Timing inaccuracy in low power PWMhttps://devzone.nordicsemi.com/f/nordic-q-a/16276/timing-inaccuracy-in-low-power-pwmI was trying out the low power PWM example and noticed the timing inaccuracy that was mentioned in documentation . For example, with low_power_pwm_0, period = 220 ticks duty cycle = 20 ticks Given a clock frequency of 32,768 Hz (perioden-USTelligent Community 13Thu, 08 Sep 2016 13:15:19 GMTRE: Timing inaccuracy in low power PWMhttps://devzone.nordicsemi.com/thread/62222?ContentTypeID=1Thu, 08 Sep 2016 13:15:19 GMT137ad170-7792-4731-bb38-c0d22fbe4515:af535504-0f76-4969-8470-33c964e3f0d9Stefan Birnir Sverrisson<p>Hi</p> <p>You need to power reset the device after you flash the nRF51. Pressing nRESET button wont make the device exit debug mode, and you will still have high current consumption</p> <p>The low power PWM uses RTC in the background which runs on 32kHz clock source. Perhaps the result you are seeing is because of inaccurate 32kHz clock source. There are two possible 32kHz clock sources, RC or crystal. Crystal is normally accurate, but RC is only accurate if you calibrate it (250ppm). See info on <a href="https://devzone.nordicsemi.com/question/81847/how-much-current-consumption-when-we-use-internal-32k-rc/">this thread</a> and in the <a href="https://devzone.nordicsemi.com/tutorials/11/">nRF51 current consumption guide</a>, section about &quot;Choice of 32 kHz clock source&quot;</p><div style="clear:both;"></div>