Tags and responsive mode

Hello,

Please could I get some clarification about using responsive mode on a tag?

I’m using the DWM1001 mdek development boards with the solder jumpers de-soldered to preserve battery as described in the online documentation.

Everything seems to work fine however I notice that sometimes the tags display some erratic behaviour. The tag position starts to jump about erratically. Not sure why this is but when I switch on responsive mode it cures the problem and then I can switch responsive mode off again. It appears something is going out of sync. The update frequency is 1 Hz.

I understand that it is best to have responsive mode off to conserve battery life, so is there a way to solve this problem without having to switch on responsive mode? Like I said it appears to work fine sometimes, but at other times goes a bit crazy.

Any advice would be appreciated.

Thank you.

Kind regards

Sunny

Hi @Sunny
the responsive mode brings you slightly better tag stability when you you are moving between anchors. But during normal operation you should not see a bigger difference.

Please see the DWM1001_System_Overview.pdf Chapter 4.6 Operation of a tag.

With the regard of erratic movement) could you post some video or some picture with floor plan with marked area where it behaves erratic?

Cheers
JK

Thanks JK.

I will see if I can get a floor plan with the marked area. When I mean erratic, probably the location jumping around a bit say within a distance of 20-30cm.

What causes this behaviour? Do tags go out of sync somehow?

As a solution, do you think it would be possible to send a message periodically (say every 60 seconds or so) to each tag on the network, to switch on responsive mode (for a few seconds) and switch off again? As this usually solves the problem.

Thanks

Sunny

Hi @Sunny
there should not be a much bigger difference between responsive and non-responsive mode. The biggest difference is when you move tag rom one location (4 anchors) to another location. The non responsive mode tries to stick with the original 4 anchors as possible and when he lost the connection then it tries to find a new ranging anchors.

Your scenario might be possible to implement via user app.

Better would be to focus for the reason why you see erratic movement while the node in non-responsive mode. The floor plan will be very helpful.

Regarding the out of sync) The measurement type is TWR (two way ranging) it can not be out of sync easily. If it will be out of sync then you will not be able to see the position at all.

Which kind of HW do you use for evaluations?

Cheers
JK

Hi JK,

I’m using the MDEK development kit with the 12 boards out of the box with the standard firmware. They work pretty well for what I am doing.

I have the anchors set up throughout my apartment (8 at the moment, which gives quite good coverage, along with 2 gateways). I am working on a small prototype system.

I have developed software that communicates with the gateways using MQTT to get the locations of the tags and then displays them graphically on a floorplan of my apartment. This is not the web app, this is my own Windows desktop application, however I believe the same thing happens on the web app (although I will double-check the next time), as they are both using the same location data via MQTT.

Perhaps I should have 4 anchors in each room. At the moment I have 8 anchors spread out across my apartment trying to maximise the coverage because I understand UWB can go through walls, although I understand they can still impact the signal (but my knowledge on this is quite limited at the moment). I am really doing this through trial and error at the moment trying to find the best placements for the anchors and seeing how the tags respond accordingly.

Cheers

Sunny

Hi @Sunny
they were two versions of PANS - 1.0 and 2.0. The first one did not support gateways and it also has incompatible UWB protocol. So I recommend you to double check if you have there the latest firmware.

I was thinking about the erratic movement and there is an internal filter so if you see erratic movement at 1Hz then the whole measurement has to be very bad. Try to connect to the module and observe the LES command output if you can see all 4 (expected) anchors an if the measurement’s are OK. .

More errors can be generated by improper AN positions. If you have the used the autopostioning feature (which is only for a demo feature) then you need to recheck the distances manually.

Beside of that your anchors should not be placed directly on walls but it should be placed as far as possible way from any wall (at least min 10cm).

The UWB signal can go trough the walls but it heavily depends on wall thickness and type. 6.5GHz is not the best frequency to penetrating walls. Its like Wi-Fi, 2.4GHz can penetrate walls more easily than 5HGz Wi-Fi.

Cheers,
JK

Thank you. Great help.

I will be reviewing all of the above to get a better understanding of this.

I think I may have tried to spread my anchors too thinly thinking the apartment is only about 8m by 12 in total and the rooms not the biggest so thinking I could get away with fewer anchors ignoring the impact of walls to a certain degree, but in light of the your comments perhaps I should try to aim 4 anchors per room regardless of room size. This way a tag would not have any doubt which anchors it will choose to range with to calculate its position. I think the erratic movements (which only happen every so often) are possibly due to placements of the anchors but will explore this.

Thanks for your help.

Sunny

ps: I’m sure I have version 2.0 of the PANS firmware.