nRF54L15 + MX25R64 Idle Power Consumption

RE: nRF54L15 + MX25R64 Idle Power Consumption

Kazi Afroza Sultana — Mon, 23 Jun 2025 12:20:15 GMT

Hello Patric,

The workaround is for System_on sample; not for system_off.

I have run your sample on another nRF54DK (I think somehow the previous DK was broken because of out-of-range supply voltage from PPK2), supply voltage 1.89V and I got average current 1.34uA.

RE: nRF54L15 + MX25R64 Idle Power Consumption

Patrick — Wed, 18 Jun 2025 17:48:54 GMT

I started looking through the nRF54L15 anomalies for the Engineering B SOC which is included in my AN54L15Q modules and found these two anomalies in particular:

Current consumption is higher than expected in sleep
Regulators have incorrect trims

The workaround for both of those anomalies suggests performing operations prior to VREGMAIN and DCDC startup. Does NCS already do this or is this something that we need to do manually? If the latter, how do we perform those operations as suggested?

RE: nRF54L15 + MX25R64 Idle Power Consumption

Patrick — Wed, 18 Jun 2025 11:20:23 GMT

Thanks for pointing out the load capacitance misconfiguration. I thought the default value looked odd but figured it was chosen as the default value for a reason, given the crystal used on the DK is also 9pf. Having said that, I changed the value to 9pf and saw no change in board power consumption. Do you have any other suggestions I might try to identify this issue?

What are your thoughts on the measured current values you observed? I'm not expecting datasheet level lows but even 7uA seems a bit high right?

RE: nRF54L15 + MX25R64 Idle Power Consumption

Kazi Afroza Sultana — Tue, 17 Jun 2025 14:50:30 GMT

I ran your sample on nRF54L15 Dk and measured current. Average current 7.15uA for source M mode.

RE: nRF54L15 + MX25R64 Idle Power Consumption

Kazi Afroza Sultana — Tue, 17 Jun 2025 14:07:47 GMT

Hello,

The default value of the LFXO is 15.5 pf which is not suitable to your crystal.

I can see in the datasheet, the load capacitances are 7, 9, 12.5pf.

So, based on that you can select any of those in the dts file.

/* For 7 pF */
&lfxo {
    load-capacitors = "internal";
    load-capacitance-femtofarad = <7000>;
};

/* For 9 pF */
&lfxo {
    load-capacitors = "internal";
    load-capacitance-femtofarad = <9000>;
};

/* For 12.5 pF */
&lfxo {
    load-capacitors = "internal";
    load-capacitance-femtofarad = <12500>;
};

RE: nRF54L15 + MX25R64 Idle Power Consumption

Patrick — Fri, 13 Jun 2025 20:59:45 GMT

Hi Kazi,

I have attached the application with build folder included as requested. Adding the lines you suggested before main resulted in a build failure, so they have been added inside of main for now, but did not result in any change to device power consumption.

I am using this LXFO, and have the internal load capacitors set to the same as the nRF54L15DK. In my post above you will also find a copy of my board files. The configuration has changed a little compared to the uploaded application, but the changes had no effect.

I did notice this other case that might have some similarities, though my power consumption is still higher than they were able to reach. There did not seem to be a clear conclusion to that case but I do wonder if the external flash is ultimately responsible.

devzone.nordicsemi.com/.../4101.system_5F00_off.zip

RE: nRF54L15 + MX25R64 Idle Power Consumption

Kazi Afroza Sultana — Fri, 13 Jun 2025 09:19:23 GMT

Hello,

There could be several reasons to get higher current than expected. One of the reasons can be the enable state of RAM retention.

You can add the following lines before main() function

NRF_MEMCONF->POWER[0].RET = 0;
NRF_MEMCONF->POWER[1].RET = 0;
NRF_MEMCONF->POWER[0].RET2 = 0;
NRF_MEMCONF->POWER[1].RET2 = 0

''LFXO is mounted and using internal load capacitors'' Is the load capacitor of your crystal same as the one on nRF54L15DK?

No UARTs or unnecessary peripherals should be enabled in the device tree. You have already mentioned this, can you share the application file with build folder so I can check if any enabled drivers or background tasks that could keep clocks or regulators on?

RE: nRF54L15 + MX25R64 Idle Power Consumption

Patrick — Wed, 11 Jun 2025 21:11:15 GMT

I have attached my board files for reference. Since my post I have removed the pull-up from the LED GPIO config. I have also modified pinctrl sleep definitions from 'low-power-enable' to 'bias-disable'.

When power profiling, I do have my jlink attached to the board, but not running. Unplugging the debugger causes a slight increase in power consumption.

devzone.nordicsemi.com/.../rsense_5F00_duo.zip