• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 64 subscribers
  • Views 185 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.

TPS7H4001-SP: Multi-Phase Current Balancing

John Lally
Intellectual 600 points
Part Number: TPS7H4001-SP
Other Parts Discussed in Thread: TPS7H4001QEVM-CVAL

Tool/software:

Hello E2E,

    The below applies to a critical long duration space flight mission.

We’re using a 4 channel TPS7H4001-SP based design to realize a 72A, 0.8V supply.

Per the current balancing concerns raised in section 4.9 of the 4 Channel Evaluation Board User Guide (TPS7H4001QEVM-CVAL), we’re working to balance the 72A load current across the 4 channels.

We have the following questions:

  1. What design elements factor into current imbalances across the channels and whose variances across the channels should be minimized, e.g.: GMps(COMP to Iswitch gm) of the TPS7H4001-SP, output inductor tolerance, etc…
  2. Can the TPS7H4001-SP be ordered to reduce the GMps variance from part to part, e.g. ordered from same lot/wafer, screening GMps to a tighter tolerance range than what’s in the datasheet?
  3. The TPS7H4001-SP datasheet shows the High Side switch current limit as 25A typical, what is the minimum limit? We need to make sure we don’t exceed the minimum limit in the presence of current imbalances across the channels.
  4. How do we determine during the design process that our 4 channel design will satisfy our design requirements without resorting to an overly conservative approach as described in section 4.9 of the Eval. Board User Guide?
    1. The testing shown in the Eval Board User Guide is useful to confirm current imbalances satisfy design requirements on a particular as built unit, it does not prove that the design will satisfy design requirements over all as built units.
    2. We’re concerned our EDU units might produce passing current imbalance test results per the recommended eval board test method, only to find our Flight Units fail because of GMps and/or inductor differences between Flight & EDU builds.
      1. Do you have any recommendations on how to prevent this situation from happening?
    3. In the 2nd paragraph of section 4.9 of the Eval. Board User Guide it details margining down the per converter current from 18A to 13.5A to 9.8A for a total of only 39A. I don’t quite follow the arithmetic & datasheet specs that were used to arrive at those numbers can you provide that arithmetic and specs ?

Thank you -John

  • 0
    TI__Mastermind 27075 points

    Hey John,

    1. The main contributor is the variation in GMPS, though smaller things like layout do contribute

    2. Ordering from the same die is an option, however screening for GMPS is not.

    3. The minimum limit is not tested for.

    4. How you do the worst case analysis for the paralleling of current is up to you and the limits you want to set.

    The math is meant to show how different uses of the electrical specifications come up with different answers for what the maximum value will be.
    GMPS is a measurement that has different numbers depending on what temperature profile you use.
    It also changes based on if one device is output 18 A, what do you assume the other values are at?

    How you want to do your WCA calculations is up to your mission.
    We have seen many in the past, so specifics often change.

    Thanks,
    Daniel

  • 0 in reply to Daniel Hartung
    Intellectual 600 points

    Thanks Daniel,


    We have some clarifying questions that we expect will help with our understanding and enable us to move forward, still a little confused on applying GMps to current imbalances and total current.

        1.) If the main contributor to imbalance is GMps variation then are the following examples approximately correct for a 4-channel design with all 4 converters at 25C, & Load Current=72A:
            a. GMps of all 4 converters =Max=50, current on each channel=72/4=18A?
            b. GMps of all 4 converters= Min=29, current on each channel =72/4=18A?
                   i. Likely diminished transient response as compared to all GMps=50?
                   ii. We observe a lowering of crossover frequency with lowering of GMps using WCA model.
            c. GMps of one converter=Max=50, GMps of other 3 converters=Min=29:
                   i.    Current on Max GMps channel = 26.28A ? (Note: We would Not attempt this, exceeds 25A typ HS current limit)
                   ii.  Current on Min GMps channels=15.25A on each ?
                   iii. Total current = (26.28 + 3x15.25)A = 72A ? (Assumes HS current limit not tripped)


        2.) Is the datasheet GMps spec. (28min at -55C to 52max at 125C) an End Of Life spec inclusive of TID=100krad, Aging (i.e. emulation of 15yr operation at 65-95C), part to part manufacturing differences etc?
            a. What’s a reasonable expectation for GMps variation of parts from the same wafer at start of life?

    Thanks for your help Daniel, -John

  • 0 in reply to John Lally
    TI__Mastermind 27075 points

    Hey John,

    1) Assuming all minor differences are also the same, yes the same GMPS would equate to having the same output current on each channel. GMPS is a gain in the frequency response, so as gain goes down, the crossover frequency does as well.

    a) Correct

    b) Correct

    c) Hitting the HS current limit aside you are correct. The difference in current is certainly proportional.

    2) The GMPS specification is guaranteed by the SMD datasheet to be tested to the QMLV specification of MIL-PRF-38535 this includes Groups A-E tests as outlined in the specification. The specification seems to line up with what you are asking about, but I would double check it for any differences your mission might have.
    (Group C testing is usually called Life test, Group A is electrical, and Group E is radiation)

    As far as variation of the device at the start of life, lets talk about that offline.
    May I email you at the email you created your account with?

    Thanks,
    Daniel

  • 0 in reply to Daniel Hartung
    Intellectual 600 points

    Yes thanks Daniel, please email me directly.  -John

  • +1 in reply to John Lally
    TI__Mastermind 27075 points

    This has been taken offline