This technical brief explains two distinct types of electromagnetic interference.

Related Information

 

Electromagnetic interference can often be ignored when we’re designing prototypes or working with development boards. But EMI is an important topic in the context of real-life electronic devices and systems, and engineers are responsible for ensuring that a circuit can operate properly in the presence of expected levels of EMI and that it does not generate excessive amounts of EMI.

I tend to associate EMI with wireless interference, and this is not too surprising considering the name: it’s called electromagnetic interference, and we naturally associate this with electromagnetic radiation. But as you have surely inferred from the title of this article, this is only one type of EMI.

 

What Is Radiated EMI?

I like to emphasize the fact that PCBs are filled with time-varying signals that propagate into space as electromagnetic radiation, whether you want them to or not. I also like to point out that every conductor is an antenna that is capable of both transmitting and receiving signals.

When you’re trying to implement RF communication, wireless transmission and reception are good things. In most other cases, you’re generating or receiving noise. When we discuss this noise in the context of its effect on a separate circuit or device, it becomes “radiated EMI.”

Circuits both generate and receive radiated EMI, and the operational environment of the device determines which of these realities presents a greater challenge to the engineer. If you’re designing a high-precision sensor board that must be in close proximity to a brushed DC motor or a wireless power transmitter, coping with received EMI will be the priority. If you’re developing an embedded device that has to meet FCC emissions requirements, you’ll probably need to focus on generating less EMI.

 

Spread-spectrum clocking (represented by the orange spectrum) can lower your peak radiated-EMI amplitude.

 

What Is Conducted EMI?

We’re all thoroughly accustomed to transferring electric signals via conductors such as wires and PCB traces, so it should come as no surprise that there is also such a thing as “conducted EMI,” i.e., interference that travels through a conductive path instead of the air.

You might be wondering about the distinction between conducted EMI and mere circuit noise. I would say that they are essentially the same thing; we use different terminology because conducted EMI is a specific form of noise.

When we talk about conducted EMI, we’re referring to noise that is generated by a device or subcircuit and transferred to another device or subcircuit via cabling, PCB traces, power/ground planes, or parasitic capacitance. The last item in this list is an important one to keep in mind: unintended capacitors are everywhere, and they will readily provide a path by which a high-frequency signal can couple from one conductor to another. I don’t include capacitive coupling in the “radiated EMI” category because it operates over very short distances and is based on electric fields rather than electromagnetic radiation.

 

Parasitic capacitance allows EMI to couple from trace to trace or from a trace to a plane layer.

 

Switching power supplies (AKA switch-mode power supplies, DC/DC converters) are a widespread source of conducted EMI. High-amplitude transient currents are created by the converter’s switching action, and these transients become conducted EMI when they negatively affect the load circuitry or the power supply that is feeding the DC/DC converter.

 


 

I hope that you now have a clear idea of the difference between conducted EMI and radiated EMI. If there is anything you’re still unsure about, feel free to leave a question in the comments section.

 

Comments

0 Comments