Regular SOS readers may have noticed Ken McBeth, a DIY analogue synth builder for over 10 years, guesting as the columnist in last month's Net Notes. We sent Paul Nagle to bonnie Scotland (where else?) to find out exactly what makes this constructor of kitchen‑table instruments tick...
In these days of software-modelled synthesis, computer sound chips and touch-screen displays, it's refreshing that there are those who still prefer to wield a soldering iron rather than program a DSP chip. Whilst browsing the Sound On Sound web page, I was drawn to a Reader's Advert which offered 'custom-built' Moog and ARP synthesizers. Intrigued, I travelled to meet its author — Ken McBeth — at his Edinburgh home, and there found a house transformed into a virtual laboratory, packed with prototype analogue synth panels, wires, knobs and oscilloscopes. In pride of place was his latest creation, the M3 rack synth — but more on that later (see the 'The Scottish Synthesizer' box). So what drives a man to hoard rare synthesizer components and schematics? And what's so great about analogue anyway? Ken explains how his hobby came about.
"When I was about 12 or 13, I was listening to prog‑rock like Emerson, Lake and Palmer, and I loved the synthesizer sounds they used. The problem was that synths were so expensive and if you're that age, where the hell are you going to get the money for one? I used to go to music shops and try things out, but it wasn't until I saw an issue of Practical Electronics with a picture of a keyboard on the front that I got the idea to build one myself. I must admit that I started building it without really having a clue what I was doing, and as I wasn't very good at following circuit diagrams, what I made just didn't work. I built about half of it then stopped."
That might have been the premature end of Macbeth Systems, but just after this, the Digisound modular synth (see SOS January 2000's Retrozone feature) was released, and had an electrifying effect on Ken. "I drooled at the front cover of Electronics Today International, which had this amazing thing on it. It used Curtis chips [on which many classic analogue synths were based — Ed] and everything you needed was supplied. At the time I thought 'Wow, this is it'. I built the power supply and that worked but I was too skint to carry on, so I got more into recording music and studying photography. I did that mostly throughout the '80s, but I still wanted to make a synth. After the DX7 appeared, you could pick up analogue synths for pennies; I got a mint ARP Odyssey for a hundred quid. That Odyssey was really where my building story started. One day, I lifted the lid, took a large-format camera and photographed the inside of it — the circuit tracks and components — and managed to get schematics of it. From that I started building one myself — an amazing experience — and I got hooked again, although I never quite got 'my' ARP working.
"Shortly after that, a guy I knew wanted his Minimoog fixed — it was out of tune. I'd never had one in my own house before, and I had it for a couple of weeks to replace some of the components. This gave me a chance to look inside, and of course I photographed it. Once I dug up the schematics that really set me off, and I started to build it, this time with some success.
"I homed in on the Minimoog because, at the end of the day, that's the sound we all want. I also tried to build an ARP 2600, but it is so fickle in its design. With Moog stuff, you can buy a lot of the components off the shelf and, unlike the ARP, most of it doesn't ask for resistors within 1 percent tolerance of obscure values. In the end, I realised that I'd have to make my own resistors, for example, to get it to work. It is beautiful though — in fact, I seem to remember reading somewhere that even Bob Moog preferred the ARP oscillators to his own because they were so stable. Stability is one of the key factors in synth design. If you think of, say, the EDP Wasp, it has an analogue filter and envelopes but the oscillators are essentially digital. When you switch it on, it's in tune, no arguments — but some Moogs can take about half an hour to warm up and settle to a constant pitch. Later Minimoogs, which are what I based my oscillators on, used a thing they call 'heated-chip technology'. To cut a long story short, they warm up far more quickly; my machines warm up in about three minutes, and after that they are stable."
"Since I started, I've made about 15 synths — six or seven of them what I would call professional units. I made a modular with a three-oscillator bank, in a 3U rack. On top there was a 4U block with one oscillator, a noise generator, Moog filter and two shapers. The bottom piece was a complex ring modulator, with noise gating, compression, and a control voltage section. A friend of mine has a rack with about 20-24 units of this stuff — I can't remember exactly how big it is!"
