ADUM5242 EMI

Hi
I have a question ADUM5242.

Ask about ADUM 5242ARZ-RL7.
Connect between the output pin (pin 8: Viso) - output GND (pin 5: GNDiso)
Although it is a capacitor, 0.1uF is connected in the recommended layout on page 13 in the specification sheet.
As a result of measuring EMI, when we set this capacitance value to 0.01 uF good result is obtained and we want to change capacitance value.
As the output load, 3 kohm is 10 kohm in parallel between the output power supply and output GND at maximum.

1. Is there a lower limit capacitance as output capacitor?
Is there any concerns such as defects when 2.0.01 uF is used?
3. By changing the output capacitor with the attached characteristics from 0.22 uF to 0.01 uF
EMI noise reduction is seen. Is there anything you can guess about this factor?
Best Regards
HOD
  • HOD:

    1. Bypass capacitors are recommended as 0.1uF to obtain the load regulation as shown in the datasheet spec tables and load transient response in figure 10. 

    2. Using 0.1uF value close to the VISO and GNDISO pins and a 0.01uF would ensure that good operation and load regulation.  Have you tried a combination of 0.01uF and 0.1uF in parallel?

    3. Using 0.01uF may have lower inductance and better impedance characteristics at the 300MHz switching frequency of the isoPower than the 0.1uF capacitor and may improve EMI emissions.  Can you share the emissions data?

    4.  Layout and location of the bypass capacitors close to Viso and GNDiso may effect emissions. The datasheet also recommends:

    In cases where EMI emission is a concern, series inductance may be added to critical power and ground traces. Discrete inductors should be added to the line such that the high frequency bypass capacitors are between the inductor and the ADuM524x device pin. Inductance can be added in the form of discrete inductors or ferrite beads added to both power and ground traces. The recommended value corresponds to impedance between 50 Ω and 100 Ω at approximately 300 MHz.

    Regards, Brian 

  • Hi

     

    Thank you for your answer.

     

    Although it is recommended to use a combination of 0.1 uF and 0.01 uF as a decoupling capacitor of the power supply, when examining the frequency characteristic of the impedance of the capacitor of 0.01 uF, the resonance frequency is about 100 MHz and the inductivity is shown at the frequency of 300 MHz.
    When examining about 1000 pF, the resonance frequency is close to 300 MHz which is the same as the driving frequency
    It is in.
    The impedance of 300 MHz is also smaller for 1000 pF. Is 0.1uF and 1000pF better as a combination?

     

    Best Regards

    HOD

  • HOD:

    I agree with you that the characteristics of 1000pF look better at 300MHz than the 0.01uF, and using 1000pF and 0.1nF may be the best choice.  Try this first, but the characteristics of the PC board may effect the results and you can also try other values between 0.01uF and 1000pF.

    Regards, Brian 

  • Hi

    Thank you for your answer.

    4, the impedance value of the inductor added to the series is 300 MHz
    Although 50 Ω to 100 Ω is recommended value and it is listed, why please tell me.

     

    Best Regards

    HOD

  • HOD:

    This is a good question. The ADuM524x datasheet was the first isoPower and data channel part, and I we have learned more about the use of ferrites on the output power and ground pins.  I would recommend you read the app note AN-1349 for the latest on ferrite resistance vs frequency, layout and location to reduce emissions.

    Regards, Brian