Item nr.

Philips FX Experimentation Project

Junk chassis became communication receiver; experimental set, learned a lot

Data for Philips FX
ProductionSouth Africa (and The Netherlands), 2016.
Price was 20€.
BandsMW (518-1606kHz), TB (1597-5060kHz), SW1 (4962-14535kHz), SW2 (14612-24995kHz); IF 455kHz. FM (71-108MHz).
TubesEF42 (RF pre), ECH81 (conv.), EBF80 (IF and det.), EBC41 (phono preamp, removed), EBC81 (audio preamp), EL84 (output), EZ80 (rect.), EM80 (tuning eye, now EM84).
1N4007 (DC supply), freq counter, pocket radio, NE555 (osc.).
CabinetThin wood. Size 49x18x15cm (chas), 61.5x25.5x16.5cm (cabinet).
PowerAC 220V, 52W.

The Design

To experiment with digital frequency readout, I searched for an old chassis on which I could mount a Frequency Counter. I wanted the chassis to have wide Tropical Band and SW coverage, and preferrably be AC powered (have transformer). I considered using a BX594A (no TB), but then found this old chassis, very suitable and very affordable, that also covers the Tropical bands and is quite spacy to work on. The chassis includes a front-rotatable ferrite antenna for aerial-less MW reception, and has an extra RF amplification stage.

Witnesses by the names on the dial, this radio was intended for the African market. I have not been able to identify the type exactly, but found some useful information. The seller of the chassis informed me that it came from a large console including a pickup, which was confirmed by the presence of a pre-amplifier (EBC41). Comparing the chassis to various Dutch models, I expect the luxury class to be 6. The chassis looks like it was actually made in South Africa and date marks on the coils show they were made in 1956, so the model year of this chassis could be 1956 or 1957. This leads me to a guessed type number like FV665A.

I got the chassis almost for free, with the dial glass damaged beyond usability (no problem for my purpose) and decided to make this a five phase project:

Obtained5/2016 from Harm Schuring; sn=43430.
Sound samplePLAY SOUND   On pirate station Europe24, on 6150kHz, I alternatingly hear German and Dutch (Studio Twente).

This Object

Phase 1: Repair

The repair of the radio chassis was not difficult. I replaced all tar capacitors (all being very leaky and way out of capacity) and the cathode elco for the EL84. I measured the tubes and found the EF42 to be dead (interrupted heater) so I put an EF41 from my stock. Then the radio produced some sounds, I measured voltages and found the mixer screen voltage to low, replaced a resistor, and everything worked. Unfortunately, on MW there isn't much to listen in anymore, but the Tropical Band on this set starts from 1600kHz, and I could hear some nice Dutch pirates there.

Phase 2: Frequency readout

While testing the repaired set, the dial chord broke, so I smoothly entered phase 2 of the project. I removed almost everything that reminded of mechanical frequency indication, and reduced the dial drive strings to one, driving the variable capacitor from the tuning axis. I soldered a small DC supply from a diode and pi filter, and connected a frequency counter. The input of the counter goes to pin 7/9 of the ECH81, connected there through 22p capacity. The counter is programmed to display the measured frequency value minus 455kHz (because the oscillator frequency is always receiving frequency plus intermediate frequency). Installing this took me about one hour and the effect is stunning! Never in my life have I been able to tune a tube radio so smoothly and so precisely! The outer knob behind the main tuning knob is a fine control, allowing to set the receiving frequency within about 0.1kHz to the desired value, even without listening! Of course this makes station seeking and identification very easy, but in addition it allows you to pre-tune a station even before it comes on air.

I was not very satisfied with the EM80 tuning eye, because it had lost all brightness, as most of these tubes have. My junkbox had nine pieces of EM84, so I rewired the socket and inserted an EM84. This gave quite visible tuning indication.

Phase 3: FM Unit

As FM unit I chose a small print of a Peansonic kit. It gets its current from the DC supply of the counter, and its output is connected to the phono input jack. So the original PHONO button now activates FM. Before I made the cabinet, the FM unit was taped to the filter caps, but in the final layout, it was on the top side of the cabinet with large brown tuning wheel in the front.

