This article briefly describes the pitfalls of proper filtering. We will understand some of the reasons why some filters that you may think will work often do not work when placed in an actual circuit. So... let's talk about why the filter failed.

Have you ever spent countless hours researching and locating the best filters that you think can meet your specific needs and have the best performance? And then found that once installed, it can hardly suppress any radio frequency radiation? One reason for this lack of performance may be that filter manufacturers have tested common mode (CM) or differential mode (DM) noise attenuation characteristics, but it is not clear from their specifications which one they use. If your emission problem is mainly CM, and the filter attenuation is specified for DM, then you will encounter the problem of successfully implementing the filter.

Another problem may be related to the standard used to test filter performance (usually MIL-STD-220). Generally, filters are characterized by their insertion loss (IL), expressed in dB. It is a measure of the load reduction due to the insertion of a filter at a given frequency. The IL of the filter depends on the source impedance and load impedance, and should not be declared independently of the terminal load/source impedance, but it usually complies with MIL-STD-220. The measuring instrument, source impedance and load impedance, input attenuator, and other components are specified to have an ideal characteristic impedance of 50Ω. Few power input circuits have the same ideal 50Ω impedance. The load impedance seen by the filter will not exactly match 50Ω. In addition, the input attenuator has a series impedance that can suppress any resonance. This is a problem because the attenuator used in the test is not present in the final product.

The current applied during the test is another problem. This test method does not require current to flow through the filter during the test, so it will not match the expected circuit of the filter anyway. If DC current flows, the inductance value in the filter may be different. When used outside of its specified current range, the choke will saturate, making it unable to provide its original expected impedance.

For these reasons, the perfect filter test situation does not necessarily follow the standard method, but is tailored to the specific EMI test power supply impedance, and uses the actual switching power supply planned for the product and runs under the expected current consumption. It should be empty Establish the insertion loss or attenuation characteristics of the filter under load and full load current levels to provide potential users with the best results and information.

Unshielded filter elements can also cause problems. When the filter component is unshielded and installed on a PCB that contains noise sources (such as switching power supplies or fast rise time digital logic circuits), noise is usually coupled to the filter component and the input connection of the filter at the same time. This unwanted crosstalk partially or even completely reduces the attenuation capability of the filter. A similar situation occurs when the input/output power lines of the filter are brought too close. This problem can be alleviated by shielding the power line filter and installing it on the wall of the device housing, and installing the input power connector on the filter housing. Keeping the input/output connections away from each other will also help.

When adding a low-pass filter to the I/O signal line, you may notice that the filter does not reduce radiation as you expected. The problem may be the presence of CM noise on every line and ground (return) path. Please note that the CM noise current flows equally on all lines, including the ground path. In this case, if a capacitor is used to suppress CM noise, it will only make the situation worse because it will bring the noise from the dirty digital ground to the clean signal line. For better results, try removing the capacitor or connecting the noisy digital ground to a clean (chassis) ground. In this case, a common mode choke may be a better solution than a capacitive grounding solution.

Don't forget the possibility of poor high-frequency response of the filter due to parasitic effects. We often overlook the fact that the nearly ideal low- and mid-frequency response of the low-pass filter will not continue to rise in frequency. Due to parasitic capacitance, the actual attenuation of the filter will be significantly reduced at the higher harmonics of the switching frequency (on the order of a few MHz). Look for ways to reduce these parasitic effects. Choose components with low equivalent series resistance (ESR) and keep the leads short, thick, and flat.

In real life, filter components will have tolerance, saturation, parasitics and coupling issues. With careful foresight, planning, and proper knowledge, these problems can be considered and mitigated so that you no longer wonder why the filter is not working as expected. Good luck in your future filtering work!

Don MacArthur is a guest writer for In Compliance Magazine. He has more than 30 years of experience in product development, EMC, testing, and product safety compliance. He has developed products for military, commercial and industrial applications.

Your email address will not be published.

Save my name, email, and website in this browser for the next time you comment.

In Compliance is the main source of news, information, education and inspiration for electrical and electronic engineering professionals.

Aerospace Automobile Communication Consumer Electronics Education Energy and Power Industry Information Technology Medical Military and National Defense