• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 81 subscribers
  • Views 1284 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

MSP430F5x and EnergyTrace in LowPower mode

peg
Intellectual 520 points
Other Parts Discussed in Thread: ENERGYTRACE, MSP430F5438A, MSP430F5437A

Hello,

I'm trying to trace my consumption with EnergyTrace (Nice Tool ;O)

But I know that my sensor get in LPM and I can verify it, measuring consumption with an external dedicated material out of JTAG debugger. I know that my sensor reach  most of the time to about 10uAmps.


Viewed with "MSP FET Flash Emulation Tool" i'can't get lower than 800uA.


I read a lot about that and here some informations:

First with my device (MSP430F5438A Rev H): I know that i can't get EnergyTrace++ (but i don't think i need it ?).

Then, I can't set the option "Low Power mode settings Enable ultra Low Power Mode LPM5 debug ?" I don't thing I need that as I'm only using LPM4 at most ?

I use the Free run mode.

In the "slau157am", page 31, The second to last Q/A talk about JTAG pin that should be correctly configure according to their states in free run mode. BUT is that could explain more than 500 uA of over consumption and Where can i get their correct configuration ?


In advance Thanks if anyone as ideas about all this.

ltra Low

  • 0
    TI__Genius 16165 points

    Hi there,

    One way to truly measure the energy without any debugger is to actually run EnergyTrace by itself without the debug session. In the latest version of CCSv6.1.2 or newer, you could now run EnergyTrace standalone without the debug session by clicking the EnergyTrace icon (Blue circular looking) in CCS. This way, we do not run the risk of introducing any debug power consumption.

    peg said:
    "Low Power mode settings Enable ultra Low Power Mode LPM5 debug ?"

    This feature has been removed from the MSP430F5x/6x family devices. It is only available in the MSP430FR5xx/FR6xx devices. Same applies to EnergyTrace++ where it is only available on all MSP430FR5xx/FR6xx family devices.


    Regards,

    William

  • 0 in reply to William Goh
    Intellectual 520 points

    Hi William,

    Thanks for the first answer.

    I tried to run Energy trace without debugger on CCS 6.1.2.

    But noway, the consumption of my "slepping sensor" (some tenth of microsAmp) stays shown by Energytrace as some tenth of miliAmps ...

    I'm JTAG alimented on the device, I use a msp430f5438a or msp430f5437a mcu.

    Notice that externally alimented and using a precise currentmeter i can see the tenth of microAmps.

    If you've got any other ideas ?


    Regards,

    Pierre-emmanuel

  • 0 in reply to peg
    Intellectual 520 points

    Hello,

    I've got some progress, I can now mesure low consumption in sleep mode using.

    - Using CCSV6.1.3

    - As required "no debugging" active

    - And to avoid any unwanted consumption on my JTAG connexion, I've only connected VCC and GND from the MS-FET JTAG

    But I still have a strange behavior.

    Consumption is not really the same that the one measured by an "Current meter" positionned between my VCC and my device. (See image joined). If anyone have an idea to explane that ?


  • 0 in reply to peg
    Guru 46710 points
    Bandwidth of your currentmeter is much higher than EnergyTrace. In other words - EnergyTrace averages measurements. After all it is never advertised as precision instrument.
  • 0 in reply to peg
    Guru 16361 points

    One thing that seems to be affecting your results is a known bug in the plotting of the power graph in CCS' EnergyTrace display: https://e2e.ti.com/support/development_tools/code_composer_studio/f/81/p/448872/1621699#1621699

  • 0 in reply to Robert Cowsill
    Intellectual 520 points
    Oh yes,
    You seem to be right ... same behavior.
    I'll soon make new measure to verify.

    How can we know if bug is known, traced and will be corrected ?
  • 0 in reply to peg
    Guru 16361 points
    Later in that thread a TI employee confirms that they've repro'd the issue and found the root cause. I don't know what the timescale is for fixing it; the best thing I can suggest is that you post to that thread and ask for an update from TI.

**Attention** This is a public forum