I bought this set on eBay for £20.50. The seller was in Devon so we collected it on the way down for our summer holiday last year, rather than risking the postal service (and of course saving the considerable cost of P&P). The tuning scale is broken and the set is filthy - these points were noted in the eBay description. Upon closer examination when I got it home I noticed that there were in fact two cracks in the tuning scale, however the second one was not particularly visible, particularly with the amount of dirt, so the seller could be forgiven for not noticing it. The set was sold as "working" but I did not plan to just plug it in!
These sets are quite awkward to take apart. First the back is removed as usual. In this instance there was an additional switch on the back that was clearly not original, which was wired to disconnect the speakers from the output transformers - ouch! I cut the speaker leads close to the speakers, then removed the access plate in the base of the cabinet so the screening wire could be released.
The chassis is held by four screws from underneath, as usual. However, unlike most sets, on this model the chassis is removed from the front. The wooden trim in front of the piano key buttons must first be removed by undoing four woodscrews from below. Then the chassis can be carefully manipulated out. The mountings for the reverb unit on the back of the chassis foul on the speaker magnets, but by moving the chassis to one side then withdrawing the other side it can be manoeuvred out past them.
Remove the knobs from the tone controls straight away (grub screw fixings), as otherwise they take the weight of the chassis when it's placed on the bench and could be broken.
Speakers and transformers
The speakers in this set are 800-ohm impedance, so a good first check is to test these with a meter - they will measure around 600 ohms (DC resistance is generally lower than AC impedance). Also check the output transformers. In my set the transformers had seven contacts each and the wiring didn't seem to make sense compared to the circuit (the circuit I had was for a slightly different model), but it all looked original and the transformers were OK as far as I could tell. If any of these parts are faulty you will probably have problems obtaining replacements.
I then removed the volume and tuning knobs (push fit), the top brass trim and the tuning scale. It then all became clear! The previously unnoticed crack had been repaired in the past, presumably with superglue or glass adhesive. Both cracks originated from the same place just above a tone control, and behind this was some badly melted and burnt plastic where the electro-luminescent backlight used to be. This had obviously failed rather dramatically and the heat had caused the cracks in the scale.
A previous repairer had fitted a piece of white card in place of the backlight and added a couple of normal 6.5V bulbs either side, connected from the heater feed to the tuning indicator and mounted on a couple of bits of stiff wire. A reasonable fix, but I'll try to devise something better, later.
I cleaned the broken edges of the tuning scale with isopropanol, and then laid the scale print side down on some tissue paper on the bench. With the pieces carefully pieced together I applied a strip of clear sellotape along the crack. I then turned the scale over, lifted the crack to open it a bit, and applied some B&G glass adhesive. I then placed the scale on the windowsill, print side down, propped on pieces of wood etc so that it was flat and the glue on the underside wasn't touching anything. This adhesive needs UV light to set, which is why the scale had to be placed with the clear side upwards. The windowsill is not in direct sunlight, and there isn't a lot of bright sunlight in March anyway, but I hoped it would set in a few days.
From a brief examination, things looked promising. The top of the chassis was dirty, but there was nothing obviously amiss. The underside was excellent! Since this is quite a late set, and also no doubt an expensive model, the components seem to be of good quality. The capacitors are those moulded beige coloured ones, which are normally very reliable. Many of the smaller electrolytics looked rather unwell, but I decided to come back to them once the set was basically running. On one end of the chassis, the tone corrector capacitors on the two output transformers were those horrible black ones that Philips seemed to like. One had a piece of casing flaking away. I replaced both of them using 1kV polypropylene types.
I connected my capacitor reformer to the main HT smoothing electrolytic. The current remained high and the voltage remained low. On closer inspection the end of the can was swollen and cracked. I wasn't going to mess around trying to reform this capacitor - it was clearly beyond redemption. The original capacitor had three 50uF sections. I had one remaining blue LCR double 50uF can in stock. I decided to use this for the first two sections, and a small radial PCB capacitor mounted underneath the chassis for the third section.
There were a number of connections to the original capacitor so I drew a rough diagram before cutting the connections close to the tags. The new capacitor was fitted in a mounting clip in place of the original, and the screw holes in the clip just about lined up with the fixing holes for the original. The connections were then remade. The third capacitor was mounted on a small piece of tag strip, mounted by a convenient nearby screw. I used a radial (PCB mounting) component since this fitted into the space and onto the tag strip easier. The leads were kept as short as possible, and the mounting is rigid enough.