Although Ken designs in various styles, I notice a definite slant towards large, grabbable knobs and plenty of panel space — a far cry from some of the diminuitive analogues that have appeared in recent years. Ken explains: "My personal preference is to go for a more traditional look, like the analogues of the '60s and '70s, along the lines of the Moog, Polyfusion and Aries systems. I like to work in standard 'U' rack sizes because it's nice and easy and fits well with other rack gear, but sometimes it's good to work outside that size with a large aluminium sheet or even plastic. I'm pretty flexible, and can fabricate most analogue synthesizer elements. With the wide range of 'cosmetic' parts and materials out there, you can build what you want and make it look how you like. I've had some pretty outlandish ideas, such as building a synth in marble, complete with 'alien' writing and sunken LEDs glowing through the stone.
"I believe it's not just important what an instrument sounds like, but also what it looks like. It inspires you to play something if it looks good. I always liked the EMS Synthi AKS and VCS3 models because of their unusual shapes and appearances — they are visually stunning.
"When starting a new design, I first take the knobs and lay them out on the panel to see how it will look. It's important where they sit because this is a musical instrument and it must have ergonomics. Finger room and a tidy layout are essential so the musician has access to the controls without leads or switches getting in the way. This is especially important in a live performance situation where too much clutter can be confusing.
"What's the name of that synth company in Cornwall? Analogue Systems? Well, on their RS Integrator system, I was interested to see that they went for the modular synth approach, with emphasis on gaining functionality over space by selecting the appropriate size of control knobs and sockets, rather like Doepfer with their A100, which is also cleverly laid out and shows what you can do in a confined area. That's fine but it's not for me. Think about the Moog Modular and other big stuff; people can't resist that big look, and they aspire to it.
"The parts used in my machines, especially the user controls, are not the cheapest parts available; for instance the 10-turn rotary controls I use are made to last, as are the rotary switch controls, and are expensive. I'd like to see my gear still functioning hassle-free in 2020, which is setting my standards high. I've learnt in the past that cheap parts used often enough wear out or fail pretty quickly. For this reason, I prefer to avoid mini-jack connectors because, in my experience, they cause trouble and after about a month they don't hold in properly anymore. On Moog modular and ARP gear, you come across switchcraft jacks. They're really expensive, but you could probably put a plug in and remove it about 20,000 times before they cause any trouble. Switchcraft jacks are too expensive for me, but you can still buy, relatively cheaply, quarter-inch jacks that behave very much like them. They're solid, positive connections and that's what I want. Sure, I'd like to sell lots of instruments, but they have to be right."
When MIDI and digital synths arrived, they appeared to strike the death knell for both analogue sound and voltage control. For a while, the older generation of instruments became unfashionable (and cheap), but as we move into the next century, it is fascinating to see that CV/Gate synths are not only surviving, but thriving; new analogue synths and modulars are appearing all the time. I asked Ken why he thought MIDI had failed to completely eradicate the older standard.
"I think CV still has some significant advantages over MIDI. Think of a MIDI system where you want to make something, some add-on component. Unless you're a genius with software you can't do it. In an CV-based system, if you want to interstage a processor between the voltage source and its destination, a lag processor or something, you can do it easily, because control signals and audio are all the same kind of voltages, but with MIDI, you have control data and note data; it's not so easy.
"I know bugger-all about getting into digital stuff and it doesn't worry me. Look at all the new analogues that are appearing, especially modulars. I've always imagined these poor Japanese designers talking amongst themselves and saying: "We've made all these superb new instruments and those stupid Europeans want the old rubbish again! What have we done wrong?' One of the things I'm keen to do is encourage lots of other people to build this stuff. If they did that, we might be able to get the bits cheaper and reduce this frantic search for elusive, ageing synths with the right components."
Ken has experienced quite a few of those ageing synths himself — perhaps this explains his passion for analogue, and dislike of digital? As he explains, his prejudices formed over many years.
