NRF_TIMER drifts consistently at 2850 ppm relative to the same code on a different DKhttps://devzone.nordicsemi.com/f/nordic-q-a/60024/nrf_timer-drifts-consistently-at-2850-ppm-relative-to-the-same-code-on-a-different-dkI&#39;m using two nRF52840 dev kits, both running with the same code. I start the timer and every 20 ms, when the timer expires, I set a new 20ms timer, and in handler(), I increment a number. When that number reaches a multiple of 256, I drive a GPIOen-USTelligent Community 13Thu, 09 Apr 2020 21:20:57 GMTRE: NRF_TIMER drifts consistently at 2850 ppm relative to the same code on a different DKhttps://devzone.nordicsemi.com/thread/244195?ContentTypeID=1Thu, 09 Apr 2020 21:20:57 GMT137ad170-7792-4731-bb38-c0d22fbe4515:44ab2731-c564-48ac-9016-6b6701db99a8Collin<p>Thanks! I added:</p> <p><pre class="ui-code" data-mode="text">NRF_CLOCK-&gt;TASKS_HFCLKSTART = 1;</pre></p> <p>And I&#39;m seeing much less drift now.</p><div style="clear:both;"></div>RE: NRF_TIMER drifts consistently at 2850 ppm relative to the same code on a different DKhttps://devzone.nordicsemi.com/thread/244156?ContentTypeID=1Thu, 09 Apr 2020 07:38:38 GMT137ad170-7792-4731-bb38-c0d22fbe4515:1b7eab7d-0179-4b7b-85b9-e9049e9e0603haakonsh<p>The HFINT clock source is an HF RC oscillator and it is not accurate at all with large variations from device-to-device. From&nbsp;<a title="64 MHz internal oscillator (HFINT)" href="https://infocenter.nordicsemi.com/topic/ps_nrf52840/clock.html?cp=4_0_0_4_3_3_0#unique_707850542">64 MHz internal oscillator (HFINT)</a>&nbsp;specs:&nbsp;</p> <table border="1" cellpadding="4" cellspacing="0" frame="border" rules="all" summary=""> <tbody> <tr> <td colspan="1" headers="d77729e19147" rowspan="1">f<sub>TOL_CAL_LFULP</sub></td> <td colspan="1" headers="d77729e19150" rowspan="1"> <p>Frequency tolerance after calibration<a href="https://infocenter.nordicsemi.com/topic/ps_nrf52840/clock.html?cp=4_0_0_4_3_3_0#fntarg_2" name="fnsrc_2"><sup>2</sup></a></p> </td> <td colspan="1" headers="d77729e19153" rowspan="1"></td> <td colspan="1" headers="d77729e19156" rowspan="1"></td> <td colspan="1" headers="d77729e19159" rowspan="1"></td> <td colspan="1" headers="d77729e19163" rowspan="1"><span></span></td> <td colspan="1" headers="d77729e19166" rowspan="1"><span></span></td> <td colspan="1" headers="d77729e19169" rowspan="1"><span>±2000</span></td> <td colspan="1" headers="d77729e19172" rowspan="1">ppm</td> </tr> </tbody> </table> <p>Your findings are expected, I suggest you turn on the HFXO as it is&nbsp;&lt; +/-50ppm on the DKs.&nbsp;</p><div style="clear:both;"></div>