20-07-2015, 09:28 PM
Let's remove the top cover (2 screws). And, in our excitement, let's remove the modulator oscillator PCB for a better view of the main board:
There's a lot going on there. The PSU is obvious enough - a nice quality toroidal transformer, as expected, along with standard 3-pin regulators. Next to that are 8 TTL ICs, and these are responsible for generating the various frequencies needed by the modulators (19kHz, 38kHz, etc). There is a row of test pins (between the tants and the 7400-series chips) that allow for quick checking (all were good in my case). If required, the trimmer allows you to fine-tune the master oscillator.
Moving left, the two ICs in sockets are CA3054 transistor arrays, arranged as modulators. You might notice that they are different to each other - that's because one had failed. Upon test, the output was basically a large 38kHz square wave - so much for suppressed carrier! Clearly something wrong with the 38kHz modulator. Luckily for me, the chips were socketed, and I had a couple of spares. Fault found
The other modulator is actually working at 114kHz, and this modulator is used to cancel out sidebands created by the 38kHz modulator. If this sort of thing interests you, have a look at the manual - it's fiendishly clever stuff
Moving left again, you can see the cores that make up the composite filter. Very complex, but ensures that the frequency spectrum at the output is correct. Again, check the manual for a well-written and lucid explanation.
Top-left is the output stage (LM318) - note the yellow wires connecting to the DIN socket and the Composite output. The SCA input is on the orange one. The peak-detect circuit for the meter is also here.
Bottom left, behind the switch-bank, are op-amps (LM356 and LM318) that are doing input processing. The modulator PCB connects to the empty IC socket.
The oscillator op-amp (LM318) is hiding behind its ceramic decoupler - the other two op-amps (LF356) are buffers for the L/R input signals. The right signal goes via the Ext switch; the left signal goes straight back to the main PCB. Note the thermistors for amplitude stabilisation - not just the large one, but there is a pair of smaller ones (that look a bit like germanium diodes) with paint dots applied for identification.
I did note that one of the electrolytic caps had some signs of leakage, and upon test the ESR had risen to a few ohms. I replace both for good measure with 105C replacements from Philips, but I didn't find any other faulty capacitors anywhere else in the unit. They are all very good quality units.
Note the date code - week 33 of 1981. Ancient, but in remarkable condition. It's clearly been looked after.
A tip for anyone who has one of these: be very careful when removing the PCB as the buttons are nowhere near as simple as you might think!
Each button has these 3 components. The brass sleeve is mostly a tight fit into the plastic, but occasionally it falls out. The brass insert is a loose fit onto the switch itself. The spring presses against the rear of the front panel, and stops the buttons popping out when they are released by the pressing of another.
I don't know why the favoured this approach. It does make the buttons feel a bit nicer than usual, but it must be very expensive and labour-intensive. When re-assembling, you have to remove the front panel (after first removing the sides and the knobs) and carefully install all the components on 17 (!) switches.
The front panel includes plastic inserts for the buttons to bear against. I did wonder about turning some around, so that the slightly discoloured outer edges would be hidden, but the inner edges already had small imperfections. Shows that someone on the production line was taking the time to select the best side to face out.
Even the side panels are works of art. They consist of two pieces, and are secured by screws covered by plastic plugs. Incidentally, all screws are stainless steel - which means they don't stick to your magnetic screwdrivers! Still, a nice touch. The more I worked with this old beast, the more the build quality won me over. It's just a lovely thing...
Really, the only thing I'd want to change is the mains wiring - as was common at the time, it's unshrouded. But, look at that mains IEC connector - screw terminals! I don't think I've ever seen one of those before...
Finally, here is the top view with the modulator PCB replaced, but just before the front panel was put back on:
Anyway, I hope this unit is interesting to some. As I say, there are almost no details about this anywhere on the web, so hopefully this thread rectifies this. As you can see, the build quality is really great, and access for service is also good (the bottom comes off to reveal the rear of the PCB).
Having spent quite a bit of time working through the calibration procedure, I think that it's now back on form. I had to think laterally for one or two steps, where the instructions call for a spectrum analyser and other obscure gear, but hopefully it's comfortably good enough for lining up vintage radios and decent hi-fi tuners. I haven't yet had a chance to play further, but will report back later.
