This follow up to my post Homebrew 1970s Hifi Stereo 25w Amplifier and FM Tuner and details the thinking behind its design, construction and restoration of my home-brew FM Tuner. I cover several faults discovered when powered up after it was last used in the 1990s.
Warning – never operate mains powered equipment with the case open as high voltages can be exposed and cause injury.
Design approach
By the late 1970s when I began this project almost all the hifi FM tuners on the market in the UK had a large tuning knob with an illuminated dial allowing manual selection from a handful of radio programmes. Tuning was cumbersome — you had to remember the frequency and tune in carefully to get a stable stereo reception without noise and distortion. If you regularly listened to several different radio stations it became more tedious to select from an ever increasing set of different programmes.
There had to be a better way to operate a radio. So while still a student, I decided to design and build an FM tuner with a series of preset frequencies for fast selection of my favourite programmes. The idea was copied from the way most colour TVs operated which provided a set of pushbuttons, each preset to a different channel. My previous Red series tuner used a simple gang of switches, but I wanted touch control as I’d seen used on a Scandinavian TV.
Modular design was key to allow the design to evolve, similar to the approach I’d used for the R2525 amplifier. I sketched a rough layout of modules as I found from some early pencil notes below. You can see the later additions in marker pen.

FM Tuner Block diagram
The design concept was to build a series of modular circuit boards each providing a single function so it was easy to upgrade to swap out elements without disturbing the whole. Almost all the commercial FM tuners available took a different route with all the components on one large board. Of course it was cheaper to manufacture a single board.
Here’s how the design developed. (You may notice I’m operating the unit powered but the warning above still remains don’t do this unless you are qualified to work on live equipment as serious injury can result. )

FM Tuner Internal modular layout – top plate removed
Module layout followed the R2525 Stereo amplifier with mains power supply on the left and radio / audio circuitry to the right. Touch control and logic display drivers are positioned adjacent to the front panel.
Power supply
On the left is, what was, a conventional mains transformer mounted far away from the audio components to reduce hum. I used separate circuit boards for the high voltage and two low voltage regulated supplies. I wanted to keep the higher 200v & 100v lines away from the 12v and 5v rails used to drive the audio and logic. This helped prevent spikes or glitches from the high voltage lines interfering with the logic gates causing spurious programme switching.
A mains transformer enabled linear power regulators rather than switched mode designs that can generate interference with the sensitive RF circuitry. I reused a transformer from an old valve radio that had 6-0-6v and 0-300v windings.
Discharge tubes, glowing a faint purple above, provided the stable 200 & 100v DC references much like zener diodes do at lower voltages. I used two tubes in series for the 200v Nixie tube supply. Notice I used small neon tubes to provide a warm orange light that compliments the main Nixie tube in the centre. The mains power neon tube, on far left, is simply fed from the AC mains via a resistor as it sits above the power switch. The other neon indicators, used for muting and manual tuning, far right, are driven from the 100v DC supply. No AC voltages allowed near the audio circuitry. A small fibre optic strand separates the neon from the front panel to produce a point source indicator.
Channel display
I needed a display to indicate which touch pad was selected and chose a Nixie tube as we had several lying around the workshop from Dad’s digital clock days at Darang. He also supplied the CµL9960 high voltage display driver along with a FJJ141 decade counters which would form the basis of the preset tuner.

FM Tuner Nixie and Touch Controls
Here’s why I didn’t pick a more modern seven segment LED display. Touch control requires a memory to store the selected channel — 1 to 6 for the preset stations and 7 for manual tuning. The decade counter provided the memory and I used the Nixie’s digit output to gate the touch control’s counter. 50Hz AC from a winding on the low voltage transformer provided the counter’s input to each touch pad. This generated a pulse every 20 milliseconds to drive the decade counter. When the counter reached the appropriate touch control’s number a simple diode gate switched off the pulses and the count stopped.
Simple. But it required fabricating the logic gate and switch for each preset, and that meant seven times. You can see the discrete circuit below the ribbon cable on the right. Additional gates were needed to connect the associated potentiometer which provides the preset voltage for each frequency. Varicap diodes in the RF tuner provided the frequency selection dependent on input voltage from the touch controller.
So why not use a 7 segment display? Well apart from the cost of buying the LEDs and driver chips the output for each channel number was not decoded to a separate channel. For example ‘1’ needs two of the seven segments illuminated and ‘7’ needs three. So I’d require another decoder for each channel number! A Nixie decade counter already decodes each number to drive a separate anode in the display.
(circuit for decade driver)
RF Tuner and IF stage

