Hot DCO Discussion!

[gs]

A polite 'intervention' - if anyone is planning on hijacking this into another argument about:

a) What constitutes a DCO .....

I couldn't resist commenting ont his, but will keep it very brief.

My biggest gripe about ppl's perceptions of DCOs is when they refer to them as "digitally controlled analog oscillators" when most of them are nothing of the sort. The only DCO design that can make that claim are Dave Smith's DCO design, as used in the Evolver series and the Prophet 08. All of the classic 80s DCOs like in the Junos, JXs, Matrix 6/6R/1000, Crumar Bit series, Akai AX80, etc. are not analog at all except for the waveshaping. The core is digital (the pulse train that creates the initial square wave). The common 80s DCO should be referred to as a "hybrid oscillator" it's neither digital nor analog, but a melting pot of both technologies.

The way to tell the difference is if the oscillation is caused by the conventional charge/discharge of capacitors (VCO and Dave Smith DCO) versus a square wave created by a digital counter.

Was that brief?

[jupiter8]

A polite 'intervention' - if anyone is planning on hijacking this into another argument about:

a) What constitutes a DCO .....

I couldn't resist commenting ont his, but will keep it very brief.

My biggest gripe about ppl's perceptions of DCOs is when they refer to them as "digitally controlled analog oscillators" when most of them are nothing of the sort. The only DCO design that can make that claim are Dave Smith's DCO design, as used in the Evolver series and the Prophet 08. All of the classic 80s DCOs like in the Junos, JXs, Matrix 6/6R/1000, Crumar Bit series, Akai AX80, etc. are not analog at all except for the waveshaping. The core is digital (the pulse train that creates the initial square wave). The common 80s DCO should be referred to as a "hybrid oscillator" it's neither digital nor analog, but a melting pot of both technologies.

The way to tell the difference is if the oscillation is caused by the conventional charge/discharge of capacitors (VCO and Dave Smith DCO) versus a square wave created by a digital counter.

Was that brief?

I'm not doubting your findings but do you have any links to back that up?
Not that i don't believe you.I'm just interested in the subject as you might have noticed.

[gs]

I'm not doubting your findings but do you have any links to back that up? Not that i don't believe you.I'm just interested in the subject as you might have noticed.

I've been deep into the DCO debate for a while now. I even got involved in the Wikipedia definition (which turned out to be a blowout of unresolved debate and confusion). The main thing I can point to is a thread on Harmony Central that happened maybe a year ago, I can't be arsed to look it up right now... maybe you can search for it, type in "DCO" and see what comes up.

In that discussion, Dave Bryce of DSI described how Dave Smith's DCOs operate and behave. Several of us confirmed that it works quite different than typical Roland DCOs from the 80s. The DSI DCOs are more 'analog' than Roland DCOs in a sense. Some kind of digital control over a charge/discharge capacitor oscillation. I'm not quite sure why it's necessary to do it that way, since if your oscillator timing is digitally controlled, why do you need to work with a charge/discharge capacitor at all? Perfect timing is perfect timing whether it's a digital pulse signal coming directly from a digi-counter chip, or a capacitor being told to do essentially the same thing.

[micahjonhughes]

Analog produced square waves aren't really square but are trapezoids as analog circuitry can't instantaneously switch states. Digital circuitry is very good at instantaneously switching states, so digital square wave are very square.

As square waves are the basic building block in many synths, the shape you start with would really affect the downstream sound. I suppose the DSI use of the capacitor is to give the waves a more trapezoid shape. So, it would not affect the timing but would affect the wave shape and harmonic content.

[Jack Spider]

Separated from the 'DCO Mono' thread, for more detailed discussion.

[nadafarms]

I didn't think we were allowed to talk about DCOs anymore because it makes baby jesus cry?

[Jack Spider]

If you whisper, he won't hear you...

[gs]

I suppose the DSI use of the capacitor is to give the waves a more trapezoid shape. So, it would not affect the timing but would affect the wave shape and harmonic content.

Very Good Point. I have also heard it said that the DSI design sounds the most like real VCOs more than any other DCO made thus far. Far more processing power to create 'analog slop' algorithms than back in the 80s (when such a thing wasn't even desired!).

I may have made a mistake when I said DSI were the only ones to use the charge/discharge cap method in a DCO. We need to take a good look at the Crumar DS1/DS2 and Jen SX-1000. Those two DCO designs pre-date the Junos if I'm not mistaken.

[adamstan]

The only DCO design that can make that claim are Dave Smith's DCO design

The other is Siel DK600/Opera 6 - here the pulse train is obtained from divider chip, driven by analogue (SSM2031) VCO. There are two VCOs, and two divider chips (one for OSC A, another for OSC B). So it behaves really 'analogue' - the tracking is perfect thanks to digital dividers, but pitch drifts slightly with temperature changes - it can go quarter-tone up after warm-up.

[synthetic88]

I read on another forum that it's entirely analog except for, "the linear voltage to exponential current converter (the most $$$ and tricky part of a VCO) with a DAC that supplies a charging current to the cap*." So that sounds pretty analog to me.

Then again, there are idiots who think that the Andromeda isn't analog because the envelopes are digitally controlled.

* MC on the Moog forum

[JUGEL]

I'm a bit fuzzy on this stuff. Where would something like my Akai AX-60 come into play? Hybrid of some kind?

[Automatic Gainsay]

I read on another forum that it's entirely analog except for, "the linear voltage to exponential current converter (the most $$$ and tricky part of a VCO) with a DAC that supplies a charging current to the cap*." So that sounds pretty analog to me.

Then again, there are idiots who think that the Andromeda isn't analog because the envelopes are digitally controlled.

* MC on the Moog forum

Yes, I remember seeing people on the Moog forum saying this as well.
Somewhere, at some point, someone very respectable was relatively harsh to me because I was decrying DCOs as not being analog, at which point he told me that some DCOs were VCOs whose frequency was digitally controlled. I don't remember who that was, but it was someone I respected enough to change my entire argument.

Unfortunately, I can only go on the respectable people I have heard comment on the topic. This is not to say you are not respectable, or wrong, gs... just that I've heard a very contrasting portrayal from others.
I personally lack the electronics knowledge to comment on the things myself.

Someone chided me on my SH-1000 video for saying that its oscillator was analog, as it came from a square wave that etc. etc. and I was like... well, if the SH-1000 isn't analog, I don't know what is.

What I do know is that DCOs are not as "organic" sounding as VCOs. While I've had a lot of DCO synths, and liked them just fine, they do not sound like VCO synths. Whether their sound is good or bad or superior or inferior is up to the listener, I suppose.

I have heard people bring up the fact that square waves are essentially digital because they are "on and off." I would like to know more about this "digital counter" of which you speak. If the digital counter is merely giving an on and off signal which defines the square wave... that could mean that while the square wave's "origin" comes from a digital device that the result is not digital... or, at least not digital in the way that digital oscillators in digital synths and software generate the sound.

As for analog square waves being trapezoidal... my oscilloscope must be crazy, then. ; )

[micahjonhughes]

AX-60 uses CEM 3394 chips. So everything is analog.

[micahjonhughes]

As for analog square waves being trapezoidal... my oscilloscope must be crazy, then. ; )

You just need a faster oscilloscope.

Think of the vertical axis of a waveform as being the location of the speaker cone producing the wave. A square wave asks the speaker to move between two points, all the way out and all the way in. A speaker must travel between the two point. It does it so quickly but must pass through 0. So if we looked at the picture of the speakers path against time, it would be a trapezoid.

[Pete]

A DCO is an osc where the frequency is controlled digitally rather than by voltage.