• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 158 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.

UCC28064A: Phase fail function

Hirotsugu Kobayashi
Mastermind 9540 points
Part Number: UCC28064A

Tool/software:

Hi team,

When VIN becomes high, the phases of A and B become misaligned.

From what point does the phase fail function start to function, and how does it correct it?

  • 0
    TI__Mastermind 40520 points

    Hello Hirotsugu-san, 

    The phase-fail function detects whether one of the phases of 2-phase interleaving stops switching. 
    It is intended to prevent the remaining phase from becoming overloaded when the output power level requires 2 phases to be operating.

    The phase-fail function does not detect whether the interleaving phase-shift is less than 180 degrees. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Mastermind 9540 points

    Hi Ulrich,

    I understand the Phase fail function. What is the cause of the phase shift as the input voltage increases?

    It is the same configuration as the circuit example in the data sheet, so where do I need to adjust it?

    Or is it possible that one of the phases is misdetected?

  • 0 in reply to Hirotsugu Kobayashi
    TI__Mastermind 40520 points

    Hi Hirotsugu-san, 

    The internal phase detector works by detecting the difference in timing of GDB with respect to GDA and generating a duty-cycle signal with respect to the Phase-A switching frequency.   The duty-cycle signal is filtered through a low-pass filter to generate a voltage.  The difference of this voltage from a center voltage (representing 50% duty, which represents 180-degree phase shift) generates small timing adjustments to the Phase-A and Phase-B switching periods in opposite directions in order to "push" the phases closer to 180-degrees interleaving. 

    Some error gets in so interleaving is never exactly 180-deg, but usually very close. 

    Additional detection error can happen by turn-on and turn-off time delays of super-junction MOSFETs. 
    High-frequency switching using low inductance makes the effect of time delays worse, especially at high line voltage where even a small dt can result in a significant dI from V/L.  Differences in boost inductance (La vs. Lb) and MOSFET delay timing (dta vs. dtb) can lead to greater difference in switching frequency between Phase-A and Phase-B, so the phase-detector must create a larger or smaller duty-cycle signal to create the appropriate timing adjustments to keep the two phases at the same frequency.    
    This leads to lower phase shift, farther away from the ideal 180-degrees. 

    To reduce this issue (keep closer to 180-deg interleaving), I believe it is important to match the two phases with tighter-tolerance inductance and less difference in MOSFET characteristics.  Also match the ZCDA and ZCDB timing, since GDA and GDB are triggered by the ZCDx signals. 
    This is especially important when the switching frequency is high, input voltage is high, and inductance is low. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Mastermind 9540 points

    Hi Ulrich,

    I understand that small inductance values ​​are likely to cause problems.

    What are the recommended values ​​when designing under the following conditions?

    　Vin 85~164V

    　Vout 390V

    　Pout 550W

    　Fsw_min 45kHz

    When calculated with an efficiency of 92%, it is about 100uH, and I am using 110uH, but would it be better to lower Fsw_min and increase the inductance value?

  • +1 in reply to Hirotsugu Kobayashi
    TI__Mastermind 40520 points

    Hi Hirotsugu-san,  

    For Vin range, do you mean "85~264V"?  164Vac is an unusual upper limit.  I will assume 264Vac. 

    If you set Fsw_min to 30kHz, L at Vin_max (= 264Vac) = 149uH, and a nominal value of 150uH can be used.

    If you set Fsw_min to 25kHz, L at Vin_max (= 264Vac) = 179uH, and a nominal value of 180uH can be used.

    I recommend not to go lower in frequency in order to maintain some margin to the <20kHz audible range. 
    Higher inductances with the same peak current ratings will be physically larger than lower inductances, but will have better interleaving phase shift.  

    Regards,
    Ulrich 

  • 0 in reply to Ulrich Goerke
    Mastermind 9540 points

    Sorry, that was a typo. 264Vac is correct.
    I'll try increasing the inductance value based on your advice.