Stephen Appleby was only one week old when his parents bought their 1956 AWA 205C (W series) Radiola "Deep Image" lowboy television set. It cost 210 guineas (1 guinea = one pound and one shilling, or £1/1/0), which was an awful lot of money back then (especially for a milkman).

His mother desperately wanted to watch the Melbourne Olympics which were due to start in November that year. What they didn’t realise until it was too late was that there was a 3-month waiting list for the set and there were only 13 days left until the Games started!

Fortunately, one of his father’s customers was an Amalgamated Wireless (Australia) Ltd company executive and so he left a note for him on his rounds, hoping he would be able to pull a few strings. The note was headed "Dr Appleby" ("Dr" in those days meant "debted to") and apparently that did the trick as the set was delivered from Rydalmere a few days later. It was installed by two men (the 21-inch set weighed a ton) and they also installed a flying bedstead antenna which was connected using 300Ω ribbon cable.

Well, Mrs Appleby was thrilled. In those days, only TCN9 was transmitting and the 16mm film was flown up from Melbourne to Sydney and then put on a telecine machine to convert it into an electronic signal for transmitting.

The AWA Deep Image set was considered to be the "Rolls Royce" model of its time. It boasted 24 valves (including four for video IF and two for sound IF), a 10-channel VHF fringe-area rotary tuner and two 9 x 6-inch elliptical electrodynamic speakers for "fidelity tone sound".

Actually, the sound system was better than that in reality. Unbelievably, each speaker was fitted in a bass reflex box inside the cabinet, along with a tweeter and crossover network.

The whole set had a limited guarantee to be "free from manufacturing defect and faulty material for a period of 90 days". And the picture tube was guaranteed for an extra 90 days.

This particular set was used daily for long hours and Stephen eventually inherited the set in 1982 when his parents acquired a new one. It was moved out to his farm in the country and he has watched it for 2-6 hours almost every day since then, right up to the present day.

Yes, it’s still going some 50 years later. Unbelievable!

Of course, the set breaks down regularly, the main problem being "worn-out" valves, especially the power output types. Obtaining the valves was always a problem and they were also costly, so Stephen was always on the lookout for them.

One time he thought he was onto a good thing when he was able to purchase eight boxed 6DQ6s from a retired technician at a good price. However, it was quite some time before he needed to use them and when he did, he discovered that all the valves inside the new boxes were secondhand and faulty.

By 2000, the set was really showing its age (44 years) and Stephen decided it was long overdue for a full overhaul. As a result, all the old Ducon wax paper capacitors and electrolytics which hadn’t been replaced were changed (some 150 components, all told). He also replaced most of the old 10% carbon resistors for 2W metal film high-stability types. The tuner was also stripped down and cleaned.

This improved the old girl enormously however the picture was beginning to get a bit dim too. But where was Stephen going to get a new AWV 21CVP4P CRT? Thomas Tubes, who re-gunned them, eventually tracked one down for him in the basement of one of their resellers in the country.

The tube was a newer type 21CBP4A Silverama which had a straight gun as opposed to the older Ion trap bent gun types. This particular tube had been re-manufactured by Thomas Tubes in 1971 and was sold for $50 plus an entertainment tax of 15%. Thomas re-tested the tube before selling it to Stephen and he installed it himself later that same year.

The set is still used daily and almost right about now is its 50th anniversary. Apart from being monochrome, it can’t tune channel 0 (the tuner is a 10-position type only) and no-one has tried using a VCR with it. However, one can imagine there would be a problem with the horizontal time constant causing line tearing.

The advent of colour TV over 30 years ago doesn’t concern Stephen who is quite content with old faithful. After all, as he points out, while the technical quality improves, the programs only get worse!

Different sets, same symptoms

Two CRT TVs came in with about the same symptoms. One was a Teac CT-M4801 and the other an LG CT-29K30E (MC8CB chassis).

Though quite dissimilar sets, they both had a bright white raster with retrace lines, almost certainly due to low voltage on all three cathodes of the CRT.

