Using 2 DWM1002 causes offset on the second node

Wireless Connectivity Ultra-Wideband
1

Hi guys,

I’m currently facing an issue and am not sure, how to solve the problem. I am using the PDoA kit with 2x DWM1002 (node) and 1x DWM1003 (tag), each firmware v1.1. The goal is to get correlated data from the nodes and combine it, i.e. the nodes are placed next to each other with an offset on the x-axis.

The phenomenon I am facing is quite strange:
When I plug in just one node I get perfectly fine values (no matter whether it’s node 1 or 2).
As soon as I open the second serial port (since I use 2 nodes for the system), the second node produces an offset on its x-values (also the y-values seem to get influenced).
I do not use any script for that but just read out the serial ports.

Even though it seems that the first serial connection influences the values of the second serial connection (but only when data is incoming), same issue occurs when i connect each node to a different pc and open a serial connection. So my guess would be, that the TDMA frame, sent out by the tag, is making trouble somehow.

Nevertheless, I’m pretty sure, something similar worked a few months ago just fine. After that, I played around with the firmware and tried different things (for instance firmware v2.0) but set everything now back to stock firmware v1.1

Attached you can find a screenshot of the output (made by a piece of code I made).
At the top you can find the x,y-values when the node was running for second with a strange output.
Then I disconnected the first node, restarted the one in the screenshot and we get adequate results at the lower part.

Any ideas?

2

Alright, I found the issue:
Both nodes need to have the same assigned address word (2 bytes). I changed it after flashing the firmware because I needed a way to distinguish between both nodes on my pc.
For me it seemed reasonable to assign both anchors different addresses and I would guessed the TDMA scheme could cause collisions with same addresses.
On the other hand, the first anchor is handing over a Ranging Config message to the tag, which includes PAN ID, the short address of the anchor, a short address assignment for the tag, and timing parameters for the ranging phase. Thus, with different addresses they probably cannot use the same Ranging Config and compete with each other.

But now I need a reliable way to distinguish between both anchors with a response I can query via serial interface. Address & PAN ID seem not to work, and any other setting leads to even more discrepancy…