Remember the days when a synthesiser was a collection of mysterious boxes linked together with yards of cable? Take a nostalgic look at the ARP 2600 with an old campaigner.

IN THE EARLY days of synthesisers, life was simple. There were only two main manufacturers offering a relatively limited range of instruments. Furthermore, these weren't being superseded every six months so there was time to actually learn to use them. OK, we didn't have polyphony, touch sensitivity, programmability and all the other things we take for granted these days but somehow it didn't seem to matter.

The two manufacturers were, of course, Moog and ARP who, ironically, no longer exist. Robert Moog needs no introduction, being the "father" of the synthesiser we know today, but Alan R Pearlman is less well-known. The founder of ARP, he tried to improve upon Moog's designs and provide a viable alternative to them.

The first ARP synth was the 2500 - an awesome modular affair seen "conversing" with the alien mother ship at the end of Close Encounters of the Third Kind. ARP's first commercial success was the successor to the 2500, the 2600. Also a modular synth, the 2600 was essentially a slimmed-down 2500.

Pearlman decided that the rotary knobs on Moog equipment (and his own 2500) were not graphically representative of their setting, so he used slider controls throughout on the 2600. The idea (and one that was subsequently adopted by other companies) was that you could literally see a sound by looking at the control panel. Another move ARP made was to eliminate the necessity of using patchcords to connect modules together. Instead, patches were internally wired but could be overridden by inserting patchcords into 3.5mm mini-jacks on the front panel. ARP also saw fit to include an audio and/or voltage mixer at the front end of practically every device - on Moog equipment, you had to patch in separate mixers, which was not only laborious but also obscured the front panel with leads.

Modular Elation

LETS TAKE A closer look at the hardware of the 2600. First off, there are three VCOs, each with a frequency range of 0.3 Hz to over 10kHz, which could be used as control sources as well as audio oscillators. These oscillators weren't as flexible as the 921 VCO's found on Moog's synths, but they were a lot more stable and consequently the 2600 didn't suffer the tuning problems of the early Moogs.

Two of the VCOs offered square/pulse and sawtooth waves whilst VCO2 boasted sine and triangle waves in addition to these. The 2600 was also (to my knowledge) the first synth to feature pulse width modulation. All the waveforms were available simultaneously on each oscillator, which is a feature I miss on today's synths. Each oscillator also had a three-channel voltage mixer for precise control of elaborate modulation.

The 2600 had a standard voltage-controlled lowpass filter with a 24dB/octave slope. It had a 5-channel audio mixer, plus the usual resonance control. The ARP filter had a very different character to that of the Moog filter - in fact, Pearlman went out of his way to design a filter that would sound totally unlike the characteristic Moog filter, and in this he succeeded. The ARP filters are as "beefy" as the Moog's design, but are a lot cleaner. Sadly there was no highpass filtering on the 2600 - though it appeared a few years later on the less expensive Odyssey.

There was only one VCA on the 2600, which could be used to govern the level of control sources as well as the usual shaping of the volume envelope. It had two audio inputs and two control inputs, one of which offered exponential control over the amplifier's characteristics, instead of the usual linear modulation option. The practical result of this was the availability of far more percussive envelope transients.

The 2600 had two envelope generators - one an Attack/Release (AR) type, the other a standard Attack/Decay/Sustain/Release (ADSR) type. All the transient slopes were variable between 2msec and 10 seconds, and were also exponential. To my ears, this gave a certain undefinable "naturalness" to the 2600's sounds. The EGs could be triggered from any external pulse as well as from the keyboard.

Other modules included a noise generator offering white, pink and low frequency noise types which could be used as a control source if you so desired. There was also a serious programmable Sample-and-Hold module. This could sample any incoming waveform and provide a stepped control voltage output. The sampled control voltage could be derived from a low frequency VCO for arpeggios or from the noise generator for random voltage changes - one interesting application was to process the voltage from an envelope generator for stepped envelope transients. The S/H internal clock could be overridden and synchronised to an external pulse, such as a keyboard or sequencer (or an early drum machine) for synchronised sample and hold effects. In all, a very flexible module and a personal favourite.

"Because the 2600's VCOs could be used as control devices in the audio spectrum, some of us were dabbling in FM synthesis 12 years before the DX7."

