Items Covered This Month Philips Matchline 33FL1880/75R (FL1.1-S AA) TV set. Shamrock SRC2102L tech. monitor Teac CT-M5928 TV set. Audioline FF895 cordless telephone.

When a 10-year old 80cm Philips Matchline 33FL1880/75R (FL1.1-S AA) TV set came into the workshop I thought it would be routine. Considering my experience with the Philips Matchline FL1 series of TV chassis, I consider myself to be something of an expert on these. Unfortunately, I soon learnt the meaning of the phrase "pride goes before a fall".

Truly, this came close to being one of my worst disasters yet.

The set came in with an "intermittent dead" label attached to it. But for the first two weeks, it behaved faultlessly, producing an excellent picture. Then, during the third week, I saw the problem for the first time. The set would start OK in stand-by but wouldn't come on. All the front seven LED display indicators were flashing, in turn, in various combinations. Sometimes the set would come on and then later stop with the same flashing indicators.

Supposedly, each of these flashing indicators represents an error code but when they all flash, what does one do? Not every single circuit can be faulty!

I obtained a service manual for the FL1.1 chassis. Unfortunately, large chunks of the manual were completely different to the set I was looking at. In the end, I settled for a mixture of the FL1.0, FL1.0S, FL1.1 and FL1.1S (and even the FL1.2). However, this particular model still had differences that weren't covered in any of them (eg, the bridge rectifier is a single D5SBA60SIN block and not four separate diodes as all the service manuals showed).

The next thing was to engage the SDM (Service Default Mode), which provides microprocessor diagnostics. This is done by shorting inaccessible test points S24 and S25 deep in the bowels of the small signal panel. I have given up trying to kill myself and the set by reaching these parts; I now always mount a self-centring toggle switch on the rear so I can go easily into Service Mode S23 and S24, as well as the SDM.

That done, I examined the SDM figures on the OSD (On Screen Display): "SERVICE. 08 10 99 23 09".

I interpreted these to represent the fault codes for the last five breakdowns, 08 being the first and 09 being the last. 08 means a PSI/Control fault and 23 means System/Ident fault. Code 09 means a video control fault. These are matched by front-panel LED display combinations. On paper, this looked great but I couldn't find what 10 and 99 meant.

Supposedly, the SDM is meant to override the LED flashing but in this case, it didn't always do this. Anyway, to refresh the error code list, I switched the set to stand-by with the remote control and put the back on before reapplying the SDM. This now gave a row of zeros, "SERVICE 00 00 00 00 00", ready for the next time the set faulted. It wasn't too long before it did, giving "SERVICE 00 00 00 09 99". I examined IC7430 (TDA4680) in the video control for 09 code but couldn't see anything wrong. The 99 code made no sense at all.

By now I noticed the code most often given on the front panel was the "mute", "stand-by" and "stereo" LED displays and so, as a last resort, I followed a fault-finding flowchart in the manual. This eventually suggested I check (Test Point) TP56 at Plug L40-1, which I did. This measured 4.4V which implied I had a SOPS (Self Oscillating Power Supply) fault.

Getting this chassis into a suitable position for servicing is precarious at best. Basically, the chassis slides out after releasing plastic catches on either side of the boards. It can then be pulled back, along with the control modules, until it clears the runners. That done, I balanced it vertically on the bench with the two boards at an angle to each other. I couldn't find brackets F, nor were the leads long enough to use the service position shown in the manuals.

Thus far, the diagnostic system wasn't really making much sense so I started making voltage measurements to see what was happening. I had 18V (TV START) on (Test Point) TP60 and 5V (STBY) on TP50. TP56 was mostly at 4.4V but sometimes at 0.79V. However, when the set was working, it was 17.5V.

I also had 280V on V1 and, in the fault condition, there was 141V on TP57. In an effort to make life easy, I replaced the SOPS drive unit module from a good working set but it made no difference.

Next, I decided to examine the protection circuits but before I could get very far, one final indignity occurred. Though the set had been completely off for quite a while, I accidentally received an electric shock from the collector of chopper transistor 7216.

But that wasn't the end of it. Without thinking carefully, I decided to discharge the main electro through the powerful chopper transistor by biasing this transistor on using my multimeter - ie, by setting the meter to the x1 ohms range and connecting it between base and emitter. This worked really well - the voltage was removed from its collector immediately. What I didn't realise was that this had also damaged the transistor, so that when I switched the set on the next time, there was an almighty bang followed by a deafening silence.

