• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 641 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.

Noise problem with the PMP8363 Reference Design

ramiro.utrilla
Prodigy 30 points
Other Parts Discussed in Thread: UCC28710

Hello,

With the aim of studying how the UCC28710 controller works, I made a custom version of the PMP8363 reference design using a slightly bigger PCB and adding test points for the analysis. I bought exactly the same components indicated in the bill of materials.

Currently, I have the circuit working pretty well, I got 5 volts at the output and it's able to supply up to 2 amps. My only problem is that the switching noise is much bigger than the one you got on your test results. For example, at a load of 2 amps your switching noise is smaller than 40mVp-p, and I got 400mVp-p, which is an order of magnitude larger.

I reviewed all the design equations, I played with the values of resistors, capacitors and so on but I didn't get significant improvement.

I also added a snubber circuit (10 ohms and 1 nF) over the secondary diode which reduces by half the noise, specially the resonance after the spike but still it's too big.

Finally, I unmount almost all the secondary, leaving only the diode and a 820uF capacitor. In this case the switching noise increase up to almost 7 Vp-p.

I don't understand why the switching noise is so high if everything except the PCB is the same. Is it possible such a big difference just by modifying the layout?

I don't know how to deal with it, if I have to work on secondary filtering or I have to face the problem on the primary.

Please, any idea, application note or comment will be welcome. I'm currently working with a 220Vrms@50Hz input, but I need it for universal range.

Do not hesitate to ask me for captures, more test information or whatever if you think it could be helpful.

Thanks in advance.

Greets!

  • 0
    TI__Expert 5900 points

    Hello Ramiro,

    First, would you check your test setting? on the probe, the area composited by the probe and gnd wire should be as small as possible. The probe would be located at C2. The bandwidth would be 20Mhz.

    Second, if the setting was right. can you attach the waveform of ripple? we may want to see whether the high value caused by the noise or low frequency oscillation.

    Regards

    Kening

  • 0 in reply to Kening Gao
    Prodigy 30 points

    Hello Kening,

    First of all thanks for your attention.

    I've repeated the measurements given your recommendations. Why the oscilloscope bandwidth have to be limited to 20 MHz?

    I attach four waveforms, all of them taking at 2A load.

    Pictures 1 and 2 were taken with 400ns/div and represent what I called switching noise.

    Pictures 3 and 4 were taken with 10ms/div. In this case it's possible to see how the switching noise is modulated in amplitude by a 100 Hz signal, that I think is consequence of a coupling of the rectified input.

    Pictures 1 and 3 were taken with 20 MHz bandwidth limitation while pictures 2 and 4 were taken at full bandwidth (500 MHz).

    Do not hesitate to ask me for more captures, test information or whatever if you think it could be helpful.

    Regards

    Ramiro

    Picture 1:

    Picture 2:

    Picture 3:

    Picture 4:

  • 0 in reply to Kening Gao
    Prodigy 30 points

    Hi Kening,

    I tested the PMP8363 reference design with the PCB provided by TI and following your test settings recommendations.

    The noise performance notably improves but still is too high in comparison with yours and the amplitude modulation effect remains.

    I attach a picture with the test setting and some waveforms with the measured noise at 2A load.

    I still don't know how to deal with this, if I have to work on secondary filtering or I should face the problem on the primary. I'd really appreciate if someone could help me with this issue.

    Thanks in advance!

    Sincerely,

    Ramiro