# Adjusting duty cycle without affecting frequency of a VCO

I've looked around the internet for a simple voltage controlled oscillator that is able to produce various waveforms in phase with one another. I asked about it in this site and was directed to a dual opamp integrator/comparator setup, of which a voltage-controlled version can be found in this schematic (from breadboardadventures.com)

I constructed this (except for the n-MOSFET switch, which I've replaced by a NPN due to lack of parts) and it seems to work as intended. However, I'm interested in controlling the duty cycle of the generator. I ran it in a simulator, and as expected, changing the ratios $$\R_1/R_4\$$ and $$\R_2/R_3\$$ (within reasonable limits as to not invert the integrator polarity) gives me different duty cycles, as well as different asymmetries for the triangle waves. Placing a pot in that position would do what I wanted, right? Well, no. Since it directly changes the slope of the integrator curves, it alters the oscillation frequency of the entire system. That is something which would be a hassle to adjust accordingly in the control voltage.

How could I adapt this circuit to adjust duty cycle and other parameters simultaneously using a single external control, in order to keep the oscillation frequency constant?

• Welcome. By default this type of oscillator can only produce a 50% duty cycle, as the output of U2 is a square wave with a 50% duty cycle. It is a simple servo-loop that is symmetric and orthogonal so it self corrects any attempt to change duty cycle. The end result is that you can only change frequency. Look into a 555 timer for easy-to-make oscillators with adjustable frequency and duty cycle.
– user105652
Commented Feb 13, 2019 at 1:10
• @Sparky256 It is true that a 555 has a very easy to control duty cycle. But as far as I know, the range of frequencies offered by it through voltage control are quite limited, and I need around 10000% bandwidth. (Also I have no idea how to change frequency without changing duty cycle for a 555, which is another problem entirely). Commented Feb 13, 2019 at 1:30

## 1 Answer

Given a perfect triangle wave, then follow that with a variable-threshold comparator which outputs your desired adjustable duty-cycle-rectangular wave.

• Great idea. The peak of the triangle gives the most narrow pulse width.
– user105652
Commented Feb 13, 2019 at 2:43
• It doesn't give the OP his adjustable slope, but yes - it'll solve the problem stated in the title of the question. Commented Feb 13, 2019 at 15:31