Well, I had done really well for myself. I not only had the original intermittent fault to deal with but a much more serious fault as well. The explosion had destroyed the main fuse, the on/off switch, the bridge rectifier (a massive 25A job), the chopper transistor (7216), two diodes (6217 and 6218) and the SCR (6228) - not to mention my ego, which was left crushed and bruised. Perhaps I should retire now?

Patiently, I diagnosed and replaced each destroyed part but some of these items were no longer available. The 2SC3973B had to be replaced by a lower tolerance 2SC3973A, and I had real trouble matching up the two 1N5602 diodes. I was lucky with the bridge rectifier but I had no hope with the SCR, which was an S0824NH (the circuit called this a BT150). Neither is available anywhere.

I tried a variety of substitutions but all I achieved was a momentary picture, before it closed down with R3227, a posistor in parallel, getting extremely hot. I discovered that the original SCRs are high-voltage high-current "sensitive gate" devices, which are now an extinct species. I must confess that, until now, I had never heard of these components!

Fortunately, I was saved by a SILICON CHIP staff member, whose modesty prevents me from naming him. He designed a little circuit that used a sensitive gate SCR (C106D) to trigger a main SCR (TYN816) substitute. I made up this neat little circuit, mounted it underneath the main PC board and it worked really well.

So now I was back at the beginning, ready to tackle the six protection circuits. There is only one way to deal with these and that is to disconnect them one at a time and monitor the base voltage to transistor 7381, which should be 0V.

It sounds simple but it's not, especially when the fault is intermittent and operating on three protection circuits. The first was the sound protection circuit, which I am sure I would have picked up earlier had I been provided with the external loudspeakers needed for this set. The problem here turned out to be IC7000 (TDA1521Q) and C2071 (which I changed to 330μF).

The second fault involved the over-voltage protection and this turned out to be D6375 (LL4148) and D6376 (LLZC18V), the latter a surface-mount zener.

The third problem had me really stumped. It was the EHT protection line off pin 13 of the horizontal output transformer, and it was very intermittent, sometimes taking several days to show. In desperation, I changed every part of the circuit and it still played up. Someone suggested desensitising the circuit by changing some of the values but in the end, I was so fed up with it that I just disconnected it and left it on test.

Well, of course, two weeks later the inevitable happened and the set died again. This time, one of the horizontal output transistors (7504) was short circuit. This device can be either a BU508AF, ON4673A, 2SD1884 or 2SC4288A but the important thing is to check tuning capacitor C2504 (470pF) and C2523 (8.2nF). This time, I didn't stuff up - C2523 measured only 6.7nF. I replaced this and upgraded C2504 to a 3kV ceramic type

I then found that the east/west output transistor (7610, 2SA1359) had also been taken out. I replaced this and reconnected the EHT protection line. My persistence was rewarded - the set now fired up correctly and was still working four weeks later. It was just as well, as it was either me or the set. I had already written the note and loaded the gun!

But I still don't know what fault code 99 means. Can anyone enlighten me?

Shamrock revisited

The Shamrock story, which started back in December 2002 but was left in limbo, has progressed - just a little.

A friend who specialises in monitor repairs very kindly offered his services and, working without a circuit, fixed the east/west pincushion problem.

He also located the smoking device that I couldn't trace. It turned out to be R495 which feeds ZD404, Q417c, U405 pin 4 and much more. Unfortunately, the smoke had also discoloured the resistor's colour code. He guessed it to indicate 100ΩW but checking in his own set suggested this could be a common fault as that, too, had been replaced with a 180ΩW 5ΩW resistor.

I was thrilled about this but sadly a new fault had developed which has brought me full circle. When the set warms up, the beam current rises and the power supply makes "chirruping/squealing" noises, and then closes down. Using freezer, I traced this to an area in the power supply around Q108/D107 (the FET driver) and IC U102, but I was unable to find the offending part. (I changed all the electros).

Yes - it's all very frustrating but I am still hopeful that someone out there might have a circuit diagram.

Teac CT-M5928 TV set

Unfortunately, that wasn't to be the end of my frustration this month. I had another Chinese built 5-year old Teac CT-M5928 come in with the repair docket saying it was dead, and that it was really urgent, as its owner was moving interstate.

