From geoff@apanix.apana.org.au Sat Nov 26 17:18:01 1994 Date: Fri, 25 Nov 1994 18:27:15 +0930 From: Geoffrey Peters To: analogue@magnus.acs.ohio-state.edu Subject: Re: CS-60,80 ... and 70 ? >> How does the the CS-70 compare to the 50/60/80 series? >> I *think* it was a later model, programmable and less colourful >> in its outfit, but can anybody give details on this one? > >I think Yamaha wanted the CS-70 (or actually CS-70M as it's called) to >be a slightly scaled-down CS-80 much more suitable for mass production, >and indeed there are similarities, but I they missed out on some of the >things that made the CS-80 special: the wighted keyboard (the CS-70M uses >a standard plastic keyboard; don't even know if it's velocity sensitive) >and the pitch ribbon (the '70 uses standard wheels). They also missed >out on the colorful front panel of the '80 (the '70 is more on par with >the SK-10/20 series in that respsect: all black with (white) knobs, and >finished weeden end cheeks). Similarities are the voice architecture, >but without the CS-80's expressive controllers I don't think they >could make it succeed. That's mroe or less right ... >But I'm trying not to put it down on its own, I know there are people on >this list who sware by the 70, and it does have some novel features, including >the small magnetic cards used for off-line patch storage. It also has a >built in sequencer; my guess is that it's a processor-based machine >contrary to the CS-80 which must be one of the few voice-assignable >synths (along with the '50 and '60) to have no microprocessor control (although >there are a couple of juicy custom digital chips for key assignment in >these machine). I also can't claim to have used the machine a lot, but I >think Yamaha were expecting more from the machine in the market than they got, >mostly because they cut down on the main features that made the '80 >successful. My sentiments precisely. >Anyone know anything about the tuning stability of the '70? It has an >autotune button, doesn't it? Yup. The CS-70M (of which I own two) is yet another unique instrument from the Yamaha stable. Polyphony is a maximum of 6 2-oscillator voices, although it is bi-timbral in two different ways; the keyboard can be split, with a different patch-pair on each half (or the sequencer plays the lower patch-pair and the keyboard plays the upper), or the oscillator pairs can each have a different patch (see below). The machine has an unusual voicing architecture, in that the machine contains 12 discrete DCO-VCF-VCA (twin ADSR) boards, all under the control of a 1 MHz Z-80 two-card-cage computer. There is only one programmable LFO, but there's another free-standing one which can't be programmed. The synth boards are odd, in that the DCO has two selectable waveforms (pulse and saw), but there's a true sine wave available in the VCA section. Also, the boards are arranged in pairs (upper & lower), but each board in the pair can have its own unique patch (however, the top board of the pair is responsible for the LFO settings for both boards). Like it's smaller sibling the CS-40M, there are not many program patch locations - 15 pairs (one for upper, one for lower), plus a pair labeled "Manual". It's possible to write to only one part of a pair at a time, so it's possible to build some very intricate dual-oscillator patches by combining parts of different pairs. Editing is not straight-forward - you can only edit one part of a pair at a time, and you have to activate it by pressing the desired patch button twice (the LED then blinks, indicating you're in edit mode). Moving any programmable feature (switch or slider) causes the programmed value to be forgotten, and the new value is read from the control being moved. Filter architecture is fairly straight forward, but because none of the programmable sliders are true pots (they're actually all 5-bit binary-encoded- decimal slider switches!!), there's only 32 positions for any 'fader', and if you crank the resonance up all the way (the VCF does not self-oscillate) and you move the frequency 'fader', you can hear the discrete steps. A bit like a Poly-61, by the sounds of it ... On-board effects consist of "ensemble" (a very lush chorus, too thick for a lot of things) and a two-speed 'tremolo' which works very well for simulating drawbar organ patches; when you togle the speed button, you can hear the tremolo gradually change speed - very realistic. There's a brilliance and sustain slider which are after-program modifiers (alters the cut-off and VCA ADSR release respectively). The keyboard is not velocity responsive (nor is the circuitry capable of it), and there's a single-channel pressure bar under the keyboard, but this only controls the free-running non-programmable LFO. Voice allocation is FIFO (firs in, first out) priority, and a non-programmable portamento (true slide) or glide (semi-tone as opposed to true portamento) is available via a slider (and rear-panel jack). The non-programmable LFO is flexible, in that it sports square, true sine and ramp waves, and it can be assigned to DCO, VCF and/or VCA via the mod wheel and/or keybaord pressure. The programmable LFO has a non-programmable fine-tune for it's speed, and also has its own 2-stage EG (AR) which can be used as a mod source, or to control the LFO speed. Also, it sports a strange Ring Mod function that has to be heard to be described :-) The onboard sequencer is solely under CPU control, so it can't be sync'ed to anything whatsoever (without a major ROM rewrite). However, it's operation is almost identical to that of the Roland CSQ-600 analog sequencer - 4 banks which can be chained or played separately, and recorded in step mode or in real-time. However, the tempo control is very coarse. You can't program in portamento like you can on the CSQ, but it is fun to use, though. Maintenance: the card reader doesn't need to be cleaned often (about every 2 years!), and some appropriately-sized bits of manilla card with the tips soaked in isopropyl alcohol will do the job of cleaning the head. Tuning, OTOH, requires access to a service manual! Sure there's an auto-tune button (that should be pressed 30 minutes after it's switched on), but the boards will drift after 6 months or so to the point where the auto-tune fuction can't compensate. Tuning *each* board requires the adjustment of three pots on each card (hi tune, low tune and spread); access to a frequency counter or tuning aid is desirable. Interfacing: there is a linear (korg-style) CV and gate output on the back which is derived from the highest note on the keyboard, and a large custom block connector called "Keycode" which is a bi-directional access point to the keyboard's scanning matrix!! There is no MIDI capabilities, but one of my CS-70M's has Robin Whittle's Universal Keyboard MIDI Interface fitted, which intercepts the keyboard scans and converts it to MIDI (as well as converting MIDI into keyboard-matrix scans), and using the Keycode socket, I can operate both synths from MIDI as a single device, effectively giving me 4 discrete DCO-VCF-VCA synth patches per note. Output: apart from the ubiquitous phones at the front, there are two phone outputs (mono/upper and lower) and a balanced output (mono, i.e. upper & lower mixed). There is an output switch by the volume knob so that the output can be switched on or off without disabling the phones socket. The CS-70 weighs in at a hefty 60 kilograms, almost as much as the CS-80, so it's not an easy synth to tote around, and thanks to it's severe lack of interfacability, is not too highly sought after. The sound, however, is unique, as almost all the audio path is composed of discrete componentry! Here in Australia, their current second-hand value is about $500, so if you can find one for about US$350, it might be worth considering. I love mine, but then your mileage may vary ;-) Geoffrey