Reading Simon Reynolds Retromania – Pop culture’s addiction to its own past – nearly stopped me from writing this post. However, well into Retromania Part 1 I realised nostalgia can inspire as well as expire. So I hope others are encouraged to try making things for themselves driven by passion and enthusiasm to create rather than buy someone else’s efforts.

Why build a stereo?

Back in the mid 1970s, after several visits to the Audio Fairs held at London’s Olympia, my passion for hifi sound pushed me to build my own hifi systems. I still have one teak cased gold panelled amplifier and radio today so thought others may be inspired by what you can achieve if you have the passion and drive at an early age.

At sixteen, I could not afford the hundreds of pounds needed to buy the much admired equipment I’d seen at Olympia but I was lucky enough to have access to my father’s workshop with tools and test equipment to hand.  I loved loud rock music, Pink Floyd, Black Sabbath, Led Zeppelin, etc. and hated distorted sound. I borrowed a copy of Mullard’s Transistor Audio and Radio Circuits for Radio Receivers, Radiograms, Record Players, Tape Recorders and Hi-Fi Equipment from the local library and devoured the chapters on high-quality audio.

Mullard’s aim with the book, like other semiconductor manufacturers cook books, was to promote designs using their devices to increase sales and company profitability. But Mullard’s approach had more depth. Chapters, such as Layout of Audio Circuitry, were devoted to improving audio design and construction regardless of whose transistors and components were used. This detail is critical to achieve quality and made a huge difference in the final results.

R2525 Amplifier modules with regulated power supply

In truth the R2525 project was the second stereo system I’d constructed after selling my first ‘Red series’ amplifier and FM tuner via Exchange and Mart − the pre Internet eBay. The small ad publication had been crucial in sourcing and selling to produce £70 needed to fund R2525. Saturday work at the local supermarket had provided the little income needed to buy the Red series components. Thank you Waitrose.

While the Red series was a success, in that it sold to a nice Australian couple who’d driven all the way from Earl’s Court on a moped to audition the system, I knew I could do better. I’d shamelessly copied designs from Practical Electronics and Practical Wireless to construct the 15 watt amplifier and separate preset FM tuner. Paired with a Garrard SP25 Mk2 and a pair of bookshelf speakers, the Red series system sounded reasonable but not quite hifi. It suffered from annoying clicks and interference from the mains every time loads switched on and off in the house. And it hissed more than the cat when the volume was turned up to my kinda levels. I needed to improve.

So with Mullard’s chapters on good earthing procedures and low noise devices I felt confident the R2525 would be hifi. As you can see from above, my design was modular from the outset. Crucially, the preamplifier was housed in an inner sanctum of aluminium to shield it from unwanted electromagnetic interference.  Steel plate, nearest the power supply, reduces the transformer’s inductive fields to lower hum levels.

Design

R2525 front panel and case

Case

Teak veneered cabinets were de rigueur for hifi in the seventies, rather like flares from the sixties. So I cut sheets of 3/8″ plywood, butt joined and covered three sides in teak veneer bought from a warehouse in London’s docklands. Underneath I sprayed matt black. Today teak is endangered and rightly restricted supply but at naive sixteen it seemed essential to achieving that hifi look. There was a cheaper vinyl plastic alternative from Woolworth’s, environmentally just as bad, called Contact with a wood grain effect. I used Contact on a loudspeaker cabinet but it never really looked the part and blistered in direct sunlight.

As other wood, SaRaiFo (Save the RainForest), veneers are now available a few words about the technique. Mastering veneer taught me to slow down and be patient. In reality I remember getting frustrated by how easy it was to tear an edge and have to take it off and start again. The trick I found was to be liberal with the PVA glue on veneer and cabinet ensuring an even coat. Always oversize the veneer, overlap edges and trim after. Veneer has a tendency to curl up when you put glue on one side so you need to dampen the non glue side first and have somewhere ready to hung it up to dry. When the glue is dry on both veneer and cabinet you can iron it in place, taking care not to split the sheet by apply too much heat.

