How to decrease the duty cycle of a square wave signal

Status
Not open for further replies.

Thread Starter

brellen

Joined Sep 24, 2021
8
I need to increase the duty cycle of a square wave signal.
The signal is 1kHz the duty cycle is 25%.
I need to push the duty cycle closer to 10%.
Any suggestions for a simple means to do this?
 

MrSalts

Joined Apr 2, 2020
613
I assume you mean...
I need to decrease the duty cycle of a square wave signal.
The signal is 1kHz the duty cycle is 25%.
I need to push the duty cycle closer to 10%.
Any suggestions for a simple means to do this?
You can use a mono stable multivibrator (74123, cd4538, or 555 timer) set to a pulse duration that becomes 10% of your continuous train of square waves. The monostable will be triggered with each positive to negative transition for some of the chips listed - one lets you pick (up or down transition). It doesn't really matter (up or down) since the period is the important part.
 

Thread Starter

brellen

Joined Sep 24, 2021
8
I assume you mean...

You can use a mono stable multivibrator (74123, cd4538, or 555 timer) set to a pulse duration that becomes 10% of your continuous train of square waves. The monostable will be triggered with each positive to negative transition for some of the chips listed - one lets you pick (up or down transition). It doesn't really matter (up or down) since the period is the important part.
Thank you very much for your reply. However, the issue is that I have no other DC or AC input in this case. I only have the PWM here.
 

Thread Starter

brellen

Joined Sep 24, 2021
8
A schematic of how the current waveform is being generated would be most helpful.
A picture is worth thousands of words.
Thank you for your reply. In my case, the PWM wave is generated by a device that is not able to be manipulated. I want to use this PWM signal as the only input to the circuit so that the output would be 10% duty cycle PWM.
 

MrSalts

Joined Apr 2, 2020
613
Thank you very much for your reply. However, the issue is that I have no other DC or AC input in this case. I only have the PWM here.
Right, you can tap into the 25% duty cycle 1kHz signal (1000 uSec period). Use that to trigger the cd4538 chip. Set the capacitor and resistor on the chip to create a 100 uSec pulse. You'll have a 10% duty cycle.

tap into the circuit's power supply that is generating the existing 25% PWM. The CD4538 can handle a very wide range of power inputs.

you'll have to disconnect the output of the existing 25% PWM generator from the power switching circuit, connect it to trigger your new 10% PWM generator - connect the new output to the power switching region of your device that the 25% PWM was using.
 

crutschow

Joined Mar 14, 2008
27,891
I asked about the output impedance of the PWM signal, since you may be able to use that signal with a Schottky diode rectifier to power the CMOS one-shot, since CMOS circuits require very little power.

LTspice simulation below of the suggested circuit using a TLC555 CMOS timer connected as a one-shot:
If you use a pot for R2 you can adjust the duty-cycle.

1632535307480.png
 

sparky 1

Joined Nov 3, 2018
585
Timer calculaton, for duty cycle shown t= 4:29

Because the output impedance of the oscillator may be less than your audio generator but you are at
at 823Ω impedance range at some point you may want to shift up to audio impedance that matches favorite audio equipment.
Being able to plug into a sound card or into analog circuits however the tube amp may be the final stage.
A lot of different impedance matching depending on the equipment or combination that you like.
 
Last edited:

crutschow

Joined Mar 14, 2008
27,891
the impedance should be 882 ohms.
For that impedance it would be better to use a CD4538 for the one-shot since it has a significantly lower operating current than the TLC555.

I do not have a working model for that device so can't show a simulation, but you would connect the CD4538 as a non-retriggerable one-shot (below).
Triggering on the falling edge is better than the rising edge, since the generated supply voltage will be at its maximum at the end of the PWM pulse.

1632540530090.png
 

AnalogKid

Joined Aug 1, 2013
9,380
To the TS:

What is the device or circuit that has an input impedance of 882 ohms?

There is a very small possibility that this can be done without an IC, but it depends on what the PWM signal is driving.

Also, what does "attack" mean?

ak
 
Last edited:

crutschow

Joined Mar 14, 2008
27,891
If the load (not the source) impedance is 882 ohms, which is a low value, then my suggested circuits will not likely work.
Need more info on both the source and the load.

I have no idea what an "attack" circuit is?
Do you mean "added".
 

AnalogKid

Joined Aug 1, 2013
9,380
Maybe, although "attack" is more of a risetime or slope term, as opposed to "sustain" which is more of a width (or % of cycle) term. Could be a language/translation issue.

ak
 

Thread Starter

brellen

Joined Sep 24, 2021
8
If the load (not the source) impedance is 882 ohms, which is a low value, then my suggested circuits will not likely work.
Need more info on both the source and the load.

I have no idea what an "attack" circuit is?
Do you mean "added".
Yes, it means "added". I also attached the schematic figure for device A and B in the last post. Please check it. I think the circuit you provided should work. I have not got the component and test it phsycially, though.
 

crutschow

Joined Mar 14, 2008
27,891
It would appear that the source impedance is 1kΩ from the output of the opamp.
If so, that will likely not be able to power any of the suggested circuits to change the duty-cycle and drive an 882Ω load (how did you derive that exact value from R7?).
 

Yaakov

Joined Jan 27, 2019
3,565
As the TS exppained, this is a cybersecurity/hacking project. It is a MITM attack at the hardware level, designed to be “passive” to the extent that is is simply placed inline and doesn’t require external power.

I have no reason to suspect the TS is not doing research of the white hat variety but in the absence of some verification (credientials, institution, maybe some previous papers, &c) I would be reluctant to help on principle.

I am bringing this up so any helpers can make their own, informed decision to do so.
 
Status
Not open for further replies.
Top