Below are the questions asked during the live event, along with their respective answers.
Q: Can multiple filters be used in series?
A: Yes, but it is preferred not to, due to unforeseen resonances that can occur between the two filters. Better to use one filter with all components integrated into one housing.
Q: You mentioned that the component values may not be as important the higher in frequency you go. Can you expand upon this?
A: The impedance characteristic of a choke increases as you go higher in frequency. Theoretically the impedance continues to increase with frequency. However, a real choke includes a certain winding capacitance. When the resonance point is reached, the impedance of the choke will reverse and decrease as the capacitance in the windings become more significant than the inductance. On the other hand, a capacitors impedance decreases as frequency increases. However due to the inductive behavior of the connection leads capacitors reach a point of resonance after which their impedance increases again. The effectiveness of grounding can decrease as frequency increases. At 1kHz a small ground wire or screw may read as 0 ohms. But at high frequency only the outer surface of the wire is conducting, and impedance will increase with frequency.
Q: So, if I’m at a test house and find my unit fails, what information would you like to have from me that will help determine the correct filter to use?
A: If you can provide the conducted emission scan ideally with and without a filter this will allow us to see the noise profile, the frequencies, and what the filter did or did not do. If you can also provide the type of equipment you are filtering and what makes up the equipment. Particular items have similar noise profiles and therefore an appropriate filter can be found. Any terminal connection preferences, voltage rating, current ratings, temperature and size constraints will help.
Q: We are deploying Android tablets in industrial environments. These tablets are using capacitive touch screens and these screens are very sensitive to common/differential pollution. However, the nature of the pollution is always different depending on the machines inside the factory. a) How to address this variable pollution? b) How to measure in a low cost/rapid way EMI? c) Could we do it with a basic oscilloscope? Thank You! Right now, we are using FN2090 series but would like to understand better the theory.
A: For this answer I would need to talk directly with you and understand the product better and the error(s) you are seeing. Even though you mention different noise pollution causing a failure, the failure might be related to the same design issue. Once the failure is better understood this can help direct us to the problem circuit. Additional filtering and/or improved grounding can then be implemented to mention a few. Grounding could include ground planes and trace routing not just wires and housing connections. Depending on the noise and error an oscilloscope may catch the noise/transient. It sounds like the issue is when you are exposed to an RF field, so to test you might be able to duplicate the noise with a hand-held radio or walkie talkie. Trigger the device next to your unit. Unsure how many wires (power/signal) your unit has that filters could be helpful.
Q: Common choke vs Y capacitors, what is the relative impact for common mode pollution filtering?
A: While a choke can provide a bigger impact, a capacitor may be all that is required. Of course, you will need an appropriately designed ground plane or trace able to carry the high frequency noise effectively. A single cap will be smaller and more cost effective than a choke. A choke will provide more impact but is larger. A choke with a cap will provide optimal mitigation.