Possible issue with radio_test on nRF54LM20A DK at very low TX power (-40 dBm / -46 dBm)

[strongman]

Hi,

I am testing the radio_test (short-range) sample on an nRF54LM20A DK. The test setup is:

• Board: nRF54LM20A DK
• Sample: peripheral/radio_test (NCS version: <v1.0.0>) [Radio test]
• Mode: Bluetooth LE 1M (data_rate ble_1Mbit)
• Channel: 2400 MHz (start_channel 0) [BLE radio test]
• TX power: -40 dBm and -46 dBm (very low output power; nRF54L series supports down to about -46 dBm) [54L radio]

Test procedure:

1. Configure BLE 1M, channel 0 (2400 MHz), very low TX power (-40 dBm or -46 dBm).
2. Start a continuous unmodulated carrier:
   start_tx_carrier
3. Stop the carrier:
   cancel
4. Switch to RX mode:
   start_rx

Observed behavior:

• After I stop TX and enter RX mode, I still see a continuous wave around -50 dBm on a spectrum analyzer, as if something is still transmitting.
• This only happens when TX power is -40 dBm or -46 dBm.
• When TX power is higher than -40 dBm, the problem does not appear: after cancel and switching to RX, no residual CW is visible.

Questions:

1. Is this a known issue or limitation of the radio_test sample or the nRF54LM20A radio at very low TX power levels?
2. Could there be any special handling required when using -40 dBm / -46 dBm TX power in radio_test (for example, additional commands or delays after cancel before entering RX)?
3. Do you have any recommendations on how to further debug this (for example, checking specific registers or using another test mode)?

If needed, I can provide:

• Exact NCS version and prj.conf used
• CLI command log from the UART shell
• Screenshots from the spectrum analyzer showing the residual CW

Thanks in advance for your support!

[Kazi Afroza Sultana]

Hello,

I am wondering which board you select for the radio_test sample for the bare metal? Because I can not find the board for nRF54LM20A ont he board list on VSCode build configuration. 

[strongman]

The board I am using is the nRF54LM20A.