EMI is the hardest thing to deal in switching power supply design. There are several factors that worsen EMI and there are also some ways to minimize EMI and able to pass the standards. One of the several ways to combat EMI is through shielding and we will discuss here EMI shielding techniques.
Before going far, let us define EMI. EMI means electromagnetic interference. This is an undesired phenomenon that all design engineers hate. Have you observed your TV reception is badly affected when you plugged in a hair blower in the same power source the TV get? This is an EMI thing. Have you heard humming sound from your radio or even TV when your cellphone receives text message or call? This is an EMI thing. My definition for EMI above is very elementary to make ordinary people understand what it is.
EMI is regulated worldwide and there are standards to follow like FCC in the US, CE in Europe, and ICES in Canada. EMI is broken into two main parts; conducted and radiated. We will discuss these fully in other topic.
From Wikipedia, shielding is defined as:
1. Shielding is a process of minimizing and controlling the coupling of radio waves, electromagnetic fields and electrostatic fields
2. Shielding is a practice of reducing the electromagnetic field in a space by blocking the field with barriers made of conductive or magnetic materials
3. Shielding is typically applied to enclosures to isolate electrical devices from the ‘outside world’, and to cables to isolate wires from the environment through which the cable runs
In practical thinking, shielding is use to prevent electromagnetic waves to escape from a noisy system or prevent them to enter sensitive systems.
EMI Shielding Application
- Noisy power devices
- Circuits and modules
- Membrane switch panel
- Ventilation apertures
- Rotating Metal Shafts that Penetrate an Enclosure
- PCB level
EMI Shielding General Concepts
Electric fields will produce forces on the charge carriers within the conductor. When electric fields are applied to the surface of an ideal conductor or simply hit to a shield material, it induces a current that causes displacement of charge inside the conductor that cancels the applied field, at which point the current stops.
EMI Shielding puts impedance discontinuity to the path of a propagating radiated EM wave, either reflecting it or absorbing it.
When an incident noise hit a shield, some portion of it will be reflected, absorbed and pass through with an attenuated magnitude. In a good shield, a weak noise wave can pass through. Below illustrates the scenario.
The effectiveness of a shield depends on the three factors; reflections, absorption and multiple reflections. Reflection means an instant reflection of incident wave when hit a shield surface. Multiple reflections include internal reflections and re-reflections of waves or noise.
EMI Shielding Effectiveness Equation
Below equation is can be used to mathematically determine how effective the shielding is.
SEdB = RdB + AdB + MdB
SEdB – Shielding effectiveness
RdB – Reflection loss
AdB – Absorption loss
MdB – Multiple reflections
In far field, reflection loss is the predominant shielding mechanism at the lower frequencies while at higher frequencies absorption loss is dominant.
For near field electrical sources, reflection loss is the predominant at the lower frequencies while absorption loss is predominant at higher frequencies.
For near field magnetic sources, absorption loss is the predominant mechanism at all frequencies
Some EMI Shielding Techniques
[stextbox id=”info”]Volumetric Shielding[/stextbox]
This is a type of shielding a unit, sub-assembly, product or device by covering in all six sides, literally, by covering a box. This is also known as “Faraday Cage”. This is effective in most cases however; the drawback is cost and thermal performance for high power application as the area bounded or covered will have poor air circulation.
Below is an example how a volumetric shielding is done. As you can see, there are six surfaces covering the unit under shield.
[stextbox id=”info”]Nested Shielding[/stextbox]
Nested Shielding is an EMI shielding using different levels of shield material. Each level is not necessarily a volumetric shield; it can be a low cost and simple shield. However, since there are several layers of shields, this technique still become effective and even more cost friendly than a single layer sophisticated shield.
In the above shielding technique, when level 1 shield fails, there are still levels 2 and level 3.
[stextbox id=”info”]PCB Level Shielding[/stextbox]
PCB shielding is specific to PCB modules. In some cases, shielding in the enclosure level is unable to address EMI on the PCB level. Whereas concentrating on the specific portion of the PCB will totally solve the issue. This is a common practice in SMPS wherein design engineers are hunting which part of the PCB is radiating or conducting noise then simply tried to shield that area.
Effect of Frequency on Shielding Performance
In order to categorize the effect of the frequency on the performance of an EMI shield, we will use the mathematical expression of skin depth. Skin depth by the way is defined as the measure of how electric current flows along the surface of a material. It can also be shortly defined as the degree of penetration into a shield. Mathematically,
Frequency – frequency of the wave
μ – magnetic permeability = μo x μr
μr – relative permeability
μo – 4π x 10-7H/m
σ – electrical conductivity in Ω-1m-1
The ideal value of skin depth must be very low that implies there is no penetration of noise through the shield and indeed an effective shielding.
From the above equation, at higher frequencies, skin depth value is low. This simply means that the current will tend to flow only at the surface of the shield material. This phenomenon is commonly as skin effect. This current will lose its energy as it travels on the surface due to the finite resistance of a particular shield material. Due to skin effect, there is no need to use a thick shield material for high frequency applications. Also in high frequency, the shield material doesn’t need to have very high permeability as the frequency is already very high.
At low frequencies, the skin depth value is high considering the same material is used. As the ideal skin depth must be very low, and then by selecting a material with a very high permeability will help. Making the shield thicker will also help. There is a danger attached in using a very high permeability material as the saturation flux density is lower. When a shield material saturates, it is not effective at all.
Saturation issue is can be work around by using more than one layer of medium permeability with a good spacing from other layers.
[stextbox id=”black”]Limiting Factors of an EMI Shield[/stextbox]
Above EMI shielding techniques are limited to the following factors:
Inherent Electrical Resistance
This is the ability of the conductor to oppose electric current that results to a partial cancellation of the incident and excited field. Resistance will vary from material to material.
Some conductors exhibit a so called ferromagnetic response to low frequency magnetic fields; the resulting field does not fully attenuated by the conductor.
Holes, apertures and bonds in a shield will reduce the field-reflecting capability of a particular shield. It will allow the current to flow around them; as a result, the fields passing through them do not excite opposing electromagnetic fields and there in no field cancellation.
The effectiveness of EMI shielding techniques is greatly dependent to the cost in most cases. In product manufacturing, cost is the most sensitive aspect. A larger and sophisticated shield will always come with a higher cost. The improvement delivered by the shield may cancel the profit of the project
The simpler the shield is, the faster it will manufacture. However, some sophisticated shields will come with a complex shape which demands specialized tools and resources. The cost of these special tools is expensive and the lead time of production is long. This will simply result to a revenue loss.
Shields do not interfere with electrical operation. However, it may block air passages to power devices and they may go thermal runaway.
For instance in a power supply, most of the times the boost inductor is causing lot of noise thus shield is applied to it. By the application of the shield, airflow cannot penetrate 100% to the boost inductor and this result to a temperature rise and the inductor may go saturation. When an inductor saturates, it will lose inductance and simply acting almost a short circuit that may cause explosion.
Due to a smaller form factor demand, the size of a shield is always a challenge to a small form factor product. It is not a good idea to consider shields late of the development stage.