How to generate a low-frequency PWM with duty cycle less than 1%

Thread Starter

Nick Long

Joined May 12, 2020
45
Hi guys, I want to produce a PWM wave with a frequency of 140 Hz and a duty cycle of 0.8%. But, the commercial Oscillators or Timers usually generate a PWM with tens of kHz. I know the schematic below can output the PWM. Is it possible to get the PWM I expected? Or do you have a better scheme? I need the power consumption as low as possible, around several micro watts of tens of micro watts. Thanks.
1621513903359.png
 

Papabravo

Joined Feb 24, 2006
16,188
Do the component values you have chosen give you the expected results?
Having a fixed frequency and a fixed duty cycle is not what I would call PWM, but YMMV.
 

KeithWalker

Joined Jul 10, 2017
1,789
As I read your question, my mind was already exploring several possible solutions until I got to your power requirements. Forget it!
 

Danko

Joined Nov 22, 2017
1,084
Hi guys, I want to produce a PWM wave with a frequency of 140 Hz and a duty cycle of 0.8%. But, the commercial Oscillators or Timers usually generate a PWM with tens of kHz. I know the schematic below can output the PWM. Is it possible to get the PWM I expected? Or do you have a better scheme? I need the power consumption as low as possible, around several micro watts of tens of micro watts. Thanks.
1621521177254.png
 
Last edited:

Thread Starter

Nick Long

Joined May 12, 2020
45
Do the component values you have chosen give you the expected results?
Having a fixed frequency and a fixed duty cycle is not what I would call PWM, but YMMV.
Yes, the circuit I provided can work. I want to learn some better schemes. Thanks for your correction.
 

Danko

Joined Nov 22, 2017
1,084
Thank you so much. I have tried the schematic I attached, and it can works.
For your circuit, the output voltage waveform after an inverter is what I want. Could you please solve this issue? Thanks.
So, use inverter. Circuit will consume 0.25uW more:

1621525921263.png
ADDED:
Inverter will functioned like buffer, so frequency and duty cycle will not affected by load.
 
Last edited:

Papabravo

Joined Feb 24, 2006
16,188
Yes, the circuit I provided can work. I want to learn some better schemes. Thanks for your correction.
The reason I asked is because using the values given I did not get the frequency or duty cycle values that you specified. I notice than in the circuit of Danko, The capacitor is approximately 167 times larger. 60 pf seems like it would be too small to produce such a low frequency. Maybe I made a mistake in analyzing your circuit.
 

Thread Starter

Nick Long

Joined May 12, 2020
45
So, use inverter. Circuit will consume 0.25uW more:

View attachment 239108
ADDED:
Inverter will functioned like buffer, so frequency and duty cycle will not affected by load.
Reversing D1 can get the same result, just like Alec_t said. If I want to use a supply of 1.5 V, could you please recommend a suitable inverter to achieve the same result? Thanks.
 

Thread Starter

Nick Long

Joined May 12, 2020
45
The reason I asked is because using the values given I did not get the frequency or duty cycle values that you specified. I notice than in the circuit of Danko, The capacitor is approximately 167 times larger. 60 pf seems like it would be too small to produce such a low frequency. Maybe I made a mistake in analyzing your circuit.
Sorry, I make you misunderstand my meaning. The schematic I provide can achieve the result I specified with modified parameters. Tuning C1 to adjust the frequency and tuning the resistance ratio of R5 and R6 to adjust the duty cycle. Also a voltage comparator with lower propagation delay needs to replace TLV3691.
 

DickCappels

Joined Aug 21, 2008
7,449
1.5V CMOS I'll be interested in knowing about those too.

Just a note: Those two little inverting triangles in Danko's schematic are not inverters -they are inverting Schmitt triggers. The oscillator part (left-most part of the circuit) would not oscillate if it were only an inverter.
 

Papabravo

Joined Feb 24, 2006
16,188
Sorry, I make you misunderstand my meaning. The schematic I provide can achieve the result I specified with modified parameters. Tuning C1 to adjust the frequency and tuning the resistance ratio of R5 and R6 to adjust the duty cycle. Also a voltage comparator with lower propagation delay needs to replace TLV3691.
One advantage of the circuit provided by Danko is that a Schmitt trigger allows for the capacitor to charge and discharge between two different points, neither one of which is GND or VCC.
 

Papabravo

Joined Feb 24, 2006
16,188
Reversing D1 can get the same result, just like Alec_t said. If I want to use a supply of 1.5 V, could you please recommend a suitable inverter to achieve the same result? Thanks.
I don't know if there are devices outside of an ASIC that work at a lower voltage than about 1.65 Volts. The 74LVC1G parts work down to that voltage and may work lower without guaranteed specifications. What is even better is that there is an LTspice library for them. It is 74LVC1G.lib and it is available from @Bordodynov

Bordodynov's Library

The library is contained in the 16M file which is the first link
 
Top