This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

MSP-EXP430FR5994: Power Consumption of Out-of-Box Example in LPM3 Mode

Part Number: MSP-EXP430FR5994
Other Parts Discussed in Thread: ENERGYTRACE

Hi,

  I have loaded the out-of-box example for the MSP-EXP430FR5994 board. Then I ran the example code with the energy trace active. After the board started and the statement '__bis_SR_register(LPM3_bits | GIE);' got executed, the energy trace showed power consumption settling at approximately 1.5 mWatt (see attached image). Now, as per my understanding, 1.5 mWatt is corresponding to 455 micro ampere (at 3.3 volts). So, my question is that this doesn't comply with the datasheet claim of 700 nano ampere power consumption in LPM3 mode (datasheet article no. 6.4 Operating Modes). Am I missing something somewhere?

  Using the launchpad in debug mode, I verified that the statement '__bis_SR_register(LPM3_bits | GIE);' is getting executed. Also, during the test, the jumpers RXD & TXD were removed, SBWTDIO & SBWTCK were inserted. Super capacitor jumpers were not populated. J7 was populated on board.

-

Thank You

-

Regards

Soumyajit

  • I don't know much about the FR5994 OOB example, but from what the Launchpad UG (SLAU678A) says I suspect it wasn't built as an exemplar of LPM3 consumption.

    The LPM3 power rating is best case, meaning that Everything is turned off. At minimum, in the OOB the UART is running, and maybe the ADC as well. [Ref Data Sheet (SLASE54B) Sec. 5.10]

    I also don't see any power gating in the schematics for the SD card, which will be a bit of a power hog even in standby.

    Example msp430fr599x_lpm3_02.c (Examples suite SLAC710A) might be a better starting point.

  • Hi Bruce,

     Thanks for your suggestion on 'msp430fr599x_lpm3_02.c'. I tried the code below from 'msp430fr599x_lpm3_03.c' in my project. But still the power consumption seems to be 3.1 milli Watts (See attached image). No SD Card was inserted in this test or the earlier one (hence no power consumption for SD card). The below code starts immediately at the beginning of the main() function. Any suggestion on attaining the actual claimed datasheet power consumption figures?

    -

    #if(POWER_CONSUMPTION_IN_LPM3_MODE_TEST)

       // LFXT, 1s interrupts

       WDTCTL = WDTPW | WDTTMSEL | WDTSSEL_1 | WDTIS_4;

       SFRIE1 |= WDTIE;                        // Enable WDT interrupt

       // Configure GPIO

       P1OUT = 0;

       P1DIR = 0xFF;

       P2OUT = 0;

       P2DIR = 0xFF;

       P3OUT = 0;

       P3DIR = 0xFF;

       P4OUT = 0;

       P4DIR = 0xFF;

       P5OUT = 0;

       P5DIR = 0xFF;

       P6OUT = 0;

       P6DIR = 0xFF;

       P7OUT = 0;

       P7DIR = 0xFF;

       P8DIR = 0xFF;

       P8OUT = 0;

       PJOUT = 0;

       PJSEL0 = BIT4 | BIT5;                   // For XT1

       PJDIR = 0xFFFF;

       // Disable the GPIO power-on default high-impedance mode to activate

       // previously configured port settings

       PM5CTL0 &= ~LOCKLPM5;

       // XT1 Setup

       CSCTL0_H = CSKEY_H;                     // Unlock CS registers

       CSCTL1 = DCOFSEL_0;                     // Set DCO to 1MHz

       CSCTL2 = SELA__LFXTCLK | SELS__DCOCLK | SELM__DCOCLK;

       CSCTL3 = DIVA__1 | DIVS__1 | DIVM__1;   // Set all dividers to 1

       CSCTL4 &= ~LFXTOFF;                     // Enable LFXT1

       do

       {

           CSCTL5 &= ~LFXTOFFG;                // Clear XT1 fault flag

           SFRIFG1 &= ~OFIFG;

       } while (SFRIFG1 & OFIFG);              // Test oscillator fault flag

       CSCTL0_H = 0;                           // Lock CS registers

       __bis_SR_register(LPM3_bits | GIE);

       __no_operation();

    #endif

    -

    Regards

    Soumyajit

  • I don't have a crystal on my Launchpad so I used LPM3_02 which only relies on VLOCLK. I also removed the LED and supercap jumpers and the SD card.

    EnergyTrace reported about 0.0017mW average, so maybe around 0.6uA.

    For a second opinion, I put it on an LPM01A which reports 0.727uA average, including 1/sec blips up to about 3uA. (I removed all the J101 jumpers for this one.)
  • Hi Bruce,

    "I don't have a crystal on my Launchpad so I used LPM3_02 which only relies on VLOCLK" - As per the MSP-EXP430FR5994 schematic, I find that the external high speed crystal is not soldered on the board but the 32.768 kHz crystal is populated on the board (You may refer to Q1 & Q2 in the schematic). So, did you mean that the high speed crystal in not there on your board or the 32 kHz one? Did you remove (de-solder) the 32 kHz crystal from your board?

