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.

UCC28881: UCC28881

Part Number: UCC28881

Tool/software:

Hi!

I’m working on a concept where, among other things, I would like to use the low-side Buck-Boost Converter configuration, by using UCC28881D Off-Line Converter from TI. See the attached schematics.

In the test, with attached measurement figures, I have an input voltage of 200V and I’m loading with constant current of 100-mA. I’ve observed that the converter, instead of operating in discontinuous mode, it is running in some kind of burst-mode. In this example I measure an output voltage higher than expected, around 13.7V, but I can easily bring it down to the desired level by tuning the feedback resistors. The good thing is, I can have a constant output voltage throughout the whole load range (0-220mA), but due to this burst mode, it will result in a high output ripple! I can see it stated in the datasheet that it might run sometimes in burst mode, without any further explanation. Can you explain why and when to expect it to run in burst mode?
I suspect it also to be sensitive to noise on the feedback line. Do you have defined any information on possible hysteresis for the feedback input signal? In the datasheet it is only given the threshold voltage (V_FB_TH).

In the attached measurements, the pink waveform is the inductor current, measured with a current probe and the cyan is the feedback (FB) signal (AC coupled to be able to see the slope, measured with a differential probe), measured at a testpoint eTP7.

With best regards,

Gunnar

  • Hi Gunnar,

    Thank you for reaching out.

    The UC28881 control law functions in a on-off manner. Whenever the Vfb is greater than the 1V threshold the pulses turn off and turns on when Vfb falls below 1V. The time constant of the output (R-C on the FB pin) could be adjusted to see if this helps but it inherently functions in this way.

    Regards,

    Harish 

  • Hi Harish,
    Thanks, I appreciate your fast feedback.

    Yes, an additional time constant will delay further the feedback threshold crossover, causing a "burst" of current and affecting the amplitude of the resulting ripple. By removing the capacitor in figure, the burst will be smaller, but not disappear. It might also be necessary to increase the output capacitance to further reduce the ripple. But I need to investigate that.

    With best regards,
    Gunnar

  • Hi Gunnar,

    Sure, please try adjusting the output capacitance.

    Please refer this equation mentioned in the datasheet which might be helpful to arrive at initial estimates.

    Regards,

    Harish