Other Parts Discussed in Thread: IWR6843AOPEVM, MMWAVEICBOOST,
Tool/software:
Seems like no other place to post this, so posting here.
Greetings, my name is Daniel Donnelly. I'm working on a sweat equity project (no initial pay). They would like to demo capabilities for 3D people tracking in a room (for home audio tuning / fall sensing), my "bosses".
What I've done so far is completely rewritten about 60% of the Industrial Visualizer provided by TI that's written in python for the xWR6843 devices. We had a hunch that it was Python slowing down and causing the app to crash. It turns out that the C++ version I wrote, probably because I used the same JSON internal data structures, turned out to be slower.
It was a PyQt5 to Qt/C++ conversion. I took me a good 150 hours of manual coding. But now that I have some auto-coding (Claude Sonnet 4) under my belt. If I ever convert Python to C++ again, I would definitelly make use of the AI tools where I can.
Anyhow, we determined that this crash is completely caused on the hardware side, and cannot be gotten out of unless you unplug/plug back in the IWR6843AOPEVM module. Then you have 5 minutes of Industrial Visualizer play time before it starts slowing down and eventually sending corrupt packets or nothing at all (I've forgotten).
So my plan is to debug the IWR6843AOPEVM's out-of-box demo using CCS. I have the code in a project, it builds, but when I use the MMWAVEICBOOST onboard XDS110, I am unable to get a breakpoint to work. You can set it when the code is not debugging. Then click debug, and they disable (temporarily only while it's running), so that's a problem, because breakpoints are my most used tool in my Debug arsenal.
Then when I plug in the XDS110 external debugger I have, I believe it's the original Texas Instruments XDS110 DEBUG PROBE (v1, but not written), with serial # TMDSEMU110-U, I can't begin to debug because of an error. It suggests I set TCLK, and I've tried both 100.0kHz as well as 1.0MHz, the default is 5.0MHz.
It says some other things too I don't quite understand, but here it is:
CS_DAP_0: Error connecting to the target: (Error -267 @ 0x0) The controller could not detect valid target supply. Check target JTAG connection and/or connection setting specifying voltage level. (Emulation package 20.3.0.3656)
If I can get this demo debugged, I would be glad to help others get a working demo out of this (I would share my bugfix guidance in the forums). I believe when researched this crashing Industrial Visualizer issue prior to this occasion, the forums said it was on the Python side, and that 5minutes is good enough for a demo before the crash. Also, this crash occurs consistently across three different client machines / three different IWR6843AOPEVM modules. They are simply wrong about this. The crash is in the hardware code. I am confident I can debug it having over 15 years of debugging experience under my belt, and I know C very well.
Claude Sonnet 4 suggests I probe the JTAG header with a multimeter, which I will try doing tomorrow. But if anyone knows just what is wrong, from experience, that would be great.
Yours truly,
Debug Master Dan