• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 39 subscribers
  • Views 552 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.

Original question:

PMP22343: PMP22343 design issue

PMP22343: Problem Ringing and waveform jiggling at D1, D2 primary switching node, huge harmonic content at output.

Stefano Alfredo La Spina
Prodigy 240 points

Part Number: PMP22343

Tool/software:

Hi.

I have built a +/- 30V PSU following suggested PMP22343 design following the specs rules however I get a very noisy output but what is worst the waveform is oscillating in a sort of manner.

Here attached the schematic:

I have tried to use same layout with 4 layer ground planes with sufficient vias

however during tests (used a 0,125mA LOAD as suggested by the TIDT158.PDF test reports provided)  I got following problems:

  • Comparing to the test output voltage ripple is higher due to spikes on the positive and negative waveform at the output caps

VDD:

VSS: even worser

  • The switching frequency is constantly oscillating between 149 and 151 kHz.

  • Primary switching node at D2 shows some sort of ringing (?) (input is +12V).

At no load applied the voltage is following:

sometimes it gets worse:

here at 125mA output:

here at 300mA output:

  • Same for Primary Switching Node D1

  • But what is worse is that the waveform observed ad D1 and D2 are constantly going back and forth in an oscillating manner. Here a magnification of the voltage:

Oscillation:

Magnification 1:

Magnification 2:

Magnification 3:

  • Finally the spectrum analysis of the output voltage of both VDD and VSS rails shows a huge harmonics of the 150kHz switching frequency:

VDD:

VSS:

Could You please help me find out what is going on?

best regards and thanks.

Steve

  • 0
    TI__Intellectual 1570 points

    Hi Steve,

    PMP22343 was designed for a 5V input, are you only noticing the oscillations with this higher input voltages?

    The power stage components should be fine in terms of voltage and current stresses, but you may need to recalculate the loop compensation. The current-mode control should minimize how much this changes, but the bandwidth was optimized for the 5V case.

    I recommend reducing the gain of the error amplifier by lowering R33 and adjusting the Ccomp and Chf values accordingly. Can you capture a bode plot as well?

    Thanks,

    John

  • 0 in reply to John Dorosa
    Prodigy 240 points

    Hi John. Thank You for the email. No the situation does not get better with a 5V input. What could be the cause of such a great harmonic content? Thank You

  • 0 in reply to Stefano Alfredo La Spina
    TI__Intellectual 1570 points

    Hi Steve,

    The jitter may be due to loop compensation, I would recommend slowing down the loop to help with that.

    In regards to output voltage ripple, it does look similar to the original test data. This topology is very similar to a flyback converter, where the output current is pulsating to recharge the outptu cap every cycle to maintain output voltage. If really low output noise is required i would recommend adding a ferrite bead or linear regulator to dampen this ringing.

  • 0 in reply to John Dorosa
    Prodigy 240 points

    Thank You John. I am waiting for an isolation transformer to come in oder to break the loop and make proper measurements. By the way when You say: slow down the loop you mean : "lowering R33 and adjusting the Ccomp and Chf values accordingly" as previously suggested?

    About the harmonic content, are there specific formulas to use in order to choose the right ferrite beard? 

    Thank you alot. Regards

  • 0 in reply to Stefano Alfredo La Spina
    TI__Intellectual 1570 points

    Correct, the gain of the amplifier is proportional to the R33 value and you will have to be careful with the pole/zero placement when changing this value.

    For choosing ferrite bead I look for a high impedance at the frequency of the ringing. This tends to prefer smaller case sizes, but you also have to consider the current stresses. Typically i will try to use the smallest possible that would not be a concern for current rating.

  • 0 in reply to John Dorosa
    Prodigy 240 points

    John thank You so much, but your answer has put me in some trouble and raised more open problems: Spectrum Analysis of the output. I was supposing that the 150kHz multiple harmonics that I see on the output were related to the switching of the internal transistor.  Now you say that it is related to the ringing effect.. could you please explain this part please? best regards

  • +1 in reply to Stefano Alfredo La Spina
    TI__Intellectual 1570 points

    Sorry for the confusion, the harmonic content at 150 kHz and increasing harmonics are due to the switching frequency. That will be somewhat inherent to any switching converter until you add additional filtering. I was referring to jitter, which refers to when the pulse width varies cycle-by-cycle (seen on the switch node waveforms). Some jitter may happen, but the 300 mA case looked like an unstable system.

  • 0 in reply to John Dorosa
    Prodigy 240 points

    Thank you, John. Will try it. Great support, as always. Regards