0
\$\begingroup\$

I am measuring conducted emission at the input of a SEPIC converter.

At first, the emissions are only about 45 dBµV at switching frequency.

But when I put my finger on the SEPIC (coupled) inductor, the emissions go up to about 65 dBµV (!).

Even worse, if I move the input cable close to the SEPIC (coupled) inductor, the emissions can go up to 70 dBµV.

If I stick a piece of copper foil on top of the inductor and connect the foil to GND, then the emissions go only up to 48 dBµV.

In a larger production, it is not feasible to use copper foil. Also, I cannot stop the cable from touching the inductor. Therefore:

  • What else can could be done to solve my problem?
  • How can this behaviour be explained?
\$\endgroup\$
  • \$\begingroup\$ What are the characteristics of the inductor? this looks like a parasitic inductance related issue, with a coupling between your inductor windings and the input cable through the parasitic inductance. \$\endgroup\$ – Sclrx May 25 '18 at 8:00
  • \$\begingroup\$ The inductor in question is a Wuerth 744878101: 2x 100 uH. Unfortunately, it's parasitic inductance is not indicated. \$\endgroup\$ – B. Kraemer May 25 '18 at 8:04
  • \$\begingroup\$ I forgot to mention an important detail: The emissions also go up if another conductive item (pliers, or even my finger) comes close to the inductor. It has not necessarily to be the cable. \$\endgroup\$ – B. Kraemer May 25 '18 at 8:05
  • \$\begingroup\$ could you please edit the additional details in your question to make it easier to answer? \$\endgroup\$ – Sclrx May 25 '18 at 8:13
  • 1
    \$\begingroup\$ I couldn't find the datasheet by that code. Is it a shielded inductor? Shields kill a bit of inductance but they stabilize capacitive coupling, besides considerations on radiated emissions. A kind of snubbed across windings might help; for example Cy capacitors are added between primary and secondary of isolated switching power supplies to avoid the secondary side is floating. \$\endgroup\$ – andrea May 25 '18 at 20:37
0
\$\begingroup\$

There are two different mechanisms at work:

  1. By putting a finger on the coil, I generate common mode noise. The finger serves as one capacitor plate, the Earth potential around me as the other. The electric field radiation is much stronger now.

  2. By putting the input cable on the coil, I introduce differential mode noise in the cable.

This behavior is not genuine to SEPIC converters, but to every switching converter. The inductor may be shielded, but that shield may not be sufficient if something touches it.

Solutions:

Make sure the cable does not touch the inductor.

Other solutions which work are a copper foil on the inductor or a metal shield around it. But the copper foil is just glued on the inductor and can detach within lifetime, thus producing shortcuts anywhere in the device. And a metal shield is big and expensive.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.