CCS/IWR1642BOOST: Battery application for IWR1642BOOST board

Hello,

I assume you are trying to use the IWR1642 EVM for the Timer settings.

a) the existing EVM dissipates power in the startup&boot, load code, and data output phases.   One of the suggestions to track the timing of these changes is to add the GPIO-0 as an output (following the MSS coding of the TIDEP-0091), the first MSS task is to set the GPIO-0 at the start of MSS programming.   The second event you are tracking is when the MSS Data Logger UART is outputting data.  some how this MSS or DSS output, would trigger the GPIO-0 set to 0 in the MSS.

b) so using an oscilloscope you can now record the AR_nRST signal 0->1 when 3.3v > 2.0v (or force this with the EVM pushbutton).   You then record GPIO-0

which starts out as a logic 0, and you record the 5v input current with a current sensor / probe.

c) in your evaluation, there is another signal called WARM_RESET for testing you can check to see how much time it takes to release from 0 -> 1 after 10used from nRESET 0->1.  Once you do this test, please replace C22 with a .022uf cap.  You will notice a lower boot time.

d) if you have a battery application, the quality of the RF power supply, can be adjusted in steps, depending on how much RF noise is acceptable for your application.   

    d1) use what you have - the PMIC_EN can be sequenced 0->1 to turn on the PMIC and mmwave sensor, 1->0 turns it off

           for reliable operation, your host processor needs to control AR_NRST_MCU, 0->1 after PMIC power is stable, PGOOD 0->1

    d2) modify the EVM to incorporate swra577 (you still need PMIC_EN control)

      note: you need to modify your CLI commands in software to include RFLDO Bypass, see mmwave sensor User Guide.

       for reliable operation, your host processor needs to control AR_NRST_MCU, 0->1 after PMIC power is stable, PGOOD 0->1

e) Like the TIDEP-0091, you can use an MSP432R401 EVM Launchpad, with the IWR1642 BOOST EVM Boosterpack

       the MSP432 has a separate 3.3v power from the PMIC 5v.  the MSP432 needs to be always powered.

f) based on your measured 5v current, and the 3.3v Trace from the MSP432 you can figure out your continuous draw, and the wake up measure, and 

send the data back to the MSP432 - note: for TIDEP-0091, the data is sent over SPI instead of UART.   

    PMIC En -> 0->1 , HOST Processor outputs AR_nRST_MCU as 0, 

    HOST Processor  waits for PGOOD 0->1, waits 1ms, then outputs AR_nRST_MCU as 0->1 (mmwave sensor start), 

    note: like TIDEP0-0091 you can hardcode the mmwave sensor commands, so that the IWR1642 can use base commands,

    the SPI/UART can add commands, this requires more SPI?Control UART communication (check mmwave SDK drivers in release package)

    IWR1642 Boot AR_nRST_MCU 0->1 until GPIO-0 0->1 TBD1 ms

    note: if SPI is used, the HOST INTERRUPT 0->1 indicates to Host Processor that communication is ready 

    IWR1642  performs software task, GPIO-0 1->0 TBD2 ms

    if IWR1642 commands come from Control UART (normal Out of Box demo) or SPI (TIDEP-0091) there is a configuration time, at completion GPIO-0 0->1

    TBD3 ms 

    IWR1642 returns serial data, or use SPI TBD4 ms

    Host Processor / msp432 receives mmwave sensor data 

    when Host Processor is done, sets AR_nRST_MCU as 0, then PMIC_EN = 0

    Sorry there is no simple answer.  You need to have the Host Processor state machine, control, and measured currents to be able to estimate the mmwave

     sensor power draw in each state.   Then you can estimate the number of measurements per day, figure out the total energy needed, then how much solar

     charging capacity is needed for keeping it operating on cloudy or rainy or non-daylight use.

     Good Luck with your project

      Radio Joe

Wow, it's a very helpful answer, Joe!
I'll try to follow all your steps and then I update the thread, depending from the result.

Thanks.

Regards.