"Almost 20 years ago, I first tried the Jupiter 8 and I couldn't believe it. I thought it sounded absolutely wonderful and looked so beautiful. I finally got one, years later and paid an arm and a leg for it. For its time, if you were a stage keyboard player, it was a lovely device — the last of Roland's great synths. I think the original Jupiter 8 had Roland's own chips and sounded even better. I quite liked the Jupiter 6 too: a bitey, brash sort of a synth — but then they started with the Junos. Basically, any time you come across built-in chorus on a synth you know that the manufacturers are having to beef it up.
"Then horrors started to appear like Korg's Poly 800, which was just awful. There was a whole bunch of pretty terrible synthesizers after that. The horror of horrors, though, was the DX7. When that arrived I loved its strange turquoise and purple look, and thought I'd get into it, but I couldn't. I couldn't do anything with it, other than play the keyboard and the sounds that were in it. I could not grasp what algorithms were — I kept thinking of HAL9000, from 2001 — and just couldn't get on top of it."
Fortunately, as Ken has already pointed out, the popularity of the new wave of digital synths resulted in a whole swathe of previously expensive analogue gear becoming affordable. And so he began collecting. "I've owned quite a bit of ARP stuff: the sequencer, which I picked up for a fiver, the 2600 semi-modular — I swapped an Ensoniq Mirage for that — and the Pro Soloist, which I was given for free. When I looked inside the Pro Soloist, it had the 'brick' — the patented filter enclosed in resin potting. There was something about it — a very expressive instrument which could do great-sounding tubas and oboes, and all because of that one filter. It even had limited touch-sensitivity. I hear Vangelis loved them." Ken's real love, though, is his MIDI'd Korg CX3 organ. "The CX3 will never leave me. If you want that John Lord, Keith Emerson type of sound, it can't be beaten."
Ever since Clavia's Nord Lead appeared, the race has been on to recreate the authentic sound and feel of analogue instruments. Ken is much less harsh about these than he is about digital synths from the '80s, feeling that the virtual sound comes pretty close, but the control interfaces still leave something to be desired: "I think about this a lot. I have a Korg Prophecy here, and it's pretty good. But linear motion of a knob is always going to be better than a stepped motion, no matter how fine its resolution. If they could make the knobs of virtual synths mimic electrical characteristics, which are not always predictable, then they'll hit it. And they will at some point. At the moment, these synths' controls are the same every time, yet in this day of technology you could maybe have 10,000 variants on each control every time it is used, randomly selected.
"Having said that, I've done recordings using lead sounds from the Prophecy, and once it was in the mix, you'd have to be some kind of genius to recognise it's not an analogue, where there are so many other harmonics happening. There's a lot of hype around these instruments, though, which makes people spend too much money. On the other hand, the same is true of old polyphonic synths — and you don't always know how long they'll last!"
Having realised that Ken is one of the more useful contacts I've made in recent years, I was keen to find out whether he intends to take his home‑designed, hand‑built synths to the next level. I asked if building individual instruments on a commission basis is the limit of his ambition, or whether he's aiming for some kind of mass production in the future?
"For the forseeable future, I see myself making two basic machines — one for general use, the other being very much professional-quality modular stuff. As always, my eye will be on quality. I'm ever-fearful of naffness.
"My new 'general-purpose' synth, the M3, took me two weeks to build, but now I could probably do it in one. My aim is to get the factory to make the PCBs and the cases with the artwork, because that's quite involved, and then I could hand-assemble what's left; I've already begun to explore this with a local firm. I am also quite open to designing individually for people but then you can't get such things factory-made. I'm considering starting with a synth very much like the M3. It will appeal to lots of people with its Minimoog looks and sound, but also it has some extra stuff the Mini didn't have.
"I make my living as an electrical engineer, but recently I've been building more and more synthesizers. I'm now at the point of deciding whether to become Macbeth Systems full time. Some people no doubt think I'm sad and anoraky, but to do this stuff you just have to work and work at it, and not give up. I think I can produce products for the market and that's what I want to do."
When I arrived at Ken's house, I was greeted by the kind of thick, squelchy, Rick Wakeman-style solo that could only come from a Minimoog. But it didn't — it came from Ken's latest creation, the McBeth M3. I was so impressed I thought a detailed look at it would be interesting, and I was right.
