ISO 11452-4 Compliant Input Filter

Is there any ISO 11452-4 compliant input filter?

At the moment, I have used a 22nF capacitor followed by 3.1uH ferrite bead for a linear current regulator. The load current is 1.2 A.

The linear current regulator has a low side MOSFET with an op amp which takes the feedback from the current set resistor.

The load current goes to zero at 200 MHz during the BCI test of 1 - 400 MHz.

Please suggest a solution? Do I need to change the value of the input capacitor or add few more capacitors?

Thanks.

It's easy to guess that you have insufficient RF filtering. Did you ever made a hand calculation about the expectable RF input voltage and current during this test?

Seeing a test failure at discrete frequencies is usually a problem of resonance, either in the wiring harness which would be normal, or in your filter which must not happen.

FvM,

Thanks. It would be helpful if you could guide to some literature explaining how to perform the hand calculations.
Besides, at the moment, I am using 22nF ceramic cap very close to the input followed by 3.1uH ferrite bead which is followed by a diode for reverse polarity protection, which in turn is followed by 2.2uF ceramic capacitor.

Would increasing the 2.2 uF to 10 or 20 uF help? I have performed some simulation which shows that replacing 2.2uF with 20uF increasing the filter attenuation.

I have only cursory knowledge of the BCI test setup. As far as I understand, it's injecting current into the wiring harness by a RF current transformer. As a first step, you would sketch an equivalent circuit to understand how the interference is coupled into your device under test, and how much common mode + differential mode signal will be expected at different circuit nodes respectively branches. I guess it's not even clear if it's primarily exposed to the power input or may be between power lines and chassis ground.

Changing 22 nF to µF is not the likely solution for >100 MHz references. More likely you have already resonances with capacitor ESL. So the solution could be using multiple capacitors in parallel and a multi stage filter. You didn't yet talk about a capacitor behind the ferrite inductor.

The following is the input configuration: 22nF||3.1uH bead - S1J Diode || 2.2uF.

I am planning to change 22nF to 100 nF and place 10uF||100nF caps instead of 2.2uF.

Besides, there is no separate chassis ground. There is only Vbattery and battery ground.

What is your opinion?

Besides, there is no separate chassis ground. There is only Vbattery and battery ground.

Click to expand...

BCI test is coupling common current (Vbat + Gnd wires arranged in parallel) or also differential mode? No separate chassis ground means the DUT is isolated from chassis? But there's at least a ground coupling capacitance?

Need to see a photo or layout screenshot to understand if the capacitors are still effective at 200 MHz.