Phase 4: BFO Unit

The chassis had an unused EBC41 amplifier, that I considered for use as a BFO. But then I found this small NE555 oscillator that could be deliverd to my house for 68cents. What, 68 cents for an oscillator including delivery from China? In a country where it costs 74 cents to send a letter to the other side of the street? Well, that certainly was not expensive. The fequency of this oscillator is controlled by two on board precision pots, but what I want is a BFO controllable from the front over a small range. So I experimented a bit with a 47k Linear switching pot and resistors. I found that the frequency can be nicely controlled over a 7kHz range if the potmeter is placed in series with a 47k resistor, parallel to one of the controlling pcb potmeters. Then I installed this unit in the chassis, fed from the DC supply that also feeds the display and the FM unit. Unfortunately, the results of this setup were very poor. I increased the filter capacity of the DC unit to 2200uF to make the oscillator more stable, and found a hint to connect the oscillator output to the shield around the ECH81 for coupling.

This still does not make a great BFO as compared to factory receivers, but I received several ham operators in the 160m, 80m, and 40m band (LSB) and 20m band (USB). One problem is that the stations may come through too strong: then the signal in the IF strip is stronger than the BFO and the sound will be very distorted. I could improve performance with an attenuator, simply a 500klog potmeter placed between the antenna and the entry. When the signals are so strong that the BFO has almost no effect, I reduce strenght with the attenuator, and sometimes still get audible results.

Phase 5: Cabinet

The old chassis has almost the shape and size of the ubiquous B5X43A-series, so I considered finding one of these cabinets. But the chassis is just a little bit wider, and will not fit such a cabinet! In the local thrift store I found a DVD rack of the right size. I had to enforce it in the corners, and saw and drill a few holes. A piece of white board forms the front, with holes for all controls and the display and tuning eye. There are two small loudspeakers in the sides, from a set of PC speakers.

How to pull the chassis

To remove the chassis from the cabinet: (1) Remove five knobs; original ones can be pulled, BFO has screw. (2) Unscrew the four screws in the corners of the front and take away front panel. Wire from antenna attenuator should be taken out through hole in the back. (3) Unsolder five leads on top of frequency display. From left to right: black, blackwhite, white, orange, green. (4) Remove bottom screws (four). (5) Take out cabinet from front, power chord comes through hole, FM unit remains in cabinet.

What I received

I hear about eight MW stations during day and more during night. I receive MW pirates between 1611 and 1670kHz. I overheard hams on the 160m, 80m, 40m, and 20m amateur bands. I receive a plethora of SW stations.

The image rejection is very poor: I can receive some station on its proper frequency, tune down 910kHz to receive it as image, and hardly hear the difference. This illustrates the limited succes of a single tuning circuit to perform its primary task, namely image rejection. I can log stations very easily and accurately, and tune back to stations in seconds. The dial turns very lightly, because the knob only moves the variable capacitor, and no longer any strings to move the dial pointer.

Most AM listeners will know the sharp, distorted sound of misaligned tuning. My digital dial allows me to precisely tell, how much mistuning gives this distortion. Roughly, on MW, tuning withing 1kHz to the station frequency gives good sound, but tuning farther away gives distortion. SW stations are a bit more narrow so they fit in the bandpass more easily. On SW you can mistune 3 to 4 kHz without serious distortion.

Given the deplorable state all the hardware is in, repairs are needed regularly. Following a lecture in 2017, I had to restring the rotating antenna chord. In March 2018, I replaced a burned resistor. I remeasured the tubes and found them in a similar state as in 2016. Early 2019, the power switch failed so the radio couldn't be switched off properly. In May I opened the radio and found a mechanical connection between lever and switch missing. Remarkable, given that the switched worked fine for two years. I screwed a metal hook to the lever and the switch worked again.

Part of Gerard's Radio Corner.
Generated by SiteBuilder on 4/6/2020 by Gerard (g.tel@uu.nl)