A couple of months ago, I was testing EMC for a medical product. As the product is enclosed within a metallic enclosure, I was concerned about potential emissions from the enclosure during operation.
This is something that I have discussed with the testing engineer, and he stated, "Francesco, don’t you know that a metallic enclosure acts as a Faraday chamber? Don't worry about it". As it was obvious that he was not as concerned about emissions above 500 MHz as I was, I replied "Since this product is handled by an operator in real life, don't you think we should have an electrical hand load in contact with the product during the test?" He agreed, and we did the test, and fortunately for me, we passed the radiated emissions test. After you have read the entire post to the end, you will get a better understanding of what I have done from an electronic point of view. Also, why adding an electronic hand to the product has drastically reduced the amount of radiated emissions.
What is a patch antenna?
A patch antenna is a very low-cost and common type of antenna that is very common these days. It is used in a variety of applications such as GPS, WiFi cards, wireless devices, etc.
There are three fundamental elements that make up the "patch" and they are as follows.
The ground plane.
The dielectric that separates the ground plane from the microstrip antenna
A microstrip antenna.
Google "patch antenna calculator" and you will find many resources that will show you how to create a simple patch antenna on your PCB.
The most common patch antenna you may have seen is a GPS antenna.
However, in this post, I don't want to discuss how to design a patch antenna, but discuss how to reduce the emissions caused by an unwanted patch antenna.
How do we create an unwanted patch antenna?
When positioning a PCB ground plane above a metal ground plane and isolating it from AC and DC, we may have created an unwanted patch antenna. In this case, the dielectric is "air", whereas the antenna is the entire structure.
In this structure, the "waves" radiate from its edge (see drawing), and its maximum efficiency occurs at half or one-quarter of its wavelength.
This "patch" structure can be driven by capacitive coupling between one or more ICs on our PCB and the metallic enclosure itself (see drawing). Therefore by reducing the coupling between the PCB itself and the external enclosure you reduce the emitting power and therefore the radiated emissions.
Therefore, as a first precaution, try to reduce the coupling between any ICs and the enclosure. A simple way to do this is to place the "noisy" IC as far away from the enclosure. Or "on the other side" of the PCB (see drawing) so we do not "energise" the patch structure and therefore we do not emit.
Another issue, which I will discuss in a separate post, is that in a product like this, where we have an insulated metal enclosure, we may also create an unwanted ground loop. I will address this in more detail in a future post or article.
Another way to reduce emissions is to "break" the patch antenna. We can achieve this by bonding the metal enclosure to the PCB ground somehow. This can be achieved in several ways.
Connect the ground of the PCB directly to the chassis at a number of points with mounting pillows
. Oftentimes, we see this kind of solution when we open a desktop computer.
Other times, however, this may not be possible due to safety concerns. What should we do if we cannot connect our PCB directly to the enclosure, for example?
You should be able to connect a Y capacitor if, for safety reasons, you can bond the PCB ground to the metal enclosure.
The self-resonance of the Y capacitor will be several decades lower than the emitting frequencies. Nevertheless, it does not matter.
By creating a preferred return path between the enclosure and the PCB GND, we are breaking the "patch" structure. It does not matter too much about the actual capacitor value, as it does not behave like a capacitor at all, but rather like an "emergency bridge".
Therefore, the smaller the capacitor value you can use the better.
It would be a wise idea to add as many bonding points as possible near any interface and any opening. This is because the patch antenna radiates at its borders, as I described above. By breaking the patch "design" near its source, you will have the highest chance of reducing emissions.
Ground the enclosure to earth.
As a last resort, you can ground the enclosure to Earth. There are times when this is an easy process. For example, if you own a desktop computer, the chassis is usually connected to the ground.
Thanks for reading this article, and I hope you have gained some insight into how unexpected issues can arise during software development