[sdiy] Synthex Oscillator

[Tom Wiltshire]

Hi All,

If I could just drag you away from the fascinating and detailed discussion of OCR (OCD is more like it by now) and PDFs for a moment, I'd like you to have a look at something for me. It's a synth. Remember them? ;)

The Synthex is a somewhat revered, certainly highly thought of synth. And it has one of the most unusual oscillator designs you're likely to see. I've been studying it in the hope of understanding what exactly is going on. My interest here is to be able to clone it using modern technology, since using a boardful of LS series logic seems a bit like overkill, although it'd be cheap enough to do and all the chips are available.

Here's the circuit for the oscillator:

http://electricdruid.net/wp-content/uploads/2017/07/SynthexOscillatorWeb.png

I've notated this a bit to make life easier.

Top-left there's a pair of LS174 latches which store a frequency divider value. The frequency divider is in the red box, and is fed by CP1, which is a 4MHz clock signal (approximately) derived from the analog master oscillators. Modulations can be applied to that clock, so that deals with that.
As I understand it, the LS161 counters run as an 8-bit divider. The frequency divider value from the latches is used as the reset value, so the counter counts up from there, and the effective division is (256-divider). The LS112 flip flop can be enabled to change the lowest bit of the division value on alternate cycles, effectively giving a x0.5 division. This is used to increase accuracy without dividing the count down further.

On the right, in the blue box, is another counter, used to determine the octave. This is pretty straight-forward. It's  a binary counter followed by a 8-to-1 multiplexer to select the required octave. The highest octave is fed directly from the clock. Ok, *almost* directly. And here's where things get interesting.

There's another little red box containing a single NOR gate. Now this NOR gate means that when pulses come in from the first divider, the clock pulse to the second (blue) counter) disappears. I understand this is known as a "pulse swallowing counter". Note that the second counter is not *clocked* by the first one, but is rather *cleared* by it (LS393, pins 2 and 12). It's clocked only from the 4MHz master clock via that NOR gate. If I'm getting it right, instead of giving you 1/255th of the input frequency, it gives you 254/255ths of the input frequency (for example) because every 255th pulse gets swallowed.

That determines the basic pitch, I think. From there we move to the green box, which is the wave shaping part of the oscillator. This consists of another 8-bout counter (2 x LS193) fed to a weighted-resistor DAC. The DAC output can be inverted by the 8 XOR gates.

 There's a lot I don't understand here. My best guess is that this is used as an 8-bit counter to generate ramps, and as an 8-bit up/down counter to generate triangles, and the XOR is also used by VCO2 to provide a basic digital Ring Mod function. What I thought was going on was that this was then followed by a smoothing capacitor (93, 10nF) and a transistor to remove that cap from the circuit so ramp edges aren't affected (BC559).

I've also had trouble working out what the frequencies are. I can't get the dividers to produce the frequencies they'd need to to get the wave shaping counter to produce ramps and triangles over the required range (256 steps for the ramp, 512 for the triangle - although I have my doubts about that too).

Just to really throw a spanner in the works, I've heard the following from Mario Maggi, the designer of the Synthex, which has thrown me back into confusion:

> I think it's not correct to call [the Synthex oscillators] "analog/digital".
> 
> Compared to similar instruments the only difference is the way oscillators are realized. They are not "digital" in the classical way of the term. Instead they are analog oscillators with a "digital controlled current source".
> 
> Looking at the schematic (sheet N. 4) you can find the following:
> 
> a) capacitor 93 is the point were the ramp waveform is generated
> 
> b) transistor 90 provide the ramp reset
> 
> c) comparator 1Na generates variable pulse width
> 
> This is absolutely identical to any analog oscillator design. The big difference is the way the current to charge the capacitor is generated, but this doesn't change the audio quality of the ramp waveform generated.
> 
> If you have an oscilloscope you can easily check this waveform on the op-amp out 1Nb pin n.6.
> 
> To see both oscillator mixed you can connect the probe to op-amp 3Oa pin n.1. This mixer is before the filter section.
> 
> In other words in this design there is nothing like "sampling rate" and "antialias filtering" that are typical of any digital design. Here, ALL of the audio chain is analog.
> 
> The term "Digital Ring Modulator" was used because it work inverting the waveform while in a classical ring modulator the waveform is also amplitude modulated. Again the inversion is an analog process.
> 
> The "digital controlled current source" realized the miracle I needed to put together 16 oscillators without any tuning trimmer on voice boards!


This contradicts my understanding of the circuit on many ways. He's suggesting that the DAC is used as a digitally controlled current source for a DCO, whereas I was seeing it as a digitally generated audio output, although with smoothing.
If it is a current source, I can't see how it gets a set value - it's a counter, not a latch.

There's lots I agree about - it's a variable sample rate system based off an analog clock, so there's no aliasing to worry about. Worse case would be a ramp wave with steps in, but (I thought) there are measures in place to limit that too. But now the designer seems to be telling me that's wrong. And there's parts I definitely disagree about - like the Ring Mod inversion being an analog process. It's done on an 8-bit signal with XORs. That seems pretty digital to me. Ok, shortly afterwards that signal is turned into an analog signal, but putting a basic DAC *just after* something doesn't make it analog in my view.

So, what insights can you offer? Is my "discrete digital oscillator driven from divide-down technology" view of the circuit right at all? Is it really a discrete DCO design, like Mario seems to be claiming? What on earth is going on?!

Thanks for any pointers,
Tom

==================
       Electric Druid
Synth & Stompbox DIY
==================