I was now in a position to show the set some electricity and see what would happen! Because of the high impedance speakers, I could not use my usual 3-ohm test speakers this time. I extended the leads with a couple of lengths of mains flex, to reach the speakers in cabinet where I had stored it under the bench. I connected a meter across the main HT smoothing capacitor, and a length of wire to the FM aerial socket.
With the power applied, there was a longer than normal delay before the HT started to rise. By luck the set was tuned to the local BBC station on FM. Reception was fairly good, although there was a lot of crackling as the controls were adjusted. The HT was about 210V whereas the service data shows about 280V. The AC to the rectifier seemed about right so I fitted a replacement EZ81 rectifier valve, which corrected this problem.
Listening closer to the speakers and placing a finger on the cones showed that one channel was doing much more than the other. Checking the voltage on the anodes of both EL84 output valves proved that one valve was doing very little and the other was doing a lot. I replaced the less active one with a good spare, which improved the situation, but the anode voltages still differed somewhat. Since the two valves share a common cathode resistor it is clear that the two valves need to have similar emission. My Mullard High Speed Valve Tester does not give accurate emission levels (it's more of a go/no-go tester), so I tried various EL84s in both positions (I have a number of used ones) to find a pair that matched reasonably well. With the pair finally chosen, the anode voltages were within 1V of each other.
While listening to the set, it was clear that something was still not quite right. The problem is difficult to describe, but basically the amount of treble being reproduced seemed to be affected by the amount of bass in the programme. I took another look at the state of the assorted electrolytics (swollen ends, crumbling deposits etc), and decided to replace all of them before going any further with faultfinding. Fortunately my employer had been having a major clearout of old components from the stores, so I had accumulated a reasonable stock of axial electrolytics in various useful values! Some of the originals had odd values, but I just replaced them with the next highest preferred value (for example 80uF was replaced with 100uF and 16uF with 22uF).
The set now sounded much better. Not only was the weird tone problem gone, but the overall quality was also greatly improved. The volume, bass, treble and balance pots all crackled, but responded well to a shot of Electrolube "EML" contact cleaner.
I tried tuning right across all four wavebands, and everything worked as it should. There were no signs of tampering and the performance was good, so I left the IF and AM RF alignment alone.
Although the FM RF alignment was clearly also correct, I wanted to receive Saga FM on the set. The tuning scale is marked from 87.5 MHz to 104MHz, but the pointer moves past both these figures, and covers a range of about 87 MHz to 105MHz. Saga FM (West Midlands) is on 105.7MHz, so the plan was to shift the alignment by about 1MHz, which would bring in Saga at the top end without losing Radio 2 (88.6MHz in this area) off the bottom.
The service data does not contain any alignment instructions, but there are only two controls on the VHF tuner assembly. Having applied marks with a fine permanent marker to indicate the current positions, I tried carefully turning each of them in turn. One affected the signal strength (as indicated by the tuning indicator) and the other caused a shift in scale calibration, as I had hoped. With the adjustments in the original positions I tuned in Radio 2, and then turned the tuning control sufficiently to move the pointer about 1cm (about 1MHz on the scale). I then adjusted the alignment adjustments to tune in Radio 2 again and to peak the level on the magic eye. I then connected the aerial socket to the large roof aerial and booster/distribution amplifier we use to receive Saga (we are outside their primary coverage area), and tuned to the other end of the scale. As I hoped, there was Saga, close to the end of the scale - I could tune just past it. Reception was excellent! The high signal level from our aerial/booster on Radio 2 etc did not bother the radio either.
The 75-ohm aerial down-lead was connected with the core to one of the 300-ohm serial sockets and the screen to the earth socket (intended for the AM aerial). A balun would probably improve the matching, but these are not readily available in the UK now.
With further use I found that a 6dB attenuator in the aerial lead improved the sound quality slightly (clarity of the treble), and reduced the background noise between stations. I also tried a 12dB attenuator, but this reduced the signal strength on Saga too much.
"Magic Eye" tuning indicator
Although the EM80 tuning indicator worked, it was rather dim, so I fitted a good used EM81 instead (I did not have any good EM80s). Although this was much brighter, it did not quite fully close even when tuned to a strong FM signal from my booster amplifier. This is probably due to the differing sensitivities of the two devices. I decided against any modifications, so that if an EM80 was fitted at a later date, it would work correctly.
From examining the circuit, it is clear that sound quality was a higher priority than price when this set was designed. The negative feedback and tone control sections contain dozens of resistors and capacitors that would never be allowed in a cheaper set. The high impedance speakers must have been more expensive to produce than standard 3-ohm types too. There are four stages of HT decoupling to prevent any hum getting through. Good quality components have been used - no wax-paper capacitors here!
