[sdiy] Favorite VCO options

[cheater cheater]

On Tue, Apr 13, 2010 at 22:05, David G. Dixon <dixon at interchange.ubc.ca> wrote:
> (Very long post)
>
> Interesting list of features, Damian:
>
>> Here are some basic features:
>> 1. DC- and AC-coupled expo FM
>> (AC coupled for feedback)
>
> I normally consign AC coupling caps to switches on the panel, because it's
> so easy to do.  This is probably the one concession I make to using panel
> components.  Hence, the board would just take whichever FM signal would be
> coming in, and the option for having this would be completely up to the
> builder.

Yep, just sayin'. Besides: the switch takes up about as much space as
a jack, so why not have a jack :-)

>
>> 2. DC- and AC-coupled linear FM
>
> Ditto.

Yup :)

>> 3. a simple scheme how to change the range. It's best if it can be
>> done on a daughter-board of some sort, so that it's easy to set up
>> your own amount of ranges, which ever ones you need.
>
> The coarse tuning ranges are determined by resistors from +15V and -15V
> going out to the legs of the coarse tuning pot.  I typically take the lower
> leg directly to -15V to get the minimum frequency for LFO purposes.  I've
> been putting 33k to 39k between the upper leg and +15V to limit the upper
> frequency to somewhere between 30 and 40kHz.  I don't like a bunch of
> inaudible headroom on my coarse tuning pot.

That's until you start using it in complex non-linearities, such as FM :-)

> The input resistor is 100k
> (standard for CV summing).
>
> The fine tuning pot is wired directly to the rails, and the range is
> determined by the input resistor.  I use 3.3M to get about one octave
> overall, although it is nice to be able to get tighter control with 10M.
> Some would choose a vernier for this function, but I'm not that finicky.
>
>> 4. a document that explains how to change the core to get a certain
>> frequency at 0V. Better yet, a formula on the silkscreen :-)
>
> The reference frequency is determined by a single resistor: the one which
> limits current into the 2164.  That one's easy!

Yep, but people won't know it. And it's a common-requested mod.

>> 5. terminals with direct core output, to use your own shapers. Shaper
>> on daughterboard for modularity: let's be hones with eachother,
>> finding the right shaper will be the next step and it won't be as
>> easy!
>
> I've already got all the shapers on the board, and I'm very happy with them.
> The feed to all of them is the output saw, not the raw integrator output.

cool. I suggest adding a place just before the shapers where you can
cut the traces and add a header/connector.

> The key is the 5V reference from REF02, which keeps all the levels very
> tight and ensures nice waveshapes.
>
>> 6. I know you like small pcbs, but don't make the design too cramped.
>> I'm good with the soldering iron, but beginners will really appreciate
>> it. On the other hand please allow components (physically) bigger than
>> you use. This also means chip sockets, big capacitors, and thick
>> (coaxial) cables (please include ground at the same spot the cable
>> would be going out)
>
> I'll probably have to get someone else to do a proper layout.  I haven't
> bothered to learn how.  My PCBs are very easy to stuff and solder, as
> everything is laid out on a strict 0.1" grid with no angled traces, but they
> are probably too idiosyncratic for most people.
>
>> 7. Mounting areas on the PCB for brackets/screws on all corners, with
>> extra space on the 'front' side to make the area of contact between
>> the bracket and the pcb a bit larger than necessary. For people who
>> don't want to use potentiometers as a way of mechanically mounting a
>> pcb.
>
> That wouldn't be an option on my current board.  All board-to-panel
> connections are done via 0.1" MTA connectors.  This makes panel wiring a
> breeze.  I always mount my boards at 90 degrees with stooge brackets.

I suggest adding holes on the other end of the board for people who
want to mount it flat somewhere, like say in a keyboard.

>> 8. sync in, sync out
>
> The circuit has variable sync in.  Sync out is the saw output (?).

