I want to read PWM signal with an A/D converter. Is it reliable to take 100 samples of adc data and average it out.Im still seeing some spurious values in-between my captured data. What would be a reliable way to acquire precise output?
You can do it directly without the ADC.What would be a reliable way to acquire precise output?
Thank you. Sorry I should have been more clear. I'm trying to read pulses with my adc which gives me random values in between. I'd like to acquire data only when there's high in the output.You can do it directly without the ADC.
Assuming a PWM that is high for a percentage of the total time:
First determine the base time by counting the time between successive low to high transitions of the PWM signal.
Then count the high time of the signal.
Compute the duty cycle as high time / base time.
Use that fraction to compute the % of full scale of whatever variable the PWM represents.
Ahh.. OK. You could just poll the PWM input and start conversions when it is high. When the conversion is done, poll again before starting the next conversion to make sure it is still high. If not, you know that the last reading should be discarded and you have to wait for the next PWM cycle to begin again. You could use the external interrupt to sync the PWM output to your code to ensure that you start the conversion near the beginning of the ON pulse.Thank you. Sorry I should have been more clear. I'm trying to read pulses with my adc which gives me random values in between. I'd like to acquire data only when there's high in the output.
Thank you. I'll try thatAhh.. OK. You could just poll the PWM input and start conversions when it is high. When the conversion is done, poll again before starting the next conversion to make sure it is still high. If not, you know that the last reading should be discarded and you have to wait for the next PWM cycle to begin again. You could use the external interrupt to sync the PWM output to your code to ensure that you start the conversion near the beginning of the ON pulse.