• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 4 answers
  • Subscribers 63 subscribers
  • Views 425 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.

BQ25713B: IL_AVG current, state machine or controller, I2C clock stretching

David Mckinney
Expert 6810 points
Part Number: BQ25713B
Other Parts Discussed in Thread: BQ25713

IL_Avg function on bits 4..3 of the ChargeOption3 (34h) Register setting?  It appears to be <10% of the 6, 10, or 15 amp settings and I would like to understand if this measurement senses and averages ACP/N 's differential voltage when Q1 is on to determine the average inductor current?  Is this correct?

s the BQ25713 endowed with an micro inside or is it a pure hardware state machine?

 

We would also like to know if it supports I2C clock stretching?

  • 0
    TI__Mastermind 25800 points

    Hey David,

    IL_AVG is a current clamp on the inductor current, as measured differentially between ACN and ACP. When you say it appears to <10% of this value, are you certain that IINDPM and VINDPM settings have been pushed out of the way? Your power supply also needs be able to handle that amount of power without crashing. 

    The charger has a mixed digital and analog state machine, but it is not a microcontroller. Internally, there is burned in memory to be read on power up and registers mapping to the digital state machine.

    The charger has a maximum timeout period associated with stretching, found in section 8.6 Timing Requirements, parameter Ttimeout, and note (1). 

    Regards,

    Joel H

  • 0 in reply to Joel Hernandez II
    TI__Expert 6810 points

    Hi Joel;

    Thank you for the quick reply.

    The Power supply is capable and here is an example of what we see on the EVM and our board:

    For the 6A step I get the foldback to occur at the  EVM at about 5.7A while on our design, it’s at 6.6A.  In the 10A step it's about 10.5A on the EVM and 11.5A on our board.

     

    What is the expected accuracy of the current settings? We are seeing 6.0A (fold back at 5.7-6.6A) 10A (fold back at 10.5-11.5A)

  • 0 in reply to David Mckinney
    TI__Expert 6810 points

    Hi Joel;

    Also, concerning the clock stretching:  All of this means that the BQ25713 shall see no I2C Clock low (due to a transaction between the master and another slave, not itself) for more than 35ms but the BQ25713 itself is incapable of generating a Clock stretch.  Is this correct?

    Note:  Still looking for:  What is the expected accuracy of the current settings? We are seeing 6.0A (fold back at 5.7-6.6A) 10A (fold back at 10.5-11.5A)

  • 0 in reply to David Mckinney
    TI__Mastermind 25800 points

    Hey David,

    The inductor current clamp is not intended to be incredibly accurate. There is not specification of the tolerance of each value, so take the values as typical. 

    The accuracy of IINDPM and ICHG are much tighter and are specified. The average current clamp on the inductor current is really a cycle-by-cycle current limit in the event the other input power management features are not utilized.

    And just to confirm, your measurement is based on the current through the inductor, correct? How are you measuring that current? Is it through a current loop? Is the charger in boost mode operation and you are measuring the ACx voltage? Or is it in buck mode, and you are measuring the system load current? 

    Regards,

    Joel H

  • 0 in reply to Joel Hernandez II
    TI__Expert 6810 points

    HI Joel;

    I actually measure the current by increasingly loading VSys with an E-Load and observe when VSys foldback begins to occur to assess IL_Avg when doing VBus=20V bucking to VSys=9V conversion.  I use a bench supply (Agilent E3633A of 20V, 10A for VBus) which has plenty of capability for these experiments.

    Question:  Can TI specify the IL_Avg's accuracy as the maximum DC I_VSys that can be supported for each of the three IL_Avg steps (6A/10A/15A) whilst guaranteeing that the IL_Avg mechanism doesn't [undesirably] kicks in?

  • 0 in reply to David Mckinney
    TI__Mastermind 25800 points

    Hey David,

    We do not specify the IL_Avg accuracy for any of the values. As stated earlier, these are only typical values.

    It would also be beneficial to directly measure the inductor current by adding a current loop in series with the inductor. In buck mode, the is constant most of the time. However, when looking at ISYS, you are looking at current after the inductor and output filer of the converter. Additionally, the Q4 HSFET2 must have a fresh pulse to its respective Bootstrap capacitor in order to operate correctly, meaning there are moments of discontinuity in order to keep that FET fully on. 

    Regards,

    Joel H

  • 0 in reply to Joel Hernandez II
    TI__Mastermind 25800 points

    Just as an addition to my comments above, the charger has not been characterized with that parameter beyond the typical value, so no direct information can be given about the the accuracy of the IL_AVG clamp.


    Regards,

    Joel H