Conducted emission voltage failed at switching frequency of boost

[will_fung]

Hi everyone,
I have a conducted emission voltage failure at the switching frequency of BOOST circuit.
below circuit is the power rail filter, and the UBATT was connected to the BOOST input. the switching frequency of BOOST is 400KHz, it has two phase and delayed 180 degree, so i get the failure at 800KHz. it exceeded 6dB.
I tried to change the 470nH inductor to 3.3uH, it gets worse. i also add more caps, such as 10uF, 4.7uF and 470nF, there is no improvement.
i aslo changed the 560uF caps to 820uF, there is no improvement too.

do you have any ideas for this CEV failure? thank you in advance.

 

[dick_freebird]

Your simulation circuit probably fails to include ESR and ESL.

Your ground may be the problem, harsh return current spikes
there can travel through case and shields. Common mode
filters usually have a ground choke too?

Your 1.2MHz breakout is even worse indicating that the
hard switching ("squareness" of edges) may be more the
problem. Fixing that will cost you some efficiency perhaps.
Or, you need input decoupling -at the powertrain- that
can shorten the input supply-switch-GND loop better
(appropriate ESR/ESL/charge capacity).

 

[will_fung]

Your simulation circuit probably fails to include ESR and ESL.

Your ground may be the problem, harsh return current spikes
there can travel through case and shields. Common mode
filters usually have a ground choke too?

Your 1.2MHz breakout is even worse indicating that the
hard switching ("squareness" of edges) may be more the
problem. Fixing that will cost you some efficiency perhaps.
Or, you need input decoupling -at the powertrain- that
can shorten the input supply-switch-GND loop better
(appropriate ESR/ESL/charge capacity).

thanks for your reply.
I will double check the ground to make the current return loop as short as possible.
that is not 1.2MHz, actually, it is 1.152MHz, it is not the 3rd harmonic of 400KHz, it is the harmonic of classD amplifier. we have fixed it.

 

[Velkarn]

Hi!

What is the power of this converter?
https://imgur.com/a/a1uZb maybe try this filter circuit
try to increase value of emi suppression cap on input, if there no improvement, so it is common mode noise perhaps... maybe try common mode choke

 

[will_fung]

Hi!

What is the power of this converter?
https://imgur.com/a/a1uZb maybe try this filter circuit
try to increase value of emi suppression cap on input, if there no improvement, so it is common mode noise perhaps... maybe try common mode choke

The converter is ISL78227, here is the datasheet.
https://www.intersil.com/content/dam/Intersil/documents/isl7/isl78227.pdf

The forward current of ABS6 is too low. I tried to increase the caps to 820uF, but there is no improvement.

 

[Velkarn]

no, i did not mean to use abs6 - this is my wrong picture cut, i mean use two inductors.
and what is converter power?
for example, here is shem of filter, i used for 100w 24v converter https://imgur.com/a/rd4aE

 

[Easy peasy]

you appear to be ignoring common mode emissions (CM) depending on your heatsinking and grounding arrangements these can easily dominate DM (differential mode ) noise in the LISN...