FM Tuner RF circuit diagram
For the high frequency RF circuitry, I wrapped some tin plate to form a box with holes cut for manual alignment of RF and local oscillator stages. The RF tuner was mounted in tin to shield it from interference. The whole tuner was encased in aluminium, again to prevent unwanted RF interference in and out. Automatic frequency control is achieved via voltage control fed via a small circuit board containing the red preset potentiometer. Any unwanted amplitude change on this control voltage, such as ripple from the power supply, will be heard on the audio output due to frequency modulation of the local oscillator. So the small circuit was added to filter noise and optimise signal to noise ratio. More on this later.

FM Tuner RF IF Decoder
For the intermediate frequency amplifier I selected RCA’s CA3053 IF amplifier, an integrated circuit buffer driving a pair of 10.7MHz ceramic filters. Ceramic resonators become standard components in these applications, replacing the more bulky inductor / capacitor resonators, bought from Ambit International – a component supplier in the seventies. Following the ceramic filters another RCA device, a CA3089E was used as an FM detector.
Stereo decoder
I selected Motorola’s MC1310P stereo decoder for the final module as it offered low distortion and was the best I could afford at the time. It still sounds better than some modern FM stereo decoders, probably helped by the multipole filters I included to remove unwanted ultrasonic frequencies – known as birdie filters. My previous Red series FM tuner had a fully discrete stereo decoder from a circuit published in Wireless World. This was a pain to set up and align as I did not have an FM stereo signal generator. The MC1310P was much easier to use.
Finally a Motorola MC14016 analogue switch was incorporated for inter-station muting while tuning manually. This avoided the usually static noise you hear between FM stations. A squelch control on the front panel allowed a variable mute threshold with a push button to disable.
Faults
As the FM tuner contained high voltage electrolytic capacitors I checked for short circuits before first powering up. I then added series resistors to bring the voltage up slowly and hopefully re-polarise the electrolytic layer if they had weakened. All seemed fine with a supply of over 350v to feed the regulators.
Touch control worked with numbers 1 through 7 selecting as I touched the pads. Some even tuned to a station but most just produced the usual static with an annoying buzzing hum. The manual tuning did not seem to work – stations just drifting in and out of frequency accompanied by the hum. The stereo light flickered on indicating the decoder may still work but the tuning indicator refused to glow at all.
The buzzing hum
I decided first to tackle the hum as I thought it might be the easiest to fix but soon found it more challenging. The oscilloscope showed 50Hz ripple on the 12v supply line which should not be there. I’d designed a zener – transistor regulated 12v supply for low ripple to prevent modulating the varicap tuning voltage that was derived from the 12v supply.
I suspected an ancient grey RS electrolytic reservoir capacitor and replaced it with a modern 2200uF axial one. This reduced the hum, but it was still audible. So next I replaced the electrolytic on the regulator’s output – the green one below. Again, this helped lower the ripple but it was still visible on the scope and audible on programme sources. The remaining two electrolytic capacitors were obviously old and should be replaced for reliability but the scope showed no significant ripple – and bridging them with modern capacitor did not improve the hum.
So what was causing the 12v supply to be modulated and cause the hum? The 50Hz waveform was puzzling. If the ripple was due to the low voltage transformer winding it should have been 100Hz as I’d used a full wave bridge rectifier feeding the 12v regulator. Maybe one of the diodes had failed? I checked the bridge and all measured correctly.
Notice also how the variable resistor next to the on/off switch has a layer of powdery residue covering it. This was evident on several fixings as well and may be due to electrolytic leakage in the sealed box over the years. It’s odd as there’s no evidence of rusting from the steel fixings. If anyone has an idea what the corrosion may be please comment. Measurements proved the 25kΩ pot was electrically okay but I decided to replace it incase the residue causes further issues.

