nRF52840 RAM retention example fails with MAX_TEST_ITERATIONS > 1

[Bob Recny]

Greetings,

We have taken the RAM Retention example and are experiencing a failure (all LEDs on) when setting MAX_TEST_ITERATIONS to a value greater than 1.  This is duplicated on a nRF52840 DK v1.0.0 and a Rigado BMD-340-EVK which contains silicon revision 1 of the nRF52840.

We also tested a nRF52840 PDK v0.9.0 and a Rigado BMD-340-EVK with Engineering Rev C of the nRF52840.  Neither of these boards failed regardless of the MAX_TEST_ITERATION value.

Items in use:

nRF5 SDK v15.0.0 - fresh download

SES on macOS High Sierra and Windows 10

Mac and Windows hosts

We did notice that if we go into the J-Link command terminal and just make a connection (connect, select MCU, SWD, speed) and let it sit there, the test would NOT fail on the Rev 1 silicon.  As soon as we exited J-Link, the test would fail again. 

Thanks,

Bob

[Håkon Alseth]

Hi Bob,

 

If you want to retain a given section in systemoff, I would recommend that you check out the example posted here instead:

https://devzone.nordicsemi.com/f/nordic-q-a/36099/ram-retention-problem

This is for SES, and uses sections, which is a cleaner and more portable approach.

 

In case anyone else is interested as well, here is one for nRF51 and nRF52832 for Keil:

https://devzone.nordicsemi.com/f/nordic-q-a/12641/nrf51-ram-retention-after-system-reset-question

 

And last, but not least, one for gcc:

ram_retention_nrf52_gcc.zip

 

The examples posted above might not fit all combinations 100 % (you might need adjustments if you use it on nRF52840 or nRF52810 on some examples).

Let me know if you run into issues.

 

Kind regards,

Håkon