CC2652R: Sensor Controller gets stuck on custom board

What is the difference between your board and the launchpad and which changes have you done in software to reflect the difference? 

Which DIOs have you defined to be used by the sensor controller? 

There is not significant difference between the lauchpad and my board. The componants are the same, I changed one decoupling capacitance as we are using a DC/DC converter for the power supply which can not cope with a output 22µF capacitance. So I used an 10µF capacitance instead. Additionally, the layout is changed as we are using 0805 sized capacitances instead of 0402.

I am using DIO26 for the external resistor needed for the TDC, DIO25 is connected an capacitive sensor. Those are the only pins I am using beside DIO6 in the main CCS program for lighting up a LED.

"Additionally, the layout is changed as we are using 0805 sized capacitances instead of 0402.": I would not expect any good RF performance using 0805 on 2.4 GHz. The parasitics and trace lengths will have too much impact. 

I suspect the DIO mapping. Could you show a snippet from Sensor Controller Studio showing the mapping and also the relevant part of the schematic?  

TER said:

"Additionally, the layout is changed as we are using 0805 sized capacitances instead of 0402.": I would not expect any good RF performance using 0805 on 2.4 GHz. The parasitics and trace lengths will have too much impact. 

True, the RF path to the antenna contains only 0402 parts. But the rest not.

Just to confirm, you are using a 7x7 packet?

TER said:

Just to confirm, you are using a 7x7 packet?

Yes, I think so. I am using this one: https://www.mouser.de/ProductDetail/?qs=rrS6PyfT74c6cwwDzpwNfQ==

Not fully sure what is casing this. To check that this is actually caused by the DIOs, are you able to change the code (for debugging) to not uses DIOs and see if you see the same issue? 

Hi,

I can use the DIOs to toggle a LED. Is that enough to ensure that the DIOs work?

When I first tested today with the originalo programm, it got stuck a bit further at waitOnTdcclk. Can it be that there is a problem with the clock for the TDC? Am I right that it uses the 48MHz external clock, if I write in my code to use the XOSC? Is there an easy way to test the clock with a programm? I couldn't find any other procedure using the clock except from the TDC

- What have you set tdcSetCntSource to? 

Note from the TDC documentation in SCS:

"The TDC counts at either 96 MHz (generated by RCOSC) or 48 MHz (generated by RCOSC or XOSC). Note that using XOSC as counter clock source can interfere with fast XOSC startup for the application and radio stack, and can potentially break radio protocol timing. "

As I said, I am using now the 48MHz XOSC. I would like to use the 96MHz RCOSC clock, but with it, my program crashes, probably because of the RTOS using it aswell. I could not find any solution for that.

e2e.ti.com/.../893843

https://e2e.ti.com/support/tools/ccs/f/81/t/920264

If you have any idea, please let me know.

But that should not be different between your board and the launchpad?