Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

RE: Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

Ziyao Zhou — Tue, 30 Dec 2025 01:41:51 GMT

I think I find the problems, the FEM is powered by VDDIO, which is not included in the current measurement pin.

RE: Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

PaKa — Mon, 29 Dec 2025 10:55:39 GMT

The PPK2 in your setup will only measure the nRF54L15 current, not anything else as far as I can see in the picture. Draw up a diagram of your connections and we can double check. That includes where exactly on the DK you tap the power to the FEM

With just the picture it's easy to make wrong assumptions so could lead to longer time for a correct answer.

With the FEM, the nRF54L15 output power "should" be 0dBm but can be anything less than 2dBm or so, if you request less than 20dBm then it will be lower.

RE: Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

Ressa — Mon, 29 Dec 2025 10:01:57 GMT

Thank you!

Thanks for clarification. According to your code, it seems you are not continuously transmitting or sending data (regardless of data rate), so it means you will have some peak instantaneous current consumption when BLE is active but most of the time( k_sleep(K_MSEC(666)) ) the radio is idle and then the current consumption will be for other subsystems rather than radio. I would recommend to monitor instantaneous current consumption to get a better overview and comparison. Nordic PPK II kits are a good one for monitoring current consumption, if you have any around please check with that. But in general as I mentioned, because the radio is not on all the time, then the current consumption will be dominant with other subsystems.

Best regards,
Ressa

RE: Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

Ziyao Zhou — Fri, 26 Dec 2025 14:51:36 GMT

Today, I conducted an experiment and discovered something quite unexpected: under maximum throughput conditions, the system with an amplifier consumed significantly less power than the one without. This result was counterintuitive. Upon further investigation, I realized that I had previously failed to fully disable the amplifier. It is not sufficient to simply turn it off in the project configuration; the overlay configuration must also be properly disabled.

// &dppic10 {
//       status = "okay";
// };

// &ppib11 {
//       status = "okay";
// };

// &ppib21 {
//       status = "okay";
// };

// &dppic20 {
//       status = "okay";
// };

// &ppib22 {
//       status = "okay";
// };

// &ppib30 {
//       status = "okay";
// };

// &dppic30 {
//       status = "okay";
// };

// / {
// 	nrf_radio_fem: name_of_fem_node {
// 	   compatible  = "nordic,nrf21540-fem";
// 	   tx-en-gpios = <&gpio1 11 GPIO_ACTIVE_HIGH>;
// 	   rx-en-gpios = <&gpio1 12 GPIO_ACTIVE_HIGH>;
// 	   pdn-gpios   = <&gpio1 2 GPIO_ACTIVE_HIGH>;
// 		mode-gpios   = <&gpio1 3 GPIO_ACTIVE_HIGH>;
// 		tx-en-settle-time-us = <27>;
// 		rx-en-settle-time-us = <12>;
//  };
// }; //54L15

I found that with FEM the average current is 7.04 mA, max is 20.91mA

devzone.nordicsemi.com/.../withFEM1300.ppk2

The average of without FEM is 17.08 mA, maxis 29.31 mA.

devzone.nordicsemi.com/.../noFEM1300.ppk2

My peripheral and central code is:

devzone.nordicsemi.com/.../central_5F00_uart_2D00_main.zip

devzone.nordicsemi.com/.../peripheral_5F00_uart_2D00_master.zip

I am using PPK to monitor the current of total nRF54L15 DK+FEM like this, the FEM is connected to VDDIO and GND.

Could you plz help me?

RE: Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

Ziyao Zhou — Wed, 24 Dec 2025 01:27:12 GMT

Thank you for your reply, but here I am measuring the power of FEM+nRF54L15, as the FEM is powered by the VDDIO and GND on nRF54l15 DK

Merry Christmas! Could you give me more information?

RE: Why the power consumption is lower when using nrf21540 (20 dBm) lower than no FEM (8dBm)

Ressa — Tue, 23 Dec 2025 13:53:31 GMT

Hi,

If you are only measuring nRF54L15 power consumption in both scenarios then it makes sense. When using FEM the output power of the nRF54L15 will be lower and most of the power consumption will be on FEM side not nRF54L15 side. So when you set it for +20dBm, the nRF54L15 output power will be less than 8 dBm and probably around 2 , 3 dBm and then FEM amplifies it to 20 dBm. So the SoC will actually consume less and the rest will be consumed by FEM.

Best regards,
Ressa