Front panel

The front panel, just like the front window of a shop should please the eye and shout “buy me”. Many hifi manufacturers used matt silver aluminium as their front window with a silk screen dressing for their brand logo and control lettering. This was an obstacle for a sixteen year old with no access to silk screen printing or metal plate shop. I tinkered with aluminium sheet in caustic soda – health and safety alert- don’t do this at home without adult supervision and in fact don’t bother as it didn’t work.

I found a stash of pre anodised aluminium sheets my father had acquired as samples when he was involved in instrument manufacturing. These were brightly coloured in blue, red, green and gold. I selected the red sheet for my first hifi design hence the red series. As there wasn’t a silver sheet, I figured gold was the next best choice so proceeded to slice off a strip for the R2525. You can now buy various anodised aluminium offcuts online for a few pounds but in the seventies supply was difficult to find.

Letraset rub down lettering provided the text and a professional looking finish over the aluminium. I selected Eurostyle extended for the main typeface as it looked modern with Helvetica for some smaller text. Letraset also sold various thickness lines which I used to try and link related controls and visually reduce the panel’s overall height. Lines provided the stylised “R” logo I designed to place along the lower section. This also filled a gap and created the symmetry.

Like veneer, Letraset needs practice to achieve good results. It works best on soft paper where the waxy transfer sticks well. On shinny aluminium results can be hit and miss with partial adhesion and missing sections of a character or line. Persistence and patience pays off! Decadry was an alternative to Letraset, although only available in smaller sheets, so common letters often ran out needing another sheet and ended up more expensive to use.

Rub down lettering was designed for photographic use and is not robust enough for everyday use. It needs varnishing – Letraset sold Letracoat 101 – a satin finish that could provide some protection. I used this on the rear panel – you can see where it has browned with age. To protect the Letraset I cut a perspex overlay sheet. Perspex or acrylic sheet is very brittle and can chip or crack so needs care when cutting and drilling. But it’s much more durable than varnish so the lettering is still legible after 50 years unlike silk screen panels that often wear thin.

Power supply

I could not find an affordable toroidal mains transformer so the R2525’s overall height was dictated by the large 3A 25-0-25v iron cored beast shown below. This together with a 4700µF smoothing electrolytic with 52v regulator and protection circuit powered the left and right channels. The preamplifier has its own 20v regulated supply fed from the 52v supply. Notice the heavy gauge wire connecting the capacitor to the bridge rectifier and the use of spade connectors to ease replacement. Separate screened cable for all audio paths reduces crosstalk and interference.

The front panel power indicator is a simple incandescent bulb, rather than a neon which was a common choice at the time. Driven from the DC supply, I could see the regulator was operating and amplifiers were powered when it was lit. The neon would not have been so helpful.

R2525 Regulated Power Supply

Regulated supply lines ensure very low hum and intermodulation distortion that can otherwise occur from ripple currents. I’d found this out the hard way with my Red series that had an infuriating low level hum with the volume turned right down and a muddiness to the bass at moderate volume.

The downside to a regulated supply, and the reason why most commercial amplifiers omitted it, is the small power loss it introduces. Without regulation there’s around 70v peak voltage available to the power amplifier so ‘instantaneous music power’ ( a much overused marketing term used at the time) could claim to be 50 watts rather than the modest 25 watts RMS per channel with regulation. You may be thinking I just want the power, as it will sound louder! But without regulation from a shared power supply the left channel modulates the right and vice versa, so you can hear a muddy distorted sound especially on rock music where channel separation is often a feature.

The power amplifiers

I used Mullard’s 25w hifi amplifier design from the book with a few modifications. The output transistors were selected matched pairs from colour TV power supplies which offered higher gain bandwidth than Mullard’s devices. However, Mullard’s lower power semiconductor suggestions provided the necessary performance. Radio Spares Vero board replaced the printed circuit board and used for each module. I considered etching my own PCBs but wanted to create mirrored layouts for left and right channels so the complexity of a one off for each seemed a waste.

Blue sleeved Mullard or Philips (Mullard’s parent company) electrolytic capacitors were used throughout except for the power supply and output that needed low ESR types. The large square heatsinks act as black body radiators to dissipate heat from the output transistors. I’d fabricated the chassis from sheet aluminium as it was easy to work so the conduction area was more than adequate. A small slot at the top of the rear panel let heat escape although I’d not thought to add a bottom vent.