The waveband switching on the circuit diagram initially doesn't make sense. However once you examine the switch itself, it all becomes clear. The contacts behind the MW button are not operated by that button. Instead there is a pivot arrangement to operate them (in the opposite direction) when the VHF button is pressed. There is a zigzag arrow on the diagram, presumably intended to indicate this. Therefore the contacts behind the MW and VHF buttons are used for AM/FM switching. The contacts on the LW and SW switches select the AM waveband.
Just to add confusion, the only circuit diagram I have is for an earlier version (without the /65 suffix). The waveband buttons are in a different order on this, with the LG (presumably LW) switch contacts linked to the FM button! Does anyone have a diagram for the B7X14A/65 version?
The tuning mechanism is, of course, typically Philips, but nicely done. There are separate drive systems and pointers for AM and FM, and a linkage on the VHF piano key causes the tuning knob and flywheel to engage the appropriate mechanism for the band selected. It works well too. The VHF button has a lot to do, and therefore needs a firmer push than the others.
Another nice detail is that the HT to the triode audio amplifiers is switched by the mains switch, so that the set is silenced straight away when switched off, rather than fading away.
At least, that's the theory! On this set, pressing the OFF button gave me MW. The mechanism for operating and releasing the power contacts did not work. I started by spraying some contact cleaner into the switch, in case something was sticking, but that didn't help.
The piano-key switch bank is fixed to the front of the chassis with two screws, and there are masses of wires and components connected to the contacts, as expected. I decided straight away that I was not going to disconnect all of these, as the risk of causing further damage was too great. So if I couldn't resolve the problem with the switch connected, I would just bypass the mains switch and provide an alternative switch (probably in the mains lead) instead.
With the chassis propped on its back edge, I removed the shield around the mains switch connections, and then removed the two switch fixing screws. This allowed the switch to be moved about 12mm out without significantly straining the connections. The end of the switch bank was covered with a piece of Paxolin, which was removed by straightening the relevant tags with long nosed pliers.
By looking into the end of the switch bank, the cause of the problem was visible. There were two adjacent plates, one linked to the button and the other linked to the moving contacts, each of which had a small hole with nothing in it. By examining and operating the mechanism it was clear that there should be a V-shaped spring between the two holes, pushing them apart. Then as the button is pressed and the holes pass each other, the spring gives an over-centre type action that operates the contacts with a snap.
Like many radio restorers, I have an accumulation of junk in the loft that "might come in useful"! Amongst this was a 1970s top-loading cassette deck, which was duly dismantled in the search for suitable springs. I didn't have to look far - the cassette lid spring looked suitable. It took some fiddling to get the spring to fit the switch mechanism and remain in place, as I had to shorten the arms then bend small hooks in the ends. After a certain amount of fiddling and swearing, it was fitted and worked well. I applied a small amount of grease for lubrication. It seemed fine after around 40-50 operations, but I was concerned that it could come adrift and become lodged somewhere. I threaded a length of tuning drive cord through the loop in the spring and round a convenient tag on the chassis, then tied the ends leaving some slack. Now, if the spring does come adrift, it won't go far and I will be able to find it.
While I was working in this area, I replaced the mains cable. I used three-core cable and soldered the earth to a chassis tag on a nearby tag strip. Before doing this I checked the insulation between the mains transformer primary and chassis with a Megger, which showed open circuit.
Before cleaning the chassis, I turned my attention to the asbestos lined heat shield between the output valves and the mains transformer. The asbestos was discoloured due to the heat of the valves and looked like it could easily crumble if disturbed. I removed the valves to clear some space then carefully released the heat shield (two screws) and lifted it out. I wrapped the complete heat shield in several plastic bags secured with tape, put it in a box, wrote ASBESTOS in large bold letters on the box, and left it on top of the dustbin for the refuse collectors. They took it. Whether they kept it separate or put it in the crusher I don't know - but I doubt that this small amount of asbestos (a 75mm square piece about 1mm think) would do any harm in the landfill.
Now, of course, I needed a new heat shield! I made this from another item in my accumulation of useful junk - the top cover from a defunct 3.5" floppy disk drive. The metal is thinner than the original, but this was helpful in my home-workshop situation as I could cut it with sharp scissors. The convenient 90-degree fold along one side meant I didn't have to bend the plate for the mountings.
The repaired tuning scale had been sitting on the windowsill for over a week, and the glue had set as hoped. The repair is visible, but not as much as it would have been if I had used superglue because the refractive index of glass glue is close to that of glass. I cleaned the front of the scale with foam cleaner then used Autoglym car windscreen cleaner to remove the stubborn sticky patches. Surplus glass glue on the front was carefully scrapped away using one of those scrapers intended for removing paint from window panes.