The fault in the Teac was simple to find and was due to C420, a 10μF 250V electrolytic capacitor, going open circuit and leaking fluid. Replacing it restored the picture.

The LG on the other hand was a "fish of a different kettle". It has no less than three electros on the 200V rail – C408, C905 & C906. I replaced the first one even though they all looked OK and I could measure 200V all the way to pin 6 (video output) of IC901 (TDA6108JF). This made no difference so I then replaced the IC. This fixed the white raster and retrace lines but I then found that the set was not responding to the remote control or the front-panel control buttons – all I got was a blank screen with a green OSD channel number in the top lefthand corner.

I don’t quite know when or where I learned about it but whenever you replace a damaged TDA6108JF, you should also check the "Black Current" TK line on pin 5 back to the jungle IC (IC512, TDA8844). There is a 7.5V zener diode (ZD904) on this line that often fails and in this case it had become leaky and was upsetting the SCL line, causing the set to lose control.

Sometimes the jungle IC itself can also be destroyed. Fortunately, in this case, it was just the zener and replacing it fixed the remaining problem.

Cheap as chips

Despite new HiFi VCRs being as cheap as chips, I still get a variety in for repair on a regular basis.

Recently, I had two 2002 Panasonics, an NV-FJ600A and a 700A, come in with the same faults. They both wouldn’t play tapes and gave FO6 error messages on their displays.

These VCRs both use the "Z" mechanism deck (just before they changed to the R4 series) and FO6 is a loading error (stop during cassette in/eject operation).

Once inside a VCR, you have to first override the end optical sensors (otherwise the ambient lighting won’t let you play tapes). And in one unit, you could see the loading motor try to load the tape around the drum but stop before it was fully in location.

The reason for this was that a number of plastic teeth in the take-up loading arm (VXL2670) were broken where it meshes with the metal loading rack. Replacing this fixed the problem but didn’t explain how it happened.

I checked the Mode Select switch S6002 but could find nothing wrong, so for good measure I also replaced the Main Lever (VML3166) which connects the main cam gear to the loading rack.

In the other machine, the loading motor refused to turn at all and it now gave an FO4 error (stop during unloading). This was because there was no 5V out of Q1002, because fuse IR004 (UNH000300A) was open circuit (this is a 300mA fuse which is in the shape of a transistor).

When the sets were fully reassembled, I soaked tested them but found nothing else wrong. Hopefully, I’ve found all the faults.

Just ask nicely

I don’t normally do car stereos as it is really a specialist job, requiring a jig and a large number of service manuals. However, I was asked very nicely whether I wouldn’t mind reassembling a 10-CD stacker (only) for a VX Calais. There was a loose round piece that looked like a CD centre platter and the owner was certain a man of my charm, wit, good looks and experience could fix it! Well, flattery will get you everywhere.

When I had removed the covers and all the loose CDs left in the machine, I could immediately see that this wasn’t going to be straightforward. What’s more, if the loose part was what I thought it might be, how would I be able to refit and align the height accurately for the laser to focus?

First, I had to order the correct service manual (1030) from Eurovox for the Maestro Series 2000 Model 92080354. When it arrived a few days later, I could see the problems mounting up. I then ordered a new M902 spindle motor assembly which was expensive – but not as expensive as the entire optical pickup assembly.

Removing it meant taking out the MD board first, followed by the servo board, until I got to the optical pickup assembly. This then had to be disassembled and all the mechanical parts transferred to the new spindle motor frame. It’s not a job for the fainthearted.

Nevertheless, I managed to reassemble the whole thing in a few hours without losing any screws or having any parts leftover. The only snag was that I had to return the stacker to the client untested, as I had no means to test it.

I haven’t heard anything back from him, so it’s a case of so far, so good!

Frustrating Teacs

The frustrating thing about Teacs of the VESTEL variety is identifying the correct circuit for the model you are working on. Even when you think you have the correct manual, there are often lots of changes in the production chassis and it is hard to know if it is the right one.

