Understanding the DMM LFO

[mth5044]

Hello!

I'm looking at LFO's used to modulate delays/delay chips, and I've gotten to the DMM and Small Clone that modulate a CD4047 to create the effect. I've isolated the LFO's below



I can't seem to find any similarities between the two.. they seem to have many aspects backwards in comparison. Anyway, I can't seem to see this as a typical voltage swing controlling a resistance, especially on the DMM. I can't seem to understand what is happening there. What is the pot on the LFO controlling? Depth? Is the diode acting as a resistor? Confused! The LFO doesn't connect to right to ground as most other examples of LFO I've seen do, perhaps that is to reduce ticking? What is up with all the additional circuitry between the depth(?) pot and the CD4047.

Does anyone know what kind of wave is coming out of the LFO before the depth part? I remember the mod being something different than most others I had played before.

Thanks!

[mth5044]

Big pic is big. Sorry about that.

[R.G.]

The central parts of the LFOs are conceptually identical, and differ only in the details of how they're hooked up and modified.

First: An opamp with a capacitor from output to inverting input, with a series resistor from some voltage to the input, constitutes an integrator. This circuit implements the math function output = integral of the voltage on the input resistor, scaled by resistor value and capacitance for how fast it integrates. If you haven't had the math forced upon you, an integrator responds to a fixed number (or in this case, a fixed voltage) by having an output that is a linear ramp. In an opamp, the voltage which drives the inputs is the difference between the inputs, so the voltage which drives the integration is the difference between the (+) input voltage (that is, bias voltage) and the voltage on the resistor to the (-) input. If the voltage is above the (+) input voltage, the output ramps down, and if the voltage at the (-) input is below the (+) input voltage, the output ramps up. If we feed the input resistor a square wave, the output ramps up and down, making a triangle wave.

And that's what comes out of the integrator in both LFOs - a triangle wave.

The second opamp in both LFOs is set up as a Schmitt trigger, a fancy term for a comparator which has hysteresis. Increase the voltage on its input slowly, and when it gets above some threshold voltage VT+, the output goes full positive, and stays there until the input is brought below some lower threshold VT-, at which point the output snaps down and stays there until it once again goes above VT+. So if you feed this a triangle wave (... like from the other opamp's output) it switches high at VT+, stays there until the waveform reverses and goes below VT-, and so on.

Here's the trick: The output of the Schmitt trigger is fed back to the input of the integrator. The integrator ramps up til it gets to VT+, the Schmitt trigger flips, and drives the input to the integrator in the opposite direction. This makes the integrator ramp back down til its output gets to VT-, at which point the Schmitt trigger flips again. The speed of ramping on the integrator is directly proportional to the voltage fed in divided by the input resistance. On the Small Clone, the input resistor is a variable 1M, and this is a speed control knob. On the DMM, it's a fixed 120K. The DMM speed is fixed.

On both of them, the output to the rest of the circuit comes from the output of the integrator, which is a triangle wave. On both, it's a variable size, depending on the setting of the pot to ground, or Vbias. That's a depth control. After the wiper of the depth pot, it goes off to be massaged to fit the needs of the rest of the circuits, which are different.

Where does that muddy it up more than explaining?

[mth5044]

That actually made it quite a bit more clear. Thank you RG

[miaumiaumiau]

I have a question... i'm still quite a noob so...
if i wanted to use this LFO and control an LED with it, where would I put the LED? would I have to add something to the 3 pins?
I wand the LED to control an opto-resistor to control a delay with the cd4047...

cheerio!

[senko]

Just find a triangle wave generator schematic online and experiment.  If you're trying to get an LED to work with an optoresistor, you're going to have to crack open some datasheets or just plain experiment (more fun).

Why are you using a 4047?  You could use a 555 timer, utilizing the CV pin to increase/decrease the output frequency.
Side note: If you're using a PT2399 for delay, using different waveforms on pin 6 does great things.  A sine wave makes some otherworldly delay sounds.