A 4U-high rackmount analogue synth with a sturdy handle-cum-knob-protector, the M3 has a generally Moog-like appearance and layout (see picture), and a sturdiness that belies its handmade origins. There are three oscillators, a mixer section, a filter and two envelopes, and although that might make it sound like a Minimoog clone, it does have quite a few departures from the Moog design — not least in its modulation routings, oscillator sync and digital ring modulator ("inspired by" the ARP Odyssey, according to Ken).
"The M3's oscillators are based on the second-series Minimoog — the first Mini sounded delightful, absolutely delightful, but it was agony to work with for tuning. The Mini's oscillators also had various shapes of pulse wave but I haven't bothered with that, going instead for a fixed square wave. On the size of panel, I had no room for a rotary pulse width, so I didn't try to cram it in."
The large tuning knob is an unusual 10-turn knob which takes each oscillator smoothly from almost stopped to supersonic over 10 full rotations. "The Freq control is a space saver. They are wide-ranging but the beauty of them is that the intervals are easy to find. 10-turner knobs are the only successful way I know to combine a coarse and fine-tuning knob."
Each oscillator has its own orange LED and when the oscillator is tuned to a low frequency, the LED blinks to indicate the speed. At audio frequencies, it just glows reassuringly. But Ken has further plans. "I'm thinking of adding tri-colour LEDs to a future filter which change colour as the filter opens..."
A selector switch determines whether each oscillator functions as an audio source or whether it becomes a source of modulation for the filter or one or both other oscillators. In the case of Oscillator 3, switching it to 'modulator mode' also turns off keyboard tracking, so it can be used for standard vibrato, filter sweeps, and so on. In all other ways, the oscillators behave indentically as modulation sources except that Oscillator 1 also has a Sync switch which routes it internally to the sync input of Oscillator 2. The mixer section has a master volume control, noise level and ring modulator level, and the filter has the usual controls, except that the keyboard tracking goes well beyond the more normal limit of 1V per octave, obtaining a great variance in filter cutoff over a much shorter keyboard area.
External modulation sources may be introduced via several voltage inputs on the imaginatively set-out rear panel. Along with the main CV input, individual CV inputs for each oscillator allow them to be controlled or modulated independently. You could even play three-note chords with a clever enough MIDI/CV convertor. Oscillators 1 and 2 even have their own direct audio outputs — ideal if you're using the M3 with a modular system. The remaining connectors are the Moog S-Trig input, the external filter audio input and a filter control voltage input. A master audio output completes the connectors.
Ken: "The filter input lets you do the Minimoog trick of feeding some signal back into the instrument, although there is no onboard input level control or overload indicator. This external input is approx 10 times gain, so a pre-amped mic or line level would be ideal. This synth could easily have been given a patchbay too, but you can make some bizarre sounds even without patching."
The filter envelope has knobs for attack, decay, and sustain, with a release on/off switch. The output envelope is of the Attack, Decay, Sustain type — simple and yet able to achieve all the classic shapes. The envelopes are impressively fast and red LEDs glow to represent each envelope's voltage output. There's also a Hold setting which bypasses the envelopes. This is handy for constant drones and when you're using the external input and just want to experiment with the cutoff and resonance settings.
Of course, the important question for many of us is surely: how close is the filter to the one on the Minimoog? Ken provides an enthusiastic explanation: "I have quite a moderate collection of old 2n3391a transistors. In the Minimoog filter, they used the 2n3392, which my ancient National catalogue explains are expiatial planar transistors. The transistor ladder in the Moog filter is important: it's a transistor that works on diode resistance. There are a bunch of capacitors and transistors in this ladder, and if you look up Towers International Transistor Selector, you can find replacement transistors for all sorts of things, and 2n3392 is in there. The replacement is the generic bc109 which is something you can get by walking into Maplin — it'll cost you 11p. That's what the modern equivalent is meant to be, but is it? Well, my take is that any time you can build close to the original, you should do it. I reckoned that instead of hoarding these 3391a transistors for some future project, I'd use them. That said, when I have used bc109s for the filter it's been excellent as well, but I just get something from this particular filter, it's given it a very warm fluffiness. If you want to know how close it is, I think it's pretty damn close!"
Mind you, he would say that...