Philips 643A "luxury" radio chassis restoration

[dazzlevision]

I recently acquired this radio, which has a rather tatty cabinet, as a source of spares for one that I already have (which has a very good cabinet). This was probably Philips’ flagship model at the time, as the cabinet is very elaborate, and the chassis has several uncommon features:

Rotatable ferrite rod aerial for LW/MW reception
Elaborate system for selection any combination of internal or external AM and VHF aerials
Extra IF stage on VHF/FM (EF85)
Dual cone loudspeaker
Output stage muting circuit at switch-off
Very complicated tone control system, which also automatically compensates for FM and SW reception tonal requirements
LW image rejection filter
Thermal fuse in the primary of the mains transformer.

The underside wiring of the chassis is the usual Philips “rat’s nest”.

The chassis had a goodly layer of fluff and dust on it, but no corrosion. After blowing out the dust and fluff with my trusty Electrolux 345 cylinder vacuum cleaner, it looked a lot more presentable. All the valves were Mullard and looked to be the originals. If my memory is correct, the chassis in my restored 643A is somewhat tarnished and a little rusty in places, so I thought I’d restore this one and do a chassis swap.

I found the piano keys were very sluggish and erratic in operation, due to dried out grease on the slide bar behind them and various other moving parts that require lubrication – a typical complicated Philips mechanical arrangement.

Once I had sorted that out, I applied mains to see what would happen (using the heatercathodeshort method). There was some life on VHF but nothing on AM. I then applied some Servisol to the band switches operated by the piano keys and “worked” them for a few minutes. I could then tune in some AM stations on two bands but not Long Wave.

The chassis had a goodly number of those (by now very leaky) Philips made black pitch covered paper dielectric capacitors. One in particular will give the EL84 a hard time if not replaced.
I powered the set down and began replacing them, starting with the one coupling audio from the EABC80’s triode anode to the EL84’s control grid, then the one in series with a resistor across the audio output transformer’s primary and finally one between chassis and one of the windings feeding the EZ80. Where possible, I fitted capacitors from the well regarded Mullard C296 Polyester dielectric range – often referred to as “Mullard mustards”. They are less visible than the “canary yellow” types available today.

I then moved on to replace all the others, switching the radio on at each stage, to see what effect it had on performance. I then replaced the ratio detector’s reservoir capacitor (down from 10 to 2uF) and the electrolytic across the EL84’s cathode bias resistor, which was made by BEC (Plessey) and was the only component in the radio that had an immediately readable date code - Apr 55.

[dazzlevision]

I noticed one historic repair. The HT feed to the ECH81’s mixer stage anode circuit was through a 2.2k resistor, decoupled by a 4.7nF ceramic capacitor. The original capacitor had gone s/c, cooking the resistor and both had been replaced. The new resistor was a 1W type and the 4.7nF had been replaced by two 3nF RS tubular ceramics wired in parallel. I replaced these items with the correct parts.

Philips used a lot of green body colour resistors in this chassis, which I find do not tend to drift out of tolerance, saving some time and work! However, not all were this type and so I had to replace a couple of (rather small) and well out of tolerance carbon composition resistors.

In the anode circuit of the EABC80 triode section, Philip fitted three large Erie “Salmon Pink” high stability resistors and the 10M grid leak was also a Welwyn special.

The set was still failing to work on LW, which was cured by replacing the ECH81 (lazy local oscillator). This also greatly improved AM and FM sensitivity, which was further improved by a replacement EBF80 IF amplifier valve.

It was at this stage that things took a turn for the worse. AM became intermittent again, sometimes going totally silent and sometimes with weak reception marred by heterodyne whistles.

After a lot of investigation, I found this was due to two issues with the SW F1 rotary switch, which is directly operated by the VHF/FM piano key. Referring to the circuit diagram extract, the switch sections SW F1 4 to 2 (which couples AM aerial circuitry to the control grid of the ECH81’s mixer section) and SW F1 11 to 9 (which connects the local oscillator section of the tuning gang to the triode section of the ECH81) were very intermittent and I had to use a fine “emery cloth” to remove the heavy tarnish on these sections.

However, there was one other fault with this switch. On all AM bands, winding S27 (on the 10.7MHz IF transformer coupling the mixer anode to the EBF80 pentode’s control grid) is short circuited by SW F3. SW F3 is mechanically coupled to the main FM switch (SW F1) by a length of drive cord. The cord is secured to a small moving metal part on SW F1, which is at chassis potential. This part was touching SW F1 contact 11 in the AM bands position, thereby disabling the local oscillator! I had to use a very fine pair of flat jawed pliers to slightly bend the fixed contact’s raised edge enough to give a reliable clearance. I now had reliable AM operation – phew!

The expensive and elaborate radio chassis with its dual cone speaker is now working very well and FM performance is particularly good.

[dazzlevision]

Here's a link to Snellings museum site, so you can see what the radio looks like (mine has a better cabinet):

https://www.snellingsmuseum.co.uk/ar...ips-model-643a