A bi-directional clock allowed you to automatically switch between two inputs (audio or control) at a rate determined by an external clock. Also on the 2600 were two voltage processors, which were basically just voltage mixers/inverters and also a source of +/—10 volts DC for manual control of modules. In the same section was a "lagtime processor" which added a variable amount of portamento to an incoming control voltage. This module could also be used as a very basic lowpass filter.

An external input preamp allowed processing of external audio signals, and an envelope follower generated an envelope voltage proportional to an incoming audio signal, which could be used to control modules (for example, using this with the VCF gave a dynamically controlled filter sweep) or to trigger devices such as the EGs or S/H circuit. There was also a ring modulator for the creation of clangorous sounds, but it could also double as a VCA. The 2600 also had a simple stereo output mixer with a pleasing (but noisy) stereo spring reverb. And if all that wasn't enough, it even had a built-in amplifier and speakers.

All the modules were designed and built to a standard more akin to test equipment than a musical instrument. This resulted in precise control over sounds as well as an exceptionally clean audio output; in fact, apart from the reverb, the 2600 is quieter than most of today's digital delights. That's progress for you.

Originally the 2600 came without a keyboard although an optional "cheap 'n' nasty" affair was available upon request. Then came the "rev 2" 2600, complete with a four-octave duophonic keyboard which was a great improvement on the original. It had a built-in glide function and LFO so that you didn't waste a valuable modular VCO for vibrato effects. All very nice, but the rotary pitch-bend control was something of a dog, and was nowhere near as "musical" as the wheels found on Moog's designs. The ability to play two notes at once was a first for ARP and seemed quite a luxury in those early monophonic days. Interestingly, the keyboard was light enough to be worn around the neck as a remote keyboard.

Interfacing was easy and the 2600 (and all other ARP gear) was compatible with older Roland, Sequential and Oberheim equipment. If you intended hooking up to Moog, however, the story was quite different, as Moog were using "switch triggering" (S-trig) as opposed to ARP's voltage trigger (V-trig). The combination of a Roland MC4 Microcomposer and 2600 gave you an absolutely evil compositional tool, and was well worth the trouble of sorting out this pre-MIDI language barrier.

So there we are: an impressive line-up of modules that were more versatile than the basic Moog Series 15 but falling short of the capabilities of the more elaborate Series 35.

The Bottom Line

SO WHAT DOES it sound like? To these ears, wonderful. The 2600 could deliver strong basslines, searing leadlines and an unlimited range of special effects. The fact that practically everything could be connected to everything else with a minimum of fuss gave the machine a flexibility sadly missing from today's instruments. Because the VCOs could be used as control devices in the audio spectrum meant that some of us were dabbling in FM synthesis some 12 years before Yamaha gave us their DX synths.

Like most old analogues, the 2600 was littered with dedicated control knobs allowing you to alter any aspect of a sound instantly and hear the results in real time. In this respect the ARP is also graphically representative, thanks to those slider controls, and you can see at a glance what's going on. And don't let anyone tell you that programming a modular synth is hard work - today's parameter-access synths are slow and tedious to program by comparison.

Sadly an ARP 2600 is hard to come by these days, but they are available if you keep your eyes peeled. When they were new, they were selling for £2600 (coincidence?), although the price fell to around £1850 just before ARP went under. Nowadays a second-hand model will cost you somewhere in the region of £450-600. Don't worry about repairs - there are plenty of synth repairers that can fix them, and as components aren't too specialised (no VLSI technology here), you should have no problems. If you're at all serious about synthesis, a modular system will go a long way towards helping you understand what's going on, and if you're contemplating spending your cash in this way you could do worse than provide a good home for one of these classic instruments. If your funds won't run to the 2600 but ARP's sound intrigues you, check out the Odyssey or the humble Axxe - they both share circuitry with the 2600, there's just less of it.

If you need any further recommendation, give Stevie Wonder, Josef Zawinul, Michael Boddicker, Tony Banks or Steve Porcaro a call and ask them what they think. If they're out ask me; I've had mine for eight years and, even now, I still write pieces of music entirely on the 2600, resisting (temporarily) the attractions of an Ensoniq ESQ1, Akai S900 and a Yamaha TX802... I can run the 2600 over MIDI with a Roland MPU101 MIDI-CV converter and so can take advantage of touch-sensitivity, better pitch-bend and a mod wheel that can be assigned to any of the 2600's modules. I'll admit the 2600 has its limitations - it's only monophonic (duophonic at best) and it certainly predates programmability - but don't ask me to sell mine. I'd rather listen to a Flock of Seagulls.