Other Parts Discussed in Thread: CC1352R, SYSCONFIG
Hi all,
I've been using the TI15.4 stack for a while now, with the old SDK CC13x2-CC26x2. Now I've been porting my applications to the new SDK and I'm having trouble. I thought it was my application but it's not.
To try and figure out the issue I've imported and tested the TI15.4 Stack sensor example, for both CC13x2-CC26x2 SDK V5.20 and for CC13xx-CC26xx V6.10.
The only changes I did (on both of the projects) was selecting frequency of 868MHz, changing the application for 60s reports, 10s polls, and 14dB Tx forcing VDDR, activating the opts: CUI_DISABLE, AUTO_START and disabled TEMP_SENSOR, and LED jumpers off as well as all unnecessary jumpers to be able to obtain a clean current measurement.
I've have ran both examples on the same CC1352R launchpad, one after the other. While having a collector example running on another CC1352R (collector project is kept the same during both tests).
I'm, able to achieve an average 1min current of 38uA (or 5uA for 1s avg) on the TI15.4 stack sensor version from the old SDK, and I'm getting 120uA (or 85uA for 1s avg) on the TI15.4 stack sensor version from the new SDK, same code (your example code with the changes I mentioned).
What is going on here? why the higher consumption?
UPDATE:
I found out the issue to be related with UART. I forgot to mention that in both tests I have an external board in low power mode (another launchpad without jumpers, connected to the TI15.4 stack sensor launchpad only through 3V3, GND, TX and RX pins, running a simple UART reply code). The external launchpad is simulating a simple low power UART physical sensor, while having the main launchpad running the TI15.4 sensor code and reading data from the UART physical sensor whenever a report event happens. Using the TI15.4 stack sensor code from old CC13x2-CC26x2 SDK V5.20 it works with minimal consumption. However when using the TI15.4 stack sensor code from CC13xx-CC26xx SDK V6.10, the consumption is way higher. It seems that somehow the new SDK/drivers dont completely deactivate the RX pin, and since it is connected to the TX pin of the other board, it creates a residual yet significant consumption. Even by removing the UART/UART2 driver from sysconfig, the consumption is still present on the demo from CC13xx-CC26xx SDK V6.10. If I remove the external launchpad that is simulating the UART physical sensor and all 4 connections, then both TI15.4 stack sensor from both SDKs consume the same thing. Therefore it is related to the difference in the SDK/Driver behaviour regarding the UART TX/RX pins.
