Xiao Seeed nRF52840 Sense - problem reading microphone data

Question

Hi all,

I am having difficulties reading microphone data using built-in microphone at Xiao Seeed nRF52840 Sense. I have configured the overlay file as suggested here. Then used one of the examples - https://gist.github.com/hlord2000/52949ffc1650bb91f2170a4ac160ebae. The project file includes the following switches:

CONFIG_GPIO=y
CONFIG_FPU=y
CONFIG_PWM=y

# Enable audio subsystem
CONFIG_AUDIO=y
CONFIG_REGULATOR=y
CONFIG_REGULATOR_FIXED=y
CONFIG_AUDIO_DMIC=y

# Enable Logging and RTT Backend
CONFIG_LOG=y

The code is compiled ok

west build --build-dir /Users/jaroslavpsenicka/Projects/nrf/blinky/build /Users/jaroslavpsenicka/Projects/nrf/blinky --pristine --board xiao_ble/nrf52840/sense --sysbuild -- -DCONF_FILE="prj.conf" -DDTC_OVERLAY_FILE=boards/seeed_xiao_nrf52840.overlay

... 

-- Zephyr version: 4.1.99 (/opt/nordic/ncs/v3.1.1/zephyr), build: ncs-v3.1.1
[181/181] Linking C executable zephyr/zephyr.elf
Memory region         Used Size  Region Size  %age Used
           FLASH:       66084 B       788 KB      8.19%
             RAM:       45824 B       256 KB     17.48%
        IDT_LIST:          0 GB        32 KB      0.00%
Generating files from /Users/jaroslavpsenicka/Projects/nrf/blinky/build/blinky/zephyr/zephyr.elf for board: xiao_ble
Converted to uf2, output size: 132608, start address: 0x27000
Wrote 132608 bytes to zephyr.uf2
[10/10] Generating ../merged.hex

and flashed using my Raspberry Pico-based OCD

openocd -f interface/cmsis-dap.cfg -f target/nrf52.cfg -c "program build/merged.hex verify" -c "reset run"

When executed, the microphone does not react to sound and produces output like so:

[00:51:40.001,464] <inf> dmic_sample: RMS: 0.015731
[00:51:40.101,470] <inf> dmic_sample: RMS: 0.015731
[00:51:40.201,477] <inf> dmic_sample: RMS: 0.015731
[00:51:40.301,483] <inf> dmic_sample: RMS: 0.015731

The code runs OK and I can attach the debugger and control the execution, however, the data seems to be stale - after a short period right after the restart:

[00:00:01.330,841] <inf> dmic_sample: RMS: 0.029410
[00:00:01.430,847] <inf> dmic_sample: RMS: 0.023021
[00:00:01.530,853] <inf> dmic_sample: RMS: 0.019408
[00:00:01.630,828] <inf> dmic_sample: RMS: 0.017508
[00:00:01.730,865] <inf> dmic_sample: RMS: 0.016566
[00:00:01.830,841] <inf> dmic_sample: RMS: 0.016117
[00:00:01.930,847] <inf> dmic_sample: RMS: 0.015907
[00:00:02.030,853] <inf> dmic_sample: RMS: 0.015811
[00:00:02.130,828] <inf> dmic_sample: RMS: 0.015767
[00:00:02.230,865] <inf> dmic_sample: RMS: 0.015746
[00:00:02.330,780] <inf> dmic_sample: RMS: 0.015738
[00:00:02.430,786] <inf> dmic_sample: RMS: 0.015733
[00:00:02.530,822] <inf> dmic_sample: RMS: 0.015731
[00:00:02.630,859] <inf> dmic_sample: RMS: 0.015731
[00:00:02.730,834] <inf> dmic_sample: RMS: 0.015731

I briefly experiemented with the regulator feature - assuming the microphone is not properly powered - is uses pin P1.10 to power the mic, as can be seen here: https://files.seeedstudio.com/wiki/XIAO-BLE/Seeed-Studio-XIAO-nRF52840-Sense-v1.1.pdf. However, adding an alias and explicitly enablig the regulator did not make any difference.

    aliases {
        mic-reg = &{/msm261d3526hicpm-c-en};
    };

and the code

	const struct device *const dmic_dev = DEVICE_DT_GET(DT_NODELABEL(dmic_dev));
	static const struct device *mic_reg = DEVICE_DT_GET(DT_ALIAS(mic_reg));

	int ret;

	LOG_INF("DMIC sample");

	if (!device_is_ready(dmic_dev)) {
		LOG_ERR("%s is not ready", dmic_dev->name);
		return -ENODEV;
	}

    regulator_enable(mic_reg);
	k_sleep(K_MSEC(50));   /* REQUIRED for this mic */

So I wonder whether some of you nRF gurus have tried using the built-in microphone with this board.
Could be a hardware problem? My next attempt is probably an external microphone, which might be a little bit over my head, to be honest.

Thank you in advance,

Jaroslav