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.

Ripple Noise at high line for UCC28711D

Other Parts Discussed in Thread: UCC28711

Hi TI,

I found a higher Ripple & Noise at high line 264Vac by using UCC28711 to supply +15/-15V outputs at 0.75A each.

Why is this so? Any way to solve this?

Attach the Ripple Noise waveforms at 264Vac and at 90Vac for your reference.

Thank you.

Regards,

KK

+15V R&N Waveform at 90Vac 50Hz 0.75A load each.

-15V R&N Waveform at 90Vac 50Hz 0.75A load each.

  • Hi KK,

    Your post indicated that there was a R&N issue at 264 V, and that plots were posted of R&N at 90 V and at 264 V. However, the plot are both labelled as 90 V, one on the +15 V output, the other on the -15 V output.

    Can yoo please clarify, and add plots of R&N on both outputs, at 90 V and again at 264 V? (i.e. 4 plots).

    Can you also confirm the size of the bulk capacitance in uF so we can estimate the ripple?

    Thanks,

    Bernard

  • Hi Bernard,

    Sorry for the wrong typing for the title of second waveform.

    It should be '+15V R&N Waveform at 264Vac 50Hz 0.75A load each'.

    The bulk capacitance used is 100uF.

    Here I attach 2 more waveforms at -15V output as per your request.

    -15V R&N Waveform at 90Vac 50Hz 0.75A load each.

     

    -15V R&N Waveform at 264Vac 50Hz 0.75A load each

    Thank you.

     

    Regards,

    KK

  • Hi TI,

    I am still waiting for your reply.

    Thank you.

    Regards,

    KK

  • Based on the ripple waveforms, it can be seen that the high ripple is at the AC frequency, so it looks like ripple break-through for some reason. The intial post with the mis-labelled plot seemed to indicate that this was happening at low line only, so it seemed to indicate that the controller was hitting an "end-stop" of some kind.

    However, the clarification and further plots now confirm that there is no issue at 90 V, but only at 264 V.

    Our suspicion is that the gain of the line feed-forward peak current adjust might be too large, causing the peak current to get reduced too severely as line voltage increases - this effect will result in too little peak current flowing vs. the control law demand level, forcing the frequency to increase to compensate. At heavy load, this could cause the frequency to hit the maximum end-stop, resulting in ripple break-through as seen.

    Another possible mechanism for this ripple could be caused by the CC-mode inception point, which will decrease with increasing line voltage if the line feed-forward gain is too high. So what may be happenning here is that the CC is coming in too early at maximum line.

    This could be checked by dropping either the load or the line voltage slightly - if the issue is due to CC mode, then the ripple should suddenly disappear at a slightly lower load or a slightly lower line voltage.

    Similarly, if the DC output voltages are monitored, and the load is increased beyond 0.75 A on both outputs (using load set to CR constant resistance mode), both outputs should roll-off just beyond the 0.75 A level, indicating operation of the CC limit.

    I would suggest that the value of the series in series with the CS pin (RLC) should be checked against eq. 27 in the UCC2781x datasheet. If this resistor is too large, the peak current will be reduced too much as line voltage increases.

    If the issue remains after RLC is adjusted, then the Rcs resistor should be increased slightly (say 10%) to compensate for the fact that the internal Iocc limit in the IC does move slightly depending on the RLC value. With the new increased Rcs value, a revised RLC vlaue should be calculated agian using eq. 27.

    I hope this information will help to solve your problem, please be sure to post your results so we can see that the issue is resolved or not.

    Thanks,

    Bernard

     

  • Hi Bernard,

    I powered up the board and found that the AC ripple issue is gone without any modification made.

    I adjusted and checked the line voltage and the probe on ripple measurement but still can't find why.

    It is so strange.

    Regards,

    Kok Khuan