Sync out should be, from what I understand, the pulse from the
comparator when it resets.

>> 9. rather than having terminals that you'll solder cables to all over
>> the place, try to have them clumped up so that it's easier to route
>> the cables and maybe even make a pin connector to simply disconnect
>> all cables at once. Yes, that can take some jumpers, but it's worth
>> it, and if someone doesn't want to use jumpers, they can just use the
>> hole where the jumper would start.
>
> I love connectors.
>
>> 10. two-wire signaling. This thing will be quite a bit complex, no
>> reason not to carry ground around with the signal.
>
> Not sure what you mean here.  It would actually be a snap to bring the
> signals out with a ground wire next to them, but I don't really think it is
> necessary.
>
>> 11. specify the power consumption after you build it
>
> Probably won't bother...  If it's more than 20mA, I'd be very surprised.

Try and see. You'll never know!

>> 12. Make sure to mention me on the PCB ;-) very important ;-)
>
> Is that "cheater" or "Mr. Cheater"?

Just 'Mr. cheater' with a small 'c' is perfect ;-))

>> Here are some of my specific requests which I think would/could be useful:
>>
>> 1. the ability to correct frequency via diode corrections on multiple
>> frequencies. Maybe you could put in space for 4 stages; whether
>> they'll be used or not doesn't hurt, they're usually not that big as I
>> understand.
>
> Not gonna happen...  If tracking is within +/-0.05% in the musical range,
> why would you need to fix it?

Hey, some people are pedantic! I guess the question is: how pedantic
do you want them to be with the VCO? The other question is: is it that
much work for you? (I don't know)

It's not a biggie, though!

>> 2. This one is a very big one for me. Direct input into the comparator
>> reference voltage. To connect it to your usual reference use a jumper,
>> or a trace that can be cut with *two* through-holes on each side of
>> the place you'd cut at. You could use one pair of terminals and take
>> it through a normalized jack; this jack could be very very interesting
>> I think :-) Better yet if we can switch between 'normal reset
>> voltage'/'sum of normal voltage +
>> jack_input*amount'/'jack_input*amount' (actually the middle one is the
>> most interesting, but the last one gives the most control). For
>> example using a DC voltage from somewhere else far away on the modular
>> could result in usual 'vintage' drifty behavior.
>> Why two holes on each side? This way if someone cuts the trace to
>> experiment they can later put in a jumper without having to cleanly
>> desolder the holes already used for the wires.
>
> On my current PCB, there's a 390R resistor between the reference and the
> comparator, for reasons having to do with the sync circuit.  One could jack
> in here.  However, I'd rather recommend feeding noise to Linear FM for this
> purpose, which would have exactly the same effect and is already accomodated
> in the existing circuit.

Linear FM has a very different effect than modulating reset position:
it changes the slope of the wave keeping the peak-to-peak voltage the
same. Modulating the reset level, on the other hand, keeps a constant
slope but the peak-to-peak voltage changes. This creates a different
sound. It's an even bigger difference with other shapes: because of
the way the waveform is imperfect, the shaped waveforms will have
glitches and other stuff - which makes it real cool. This lets you get
a controllable version of one aspect of what makes vintage oscillators
sound so nice and fat.

>> 3. core on daughterboard. Terminals only carry current input (from
>> expo and mixing stage for linear FM) and vco output (to route to
>> shapers etc)
>>
>> This would open a lot of possibilities:
>> a) using different cores. You can use one 'main board' and then if you
>> want to try out a different core you just desolder the old core and
>> add the new one. You could experiment with saw, triangle, trapezoid,
>> multicore, with soft sync, without soft sync, hard sync or not, etc...
>
> Why bother?  You can easily shape the saw into all those other forms (and
> the VCO already does all except trapezoid, but this is easy to do with a
> very small off-board variable clipping circuit for the triangle output), and
> the VCO already has variable sync with a panel pot.

