• Locked Locked
  • Replies 3 replies
  • Subscribers 62 subscribers
  • Views 972 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.

TPS65150 EMC

William Liu
Prodigy 190 points
Other Parts Discussed in Thread: TPS65150

We are using TPS65150 in automotive application. We can't pass the EMC test because the switching noise. We are using 2-layers. 

Could you please give us some comments about the layout ? How to modify the layout to make the noise smaller ?

Attached picture is the layout.

We measure the voltage difference between Boost power ground pin and power pad, we can see big switching noise(1.2M). We are using "ground ring" to avoid the radio noise, and also using snubber to damp the ring. There

is not much improvement. But when we remove the negative output, the result is much better(attached picture). 

 is 

  • TI__Guru 58225 points

    Hello William,

    It is almost impossible to review a layout without schematic and placement. If you could provide these, it should be possible to check the layout.

    Can you pass the EMC test when removing the negative output? If yes, it might help to add a series resistor of a few ohms to DRVN. It is interesting that you get higher spikes when the negative charge pump is connected. I think this means that the spikes are on the GND plane and not on the signal. Your GND connection is cut a lot.

    You could as well try to use Zener diodes for the charge pumps if you are not using some already. The spiking is for example higher when using BAV99 compared to BAT54.

    Another possibility is adding snubbers. Sometimes it seems to be helpful to have a small cap directly at the cathode of the boost diode to GND. If you have a snubber on the SW node available you can check this as well.

    Best regards,
    Brigitte

  • in reply to Brigitte
    TI__Prodigy 190 points

    Thanks so much for your reply. I have added 20ohm in series. The result looks much better. 

    I have another question that : Is there any consideration about the charge pump diode ? Reverse recovery current ? Recovery time ? 

    As customer is using SMAB33 in their circuit, the result is better when customer change the diode to KDS160, It seems the reverse recovery current is 2mA(SMAB33) and 0.5uA(KDS160), does it have so much influence while the current is not large ?

    Thanks so much !

  • in reply to William Liu
    TI__Guru 58225 points

    Hello William,

    When there is a recovery current of 2mA, this is already 10% of an output current of 20mA, so 2mA recovery current is high for a charge pump. Normally we are using BAV99 or BAT54 for the charge pump diodes.

    Best regards,
    Brigitte