Does the definition of duty cycle indicate only the percentage of time that the level is asserted? For example, if a 10 Hz square wave is de-asserted for 5 of those 10Hz cycles and asserted for 5 of the 10Hz cycles, would that represent a 50% duty cycle? If I represent each of the de-asserted 10 Hz cycles in a single second as a 0 and each of the asserted 10 Hz cycles in that same second as a 1, I could represent a 50% duty cycle as 0000011111. But if it is only the percentage that matters, I could also get 50% by the sequence 0101010101. In the definition of duty cycle, is it necessarily the case that the asserted periods be consecutive and that the de-asserted periods also be consecutive?

  • 2
    \$\begingroup\$ duty cycle percentage is measured within a single cycle of the signal ..... high for 5 seconds and low for 5 seconds equals 50% duty cycle ..... high for 1 second and low for 9 seconds equals 10% duty cycle \$\endgroup\$
    – jsotola
    Aug 5, 2018 at 1:24
  • \$\begingroup\$ I am tempted to write a one word answer: "Yes." It probably wouldn't be well received, but it would technically answer your question perfectly. en.wikipedia.org/wiki/Duty_cycle - For example 1010 would be a duty cycle of 50%, as would a repeating signal of 1111 1011 1111 1111 0000 0000 0000 0010 (and countless other permutations). \$\endgroup\$
    – Dampmaskin
    Aug 5, 2018 at 1:35
  • \$\begingroup\$ When discussing duty cycle, there is generally an assumption that the frequency is fixed. So the percent is always calculated over a single period. If the output is de-asserted for an entire period, you might refer to that as a skipped pulse. Some regulators will go into a skip-pulse mode under certain conditions. \$\endgroup\$
    – user57037
    Aug 5, 2018 at 3:09

1 Answer 1


If you had the pattern 0101010101, you’d still have a 50% duty cycle. However, the frequency would be higher. If 0000011111 represents 10 Hz, then 0101010101 would represent 50 Hz.

  • \$\begingroup\$ Or is it 10Hz and 1 Hz? \$\endgroup\$ Aug 5, 2018 at 3:31
  • \$\begingroup\$ @TonyEErocketscientist assuming each bit represents 10ms, then 0000011111 represents 50ms on/50 ms off = 100ms cycle = 10 Hz and 01 (repeated as many times as desired) is 10ms on/10 ms off = 20 ms cycle = 50Hz. \$\endgroup\$
    – DoxyLover
    Aug 5, 2018 at 6:33
  • \$\begingroup\$ I want to electronically control a normally closed purge control solenoid (Honda Part No.: 36162-PAA-A02 / DENSO 136200-1570 PCS14). With a warm engine and sufficiently high RPMs, the car applies a PWM duty cycle to the valve. Can a test tool drive a PWM duty cycle? Power application must be long enough to actuate, but not long enough for current to cause overheating. Most consumer PWM devices have too high a frequency (e.g., 13KHz PWM DC motor speed control). An adjustable rate in 0.5 Hz to 5Hz range should work. Perhaps I should skip PWM and use an adjustable rate 12V flasher relay? \$\endgroup\$
    – John G
    Aug 6, 2018 at 0:11
  • \$\begingroup\$ For clarification, my initial question did assume each bit represents 10 ms. I now understand that “0000011111 represents 50ms on/50 ms off = 100ms cycle = 10 Hz”, while “0101010101 would represent 50 Hz”. I gather that people seem to use PWM for (1) dimmer control for LEDs, and (2) speed control for DC motors. For dimmer control, is PWM simply providing a lower average voltage? What is the concept for speed control of a DC motor? Given the basic nature of my questions, is there a book or subject matter that I should look into for more information? \$\endgroup\$
    – John G
    Aug 6, 2018 at 1:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.