    "I also removed the LED" - Did you remove the RED LED (LED1 on schematic) by removing the J7 jumper or you de-soldered the GREEN LED (LED2 on schematic) as well?

    I will try to replicate your setup and see if I can attain 0.0017mW as is the power consumption in your case.

    -

    Thanks

    -

    Regards

    Soumyajit

  • I have not physically modified my board. I removed the LED (P1.0) jumper. I removed the supercap jumper. I removed the SD card. I also removed the TXD/RXD jumpers (which you mentioned earlier).

    I defer to you on the 32kHz crystal, since I don't have my board here. I saw an empty space where I expected to see a 32kHz "can", but maybe that was an illusion. Watch out for DNIs ("Do Not Install"). Also keep in mind that some Launchpads populate the 32kHz crystal but not the caps, to allow GPIO access to the XIN/XOUT pins. Using the VLOCLK example sidesteps all these questions.
  • Aha! The crystal I never knew I had.

    msp430fr599x_lpm3_03.c, same test conditions: EnergyTrace shows 1.0uA. LPM01A shows 0.989uA with 1/sec blips up to about 3.25uA.
  • Hi Bruce,

     Thanks for your help in the matter. Would you mind running the attached code in your launchpad? I have attached the code I have used to test at my place. This code I have tested at my end resulting into a power consumption of 2.6 mW. I am curious to know whether you will also get the same results.

    -

    Thanks

    -

    Regards

    Soumyajit

    BlinkLED_MSP430FR5994.zip

  • I was unable to run this program in the debugger. It starts the Launch, then reports "Disconnected: Unknown", and I don't know what that means.

    Based on inspection, I think the problem is that you don't have a WDT ISR (WDT_VECTOR) defined. I see some code, but it's compiled out based on

    > #if(POWER_CONSUMPTION_IN_LPM3_MODE_TEST == 0)

    If it were included, it would just spin, which might result in your 2.6mW.

    There's quite a lot (80%?) of dead code here, and I'm not sure where you're headed with it. If you want to run lpm3_03.c, just create a new project and copy/paste the .c file into it. Resource Explorer might have an importable project.
  • Hi Bruce,

      Following your suggestion, I copied the 'lpm3_03.c' file content to a new CCS project. Running the code, the power consumption still remains at 2.6 mW. Am not understanding what might be going wrong. LED P1.0 jumper is removed as well as RXD & TXD jumpers. Super Capacitor jumpers were removed too. No SD card in SD card slot.

      I am attaching the code/project I just used in my test. Requesting you to run the same at your end and see if still you are getting the low power consumption (which I am not getting at my end).

    -

    Thanks

    -

    Regards

    Soumyajit

    msp430fr599x_lpm3_03.zip

  • Running this on a Rev 1.2 Launchpad (CCS v8.1) with SDcard removed, supercap jumper removed, LED (P1.0) jumper removed:

    Using Run->Free Run, I see Mean of .0028 mW, 0.9 uA over 10 seconds.
    LPM01A says Mean of 0.986 uA over 10 seconds.
    Using Run->Resume , I see Mean of 0.485 mW, 147.9 uA over 10 seconds.

    I threw in the Run->Resume experiment just in case you were doing that. The numbers don't match yours, but Free Run does make a visible difference.
  • Hi Bruce,
    It seems to be a dead end for me. I failed to attain the desired results for the low power consumption of this MSP MCU on the launch pad. Will try to arrange experimental setup with actual MCU chip on a custom made PCB and run the test again. Hope power consumption figures will be as expected.
    Thanks for all your help Bruce.
    -
    Regards
    Soumyajit
  • I will initiate a new thread for my findings with the custom PCB based MSP MCU power consumption evaluation.
    -
    Thanks
    -
    Regards
    Soumyajit
  • Hi Soumyajit Das,
    Do you need additional support from our side? I will close this thread
    Regards
    Kostas
  • Hi Kostas,

     Thanks for your comments. I am still waiting for the PCB tests (power consumption related) to be conducted. It's of severe priority that the MCU consume the power as per figures indicated in the datasheet. This is because the entire project depends on the low power consumption of the MCU (along with additional circuitry). In this regard, I will need to solve the power issue at all cost. I am expecting that the power consumption tests will be over by mid February 2019. You may go ahead closing the thread (in this case I will open up a new thread to discuss the power consumption issues faced in the custom PCB).

    -

    Thanks

    -

    Regards

    Soumyajit Das

**Attention** This is a public forum