FM Tuner Power Supply with faulty 2000uF capacitor
Returning to the buzzing, my attention switched to the high voltage supplies to see if these were to blame. The large electrolytic above the transformer was an obvious candidate to replace although it measured fine. It was a double 32+32uF at 450v and I had no modern replacement so I connected a single 32uF in parallel with each. This made no significant difference to the audible hum.
I noticed a thyristor in series with the high voltage feed. Then I remembered the notes I’d found detailing a modification to reduce heat dissipation from the 200v supply. The transformer I’d used provided too higher voltage ~ 350v and so I’d added series resistors which heated up inside the sealed case. So I’d borrowed the idea of a light dimmer switch in an attempt to reduce voltage and power dissipation. I temporarily disabled the high voltage supply by removing a fuse and the buzzing stopped. This is what was causing the 50Hz ripple but in a strange way.
So why had the thyristor on the 200v supply started to cause the interference on the 12v rail? I’d made recordings in the 80s on my reel to reel Akai 4000D (subject of a future post) from my FM tuner and there was no buzzing hum present — how could the the interference have developed over the years of not being used?
I decided to revert the thyristor circuit with a simple resistor dropper so the audio was free from the hum and I could continue with other faults and return to the vexing problem later.
Manual tuning
I’d chosen an industrial 10 turn helical potentiometer for fine manual tuning over the waveband – one left over from my father’s store. Instrument quality components should be reliable but the ohmmeter proved the wiper was open circuit after years of not being used. This explained the frequency instability as the voltage to the varicaps drifts due to high resistance.
The helical pot was sealed and prevented application of the usual switch cleaner so I replaced the potentiometer with a similar multi turn helical I’d found in an old electronics shop in Budapest and stable tuning function was restored.
Tuning indicator
A quick check with a voltmeter identified the red LED I’d used to indicate signal strength had failed. It was over 50 years old and had likely suffered from over zealous soldering. Easy to replace and I could choose from a far greater range of colours than the red that was available back in the mid 70s. I decided to replace like with like and selected red to keep harmony with the Nixie and stereo indicator.
I adjusted the six channel presets on the back panel selecting various radio stations from BBC R2 to Smooth FM. Oddly, programmes above around 104 MHz would not tune in. Back in the mid 70s there were fewer FM radio stations available. I recall the usual sequence from 88 MHz were BBC R2, R3, R4 then Radio London, Capital Radio and LBC. The spectrum was not as packed and I guess nothing much above 100 MHz. Public services such as the police and ambulance service could be heard higher up the frequency range here in the UK.
I looked for a portable radio in the workshop and found a Beolit. It too stopped around 104MHz as did several Roberts radios. Some local stations like BBC Radio Surrey, Radio Kent transmit around 104MHz but there are several commercial stations up to 108 MHz. My FM tuner for now won’t receive these.
Finally the FM tuner was restored after so many years. Listening tests to various programme via the main house aerial sounded good and compared well with the JVC R-S5 I had as a comparison.
I will return to investigate the thyristor in the high voltage supply and see what’s needed to increase the frequency range another day.

R2525 Stereo Amplifier and FM Tuner
04/12/2024 at 4:19 pm
How long did it take to make the radio? Was there a kit like you could buy in the sixties?
04/12/2024 at 11:10 pm
In truth I don’t remember how much time it took to make, my guess is around six months. Back then I spent evenings and most weekends on projects along with school work and later degree course work. So over that time I had to make the ply and teak cabinet – say 5 days, chassis fabrication and front panel – another 5 days leaving around 30 days for the circuit boards and wiring. Will look to see if there are dates mentioned in the notebook to given me a better idea.
There were kits and prebuilt boards and modules available. Ambit International sold them along with some of the RF & IF components I bought. Heathkit also supplied a complete kit of parts including cases but these were too expensive for me.