All the best,
Mark
There's a lot going on there. The PSU is obvious enough - a nice quality toroidal transformer, as expected, along with standard 3-pin regulators. Next to that are 8 TTL ICs, and these are responsible for generating the various frequencies needed by the modulators (19kHz, 38kHz, etc). There is a row of test pins (between the tants and the 7400-series chips) that allow for quick checking (all were good in my case). If required, the trimmer allows you to fine-tune the master oscillator.
Moving left, the two ICs in sockets are CA3054 transistor arrays, arranged as modulators. You might notice that they are different to each other - that's because one had failed. Upon test, the output was basically a large 38kHz square wave - so much for suppressed carrier! Clearly something wrong with the 38kHz modulator. Luckily for me, the chips were socketed, and I had a couple of spares. Fault found

The other modulator is actually working at 114kHz, and this modulator is used to cancel out sidebands created by the 38kHz modulator. If this sort of thing interests you, have a look at the manual - it's fiendishly clever stuff
Moving left again, you can see the cores that make up the composite filter. Very complex, but ensures that the frequency spectrum at the output is correct. Again, check the manual for a well-written and lucid explanation.
Top-left is the output stage (LM318) - note the yellow wires connecting to the DIN socket and the Composite output. The SCA input is on the orange one. The peak-detect circuit for the meter is also here.
Bottom left, behind the switch-bank, are op-amps (LM356 and LM318) that are doing input processing. The modulator PCB connects to the empty IC socket.
The oscillator op-amp (LM318) is hiding behind its ceramic decoupler - the other two op-amps (LF356) are buffers for the L/R input signals. The right signal goes via the Ext switch; the left signal goes straight back to the main PCB. Note the thermistors for amplitude stabilisation - not just the large one, but there is a pair of smaller ones (that look a bit like germanium diodes) with paint dots applied for identification.
I did note that one of the electrolytic caps had some signs of leakage, and upon test the ESR had risen to a few ohms. I replace both for good measure with 105C replacements from Philips, but I didn't find any other faulty capacitors anywhere else in the unit. They are all very good quality units.
Note the date code - week 33 of 1981. Ancient, but in remarkable condition. It's clearly been looked after.
A tip for anyone who has one of these: be very careful when removing the PCB as the buttons are nowhere near as simple as you might think!
Each button has these 3 components. The brass sleeve is mostly a tight fit into the plastic, but occasionally it falls out. The brass insert is a loose fit onto the switch itself. The spring presses against the rear of the front panel, and stops the buttons popping out when they are released by the pressing of another.
I don't know why the favoured this approach. It does make the buttons feel a bit nicer than usual, but it must be very expensive and labour-intensive. When re-assembling, you have to remove the front panel (after first removing the sides and the knobs) and carefully install all the components on 17 (!) switches.
The front panel includes plastic inserts for the buttons to bear against. I did wonder about turning some around, so that the slightly discoloured outer edges would be hidden, but the inner edges already had small imperfections. Shows that someone on the production line was taking the time to select the best side to face out.
Even the side panels are works of art. They consist of two pieces, and are secured by screws covered by plastic plugs. Incidentally, all screws are stainless steel - which means they don't stick to your magnetic screwdrivers! Still, a nice touch. The more I worked with this old beast, the more the build quality won me over. It's just a lovely thing...
Really, the only thing I'd want to change is the mains wiring - as was common at the time, it's unshrouded. But, look at that mains IEC connector - screw terminals! I don't think I've ever seen one of those before...
Finally, here is the top view with the modulator PCB replaced, but just before the front panel was put back on:
Anyway, I hope this unit is interesting to some. As I say, there are almost no details about this anywhere on the web, so hopefully this thread rectifies this. As you can see, the build quality is really great, and access for service is also good (the bottom comes off to reveal the rear of the PCB).
Having spent quite a bit of time working through the calibration procedure, I think that it's now back on form. I had to think laterally for one or two steps, where the instructions call for a spectrum analyser and other obscure gear, but hopefully it's comfortably good enough for lining up vintage radios and decent hi-fi tuners. I haven't yet had a chance to play further, but will report back later.
All the best,
Mark







