Funk-O-Matic ( a Swiss Army Knife for amplitude manipulation of audio and CV signals )

[Jos]

Funk-O-Matic

Sculpt-O-Sound presents Funk-O-Matic, a selective amplifier, compressor, attennuvertor utility, distortion unit, wave shaper, wave folder and what not. In technical terms is is called an adjustable function network, or as @Bas commented “a Swiss army knife” for amplitude manipulation of audio and CV signals.

Introduction

While paging through an old text book on electronics ( U. Tietze, Ch. Schenk, Halbleiter-Schaltungs-Technik, 1983 issue ) I found this:

This is an amplifier which can be set to have a different gain for different input voltage levels.

The circuit consists of 5 sections.
Section 1: P1 and OV1 : thresholding and gain
Section 2: P5 and OV2: thresholding and gain
Section 3: R + P4: gain
Section 4: OV2 and OV3: subtraction of signals
Section 5: 10R + P3: DC offset

Sections 1 and 2 can be repeated.

How does it work?

You can set the gain and the voltages at which you want the circuit to have a certain gain. It works as follows.

P4 determines the gain for input signals at zero crossing. If P4 is centered then Ua = 0. Depending on the direction of turning P4 this results either in a copied or inversed output voltage (multiplied by a gain factor).

P2, P4 and P6 can be used to send the resulting signal of the sub circuits to a difference amplifier build using OV2 and OV3. The output voltage Ua = gain * (Upot-top - Upot-bottom)

Upot-top is the voltage a the top of the pots (P2, P4, P6) in the circuit.
Upot-bottom is the voltage a the bottom of the pots (P2, P4, P6) in the circuit.

OV1 and OV5 function as selective amplifiers. The signal they output depends on the setting of P1 and P5. If the input voltage Ue is higher than Uk1, the diode D1 will start to conduct a current and U1 will be -(Ue - Uk1) (for Ue>=Uk1>=0). OV5 does the same for negative input voltages. If Ue is lower than Uk1 then OV1 will not produce and output voltage ( assuming ideal op-amps here ).
So OV1 will add something to the signal Ue that is send via P4 to the output depending on the threshold set using P1.

This circuit can be used to change a waveform by selectively amplifying or dampening parts of it depending on the thresholds set. By using multiple instances of OV1 … P2 and or OV5 … P6 you can select more amplitude levels in the original signal you want the signal to be changed at.

Show us the signals!

OK, after spending “a thousand words” to describe the workings, here are some pictures, that may clarify what it all means. The yellow signal is the original input signal the blue one is the output signal. The zero level of the blue trace is at the line pointed to by the blue triangle on the right hand side of the display.

In this first example you can see there are 2 amplitude levels where the gain factor is changed.

This amplitude level can be chosen for the positive part of the signal independently of the amplitude level of the negative part of the signal, as can be seen here. Only the positive part has been influenced.

Here the negative half of the signal is inverted (using P5 and P6):

If you use an external oscillator in stead of -E or +E you can get wild stuff like this:

Implementation

I implemented the circuit using the same amount of op-amps as can be seen in the schematic.
I used 10K Ohm for R and one TL074 for the op-amps, 1n4148 for the diodes and 100 k linear pots. I printed a euro rack style front panel and added switches for connecting P1 and P5 to either -12V and 12V respectively or to CV inputs. And using double switches I added some LEDs to signal what source the voltage for P1 and P5 is used. Given the circuit diagram above it should be simple enough to add this by anyone with a bit of experience in reading / drawing schematics. Oh, and I left out P3 in my implementation which can be used to give the output signal a DC offset (or compensate for a DC voltage).

For audio and CV signals

The circuit can be used to process audio and CV signals. I added a second output via a non polar 10 mu-F capacitor so that I can get rid of any DC value if I want to use the circuit as an AC signal source as well.

Audio!?

Well, there is a bit of that here:

Funk-O-Matic

Afterthought

As I was browsing YouTube the other day, I could not help but see a lot of similarity between Funk-O-Matic and the Doepfer A136 Distortion/Waveshaper Eurorack module. You think Dieter studied electronics from the same textbook I was reading?

[BlackDeath]

Very cool! Any chance of a simple schematic in Sam format?

[Caustic]

@Jos Very informative post, thanks a lot!

@BlackDeath, conversion from other formats to kosmo is a matter of just hooking the circuit gubbins to a bigger panel.

[BlackDeath]

No, I mean in the stripboard layout style, like the illustration here: LMNC Super Simple Mixer [Strip Board Layout + BOM]

He posted an abstract circuit diagram, but nothing buildable for script kiddies like me.

[Jos]

The diagram is abstract in the sense that it does not name any component values. But you can use standard components (see text above).

