Probably the number one cause of failures to comply with the relevant EMC standard the first time is due to failing Radiated emissions testing. Most good EMC testing laboratories will start a compliance testing program with this test first. This is to try to ensure that any required EMC modifications are fitted to the product for the rest of the testing phase and minimize any possible required re-testing.
The severity of the failure, product design, any practical restraints and the stage of the product development can impact what the best solution in the way of a modification can be applied.
It is best practice to attempt to design the product with EMC in mind at the development stage of the product. The earlier the compliance is considered the more time it allows to optimize the product. Primarily compliance testing (pre-compliance testing) can help identify issues at an early stage and allow modification prior to producing factory quality final boards/products.
We will assume in this article that all consideration to the design stage has been noted and that the basic bullet points have been applied to the PCBs and associated components. And we will focus on the retrofit modifications that are commonly used to achieve compliance when the product is tested and fails formal compliance testing.
Ferrites for cable radiation
You probably already own a product that has ferrites in it or on the cables, as they are a common feature of many products. Ferrites are a passive component that has ferrimagnetic properties which are manufactured using a ceramic compound consisting of a mixed oxide of iron and small quantities of other materials such as nickel. How they are manufactured and the process used plays a critical role in the performance, so not ferrites are the same. Typically they come in the form of surface mount, solid core and split cores (snap-on cores). These variations allow forms to allow different uses at different product design stages. Ferrites do not conduct electricity and have high electrical resistance, with high magnetic permeability.
Ferrite beads are used as inductors resulting in an impedance and work by forming a low-pass filter that filters high-frequency electromagnetic interference (noise). The energy is either reflected back through the cable to the source or dissipated as low-level heat (usually not a noticeable amount of heat).
So how to select what ferrite is needed? Firstly the frequency range should be considered, by mixing metal oxide with other elements along with the manufacturing technique the performance and effective frequency range can be dictated. Once a suitable and usable range has been determined by the ferrite material choice. Then the other characteristics and the effectiveness can be compared for each ferrite available in the range of material. Once this has been selected the verify that a suitable package or fastening (such as a snap-on) is available. Most ferrite manufacturers provide data sheets with ‘Frequency Vs Impedance’ charts, this is a great tool to look for the best performance. Fit the most suitable ferrite to the suspected source of radiation such as a cable acting as an antenna a perform the test again and verify the impact. It is generally best to select a good tight-fitting ferrite if possible, adding additional cable passes through the ferrite may help further reduce the emissions. Always re-scan the frequency range as the impedance change and then the effective shortening of the cable as an antenna can affect the frequency at were the EMI noise was initially observed.
A quick tip to determine if the issue is cable radiation is to remove the cables one by one if possible whilst viewing the spectrum. Sometimes removing the cable and getting a reduction can also just be because you are not exercising that portion of the circuit, which could be the culprit.
Split-core, snap-on ferrites can be a quick and easy fix for marginal failures, they also usually come in a plastic protector case and some like the Wurth ferrites require a special key to remove them. EMC Bayswater has a large array of ferrites on-site if you require them along with other inline filters such as AC & DC EMI filters used for conducted emissions, USB 2.0 and 3.0 inline filters, D-Type filters and more. Wurth has done an excellent job of stocking our facility and also providing support to many of our customers.
Here is some useful links: http://www.wurth.com.au
https://en.wikipedia.org/wiki/Ferrite_bead
Faraday cage for cabinet radiation
If the radiation is emanating from the PCB and associated components directly then possibly changing or modifying the enclosure may help reduce the RF emissions. Metal enclosures ideally will have all their sides electrically bonded together with either RF gasket or mechanical fastenings to ensure a true Faraday cage. It is very easy to attach two metal plates together with no gasket or electrical bonding and create a slot radiator. If the enclosure is plastic producing a Faraday cage can be achieved by installing a metal can over the noisy PCB sections or by using a metal spray coating internally. Also, note it may require providing a low impedance bond from the Faraday enclosure to the earth.
Please check out our shielding effectiveness testing page for further details about shielding testing and also our article about effective uses of electromagnetic shielding.
Port connector bonding
As part of both cable radiation and enclosure, radiation is important that any port connector entering or leaving the enclosure is correctly bonded. In some instances, the correct method may be to isolate and float the connector but in most instances, the most effective method is to ensure a low impedance path to the enclosure. This may also help with the port impedance matching, the higher the mismatch between the driver and the receiver of the signal the more the emissions. This is particularly noted with common mode currents travelling back along with the shield of shielded cables. Hence why it is also critical to ensure that correct impedance termination of ports takes place. Not to be forgotten whilst on the same subject, the cable length between the port and the termination will add inductance this can also greatly affect the matching and cause significant emissions as seen with HDMI applications. For further reading of HDMI specific Radiated emissions issues please read our article.
Another more technical solution can be the use of jittering on clock signals. Some more advanced integrated circuits (IC) now have the ability to oscillate the operating clock frequency If used correctly this may reduce the final measurement if using measuring detectors such as a CISPR Quasi-peak of CISPR average to below the limit.