The printing on the rear of the glass was fairly solid, so the clear parts were very carefully and gently cleaned using cotton buds moistened with water. Note that great care must be taken if attempting to clean the rear of tuning scales - in many cased the print will be soft and poorly adhered, making cleaning without causing further damage impossible.
By holding the scale up to the light, areas of damaged print along the line of the repaired crack were visible. I touched this up with black and gold enamel model paint, as appropriate. There were a few other minor chips and blemishes in places, which I also touched up.
The original electroluminescent backlight would have been a flat rectangular section that acts as the backing plate behind the tuning pointer. This would have fitted into plastic supports at either end, one of which had suffered some heat damage as mentioned previously.
I cut a piece of glass (from an old photo frame) to fit in place of the long-gone backlight. After careful cleaning with white spirit, I sprayed one side with glass etch spray. With the glass fitted temporarily in place (with the sprayed side to the rear), the tuning scale glass also temporarily positioned, and the dial lamps held behind, the etch effect did diffuse the light somewhat but not as much as I had hoped. I cut out a piece of translucent plastic from the lid of a component storage box to fit in behind the glass, which considerably improved the effect. The mountings for the original backlight are wide enough to accommodate the thickness of the glass and plastic together.
I moved the lamps around to get the best effect without any shadows from components. Having established the positions, I made two mounting brackets from the leftover bits of the floppy drive cover. One was mounted under the same screw as one of the beehive trimmer capacitors, and the other was mounted on a convenient spare hole.
The lamp holders were the ones fitted by the previous repairer. However I rewired them tidily and took the connections back to the relevant tags on the mains transformer. I also fitted a couple of new 6.5V 0.3A bulbs as the existing ones were somewhat blackened. I replaced the bulb used to light the green Philips logo at the same time.
This repair is obviously not as good as the original electro-luminescent backlight would have been, but it's a considerable improvement on the piece of cardboard fitted by the previous repairer. I don't actually know if the original backlight was white or some colour, but it looks OK in white to me. I have seen one mention online that it was green.
The valves were removed, then the chassis was thoroughly brushed using a selection of small paintbrushes to remove all the dust. There was an even film of dirt on the whole chassis, however there were many delicate parts on the top of the chassis so I decided that it wasn't worth risking damaging these by cleaning. If I were to just clean the more accessible parts, the result would be patchy and would look worse than not cleaning at all. So I just cleaned the valves and left it at that. Nobody can see the chassis when the set is assembled anyway!
The buttons on the two switch banks and the balance control knob were cleaned in place using foam cleaner. A couple of the buttons were loose so these were removed for cleaning then fixed back in place with superglue.
Cabinet and trim cleaning
The plastic speaker grills on the ends of the cabinet were absolutely filthy, due to nicotine. I removed them (they are held with two small nails at the back and pushed into a slot at the front) for cleaning.
I put them in a sink of warm soapy water and started brushing them with an old nailbrush. The water rapidly turned the colour of black tea. If this is what years of cigarette smoke does to inert plastic, imagine what it does to live lungs! With the water changed I finished cleaning the speaker grills, and also cleaned the volume, tuning and tone control knobs.
I cleaned the brass strip from above the tuning scale using Brasso, taking care not to damage the small self-adhesive MADE IN HOLLAND label below the PHILIPS badge.
I brushed and vacuumed all the dust out of the cabinet and from around the edges of the speakers. I then cleaned the varnished sections with Rustin's Surface Cleaner. This is designed for removing grime and wax from varnished wood, and did an excellent job of removing the nicotine and other muck. The varnish underneath was in excellent condition (probably protected by the dirt) and needed no further polishing. There were a few chips and the odd small scratch, so I applied Rustin's Dark Scratch Cover to these, which made them tone in with the colour of the finish so they were less visible.
This is when you know you are nearing the end! If all the previous work has gone to plan, reassembly should be uneventful, as indeed it was in this case. I replaced the rubber pieces on the tuning scale mounts with pieces of sleeving.
There were originally various pieces of black foam along edges where pieces fit together, and these had all crumbled away and been removed during cleaning. I replaced many of these with pieces of draught excluder, in some positions blackened with permanent marker pen.
This is an excellent set for daily use. The sound quality is, I think, better than many modern midi-systems. The set is now installed in our spare bedroom/study and provides entertainment from Saga while scanning piles of service data!
As discussed on the previous Reverbeo repair, the reverb feature is not worth using. I haven't found any music that is enhanced by it.