I had a Teac CTW8250S widescreen CRT TV come in and a label on the back indicated it had an 11AK19PRO chassis. By contrast, the labelling on the chassis indicates it is an 11AK19P4, while the service manual lists it as an 11AK19P3! So which one is correct?

The problem with the set was that it was dead and it turned out to be the line output transistor (Q605, BU2508D) and a few dry joints. However, these faults masked another problem which appeared when the picture was restored. There was now excessive width and pincushion distortion.

Initially, I thought I’d cracked it when Q603 (BUK444) was found to be short circuit and R629 (2.7W) open circuit. However, replacing these made no difference.

Next, I started checking and changing all the tuning capacitors in the line output stage, as well as checking the inductance of the coils. All were OK. I then tried going into the Service Menu by pressing the Blue INSTALL button on the remote, followed by 4, 7, 2 & 5. That done, I tried adjusting the width and east-west controls for different aspect ratios but again it made no difference.

By now, there were only a few more components left to check in the east-west correction circuit. In fact, many of the parts shown on the circuit aren’t even fitted!

There is the EEPROM where the data is memorised, the microprocessor and the jungle IC (IC401, TDA8844). The voltage on pin 45 of IC401 was 3.5V and when this was shorted to ground, it caused the width to come right in, implying I was on the right track.

I changed the EPROM (IC502) first, as it was the smallest and easiest. The microprocessor (IC901) was also easy as it was in a socket and I fitted one from a scrapped chassis. It was the jungle IC that was the most difficult as it was a 52-pin high-density IC fitted under a number of metal screening cans.

However, all this effort proved useless. The geometry stubbornly remain-ed the same.

It was a clever friend of mine who worked out the cause and the solution. The east-west FET (Q603, BUK444) has different specifications according to different manufacturers and I had paid no attention to the manufacturer of the one I had fitted. This FET is DC biased by R606, which in my set was only 15kW. However, in other sets, this resistor can vary up to 82kW.

In this case, adding an extra 33kW in series with the existing 15kW resistor (making 48kW) biased the FET into its operating range, in turn allowing the menu to make the appropriate adjustments. And that fixed the problem.

I must admit that I have often heard of transistors sometimes requiring circuit changes to allow for their gain but this is the first time I have had to do this for a FET, especially as the replacement had the same type number on it.

Customer abuse

More and more LCD displays are coming in, usually for customer related problems.

Recently, I had a 2004 Super Multifunctional TFT Colour LCD TV come in dead. This had an external 12V 13W regulated power supply but the client had tried to run the set directly from a car cigarette lighter socket which, as everyone knows, is nominally 13.6V but can vary up to 16V. What he should have used was a regulated DC-DC supply adapter (ie, one with a 12V IC regulator).

Anyway, we both agreed it had probably blown a fuse. Unfortunately, no-one knew who the agent was for "Super", let alone whether there was any circuit for a Type SP-13A.

I removed the covers after I eventually found two concealed screws (one under the serial number label and the other in a deep hole at the base) and searched and searched for a fuse. I could find the 12V coming into the unit in one or two places but nowhere else.

Eventually, after a lot of searching, I found a glass tube with wires coming out at each end – rather like a lilliput/miniature festoon style globe. This was open circuit but the major clue was the "FS1" marked on the PC board next to it (yep – mind like a steel-trap). I took a punt and soldered a strand of wire (from a multi-strand cable) across it and that did the trick, the set now coming on.

Everything was working OK except that the red Standby LED wouldn’t always come on, although the green LED was OK. However, it was beyond my means to economically find out why this was happening.

At least the set was otherwise performing OK and so we let it go at that.

Sony rear-projection set

The 1998-2000 Sony KP-EF41 (RG-2 chassis) rear-projection CRT TV has been a reliable and popular model. It was a further development of the RG-1 chassis fitted in the KP-E41 series and had an interesting one button automatic convergence system.

Now eight years old, we are beginning to see a few of these sets which, like all rear projection units, can be difficult to fix. And from bitter experience, I will only repair them back at the workshop where I can raise the whole set to waist height. Gone are the days where I will crawl around on the floor, peering into dark holes.

