• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 4 answers
  • Subscribers 62 subscribers
  • Views 240 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

UCC28911: Flyback converter works only if GND of UCC28911 is connected to PE via a capacitor

Patrick S.
Prodigy 10 points
Part Number: UCC28911

Hello,

I use the UCC28911 as shown in the schematic, but it only works if the GND pin of the controller is connected to PE via a capacitor. The VS voltage at a successful start is shown in figure 2 and 3. A failed startup is shown in Figure 4 and 5.
It seems that the greater noise on the VS pin, when not connected to PE, triggers the OV protection.

Is there any way to fix this problem without requiring a PE connection for filtering?

Thanks in advance.

Figure 1

Figure 2: successful startup (PE connected to GND via a capacitor)

Figure 3: successful startup (zoom)

Figure 4: failed startup

Figure 5: failed startup (zoom)

  • 0
    TI__Guru 66135 points

    Hello,

    I just have received your inquiry and I am reviewing it.  I will get back to you within 24 hours.

    Regards, 

  • 0 in reply to Mike O'
    TI__Guru 66135 points

    Hello,

    What do you mean by PE?  Could you clarify?

    Regards,

  • 0 in reply to Mike O'
    Prodigy 10 points

    Hello,

    By PE I mean the protective earth conductor of the mains.

  • 0 in reply to Patrick S.
    TI__Guru 66135 points

    Hello,

    You should not have to connect the bulk capacitor to earth ground.

    The UCC28911 does use primary side regulation fault sensing which is susceptible to noise which you have discovered.  The most common issues with this is aux winding noisy which you have discovered.  The following link will bring you to an application that discusses how to resolve these issues with filtering and snubbing. https://www.ti.com/lit/pdf/sluaac5

    I hope this application not will help resolve your issue.  If it does could you please select resolved in the post?

    Regards,

  • 0 in reply to Mike O'
    Prodigy 10 points

    Hello,

    Thanks for your quick reply. I will go through the document as soon as possible and will select the post as solved if this application note solved the issue.

    Regards

  • 0 in reply to Patrick S.
    TI__Guru 66135 points

    Hello,

    Thanks for letting me know and thankyou for choosing TI.

    Regards,

  • 0 in reply to Mike O'
    Prodigy 10 points

    Hello Mike,

    I observed that the Vaux signal looks very clean, but the VS signal has a lot of ringing, which is strange because VS should only be a scaled value of Vaux due to the voltage divider (during the MOSFET off time).
    I tried to filter the Vaux signal anyway, as mentioned in the document, by adjusting the snubber circuit on the primary and adding an additional snubber parallel to the output diode (D117), but this showed no effect on the VS signal.
    Finally I was able to fix the problem by adding a 1nF capacitor between the primary and secondary grounds, but I still don't understand what is causing the problem.

    Do you have any idea why the VS signal has a much stronger ringing than the Vaux signal?

    Regards

    Figure 1: Vaux

    Figure 2: VS

  • 0 in reply to Patrick S.
    TI__Guru 66135 points

    Hello,

    The Aux signal appears to have ringing after the FET has turned off for 1 us.  That high frequency ringing could cause the design to shutdown based on the tlk_reset blanking period of the device.  Adding and adjusting the snubber and RCD clamp will reduce this noise.

    In regards to your VS ringing.  I would double check this to make sure it is real.  It should track the aux winding except below ground where the signal is clamped.

    One last thing when you are trying to figure out what caused the device to shut down.  You need to trigger on VDD, Vout, CS and Aux and look at the last three switching cycles to see what caused the controller to shut down.  Your waveforms look steady state and are not showing where the device shut down.

    Regards,

  • 0 in reply to Mike O'
    Prodigy 10 points

    Hello,

    With the adjusted and added snubber the ringing of the Aux signal gets reducet to about 400ns (Figure 2).

    Regarding the triggering: all plots are showing the last switching cycles. In Figure 1 we can see that the OVP is tripping the controller (Naux:Ns=1:1; Vout,ovp=14.7V). Pre-Load resistor is small enough.

    I double checked the VS ringing and it is definitely real (Figure 3). 

    I guess I found out the real cause of the malfunction. As you can see in Figure 4 and 5 (voltage between GND of controller and GND of secondary side) there is a ringing with the same frequency as at VS. In the PCB layout the ground plane of the secondary side ground is also present in the upper inner layer underneath the UCC28911 and therefore the noise of the secondary ground is coupling to the VS pin.

    To validate this i will have to wait for the next PCB revision. Until then I will use the filter capacitor between the GNDs.

    I will select the post as resolved for now and will report you if the new PCB revision fixed the issue.

    Regards

    Figure 1: Vaux

    Figure 2: Vaux

    Figure 3: VS

    Figure 4: Voltage between GND of UCC28911 and GND of secondary side

    Figure 5: zoom of figure 4

  • 0 in reply to Patrick S.
    TI__Guru 66135 points

    Hello,

    I am reviewing your inquiry and will get back to you shortly.

    Regards, 

  • 0 in reply to Mike O'
    TI__Guru 66135 points

    Hello,

    Your VS ringing is greater than the aux winding ringing.  These generally track.  Your ringing on the VS pin may be caused by poor layout.

    There may be too much trace with from your VS resistor divider (RS1, RS2), aux winding. Please note that the trace generally adds 10 nH per/inch.

    If these traces are too long it could cause issues with ringing.  You should try to keep these traces as short as possible.

    The UCC28911 gives recommendations on layout an example you can reference.  Notice the grounding and the position of the VS divider resistors (RS1 and RS2) and the aux winding.  They are very short.  Modifying your layout by these guidelines could remove your issue.

    Regards,