ncs3.10 NRF_SAADC_GAIN1_2 has opposite gain effect on nrf52832 & nrf54L15 while data sheet states same words

[listplot3d_dev]

I was migrating from nrf52832 to nrf54L15, and found that,

ncs3.10 + NRF_SAADC_GAIN1_2 has opposite gain effect on nrf52832 & nrf54L15.

More specifically:

* on nrf52832, using NRF_SAADC_GAIN1_2  will have sampled data value* 1/2

* on nrf54L15 , using NRF_SAADC_GAIN1_2 will have sampled data value *2

* similar behavior applies for other NRF_SAADC_GAIN_X macro

if the inconsistency mentioned above is not about to change in short term, I supposed it would be good to document it some where.

[Simonr]

Hi Shawn

Thank you for your report. I have confirmed that this is an issue and have reported it internally. This seems like a bug and the intended behavior is that this should be the same for nRF52 and nRF54L series devices.

I'll keep you up to date on the internal bug report when I get updates there.

Best regards,

Simon

[listplot3d_dev]

Hi Simon, 

Thanks for you response. There is some detailed context:

the board used for report, was getting ADC wired with AIN1 and AIN2, and the difference between AIN1 and AIN2 get me confused. So testing was done with front amplifier shutdown, in which case ADC input is only circuit noise and only MCU is running. So the report in my post was actually about the relation between gain setting and noise signal voltage, with 3~4 times of tests.

Later I found this post:https://devzone.nordicsemi.com/f/nordic-q-a/120212/battery-voltage-reading-with-a-voltage-divider-on-the-nrf54l15-dk?ReplyFilter=Answers&ReplySortBy=Answers&ReplySortOrder=Descending, then AIN4 & AIN5 are used instead of AIN1 & AIN2, with a new board.

Then with new board, ADC works pretty well with same understanding for GAIN on both nrf52832 and nrf54L15, at least with my current configuration(measurement range of 1.8v and 3.6v). 

Currently I don't have the old board to re-check,  but  guess the phenomenon might be related to noise condition. So the description in original post maybe misleading, as it didn't state the detailed procedure.

So to me personally, this issue is resolved. With new information collected, hope it can get some points clarified.

Thanks & Best Regards

-Shawn