If these sets reside near beach-side suburbs, the insulation of the HV block often fails and arcs. The HV block is basically a 3-way splitter for the 34.5kV EHT to the CRTs, as opposed to the flyback transformer (T504) which can also fail as a result.

Recently, I had one of these sets in (KP-EF41SN3) and you could hear and see the arcing from one of the EHT cable sockets which had developed a crack. Unfortunately, by the time we had diagnosed the fault and unplugged the HV block, more damage had been inflicted and the set was now completely dead.

It didn’t take long to discover that the line output transistor (Q502, 2SD2539) had gone short circuit and I initially thought that that, along with the cracked socket, would be the end of my troubles. Unfortunately, I wasn’t so lucky because after replacing them, the set was still completely dead.

My next step was to take a look at the power board (G). First, I removed and checked C6030 (0.039mF, 1kV) as it sometimes causes strife. This one was OK so I linked CN6007 to CN6006 and pins 1 & 3 of CN6008 to feed the standby 7V rail to the Power-on control. In addition, I hung a 100W globe between the +135V rail (on CN6011-1) and ground.

That done, I found the power supply was working perfectly OK, with correct voltages on all rails. I then refitted the power supply into the set and this time shorted pin 4 of CN6008 (or the link on the board next to CN807) to ground, which is the Protection Line. This let the set start up but I was reluctant to leave it on for any length of time as protection circuits are there for a good reason.

The protection line meanders across the entire set, so I started disconnecting it to various circuit sections in an effort to narrow down the fault. I soon found that unplugging CN506 to the D board would allow the set to come on.

Once again, I was blessed by having a scrapped set on hand from which I could swap modules. This really can save a lot of time with a problem like this. Anyway, I swapped both the D and AG boards and all the plugs one by one until I realised that the protect circuit wasn’t being switched on by them. Instead, replacing the E board cleared the fault, so the service fault lay there.

This was getting confusing and it was time to carefully analyse all the functions performed by CN506, which is a 10-pin connector. Three of the pins are ground and two are ±15V which all measured OK. That left the PROT, ABL (Automatic Brightness Limiter), HD (horizontal Drive), PIN (Pincushion Correction) and HP (Horizontal Pulse) pins. I figured that a signal from the E board was turning the microprocessor protect mode on.

However, before getting involved in any fancy convoluted theories, I managed to drag myself back to basic principles – when in doubt, measure the B+, or in this case, all the supply rails on the E horizontal deflection board. This I did, and I initially thought that they were all OK. Then I remembered that this was a KP-EF41SN3 model and its E board is different from that in the SN2 and SN1 models as described in the service manual I was working from.

In the SN3 model, the ±13V rails are derived via IC regulators IC807 and IC814 from the ±15V rail – not from the flyback transformer. And when I measured the output of IC814 (NJM78L12A), there was nothing. Fitting a new one gave only a low output and the device was getting hot and that meant there had to be a problem on this rail, even though it didn’t measure short circuit.

After a lot of painstaking circuit tracing, I found that Q805 (2SC3311A) had overheated (it had a burnt appearance) and had gone short circuit. This transistor functions as a buffer stage in the ABL vertical blocking line to the shading control (IC803). Replacing this restored the +13V rail but just why this particular fault triggered the protection line is a mystery.

For good measure, I also fitted a new HV block and at last got a picture. However, I still wasn’t out of the woods because the convergence was way out, especially along the bottom, and I was unable to adjust it.

Replacing both the convergence ICs (IC809 and IC810, STK392-150) improved the situation a lot but there were still problems, especially at the bottom of the picture. Replacing IC803 (CA0007) finally fixed that problem.

Sorting out plugs CN503 and CN504 to the green and red CRT horizontal deflection circuits eventually allowed me to get the static convergence and auto convergence to work properly. However, I had to connect them in the opposite way to that shown in the circuit.

The final touches to the dynamic red-green convergence had to be made in the service mode (GH SIZE, LIN and RH SIZE), with the data saved and written onto the EEPROM afterwards. I’m sure that you will now understand why I insist that these units be repaired in my workshop!