What I meant was the multi-core trapez VCO core, there was a thread on
it last year. A triangle core has less tracking error in the high
frequencies, and the trapez vco core even less. And there are other
things you can do :-)

Besides: the 'why bother' has more answers below, quoted by you :-)

>> b) using multiple cores. Instead of using a core in the spot the
>> daughterboard goes, the user could use a 'splitter board'. That
>> splitter board would have multiple spaces for cores, as well as
>> current mirrors for each core, and a mixer for each of the outputs.
>> I'm not saying make that board, but make it possible to use one
>> cleanly.
>>
>> c) modularity in design makes it easier for people who know nothing
>> about electronics, like me, to understand what's going on, and it
>> makes it easier to debug (oscillator doesn't work? try swapping the
>> daughterboards around)
>>
>> d) smaller pcbs could be easier/cheaper to produce, I don't know.
>>
>> e) you could use the same expo for multiple oscillators at different
>> ranges, a sort of 'driver/slave' design like in the moog modular
>
> Well, that's a completely different VCO you're talking about now.  The
> problem is that the expo part of the circuit contains most of the panel
> connections, so the limitation is really in panel real estate.  The entire
> expo + core circuit could be done on a PCB about 2" square!
>
> I always assumed that the Moog slaves were just synced.  I didn't realize
> they used common current sources.

I thought they did. Didn't they? I'm fairly sure they did :-) I should
check though..

>> 4. start phase (for sync)
>
> Not sure what you mean.  You can control the starting point of the wave by
> syncing to a gate signal.

When you sync the oscillator, the wave should be at a phase between 0
and 2pi, settable. It is *very* useful.

>> 5. soft sync
>
> Already there.
>
>> 6. linear FM offset potentiometer
>> 7. exponential FM offset potentiometer
>
> Not sure what you mean.  The 1V/oct input provides offset for Expo FM (via
> the keyboard, typically), and Linear FM causes fluctuations in the reference
> current, so offsetting this would be the same as shifting the reference
> current up or down, which might be useful in its own right, but probably as
> an on-board trimmer rather than a panel pot.

a linear fm offset pot is just for when you don't want to go all the
way connecting a dc source to a VCA/gain then to the lin FM input.
Instead of wasting 2 modules and 2 patch cords, you turn the pot :-)

same with FM offset.

It's a simple thing for you to add a header like that, but it's
difficult for typical users to figure out how to do the simple mod :)

>> 8. CV scaling (it is useful to have this separate for each VCO) - to
>> make octaves larger or smaller than 1V
>
> Why?

It can make the chords *huge*. Try it! If you have windows, try the
'oatmeal' VST synth, it has a knob for stretch tuning (= size of
octave), I think somewhere on the upper right of the GUI. Just
pressing C3 and C4 together makes a nice phasing sound. So you start
with C-F-A and it's fairly inanimated and then you press the C an
octave higher, and it suddenly changes the sound completely.

>> 9. glide; you can put as a slew limiter *after* the expo, that's why i
>> put it here.
>
> All you need to do is glide the CV.  There's no need to mess with the actual
> current (?).

It gives a different glide.

>> 10. your favorite thing here :)
>
> Well, two switches I always add to my VCO panels are "centre/egde PWM" and
> "sine/sigmoid".  Both switches simply send either triangle or saw to the
> pulse and sine shapers, respectively.  It's dead easy, and actually adds a
> lot of functionality, particularly the sigmoid, which has very interesting
> properties when fed to waveshapers.

is the sigmoid going to be the same thing as sin(x) from -pi/2 to
pi/2, repeated over and over?

I'm not sure what the centre/edge pwm switch does.

Both sound really interesting :)

>> I didn't include any requests for the shaper, because that's a
>> completely different pair of boots altogether :)
>
> Well, beyond the normal "pure" waveshapes, I think that outboard shaping
> modules are the way to go.

I think even 'onboard' waveshapers can be interesting :-)