I'm afraid I'm much more of a synth 'expert' than I am on basses (and thats stretching the definition of expert somewhat!). I started out with non-working vintage mono's and poly's buy as I say these days there are so many more choices without the sky rocketing vintage prices and inevitable issues with 40 odd year old electronics!
Bit of boring but useful info.... Please don't think I'm teaching you to suck eggs; you might already know this stuff..
Early synths such as the Moog modulars (the genesis of the MiniMoog) and offerings from Serge, Buchla etc used discrete components to build their VCO's, VCF's, VCA's and modulators. Expensive, time consuming and prone to calibration/tuning drift etc. To many people this non linearity it what gives these early instruments their sound and vibe. I mean who doesn't love two oscillators drifting in and out of tune against each other for that fat, chorus/phase sound?! Not great if you want precise tuning though! Enter Doug Curtis. He won a design competition with a semiconductor fab company an the prize was to see his design go into production. He had designed a VCO and exponential converter realised on one chip, the venerable 3340. It revolutionised the synth world almost overnight; VCO's are very temperature-sensitive circuits, most employing a tempco resistor which is usually in physical contact with the two transistors that make up the current mirror. By keeping the VCO circuit and its exponential converter on the same chip the stability of the oscillator was improved massively over conventional discrete designs.
Doug Curtis then started up CEM and designed and marketed other synth 'blocks' such as VCO's, VCA's, envelope generators and a few other bits and bobs and many manufacturers started using CEM chips (or those of another company doing the same thing, SSM) to reduce costs and improve stability. So what's this got to do with the uptake in the new synth market? Well, once FM and later other forms of digital synthesis became popular the demand for CEM and SSM chips fell through the floor and they became obsolete parts. The patents/ copyright/insert correct legalese here have lapsed on the CEM designs (and also the currently very popular Moog ladder filter design) so at lease two companies now produce pin for pin clones of the, uptown that point, hard to source and increasingly expensive classic CEM chips. Behringer owns, or has some connection I believe to one of the companies, CoolAudio, which in part explains how they can produce their own clone instruments for the price that they do, but thats a whole other can of worms!! As an aside, on the anniversary of Doug Curtis designing the 3340, the current incarnation of CEM called OnChip collaborated with Doug's widow and agreed to re release her choice of classic CEM chip. She chose the CEM3340 and so now there are three sources of the chip. Hence them popping up everywhere in synthland these days. Without the CEM chips there wouldn't have been a successful Sequential Circuits Prophet 5 (although the first rev's used SSM chips they were unreliable so they switched over to CEM). As a company they stuck with CEM and the new generation of DSI/Sequential synths still use an OnChip-sourced version of the CEM3396 for their DCO's
Roland and Yamaha did the same sort of thing after the event but they used their own designs and fab facilities while Korg used NJR to fab theirs. You might see mention of IR3109's, BA662's etc etc. Roland used the 3109 in the filter circuits of everything from the last rev. Jupiter 4's right through to the Juno's (although even they succumbed with their SH101 which had a 3340 VCO!)
To some degree all synths stand and fall by their modulation capabilities and that is something new instruments really excel at, with modern super fast processors, FPGA's etc. Pretty much since the birth of the Prophet 5 most synths operate under the control of a processor in one way or another, even if it's just to scan the keyed and front panel controls. This makes them hybrids in truth although for most the actual audio path is in the analogue domain. As time passed the processors got better as did the software and processors were now generating (usable) envelopes and LFO's, the most basic of modulation sources. In the old days and still in the world of modular synths you could created deep, multi-layered modulation CV's buy having suitable modules and patch cables. In a compact synth you're physically limited in terms of controls and connectors to implement those sort of features, however some do more than others in terms of their modulation routings in software. You might see Mod matrix mentioned? A Mod matrix has a pre-defined number of 'slots' and each slot comprises of am assignable mod source, a destination and some form of level or amount control. Obviously the more slots, sources and destinations, the greater the modulation capabilities. Arturia went b**ls out of the bath with the MatrixBrute with 16 mod sources, 12 destinations and 256 patches! I've got one and its super deep but I wouldn't recommend it for a first or only synth.
Anyway enough of me rabbiting on about synths on a bass forum...I'll be getting myself banned!
