• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 30 subscribers
  • Views 645 views
  • Users 0 members are here
Support feedback
Options
Options

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.

Notable increase in power when migrating from Tiva 123 to Tiva 129

Ryan Thompson
Prodigy 110 points

Hi,

 

We have recently migrated from a Tiva 123 to a Tiva 129 for a battery powered product.  We have compared the Tiva power consumption between the the two processors, and the Tiva 129 is consuming about 50mW more power when active (74mW for Tiva 123 vs. 121mW for Tiva 129).  The functionality between the two processors is identical, and the Tiva 123 is operating at 50MHz while the Tiva 129 is operating at 48MHz.  Note that all unused peripherals are turned off.  Being a battery powered product, this increase in power is very significant for our product's battery life.

Can you provide an explanation for the notable increase in power consumption and any suggestions for hardware/software changes to minimize or eliminate this increased power consumption?

 

Thanks for the support.

 

Ryan Thompson

  • 0
    Prodigy 110 points
    We have taken some additional measurements in the lab, and the results are shown below.

    Clock Frequency Active Power Bootloader Power
    TIVA123 50 MHz 79 mW 52 mW
    TIVA129 48 MHz 134 mW 111 mW
    TIVA129 30 MHz 107 mW 94 mW


    As you can see, the bootloader power is much higher for the Tiva 129 (52mW for Tiva 123 vs. 111mW for Tiva 129). Any help is much appreciated.

    Ryan Thompson
  • 0 in reply to Ryan Thompson
    Prodigy 10 points
    This is a very critical discovery for us and greatly appreciate any help understanding and reducing power consumption of the new Tiva 129.
    Thanks,
    -Terry Sherman
    Project Manager
  • 0
    Guru 47900 points

    Ryan Thompson said:
    any suggestions for hardware/software changes to minimize or eliminate this increased power consumption?

    Appreciate the depth & detail w/in your report.     That said - many are confounded by the (unwanted) change to the "SysCtlClockGet() function - are you quite certain that your 129 (really) is clocking @ the frequency you list?    (no disrespect intended - hundreds here have suffered from this hiccup)

    Might it make sense to "disable" one of your operating/enabled MCU peripherals at a time - and note which one(s) - are the biggest power hogs?     Armed w/that knowledge - best plan of attack may emerge.    (i.e. we attempt to break the "larger" complaint/problem into its (more manageable) constituent parts.)

    We know nothing of your board's usage and/or functionality - but if battery life is especially critical - perhaps peripherals may be enabled (only) just as/when required - and otherwise disabled.    This presents execution penalties - unknown is their severity.

    Are you properly treating any/all "unused" GPIO?    Unused pins should not float (surely you know that) but perhaps a change from "weak pull-up/pull-down to the opposite" may reduce drain...

  • 0 in reply to cb1
    Prodigy 110 points

    Hi,

    We do understand the change to the SysCtlClockGet() function, and we have confirmed the processor is indeed clocking at 48MHz.

    We have gone through an exercise to disable the peripherals, and we are confident the extra current draw is not related to an active peripheral compared to the Tiva 123.

    We took the power measurements during bootloader to try to get an 'apples-to-apples' comparison of the Tiva 123 to the 129, and in bootloader the power was still much higher for the Tiva 129.

    Ryan Thompson

  • 0 in reply to Ryan Thompson
    Guru 47900 points

    Ryan Thompson said:
    we are confident the extra current draw is not related to an active peripheral

    That confidence - minus (any) supporting detail - renders "outsider" help impotent...   Your inclusion of such "pre-measures" would have saved my time/effort...

    As to bootloaders providing "apples to apples"- unless clocking & all "active MCU participants" are very near equal - "non-apple" may intrude...