1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

LTspice FFT not giving same EMC conducted emissions graph as actual graph

Discussion in 'Circuit Simulation & PCB Design' started by Flyback, Apr 15, 2017.

  1. Flyback

    Flyback Well-Known Member

    Joined:
    Jan 5, 2007
    Messages:
    1,989
    Likes:
    34
    Hello,
    The attached LTspice simulation is a buck converter which has the same waveforms as the buck converter in the document RDR-506 by power.com……

    RDR-506 report on 1.44w offline buck converter…
    https://ac-dc.power.com/sites/default/files/PDFFiles/rdr506.pdf

    Page 23 of this report shows the conducted EMC scan of this buck converter. However, the LTspice simulation of this buck converter , using its FFT function, does not give a graph anything like this. Do you know why not?

    Pdf schem of LTspice simulation also attached


    However, Another thing is that surely the conducted emissions graph on page 23 of RDR-506 looks wrong? I mean, if you look at the schematic of the buck converter on page 5 of RDR-506, it has a really heavy PI filter consisting of a 1mH inductor and a 4.7uF capacitor either side of it………for a buck converter of just 1.44W rating, that is an extremely heavy filter….how can all the Megahertz frequencies on page 23 of the RDR-506 document be there? If you look at the input current waveform, it is very smooth, there is utterly no evidence of the 10MHz waveforms which the EMC scan of page 23 suggests. Do you know what’s going on here?
     

    Attached Files:

    Last edited: Apr 15, 2017
  2. alec_t

    alec_t Well-Known Member Most Helpful Member

    Joined:
    Jul 10, 2011
    Messages:
    9,244
    Likes:
    1,218
    Location:
    Cardiff, Wales
    The P23 graph and the LTS FFT graph are not plotted with the same units, which makes comparison difficult. I understand 'dB' (as per the FFT), but how are 'dBuV' (P23) derived?
    Another difference is that LTs values are ground referenced, whereas the P23 graph is for a floating ground situation (if that is significant?).
     
    • Like Like x 1
  3. ronsimpson

    ronsimpson Well-Known Member Most Helpful Member

    Joined:
    Apr 17, 2007
    Messages:
    7,247
    Likes:
    961
    Location:
    Loveland, CO USA
    I might use SPICE to get idea of "noise" at the switching frequency. BUT at high frequency much of the problem is so related to the PCB and parts placement. (which is not in SPICE)

    I do not trust SPICE for "noise". I trust spice for building a filter. If you add a filter you can check what the noise reduction will be.
    Remember to build real world Cs and Ls in you SPICE filter. (watch for resonant frequency)
     
    • Like Like x 1

Share This Page