If you are interested in a strip board layout, I invite you to design one yourself and publish it. I will help you to correct it should that be necessary.

[Caustic]

This would be a good exercise for newbies (like myself) to get more familiar with working with circuit diagrams. Its small enough to read on a single page, so i imagine taking this to a breadboard is doable, even for a novice. Nice of you to offer help.

[SYNTHGUY57]

yes indeed he did and so far you did good too

[Bas]

Great summary of your actions. A good starting point to learn understanding circuits.

[hatti]

I would like to rebuild the circuit and I am not sure what +E and -E means.
Is it equivalent to Vcc and Vee, so positive and negative operating voltage?

[Dud]

yes i think it’s that

@Jos

and added switches for connecting P1 and P5 to either -12V and 12V

[hatti]

Thank you!
Switches I have.
Than I will start to test it.

[Jos]

@hatti

Is it equivalent to Vcc and Vee, so positive and negative operating voltage?

Yes, that is correct. The component values are described in the ‘implementation’ paragraph. Will you implement section1 and 2 once or multiple times?

[Jos]

Let me know if you run into any issues.

[Jos]

@Jos

If you use an external oscillator in stead of -E or +E you can get wild stuff

To clarify this a bit. This refers to the voltages P1 and P5 are connected to, so in case of connecting P1 and/or P5 to some external oscillating CV voltage.

[hatti]

Thank you for your answers.
It does not work. Maybe I made a mistake, I’ll look at it tomorrow.
At the input I have applied a triangle voltage from Sam’s VCO.
Without supply voltage I have a “cut off” triangle voltage of about 0.8V at the output (blue graph), which I can’t influence by potentiometer.

If I apply the supply voltage I get no current at the output.
I think I will do the setup again.

[Jos]

Is this triangle an AC signal or is the whole of the signal above 0V? In other words, where is the 0V line in the scope image for the yellow trace? If it is at pointer 1, then only half of the circuit will have an effect on the output signal.

The output signal looks like the gain of the input signal is quite high, making it clip against the power lines?

When trying the circuit, start with the threshold pots P1 and P5 at max value (away from GND). Set P2, P4 and P6 in the middle position (this may require some practice). The input signal Ue should then be copied to the output via R and P4. OV1 and OV5 should not produce an output at that point.

Now you can choose either P1 or P5 and turn them to a lower value. Once their value (offset) is lower that the input signal’s max positive or negative amplitude, they will produce an output on OV1 / OV5. Try to work with P1 OR P5, not both at the same time to get a feeling of how it works. You will however only see the effect of P1 or P5 if you set P2 resp P6 at a positive of negative angle from their mid position.
While P1 and P5 influence the shape of one half of the signal (below or above 0 Volts) P2 and P6 determine how much of that signal is subtracted from the signal send via P4. Because P2 and P6 can be chosen to set a positive or negative value you can choose to add or subtract the signal part output by OV1 or OV5 from Ue.

You’ll notice that there is considerable interplay between the pots’ settings.

[hatti]

Something is still wrong with my circuit.

I have built it exactly as it is listed on Jos’s schematic. I left out P3 and separated P1 and P5 with switches.

All elements I used as in Implementation.

And there is something I am not quite sure:
What does the information about two resistors of 10 R mean?
Does that mean 10 ohms?

[hatti]

@Jos

When trying the circuit, start with the threshold pots P1 and P5 at max value (away from GND). Set P2, P4 and P6 in the middle position (this may require some practice). The input signal Ue should then be copied to the output via R and P4. OV1 and OV5 should not produce an output at that point.

I have made the above settings. P1 / P5 are separated. On the oscilloscope, the yellow graph is always input, blue is output. Both are on the zero line with this setting.

[hatti]

@Jos

Now you can choose either P1 or P5 and turn them to a lower value. Once their value (offset) is lower that the input signal’s max positive or negative amplitude, they will produce an output on OV1 / OV5. Try to work with P1 OR P5, not both at the same time to get a feeling of how it works. You will however only see the effect of P1 or P5 if you set P2 resp P6 at a positive of negative angle from their mid position.

Same settings as before.
The first image applies with P1 enabled, P5 off and the 2nd image with P5 enabled, P1 off.
Settings at P2 or P6 show no effect. There must be the error somewhere - my next task

[Jos]

@hatti

Both are on the zero line with this setting.

Hmm, do you mean to say that the output of the oscillator you are using is either 0 or a positive value? That would be very odd, the mean value of the signal should be 0, so it should be centered around 0 Volts. However if that is the case, put a capacitor e.g. 1muF between the output of the oscillator and the input of Funk-O-Matic and repeat the test.