If you have never taken a product to be tested for electromagnetic compatibility (EMC), that first trip to the laboratory can seem overwhelming. What exactly happens at the EMC lab? How can you prepare?
My new miniseries “A Visit to the EMC Lab” aims to answer some of the most frequently asked questions to help give you the best chance at success. First, we’ll get an overview of how labs test EMC. Later, we’ll get into the weeds on emissions and immunity testing. Finally, we’ll cover what to do if your design is one of the 50% to fail its first round of EMC testing (step one: Don’t panic!).
EMC testing intends to subject products to specific types and levels of interference and to do so in particular, repeatable ways. To ensure a product will perform as expected in its intended environment, the lab will test its response to (and its effect upon) the surrounding electromagnetic environment.
The actual testing can vary from minutes to hours, but this is often only a small part of the entire test effort, from both the product and the lab perspectives. For example, the laboratory must confirm test waveforms before testing to ensure that nobody walks out of the lab believing their product meets EMC requirements, when in fact the test equipment was faulty. Upon arrival, the lab must establish a baseline—called a “noise floor”—before switching on the product to see what it contributes to these baseline emissions.
On the product side, you will want to agree to a test plan weeks ahead of time, to help streamline your experience at the lab. After all, you are paying for your time there, and often quite handsomely…
EMC testing typically requires specialized equipment that can be very expensive. Some anechoic chambers, like the one pictured at the top of this post, can cost over a million dollars. There is, however, some justification for those costs. It is becoming increasingly difficult to create a quiet electromagnetic space these days. Rising EMC standards are placing more requirements on labs to demonstrate qualities such as electromagnetic quietness before full-on testing.
With all these factors combined, you can expect to pay over $1500 a day for standard EMC testing, so make it count.
EMC is ultimately about ensuring a product will perform as expected in its intended electromagnetic environment. Therefore, the lab will subject the product to testing that demonstrates its ability to do so.
Generally, this includes testing all the external ports and power supply inputs, as they are the likely means of coupling between unwanted energy and the product. Because electrostatic discharge (ESD) can also trigger EMC issues, testing often includes surfaces, too—anywhere that a user may lightly touch with their charged fingers or contact while plugging in a cable, for example.
Then there is radiated emissions testing, which doesn’t involve contact with the product during the test at all. Instead, antennae measure the electromagnetic energy emanating from the product at distances from 1m to 10m away from the device.
Testing equipment varies based on the product or equipment under test (EUT). Some equipment serves a bespoke purpose and will only be used for certain kinds of testing (these are generally the coupling devices), while other equipment may be more general (for example, signal generators, amplifiers, spectrum analyzers, and antennae).
Let’s look at an example EMC product test. Take lightning phenomena for example. Since lightning energy is likely to arrive through cabling that has been struck outside a building, systems with cabling undergo lightning strike simulations with a surge generator and a bespoke coupling device known as a CDN. These devices were specifically developed for EMC testing. CDNs couple most of the energy from a simulation into the equipment under test (EUT), but they also decouple (or attenuate) the energy somewhat, dispersing it into auxiliary equipment (sometimes referred to as the “load”). For this example, the CDN and surge generator would be unique to testing a product’s lightning immunity and would not be used for any other tests.
Now to the fun bit: How does the lab decide which products pass the test? This is where the Functional Performance Specification Classification (FPSC) or performance criterion of the product comes into play. In short, this criterion is the expected performance of the product during and after the test. If actual performance deviates from this agreed quality level, then the product will fail.
Who determines the performance criterion? This can come from several sources: A customer demands it, a manufacturer hopes the industry accepts it, EMC regulations demand an exact performance for such a known product or product family, or the lab itself deems the criteria reasonable. The crucial aspect is that all the details of the product and test are laid out for subsequent observation.
After all, if the product doesn’t pass, designers will need to understand which details failed to line up so they can improve EMC performance for the next visit. We will take a closer look at some of those details in an upcoming post.
You can find the second blog in this miniseries here.