Mullard’s 25w power amplifier design was class A/B, or push pull, rather than class A, which can generate a lot of heat. Setting the output stage bias in such circuits is crucial to reduce crossover distortion and avoid a harsh transistor sound. Annoyingly, I’d mounted the preset to control bias current under the heatsink so I had to remove them each to adjust.

R2525 Power Amplifiers

The preamplifier

Again, I used Mullard’s high-quality pre amplifier design from their book with some more modifications. As mentioned, I designed a separate 20v DC regulator to reduce intermodulation distortion that can be introduced from the power amplifier with a shared supply line. I decided to use BC184 devices apart from the phono input stage which required Mullard’s very low noise high gain transistors.

R2525 Preamplifier

A mixture of red blob tantalum and radial aluminium electrolytic capacitors allowed a compact layout. Mullard’s blue axial caps were too bulky. Individually screened cables are used for all audio signal paths minimising cross talk and interference. The balance control is rather odd with a pull switch to select mono operation – not very intuitive but not often used.

The monitor switch was added for an Akai 4000D, a three head reel-to-reel tape recorder that allowed real time playback to check levels while recording. Later on I connected a cassette recorder via the Aux / Tape input and used a standard 5pin 180 DIN socket.

What did it sound like after 50+ years?

Remarkable, but it took a while to hear…

After cleaning all switches and pots then attaching a pair of speakers to the unmarked terminals I powered up the R2525 with trepidation. It was silent, but not in a good way.

I checked the rear mains fuse, a 1.25A anti surge, and it had blown. I found another from Dad’s component draws and powered on again hoping for the best, knowing I should have used a Variac to apply power gently. The front panel light glimmered indicating some life but again no sound from any control, which seemed odd. I couldn’t believe the contact cleaner had worked such magic and the volume control was crackle free.

I connected a portable CD player via a jack to DIN lead I normally used for a Beomaster 3300. Still no audio. Puzzled, I traced the wiring from speaker terminals via the output coupling capacitors to the power amplifiers. All seemed okay, the terminals were indeed connected although I’d not lettered them as I had the DIN output connectors which boasted Output Left & Right 8 ohms 25w each RMS.

R2525 Connections

Closer inspection of the DIN outputs I found one, a 4 pin, had no wires connected and the other, a 5 pin 180°, had 4 screened cables connected. Hmmm, after 50+ years my memory of what I’d done was hazy. I guessed the DIN output connections were too fiddly when trying different pairs of speakers, which I liked to do. Also I recall building a matching sound to light modulator that required an audio signal from the amplifier, so maybe I’d adapted a DIN socket to allow upgrades and additional devices?

I traced the screen cables from the 5 pin DIN and realised the power amplifier inputs were connected along with the preamp outputs. It needs a simple jumper lead between each for the amplifier to work as intended. As my jack to DIN lead had the correct wiring for the power amplifier input, I tried connecting it directly to the CD player’s headphone output so I could use its volume control in place of the preamplifier’s. It worked, finally I could hear music and I wondered how long it had been since I’d heard my amplifier work.

Encouraged by the sounds I wired a DIN plug with a pair of jumper leads so I could try using the preamplifier controls and its inputs. I plugged the CD player into the DIN socket labeled TAPE IN/OUT and switched the input selector to Tape. Silence, apart from a noisy crackle from the volume control. I tried the Tape Monitor switch but no music played. After checking all the amplifer’s controls I found the balance and bass were noisy although no where near as much as the volume control which crackled persistently between the 10-11 o’clock position.

Puzzled by the missing audio I tried the other DIN socket marked AUX / TAPE but still no sound. The RADIO and MAG PU inputs were both 5 pin DIN but 240° so my lead would not fit. I’d added a pair of phono sockets above the DIN inputs – presumably to connect my Rega Planar 3 which did not have DIN connections. I found a jack to phono lead and tried connecting the CD player. I turned down the volume to near minimum to avoid overloading the sensitive pickup input and switched to Pickup. Music at last through the preamplifier!

