. I look forward to seeing the pictures of the cleaned up and finished set.

As an aside, I wonder how it compares, performance wise with its contemporary set from the Thorn stable, the 980?

Hello Tas,

In point of fact, I have another fault on the set, as I hadn't paid much attention to the sound department. It is OK on UHF/625 (apart from the usual intercarrier buzz on early dual standard sets - and always on the Thorn 1400!). However, on VHF/405 the sound is quite low level compared to UHF/625 and when the volume is turned up, there is a background hum audible (not mains hum or 50Hz ripple, I think).

With my justifiable suspicion of STC "GD" series Germanium diodes, I looked at the sound detector part of the circuit diagram to see if any were used in the AM sound detector section and I saw that one is used as the sound interference limiter diode.

I up-ended the chassis and looked out for an STC GD6 but there wasn't one! It was actually a Mullard OA81 (factory fitted) but it is open circuit (at least on my Fluke DMM when tested in circuit on both "Ohms" and "Diode Test"). I will confirm it as faulty after I have removed one end to test it out of circuit, but it's highly likely to be u/s. Whilst it is out, I will also check the high value resistors that form its bias circuit, as they are quite likely to have gone high or o/c.

Concerning the VHF/405 performance of the VC11 when compared to the Thorn 980, I think the VC11 is better. I have worked on several examples of both chassis and I consider the VC11 to have a higher overall sensitivity (probably due to the use of two frame grid valves in the vision IF stages; EF183 & EF184). I would also say the line timebase synchronisation is better (and more stable w.r.t. frequency drift) on the VC11, due to the use of the PCF802 as sinewave line oscillator and reactance stages.

I also think the VC11, being hand wired, is more durable than the 980. The 980 uses one large PCB, with its high power wirewound resistors mounted directly onto it and with little if any stand-off spacing. Hence the PCB got rather dark and crumbly in those areas.

However, the 980 is lighter! It also has the Thorn Jellypot LOPT, which is much more reliable. I consider myself very lucky that the "dodgy" LOPT in my VC11 has recovered from the brink! For the customer, I think the 980's initial price of 39 and a half Guineas was the clincher!

To be continued...........

Regards,

Dazzlevision

[dazzlevision]

Hello,

Well, the OA81 sound interference diode was indeed o/c and fitting a replacement brought up the sound level considerably.

Whilst the set was running, it “died”. This was due to the 56 Ohm 6 Watt heater chain series ballast resistor going open circuit; so I fitted a replacement.

Looking at the circuit diagram, I noticed that a GEC GEX14 diode is shown as the AM (405) sound detector. Having also had trouble with this series of elderly Germanium diodes in the past, I thought it best to check it out. It is inside the AM sound detector IFT, which I removed. Looking inside, the diode is in fact a Mullard OA81 (factory fitted). Even so, I disconnected one leg to test it properly. It had a Vf of over 0.8V, which is very high for a Germanium p-n junction (typically 0.3V), so I replaced this diode as well. After reassembly, the AM sound was much the same.

At this stage, with the volume control advanced only a little from minimum (say 45 degrees), there is a hum/buzz on AM (405) sound, which disappears if the volume control is advanced a bit more (say another 45 degrees). I have found that by pressing on the metal cover for the loudspeaker’s magnet, it clears the hum/buzz (which sounds a bit like vision buzz on sound). I think the next move will be to remove the front moulding, which supports the CRT, loudspeaker and the two Silicon rectifiers (HT rectifier and Heater chain wattless dropper), to inspect the speaker in more detail. This will be a bit of a chore, but it will also give me the opportunity to fully clean the front moulding.

I will also have a close look at the earthing tags punched out of the chassis, used for chassis connection of components. I have known these to become high resistance (in particular, in the "hand wired chassis" decoder of the CVC1 and CVC2 colour chassis). Such h/r chassis connections can give rise to all manner of obscure faults. I sometimes find that when I desolder the wiring to such earthing points, the metal plating comes away, leaving me with bare steel, which is not easy (impossible?) to solder to! In such cases, I have drilled a small hole in the chassis adjacent to the tag and fitted a solder tag that has a shakeproof type of circular hole through which the screw passes, in order to give a more reliable electrical connection (and a self tapping screw).



To be continued…………..


Regards,

Dazzlevision

[dazzlevision]

Hello,

Well, I have inspected the chassis earth lugs and they all seem to be OK.

So, I removed the front moulding, complete with CRT. This is not the easiest of tasks with this set (gaining access to two of the screws holding the controls/speaker metal ”chassis” to the cabinet front moulding is very poor).

As soon as I could see the speaker cone, the likely cause of the hum/buzz at low settings of the volume control became clear. There is a 1” piece of sleeving (as used by KB sets of this era) stuck to the edge of the speaker cone!

I have now removed the sleeving and the speaker voice coil doesn’t seem to be rubbing against the magnet (another cause of distortion at low volume levels).

I’ll reassemble the front to the main chassis and let you know if that has resolved the sound problem.

Looking at the picture of the chassis, you will see two short lengths of black wire on the front left hand side. The piece with the tinned Copper braid is the dag earthing connection and fits between a 1" square cork pad and the CRT whilst the other is held onto the rimband by a piece of tape. This is how it should be, according to the KB service manual. In this set, the two wires were transposed. I will correct this when I reassemble everything.

Regards,

Dazzlevision

[dazzlevision]

Hello,

Well, after giving the cabinet front moulding and cabinet shell a good clean with Greygate plastic polish (to remove some ground in black marks) and then foaming cleanser, it has all come up a treat. The control knobs, with the exception of the UHF channel indicator, which has the numbers printed on a paper strip, were cleaned with foam cleanser and a stiff brush.

After reassembly, it looks much better (cabinet shell still off). I have removed the two incomplete / badly corroded telescopic aerials fitted at the rear of the cabinet shell. If I cannot find any suitable replacements, I shall have to blank off the two holes for safety reasons. I think a couple of black PVC blind grommets will do nicely.

I decided to have another look at the patterning on UHF/625 when the VHF turret was set to any Band 1 channel. I discovered that altering the VHF fine tuner affected the patterning, so I measured the HT feed to the VHF tuner's local oscillator on 625 (when it should be absent). Guess what, it wasn't! This was due to someone in the past adding a wire bridge across the 405 contacts of S10 in the 405/625 changeover slider switch!

After removing this, the lack of VHF local oscillator HT meant the set worked happily on UHF/625 at all settings of the VHF turret. Presumably, it had been modified in this way to receive 625 programmes on a wired distribution system where the UHF channels had been down converted to VHF ones.

With no aerials connected, there is quite noticeable snow (noise) on the screen, indicating a good level of RF and IF gain (which I would expect from this chassis, when operating correctly).

I'll post some pictures of the completely reassembled set and 405/625 pictures tomorrow.

Regards,

Dazzlevision