0
\$\begingroup\$

I struggled to compare the EMI filter needed for 24V to 5V non-isolated BUCK DC/DC (running 300KHz to 500KHz) to ensure it passes CE conformance. I wish to keep switching noise within the DC/DC area as well as filtering out external noise entering into DC/DC.

I have two choices: (a) Simple SMD ferrite beads and COG cap as pi filter (b) SMD X2Y filter (1206J1000103MXTE03). It claims benefits superior to (a).

I have spoken to support and they have very little application note to demonstrate the benefits of the X2Y technology. No PPT presentation. enter image description here Has anyone use X2Y and how they compare with C-L-C pi filter.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ I have just designed these in for a tech demo, well the Johanson's variant, Have a read of their application notes johansondielectrics.com/x2y-filter-decoupling-capacitors \$\endgroup\$ – JonRB Dec 19 '19 at 22:34
  • 1
    \$\begingroup\$ I too have been wondering if anyone has any actual experience with this. They have a ton of data for motors, but not as much for EMI filtering. \$\endgroup\$ – Aaron Dec 19 '19 at 22:35
  • 1
    \$\begingroup\$ These Caps have much better RF SRF ESR characteristics when used properly. What is the problem you want to solve? Ripple ? how much? Not a pun, but this is an undefined XY problem Johanson's parts are same as my answer . TY @JonRB \$\endgroup\$ – Tony Stewart EE75 Dec 23 '19 at 16:10
1
\$\begingroup\$

Filters are not trivial or simple with mutual coupling , unknown source impedances, complex loads and variable Q resonance and anti resonance.

Your schematic is already a 9th order LPF and non of the ESL are included.

The 10nF part you selected has an SRF of 100MHz and that is due to the ESL of about 1nH due to its small size. But if you add 0.5nH/mm traces that will lower the SRF and possibly react with all the other reactances. Good or bad all depends what emissions you have and need to control.

enter image description here

This is not a solution but demonstrates a lossy ferrite bead and cap with a SRF and a lower PRF anti-resonant peak. You must get s parameters & simulate or characterize your approved vendors and test with suitable test methods on your layout or experiment with other modelling tools such as Falstad's or using a scope with impulse testing using AC couple 50 ohm semi-rigid coax terminated at DSO with 50R.

enter image description here

For example here was a perfectly flat group delay Bessel filter destroyed by mismatching the load R. enter image description here

\$\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.