Variable mark-space (pulse width) and frequency multivibrator

[neazoi]

Hi,
I would like to be able to make a variable mark/space multivibrator like this one from https://www.nutsvolts.com/magazine/article/bipolar_transistor_cookbook_part_6
However I would like the frequency to be variable from a few Hz to about 60Hz or maybe a bit more.
How can I accomplish this frequency variation?

On another point, how about "rounding" a bit the edges of the signal of it, is there anything I can to to accomplish this?

 

[betwixt]

Rounding edges is easy, just add an RC filter at the output.
Making it variable frequency is not so easy, within a limited range you can make R5 and R6 a ganged variable resistor but ideally you would also make C1 and C2 variable as well. Not practical at those values.

Over a limited range you could consider using a variable frequency oscillator to trigger an adjustable monostable to produce the pulse width you need. Again, this is a trivial task for an MCU.

Brian.

 

[BradtheRad]

Vary frequency and duty cycles using potentiometers. With some finagling you might manage to obtain a 10x frequency range.

Somehow the simulation puts slight rounding on leading edges.

 

[neazoi]

Vary frequency and duty cycles using potentiometers. With some finagling you might manage to obtain a 10x frequency range.

Somehow the simulation puts slight rounding on leading edges.

View attachment 177718

What are the potentiometer pairs for the frequency and the width?

 

[neazoi]

What are the potentiometer pairs for the frequency and the width?

hm... on ltspice it did not work...

 

[BradtheRad]

Look at your time step. It needs to go with slow frequencies. My simulation time step is 20 uSec.

Start the 50k pots at 30k or 40k initially.
Oscillations stop if their resistance is set too low.

 

[KlausST]

Hi,

it´s easy with a dual comparator or a microcontroller.
But usually you don´t want IC solutions..

Klaus

 

[neazoi]

This circuit from that page https://www.electronics-tutorials.ws/waveforms/monostable.html mentions "the frequency of the output signal is always the same as that for the trigger pulse input, the only difference is the pulse width".

I am thinking that I could use this circuit with a potentiometer on Rt, to vary the width, and having this circuit triggered by a variable frequency multivibrator.
But before I experiment with this I need to know something about the trigger signal.
Author mentions that trigger signal must be a negative pulse. Negative, referenced to the ground? Or just a high to low transistion?
Because, if a multivibrator is used, its output signal (trigger input to the next stage) will be a square with +vcc and 0v levels. Can this +vcc to 0v transition be used as a trigger?

 

[Warpspeed]

How about one of the function generator chips such as an XR2206 ?
Variable frequency, variable mark space ratio, and if you want rounded edges it has a sine wave output.
https://www.alldatasheet.com/datasheet-pdf/pdf/80496/EXAR/XR2206.html

 

[betwixt]

Author mentions that trigger signal must be a negative pulse. Negative, referenced to the ground? Or just a high to low transistion?

I would say the opposite. The trigger pulse is used to bias the left side transistor into conduction so it starts the monostable action. The diode blocks any negative voltages and is there to prevent leakage back into the trigger circuit.

When the trigger voltage (a square wave from the waveform shown) is low, that is ground or slightly negative, C1 discharges through R4. On the rising edge, as C1 charges, the current through it raises the voltage across R4 and if high enough makes the diode conduct. The waveform with positive and negative spikes is correct but it is only the positive going edges that trigger the timer.

Brian.

 

[neazoi]

I would say the opposite. The trigger pulse is used to bias the left side transistor into conduction so it starts the monostable action. The diode blocks any negative voltages and is there to prevent leakage back into the trigger circuit.

When the trigger voltage (a square wave from the waveform shown) is low, that is ground or slightly negative, C1 discharges through R4. On the rising edge, as C1 charges, the current through it raises the voltage across R4 and if high enough makes the diode conduct. The waveform with positive and negative spikes is correct but it is only the positive going edges that trigger the timer.

Brian.

So a square wave trigger input (0-5v) is enough to trigger tbe circuit every rising edge. Thanks Brian