nRF9151 - Low Power Mode

cptel · 2025-08-18T18:13:39+00:00

is the ‘active_sleep’ example on GITS the best example to put the nRF9151 Feather into low power mode? what wakes it up to normal power mode given GPIO is a...

jaredwolff

@cptel here’s how I hook it up:

If you don’t have a JST connector with wires, you can equally connect directly to the VBAT and GND pins. I typically feed it 3.6V as that’s nominal for a LiPo battery.

AchimKraus

cptel Objective: Create a current profile of the entire device using the PPK II to select the appropriate battery size.

The battery will drive the device either via the JST2 or GND/VBAT. If you want to estimate the battery size, use that for power in.

Assumption: I would have to connect the PPK II in-line with the main power source (USB)

Maybe there is a misunderstanding. A battery powered device will use VBAT as main power source, not USB. USB is then only used as power source if the battery needs to be charged.
There may be other “exotic” setups, where you don’t use the “direct battery support (GND/VBAT)” of the feather, but I don’t think, you want or need that.

cptel

quick question (hopefully)…
if I have the nRF9151 Feather connected to USB (primarily for debug/data output) and VBAT connected to 3.6V power source, which power source will it use? VBAT or VBUS?

jaredwolff

VBUS

cptel

thx for details and all the answers! got baseline current profile.

cptel

i finally got a chance to test out Deep Sleep.

Deep Sleep works as expected - it enters low power mode and consumes about 1uA on average and stays in Deep Sleep state.

Deep Sleep Hibernate feature enters low power mode as expected but it stays there for about 10 seconds irrespective of value I set. I’ve tried 30000 and 60000 but it only sleeps for 10 seconds:
ret = mfd_npm1300_hibernate(pmic, 30000);
I tried the same with your sample code and similar results - stays in deep sleep state for about 10 seconds.
Thoughts?
.

jaredwolff

The deep_sleep sample on the v3.0.x branch seems ok:

Zoomed in, I get some small spikes but average is 480nA.

cptel

Deep Sleep/Ship Mode works as expected. No issues with it.

The “hibernate” code timeout is what my question was related to: ret = mfd_npm1300_hibernate(pmic, 30000);

jaredwolff

Oh your right. I changed the macro flag in a different repo. Let me try again.

jaredwolff

That’s with hibernate at 60s

Average current is a little higher (expected since the lpclk/timer is running)

Seems to come back at the expected time.

Did you modify the deep_sleep code? Or are you testing in your own project?

jaredwolff

I also assume USB is unplugged and you’re supplying battery power?

cptel

jaredwolff yes

cptel

the only modification I made was to change the #if false statement to #if true to test the hibernate code

#if false
    /* set hibernate mode and power down */
    ret = mfd_npm1300_hibernate(pmic, 60000);
    if (ret < 0) {
      printk("Failed to hibernate. Err: %d\n", ret);
      return ret;
    }
#else
  /* Put into ship mode */
  ret =
      mfd_npm1300_reg_write(pmic, SHIP_BASE, SHIP_OFFSET_TASKENTERSHIPMODE, 1U);
  if (ret < 0) {
    printk("Failed to go into ship mode. Err: %i\n", ret);
    return ret;
  }
#endif

cptel

I was providing VBAT with 5V from nRF9151 Dev Kit with USB disconnected. I am expecting actual batteries mid-week and will try it again with 3.6V battery.

jaredwolff

I was providing VBAT with 5V from nRF9151 Dev Kit with USB disconnected. I am expecting actual batteries mid-week and will try it again with 3.6V battery.

The PPK2 allows you to provide power (it has it’s own adjustable power supply). It’s on the left under Source Meter. But yea it always helps to test with an actual application battery!

cptel

When in HIBERNATE mode, can a Timer running on nRF9151 wake itself out of hibernate mode? I would think not as the functionality is controlled by PMIC and not the nRF9151. Thoughts?

jaredwolff

Electrically, hibernate and ship mode do the same thing. Power everything down.

cptel

I keep coming back to this as I have not been consistently able to put the nRF9151 Feather in deep sleep mode. Let me explain….
1) I tested with the ‘deep_sleep’ sample code and module enters deep sleep mode as expected.
2) I took the same baseline code and added my code to it - module enters deep sleep mode. However, it cannot open: littlefs() - ‘issue #1’

[00:00:00.252,716] <inf> spi_nor: w25q128jv@0: SFDP v 1.5 AP ff with 1 PH
[00:00:00.252,746] <inf> spi_nor: PH0: ff00 rev 1.5: 16 DW @ 80
[00:00:00.252,929] <inf> spi_nor: w25q128jv@0: 16 MiBy flash
[00:00:00.259,185] <inf> lis2dh: lis2dh@19: int1 on gpio@842500.19
[00:00:00.259,246] <inf> lis2dh: lis2dh@19: int2 on gpio@842500.12
[00:00:00.260,986] <inf> lis2dh: fs=2, odr=0x4 lp_en=0x0 scale=9576
[00:00:00.270,355] <inf> littlefs: littlefs partition at /lfs
[00:00:00.270,385] <inf> littlefs: LittleFS version 2.9, disk version 2.1
[00:00:00.270,416] <err> littlefs: can't open flash area 0
[00:00:00.270,416] <err> fs: fs mount error (-19)
[00:00:00.270,477] <err> littlefs: Automount /lfs failed: -19
*** Booting nRF Connect SDK v3.0.1-9eb5615da66b ***
*** Using Zephyr OS v4.0.99-77f865b8f8d0 ***

3) My original code is based off of the ‘Blinky’ sample code. Everything works well except I get an error when the following code is executed:

ret = mfd_npm1300_reg_read(pmic, NPM1300_BUCK_BASE, NPM1300_BUCK_BUCKCTRL0,
                             &reg);

with the error: Failed to set VBUSINLIM. Err: -5 .

of-course, the same line of code works well when using the deep_sleep sample code baseline.
NOTE: my I2C connected peripherals: accelerometer, TOF sensors work just fine and the LEDs blink as expected.

QQ: Besides making sure all “CONFIG_….” values are set correctly in prj.conf and circuitdojo_feather_nrf9151_ns.conf, are there any other files I need to touch to make this work?
I am at a loss as to why one baseline code works and device enters deep_sleep mode as expected but file-system can’t be loaded while other baseline code works and doesn’t enter deep_sleep mode but file-system works just fine.
HELP!

jaredwolff

@cptel since you’ve added some I2C peripherals, what does the signal integrity look like on those lines? The pull ups are fairly light (20k) on the 9151 Feather.

I’m not sure why LittleFS would fail to initialize other than it’s not being configured properly. Or you’re trying to use it before it is configured properly. Keep in mind you can only have one interface per interface number. UART1 I2C2 SPI3 Though, this would get caught at compile time.

Make sure you’re accessing the peripherals, ideally, from the same thread. If you’re trying to use INIT functions make sure you set them to the least priority so all system code can initialize first.

Otherwise you’ll have to share more code on what you’re trying to accomplish. It’s hard to diagnose with limited info.

cptel

Adding CONFIG_PM_OVERRIDE_EXTERNAL_DRIVER_CHECK=y config option in my prj.conf seems to have fixed the LittleFS mounting issue.

There appears to be some gremlins hidden in the pmic as it sometimes wakes up at configured 60 second interval and other times at between 10 - 13 seconds. See this pic. Getting close (fingers-crossed)

« Previous Page