As usual, I felt pretty confident this was going to be quick and easy; after all, I had a circuit diagram and plenty of experience with these models, plus lots of parts in stock.

With the back off, it didn't take long to find that the problem was due to no horizontal drive from pin 26 of IC102 (TDA8305A) to transistor Q401. Actually, it wasn't strictly true that there was no drive but the oscilloscope showed only a brief burst before the signal died almost immediately. Waveforms were also present on the horizontal driver and output stages but, as expected, they too collapsed along with the drive.

All the voltages from the separate power supply were correct at 143V, 18V and 18V, as were all the 12V and 5V rails. Obviously, as I had no horizontal output stage operation, I could not yet check the voltages emanating from there.

All I had to do, it seemed, was find out why there wasn't a consistent output from the jungle IC. This is a very popular Philips IC and is also used in their sets, notably the GR1 AX. Being a jungle IC, I was only really interested in about a dozen pins pertaining to the horizontal drive output circuit.

I had a very similar model soak-testing in the workshop (AWA6850SH) and I quickly jotted down all the pin voltages of this IC and then compared them with the faulty set. Naturally, because the faulty set wasn't working, not all voltages could be compared.

Pin 7 was for the 12V input, which was correct, with no appreciable ripple. This 12V also supplied the vertical oscillator on pin 2 and the horizontal drive on pin 26.

I established that the horizontal drive was still present on the good set , even when its horizontal output stage was disabled. This was important because it may have been that the 12V from the secondary of the horizontal output transformer may have affected the horizontal phase on pin 28 which it eventually supplies. I then found that the faulty set couldn't supply substantial output with or without the horizontal drive circuit being connected to the horizontal driver transistor (Q401).

I removed the 28-pin jungle IC and fitted a socket in its place. I then swapped it with the IC in good set but it made no difference. Back to the drawing board.

Just in case there was something wrong with the 12V supply (I didn't really think there was), I connected an external power supply to R125 and then added the 12V supply from the horizontal output transformer. Again, nothing changed.

By now I was beginning to feel nervous and time was ticking by. I measured every component (15) attached to pins 23-28 and also replaced all five electrolytic capacitors around the IC. Once more I drew a blank.

It was at this point that I had some good luck. I was about to check all the waveforms around the IC and was in the process of winding up the external power supply when I noticed that, in the range from 7-9V, the horizontal drive would come on and actually stay on! I then reconnected the horizontal output stage and repeated the operation. This gave sound and a picture, complete with colour - in fact, almost everything was working.

But this was all quite unexpected and had me really confused. I spent some considerable time experimenting with the power supplies to pins 7, 2, 28 & 26, as well as to pin 11. In the Philips set, the latter is used to start the horizontal oscillator but in this set, there is only a resistor to chassis.

Unfortunately, none of this was leading anywhere. The best I could do was fit a 33Ωresistor from the 12V supply to pin 7 to make the set work. If the voltage increased beyond 9.5V it stopped. And I did start to notice other deficiencies with the set's operation. First, the sync was somewhat weak. Second, the AGC was prone to make the tuner IF stages overload and the RF AGC control apparently did not have any effect.

Third, the automatic tuning system did not stop after a station was tuned in. And finally, the colour would occasionally change its phase - even when using a colour bar generator connected to the AV input sockets. However, all this could be because the rest of the IC was underpowered.

But time was up - the client just had to have the set back. I told him of the developments and, somewhat disappointed, he just collected the set and left. I was equally disappointed - I had given it my best shot but it wasn't good enough.

Afterwards I went over my notes to try to see if I had overlooked something on the sandcastle line to pin 27. This was a 2-way connection, with sandcastle pulses going out and horizontal output pulses coming in from R419, R420, and C433 to the horizontal output transistor (Q402).

It's too late now of course; the set has gone but I sure would like to know what I missed!

Now, on a brighter note, here is a contribution with a happy ending. It comes from J. B. of Hampton Vic. This is how he tells it.

Audioline cordless telephone

Following a recent electrical storm in our area, I ended up with an Audioline FF895 cordless phone on my bench. This is an upmarket phone, with multiple handsets and lots of features. The reported failure was that the phone would not connect to the phone line.

Examination revealed that the handsets appeared to be working OK and that the base station had most of its functionality, the main problem being that it could not get a dial tone.

