Shielding Help

Discussion in 'General Electronics Chat' started by blazin, Mar 8, 2014.

  1. blazin

    Thread Starter New Member

    Jun 4, 2012
    Trying to attenuate an EM field generated by equipment with the following waveform outputs:

    240kHz sinusoid repeated at 39kHz 8% duty cycle, 3500 Vpk-pk, 120 W peak.

    390kHz sinusoid randomized burst centered at 28kHz, 21-35kHz range, 9000 Vpk-pk, 120 W peak

    High voltage, relatively low current

    Spectrum Analyzer with near field e-probe shows frequency components all below 100MHz centered at low frequencies listed above.

    The EMI generated by this equipment is disrupting a SERDES application causing the pair to lose lock. Copper tape shield attempt did not resolve the issue.

    Source to shield distance will be approximately 40 mils, shield to system distance another 20 mils. Shield thickness was 1 layer of tape? maybe 1mil? A thicker shield? ie 10x 20x? different material?

    Any suggestions? Thanks!
  2. Papabravo


    Feb 24, 2006
    I cannot visualize what you are describing.

    How is it that sinusoids have duty cycles? I'm not familiar with that situation.
    How does copper tape work? Does it surround the signal conductors? Is it grounded? At how many places?

    Does SERDES stand for serialize deserialize? What kind of drivers and receivers are you using?

    I guess a picture is worth several thousand words.
  3. alfacliff

    Well-Known Member

    Dec 13, 2013
    I had a similar problem with a new machine here, a vfd was putting put huge spikes common mode on everything. a change of data lines to the fiber origionally specified worked. someone hadnt read the specs and had put in cat 5 cable.
    the spikes were due to the current draw while charging the caps in the vfd. otherwise known as power line distortion.