It sounded rather muffled and too much bass due to the RIAA equalisation being apllied via the phono inputs for magnetic cartridges. By lowering the bass and boosting the treble the jazz CD bagan to sound half reasonable and I was transported back to the smooth sounds my amplifier could produce. I really needed an analogue source to rekindle the sounds of the seventies.

Reconnecting with Vinyl

There wasn’t an easy way to connect one of my turntables to the R2525.  The Rega’s output was too low since it was now fitted with a moving coil cartridge. At the time it had a Shure M75 Mk 3 moving magnet cartridge which I do still have hidden somewhere. And the Beograms both have 7 pin DIN plugs. So I decided to replace the R2525’s non standard DIN socket with a 7 way that I had lying in the component drawers. This way the R2525 could be used with any vintage turntable with DIN connectors.

Guess the RCA Phono v. European DIN connection standards are a bit like the VHS v. Betamax wars that followed. VHS won but there are still many casualties out there!

The DIN socket I found had a metal surround, like most of the others I’d fitted back in the 1970s but curiously the tape socket was plastic. After plugging in the Beogram 3300 and spinning up ELO Secret Messages I remembered why a metal was not a good idea. It hummed. Yes, the dulcet tones of Jeff Lynn and his crew shinned through but at any reasonable volume the amplifier hummed.

If I jiggled the Beogram’s DIN plug the hum stopped and was due to the way the plug was wired. An earth loop is formed when the DIN plug’s metal screen is connected to the chassis along with the centre pin 3.

The only way around this dilemma is to fit an insulated DIN socket or mount the metal version on an insulated panel – as I’d done for the phono sockets. Doh, how can I be so stupid after 50 years… Just to note, the Beogram is not earthed – it has a double insulated twin mains lead – the hum is due to electrostatic leakage.

In the meantime I jiggled the plug, turned up the volume and marvelled at the sounds from the seventies.  I’d just been repairing a Pioneer SX-850 see Pioneer SX-850 Common Faults and Fixes, manufactured at a similar time, which I like to think sounded similar.

I’m still impressed with the sound quality of the R2525 and amazed I managed to make it while sill at school studying for A levels and out partying Friday nights!

Faults after fifty years

After listening to the amplifier from various sources several niggles emerged. My tinnitus was becoming more noticeable! In fact I blame the tinnitus on listening to rock music at ridiculous levels back in the day. And there’s no warning level via the headphones on this older analogue kit so it’s all too easy to damage your hearing.

The amplifier’s volume control was difficult to turn and crackled despite copious amounts of switch cleaner and isopropyl alcohol so needed replacing. I ordered one from CPC Farnell and this helped, but there was still an annoying rasp like noise when the volume was turned up with no signal. Although this is an unlikely use case, it should have been silent and suggested there’s a DC voltage across the potentiometer. The multimeter confirmed around 225mV was present on the right channel and only a few millivolts on the left.

The electrolytic coupling capacitor was most likely to blame so I added another in series at the volume control to check. This introduced a continuous crackle and the scope showed a varying voltage presumably as each capacitor tried to charge. So I added a resistor to ground between the two and the noises stopped. I need to open the preamp and change the offending capacitor.

I know from the BeoCenter 9000 CD Fault that the small blue Philips electrolytics can develop age related faults so I may swap these out to avoid later problems. I generally avoid changing working components due to PCB damage but my Vero boards are far more robust.

I also changed the Erie 4700µF reservoir capacitor as it was only rated at 63v and ran with a quiescent of over 70v. I replace the Erie with a new old stock Gould electrolytic that has a 75v working rating. With the same dimensions it fitted the mount without modification. Modern equivalents are often smaller so it can be difficult to find the right diameter. It’s tempting is to select a higher capacitance with larger diameter but this changes the original design and places greater surge current stresses on the bridge rectifier and fuses so can introduce more problems.

FM Tuner

While the amplifier sounded rather good, plugging in the FM Tuner was a disappointment.

Although the nixie display worked, showing the selected preset channel, only the odd station tuned in. The stereo light, a red LED briefly flickered. A persistent rhythmic tick, tick, tick sounded when a station drifted into frequency so it needs some investigation. See my follow up post Homebrew 1970s FM Tuner Faults and Fixes for more on the FM radio tuner.