I was also armed with the knowledge that this phone had been taken to two other repairers and they had declined to repair it.

I had no circuit information about the phone and not being very familiar with their workings, simply carried out a few basic checks. The plugpack was OK but I thought I was on to something when I measured the input resistance (DC) of the phone line connection and found it open circuit.

At this point I suspected that what ever had failed was the cause of the open circuit. And when I say open circuit there was no reading on my multimeter. As this device has a mixture of surface mount and through- hole components, I had to use a couple of pins on my multimeter to probe for some of the connections.

Tracing the circuit, I found my way through various chokes and filter components, all of which checked out OK. When I started to find various 10MΩ and higher value resistors, I was starting to think this high impedance was normal. I tested several diodes and some transistors but was having no luck at all.

I have a theory, which I call the "black box syndrome", which states that if you do not understand any component, or what it does, then that is the component you are most likely to suspect as being faulty. This, of course, is questionable and really means that if you don't know what the part does, you won't know how to test it and thus eliminate it as a suspect.

I try at all costs not to apply this theory as it is usually wrong but I was staring at a TEA1110AT IC and had no idea what it does; only that it was in the right area.

This is a Philips device, so I got straight onto their website and downloaded the data sheet for this chip. At least I could find out what this device does and work out whether it was causing the problem.

It turned out to be a "Low Voltage Versatile Telephone Transmission Circuit With Dialler Interface". What a mouthful. Still, it sounded like a good culprit. What's more, Audioline appeared to have used a circuit configuration that was very close to that shown in the data sheet.

This IC is powered from the phone line via a bridge rectifier, 10Ω resistor and a 10V zener. With the base station in the stand-by condition, there was no power to the IC but when the hands-free speaker was selected, I had 10V on pin 1 (LN).

There were a couple of transistors controlled by another section of the phone between the bridge rectifier and 10Ω resistor. These were switched in when the base station was off-hook. The data sheet suggested 619ΩW between pin 1 and pin 14 (Vcc) but in this case, it was 1kΩ.

When the voltage at pin 14 was measured, there were only a few millivolts. A check to chassis found about 25Ω at pin 14. Maybe there was something in the black box syndrome after all. There were several other components connected to this pin and I wanted to isolate the IC from these components. The IC is a surface mount type, so with the aide of a magnifying lamp I was able to unsolder pin 14 and check the resistance without the IC. This turned out to be 80kΩ or so, so the problem was in the IC.

I phoned several suppliers and soon found that the TEA1110AT is not stocked by anyone I normally deal with. A call to Philips lead me to a few of their agents who either did not want to know me (I'm not a manufacturer) or would only sell me a reel (1170 pieces). The next problem was that there were none in the country. Maybe those other repairers have seen this problem before and knew that the necessary parts were unobtainable.

After making further calls, I eventually found a very helpful lady at Philips who was able to help. As a result, I decided to remove the faulty IC and now have great sympathy for anyone who has to repair devices with surface-mount components. However, I eventually managed to remove it without any damage to the PC board.

With the IC on the bench, I checked each pin and found only 5Ω between pins 14 & 2. Because pin 2 has a 20Ω resistor to chassis, that accounted for my measurement of 25Ω. A look at the internal circuit for the IC suggested that there was a very convoluted path between pin 14 and pin 2. I was feeling confident as there is no way it should be so low, so things were looking good. All I had to do was sit back and wait for the replacement part.

While waiting for the part, I had an opportunity to measure the line resistance of another FF895 and found it to be open circuit (higher than I could measure), so this was the normal condition. When the IC turned up all I had to do was solder it in - sounds easy doesn't it?

I tinned each of the PC board pads, then gently held the device in place with tweezers and tacked one pin, aligned it, then removed it, tacked it again and realigned it! Once in location, soldering the remaining pins wasn't all that difficult but I did spend some time checking each pin with a magnifying glass to make sure I didn't have any solder bridges.

Reassembly involved holding one's mouth correctly to align the circuit board and I have to admit to being relieved that the phone now worked correctly.

Weakest link

I thought the failure was interesting in that the component that failed was not the first stage - there was a lot of hardware in between. The surge had obviously come down the phone line but this phone has various chokes and filters, so it appeared to be well protected. Like all things I suppose, it is always the weakest link that fails.

Share this Article: 

Digg this story

Seed Newsvine

Post to Reddit

Save on del.icio.us

Email