• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 61 subscribers
  • Views 4247 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.

Output impedance of EMI-filters

Constantine Gershunsky
Prodigy 110 points

Hello everyone,

I design an EMI filter to which 3 identical DC-DC converters are connected in parallel and I want to ask you how to calculate output impedance of EMI filter for this case? 

P.s. Each of DC-DC converter's have internal LC-filter.

  • 0
    Prodigy 5 points

    Hello Constantine,
    This is a good question.

    Here is a good paper that sums it up.
    www.ti.com/lit/an/snva538/snva538.pdf

    The filter output impedance will be the same no matter how many converters are connected to it.
    Each converter will typically have a NEGATIVE input impedance, and impedances in parallel will act like resistors in parallel.

    The main goal is to make sure the EMI filter output impedance is much less than the summed converter impedance.
    See figure 3 in SNVA538.

    Each filter upstream must have lower output impedance than what is connected to it.
    So the Main filter must have lower output impedance than the summed input impedance of the DCDC's internal filter.
    You must be able to quantify the DCDC internal filter details.

    It is probably best to model the filter as above so you can see the impedance vs frequency curve.
    You can use TI-TINA spice based simulation program to model this.

  • 0 in reply to Ed Walker
    Prodigy 110 points

    Thanks for paper's!

    Maybe I'm wrong, but the internal filter is the last stage of the emi filter, and in the case of three dc-dc converters with parallel connection, we have the last stage of the emi filter, which consists of three parallel LC filters, right?

  • 0 in reply to Constantine Gershunsky
    Prodigy 5 points
    Yes Constantine, there will be 3 LC's in parallel on the output of the main input filter.
    So the impedance of the main single input filter must be a magnitude lower than the parallel impedance the 3 converters present.
  • 0 in reply to Ed Walker
    Prodigy 110 points

    Well, I see. So, equivalent circuit of 3 LC's in parallel will be L1C1 filter where L1 = L/3, C1 = 3*C?

  • 0 in reply to Constantine Gershunsky
    Prodigy 5 points
    Yes I think that would be correct Constantine.
    3 Inductors in parallel = L/3, 3 capacitors in parallel = 3C

    I look at it as frequency dependent impedances in parallel.
    1/Zeq = 1/Z1 + 1/Z2 + 1/Z3
    For 3 of the same LC in parallel Zeq = Z/3
    So if the impedance curve is given for the converters, just divide it by 3.

    One thing to consider would be that the L and C actual value will have some tolerance.
    This would tend to vary the Q of the equivalent parallel LC combination, which should be good.
    However if all the Q's line up exactly it may cause an unstable input filter design.
    Parallel 3 LC Circuit Q can be lowered by adding some series resistance at the main input filter output.

    Have a great day.
  • 0 in reply to Ed Walker
    Prodigy 110 points
    Thank you for your help!