a timbre modulator

[OctaveB]

Hey guys,

It's been a whole afternoon that I am trying to breadboard this little "timbre modulator" circuit from a Barry Klein book (1982).
I replaced the CA3080 with a half of a LM13700 and the LM307 with a TL074. And I power it from +12v/-12v.
This circuit looks very interesting but it's not, because I DON'T UNDERSTAND IT, and, even more annoying, I can't make it work 

The most curious thing to me (it must be noted that I am an amateur) is that, instead of a conventional -12v in pin4 of the CA3080, there is this little piece of circuit that I have going from -1.5v to -1.8v on my multimeter depending on the voltage I give to the "ENV IN", which I find odd. I would expect something around -12v to make the OTA work, wright...?

Another thing to note, but I doubt this is a problem here, is I replaced the 2 zener diodes with 2 LEDs.

Maybe someone could enlight my deep obscure bloody darkened path ?

Here is the schematics :


Thanks a lot

[ElectricDruid]

I'll have a go, but that is *not* a simple circuit to understand since it's doing lots of atypical things.

Let's start with the simple bits:

The right-most LM307 is a basic inverting op-amp mixer. It's taking the dry signal, and two phases of "timbre modulated" signal, and mixing them all together again. Presumably the output is pretty loud, because they've decided to cut it in half with the 2 x 2K resistors on the way out.
The two "shape" channels allow you to individually tweak the positive and negative sections of the waveform. The silicon diodes are going to give a certain amount of cross-over distortion, but if there's sufficient gain earlier, this might not be that siginificant. Then the zeners are going to clip the level.

Working backwards to the left, we have another LM307 in the centre, with 560K/100pF wrapped around it. This looks like an inverting op-amp with the input resistor removed, because that's essentially what it is! It's an I-to-V convertor. It takes a current I, and gives you a voltage -V (note the inversion). It's there because in common with many(all?) OTAs the 3080 or 13700 output a current not a voltage. Often you seen a simple resistor to ground doing the job of converting a current into a voltage, but this is "the proper way" of doing the same thing.

Continuing backwards, we have the OTA stage itself. Now, this isn't really doing anything much, except for those weird pin 4 and pin 5 connections. So, let's have a look at it.

Notice that there's a Envelope Input over on the left. This is a synth circuit and expects that you can feed an envelope signal in from somewhere. It doesn't say what level, but 0-10V is pretty common for the era of this circuit.

Now, this is the bit where I really don't know what it's doing exactly unless I sit down and do some serious analysis. My first guess is that it inverts the envelope voltage to give a negative supply voltage which varies between 0V and -10V. That on its own will modify the amount of distortion that the input signal sees depending on the envelope. But the signal is also fed to the Iabc bias input, which changes the gain of the circuit, presumably to increase the effect.

So what you've got is a voltage-controlled waveform distortion unit, with knobs on! For guitar use, you really need to add and envelope follower and feed it to the envelope input.

That's my initial reading of it, for what it's worth. It's an odd circuit and I'm not claiming that I've got it all right just yet, so if anyone else would like to step in, I'm all ears!

[PRR]

> instead of a conventional -12v in pin4 of the CA3080, there is this little piece of circuit that I have going from -1.5v to -1.8v

The '3080 is not a conventional circuit. It will work fine with very low voltages. Often the simplest way to drive a current through the '3080 is to put it inside a current source.

The LM13700 is similar but has a different voltage on the Iabc pin.

The '307 opamp may be critical in this low-voltage application.....

[antonis]

Can't see the reason for two SHAPE pots..
A single pot on any side and a fixed resistor on the other side should do the job..

[puretube]

ENV-controlled VCA with distortion (limiting). The latter controlled separately for the pos/neg excursion of the wave with the respective "shape"-pots.

[idiot savant]

The odd arrangement of the 3080 is known as the "Jung Trick"

https://electronotes.netfirms.com/AN399.pdf

I've used that arrangement a few times, but it doesn't really offer a huge advantage over the more traditional approach.

[ElectricDruid]

The odd arrangement of the 3080 is known as the "Jung Trick"

https://electronotes.netfirms.com/AN399.pdf

I've used that arrangement a few times, but it doesn't really offer a huge advantage over the more traditional approach.

Hey, nice work remembering the reference! I knew I'd seen it somewhere!

[jimbeaux]

The original article appeared in Electronotes #72 Timbre Modulator - Option 1

http://electronotes.netfirms.com/EN72TM.pdf

And an updated circuit (Scott Bernardi)

https://electro-music.com/wiki/pmwiki.php?n=Schematics.TimbreModulatorBasedOnElectronotes

Links to PCB layout and BOM

https://www.bernacomp.com/elec/og2/og2_timbremod.html

[iainpunk]

i hate like this already, its a kind of wave-folder.
oh, god, here we go, my love-hate relationship with wave-folders rears its head again.

atleast, its wave-folder like when the ''add input'' is connected.
when a signal gets some inverted gain, and then experiences crossover distortion, and gets mixed in with the original signal, the flats in the middle of the wave, around the zero crossing, will be 'taken' by the middle slopes of the original wave, which are in the opposite sloping direction compared to the wave that got inverted.

the gain of the OTA determines the amplitude of the post-fold-wave and the ''folding point'' of this circuit, the 100k pots also vary the amplitude, but have no bearing on the 'folding point' of the wave. the folding point is determined by the amount of gain it takes to overcome the Uf of the crossing diodes, in this case the 3v zener is probably a 3.3v adding the 914, you can expect a 4v drop.
higher gain crosses the diodes at a lower input voltage, and thus folds the wave 'earlier' while the opposite happens with a lower gain.
im not fluent in OTA, so i don't know if the gain rises or falls with a larger input signal. i assume [yes this one] that the gain wil be lowered by the input signal being larger, to compress, and have the ''folding point'' move along with the amplitude to make it work more ''evenly''. to have the folding point be more consistent in time, instead of in amplitude.

if the 'input add' is loose, its a crossover distortion device that possibly also works as a compressor or (if my previous assumption is erroneous) an expander.

i have seen and experimented with this method of wave folding, (but without the variable gain) and found it unsatisfactory as its gain only affects the post-fold-part of the wave and also affects the folding point. the reason i fell in love with wave folding was its 'envelope filter'-like qualities, which (if my assumption is correct) is deliberately reduced in this circuit.
however i do like the feature of being able to limit the amplitude of the post-fold-part individually, instead of both at once. guitar's wave forms are naturally asymmetric, and being able to treat both sides differently appeals to me.

very cool circuit nonetheless

cheers, Iain

ps. is it just me or am i seeing a lot of wave-fold related topics pop up on this and other forums.

[OctaveB]

Thanks, I learn a lot in this topic, this is very interesting. And also it is a very useful article about this "Jung trick". I have been on this circuit all day and it's much more clear to me now.
I've had cool tones but the whole thing is hard to tame, the scope goes woooow pretty quick. But I heard it is hard to tame OTA's in general.

About the rise or fall of the gain of the OTA depending on the input level, what I notice experimentally is that it always seems to rise when the input signal rises (pretty quickly). But maybe I am off topic.

Cheers