DWM1001 gives blanks using "lep" command after every few hours

Hi,
I have two tags which are in shell mode and giving continuous output from the “lep” command. Upon analyzing the output, it is seen that both tags almost simultaneously give blanks to the terminal for anywhere between 5 - 15 seconds.

The tags do not recover at the same time, but the recovery time is within the range mentioned. For example, if both the tags blank out at X seconds after the “lep” command is started, one tag will recover and start giving data at X+6 seconds, while the other will recover at X+11 seconds.

The issue occurs quite sporadically and is spotted every few hours when the command is running continuously. Sometimes it occurs after 2 hours, sometimes after 4 hours, the frequency of occurrence looks to be random. But the duration of the blanks is long enough to be a problem.

Details of the setup:
I am using 6 anchors which are roughly placed apart by 10m. The tags have good line of sight.
Both tags are the DWM1001-dev board connected through USB to a Raspberry Pi 4B, which is powered by a 5V 2.1A USB-C power source.
The update rate is set to 10Hz/100ms and stationary detect is turned off.

What could be the issue here? How would you suggest fixing it?

Regards,
Abhishek

Could this issue be due to the initiator of the anchor network failing? I feel the tags are not the issue since both of them started giving blanks at the same time.

Please suggest a course of action to diagnose/fix this problem.

Regards,
Abhishek

Hi @abhishekd
it seems like a network issue.

Questions:

  1. How do you power anchors?
  2. Could you connect one of anchors to the RPI and log the shell output. If there will be a network outage the anchor shell should report it.

Cheers
JK

Hi @leapslabs

In response to the questions:

  1. The anchors are each powered by a small SMPS that converts 230V AC mains power to 5V DC. The AC supply has UPS power backup as well in case of a power loss.

  2. For the logging of the shell output, do I need to log the output of any particular command or just log the plain shell mode output with no commands running? What sort of error am I looking for?

Regards,
Abhishek

Hi @abhishekd
just the plain shell should be OK.

Cheers
JK

Hi @leapslabs

I connected a new anchor to the network, set it up properly and monitored its shell output as you asked.

Every single time that there is a simultaneous blank out from the tags in data, when the tags all recover, the anchor shell shows the message “[abcdef.xyz INF] uwbmac: connected” after a few seconds (where abcdef.xyz represents the different timestamps, such as 015282.980).

Is this intended behavior? Or does it indicate a network issue, like you suspected? What steps can I take to fix this, if it is a network issue?

Regards,
Abhishek

Hi @abhishekd
yes this means some issues with your UWB network (definitely unintended). How far is the Anchor from Initiator?

Cheers
JK

Hi @leapslabs

The anchors (all in the network, including the one connected to the PC) are about 10-12 meters away from the initiator. They are placed in a large open room, giving them good line of sight to each other.

Regards,
Abhishek

Hi @abhishekd
ok. I see three possibility why is this happening:

  1. There is a HW malfunction - try to preplace ANI with different node
  2. There is a power issue - try to replace the ANI power adapter with different one
  3. There is another UWB network running - When two network collide then the colliding network goes down and try to resynchronize wit the new one.

I see the option 2 as most probable.

Cheers
JK

Hi @leapslabs

Thanks for your response. I will try out these changes, run tests and let you know of the results.

Regards,
Abhishek