• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 63 subscribers
  • Views 374 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS65023 Workaround for DCDC1 reset needed?

Charlie Sweet
Prodigy 140 points
Other Parts Discussed in Thread: TPS65023, AM3517

Hi Folks,

TPS65023 datasheet indicates reset condition of DCDC1 is 1.225V if none of the workarounds are used (see page 43 of datasheet SLVS670J revised Sept. 2011) .  I am not using any of these workarounds but measure 1.204V at VDCDC1 output on all 12 boards.  VCC reaches 5V no more than 5ms prior to DCDC_EN being valid.  

Do I still need to use one of these workarounds?

Any feedback on the conditions when these workarounds are not needed would be helpful, especially timing for Workaround #1.

Thanks,

Charlie

  • 0
    TI__Genius 16870 points

    Charlie, 

    There is no timing minimum that you can use to say when a workaround is required and when it is not. However, if you are ok with the plus 2% on the DCDC1 output, you do not require a workaround. Do you have a limiting enable factor preventing you from doing workaround #1? 

  • 0 in reply to Michael Green
    Prodigy 140 points

    Michael,

    Workaround 1 does not appear to be an option.  I am sequencing the processor supplies based on the AM3517 recommended sequence.  As a result, I can't just tie DCDC1_EN to VINDCDC1.  The sequence is DCDC3 (1.8V) first, DCDC2 (3.3V)  second then DCDC1 (1.2V).

    We are having some issues with load changes causing transient DCDC1 supply voltage shifts so we would prefer to have DCDC1 centered at 1.2V.  AM3517 core voltage range limits dominate with its recommended range of 1.152V and 1.248V.

    Please confirm that there have been no die changes that would explain why my PMICs come up at 1.2V rather than 1.225V.  Are you confident that another lot of PMICs would also come up at 1.2V?

    Thanks,

    Charlie

  • 0 in reply to Charlie Sweet
    TI__Genius 16870 points

    Charlie, 

    It is a well documented and understood errata that the DCDC1 should default to 1.225V. You might have a voltage drop between the regulator and the load. This could explain why you are seeing 1.204V on DCDC1 instead of 1.225V. Since, you are powering the AM3517 processor, I would recommend workaround #2.

    You can set the DCDC1 to forced PWM mode before you enable it. This will make the DC voltage accuracy +/-1 %. Thus, during the time that you are +2% your dc voltage accuracy will remain under the +4% limit you have.  This would yield an accuracy between 1.211V and 1.237V. 

    Then, bump the voltage down to 1.2V via I2C. and finally enable the PFM mode again for light load efficiency. 

  • 0 in reply to Michael Green
    Prodigy 140 points

    Michael,

    Thanks for confirming the errata. 

    I do not believe there is a measurement erro and tried slowing down the sequencing by a few ms between the 1.8V supply and 3.3V supply and a few more ms delay until the 1.2V supply is enabled to get VDCDC1 to come up at 1.225V.  I am using a BK5390 precision multimeter (4 digit) directly across the output capacitor and find a value of only  1.2026 V.

    Since I do not have confidence in the startup level for future PCB builds, the I2C registers will need to be programmed.

    I can't set the DCDC1 to forced PWM mode before enabling it because the power-up process takes place before the processor boots up.  I can set DCDC1 to forced PWM mode after processor boot and then bump the voltage down to 1.2V via I2C.  If light load efficiency causes significant increase in power dissipation, I will enable PFM mode.

    Any other explanations for the DCDC converter not showing the documented errata would be appreciated.

    Thanks,

    Charlie

  • 0 in reply to Charlie Sweet
    TI__Genius 16870 points

    Charlie, 

    You will not need the forced PWM command since, that was to keep the error % of the DCDC1  low until you bumped it down to 1.2V. Since, you are booting up into 1.225V before yu can write anything to I2C, you are better off bumping the voltage down with your fist I2C command. Then there is no need to use the forced PWM  mode as you will well with in your plus and minus limits.