Unreliable Ranging Behavior and Inconsistent Frequency on NORAB126 (NCS v2.9.0)

[Luis Santos]

Hello,

I'm working with Distance Measurement (nrf_dm) using the NORAB126 module on nRF Connect SDK v2.9.0 and nRF Connect SDK v3.0.1.

I'm facing an issue where, occasionally and randomly, the devices stop performing ranging between them. Once this happens, the only way to resume distance measurements is by restarting one or both devices. I've tried adjusting the following configuration parameters, but the problem persists:

• CONFIG_DM_INITIATOR_DELAY_US

• CONFIG_DM_MIN_TIME_BETWEEN_TIMESLOTS_US

• CONFIG_DM_TIMESLOT_QUEUE_LENGTH

• CONFIG_DM_TIMESLOT_QUEUE_COUNT_SAME_PEER

• CONFIG_DM_RANGING_OFFSET_US

Despite these changes, the behavior remains the same — ranging between peers stops unexpectedly and only resumes after a reset.

Additionally, I noticed that the distance measurements occur at around 4 times per second when the devices are close. However, the measurement frequency appears to vary randomly.

My questions are:

1. Is there a known issue that could explain why ranging suddenly stops and only recovers after restarting the device(s)?

2. Is there a way to set or control the number of distance measurements per second (i.e., define a fixed ranging frequency)?

Any insights or recommendations would be greatly appreciated.

[Simonr]

Hi

Elfving is currently out of office, so I'm taking over this case in his stead. I'm sorry about the late reply, but due to staff shortage in the summer holiday period that must be expected I'm afraid.

1. Using the SYNT clock for distance measurement should be just fine, but has not been tested as far as I know since the SYNTH option is the "least common" option for the LF clock source.

2. The main advantage of using a SYNT clock source is that you save space on your PCB while still having the option for a good accuracy LF clock source as it will assume the accuracy of the HFCLK. If high accuracy is required as it seems to be here, the HFCLK must be generated from the HFXO.

The main disadvantage is that the power consumption will increase significantly. Since it is synthesized from the HF clock, the HF clock will need to be running whenever you use the SYNT. If you synthesize it from the HFXO, the expected current consumption this will require is ~350-900µA based on what crystal is used in your design. (Numbers are from a ballpark range from the ISTART_X32M numbers in page 102-103 in the nRF5340 PS.

In summary, our recommendation is that if you can spare the room on the PCB for the external 32.768kHz crystal oscillator, we strongly recommend that use the XTAL option.

Best regards,

Simon