When new audio gear arrives, the box is opened, the manual briefly scanned, but then many people just start making noise. Even the most meticulous reader of manuals will typically skip the preamble and “declaration of conformity” pages at the front of the manual. Yet locked within these pages are much of the blood, sweat and engineering tears shed during the design of your favorite pro audio products.
This month we will peek behind the engineering curtain at some of the various certifications end users take for granted, but companies are keenly aware of, during the product design cycle. Depending on the product, certifications can take a meaningful fraction of a year to complete. Further, after a design is certified, any changes or modifications require re-submission to the certification body.
For this article, let’s take a brief look at how certifications are structured, highlight a few testing techniques and put some context to how this influences the engineering process. While the content in this article will be familiar to pro audio manufacturers, it is by no means unique to the discipline. The insights and lessons apply to virtually every product in your shop, venue and home.
Alphabet Soup
Perhaps we should state at the beginning that the certifications we are speaking of here do not mean that a product performs, in an audio sense, like the datasheet says. Certifications are generally not concerned with the performance metrics specific to a product class, but instead are focused on safety, fault tolerance and electronic emissions. A product passing the certifications and operating safely could still have dreadful sonic performance, but should not readily catch fire, electrocute the end user or cause radio interference.
Following the certification road to safe, fault tolerant products will take us along a path littered with abbreviations and acronyms. The dizzying alphabet ahead hints at the complexity in navigating the certification process. We start with some highlights of the testing structure in the United States.
In the U.S., safety standards or codes are proposed and ratified by private and public organizations (e.g., ANSI, ASTM, NIST, IEC, FCC, UL, NETA, CSA, etc.). These standards are then given the force of law when requirements for compliance are codified at the federal, state, county, city, or municipal level. Compliance to the standards is then tested by private organizations approved by the Occupational Safety and Health Administration (OSHA) called Nationally Recognized Testing Labs (NRTLs). NRTLs both certify the product and manufacturer of the product. After a product is certified to the relevant standards and the manufacturer (or final assembler) is qualified, the facility is approved to apply the mark of the NRTL. The mark is the specific identifier of that NRTL.
NRTLs include UL, CSA, ETL, TUV, MET and NSF. NRTLs compete with each other on price, customer service, turnaround time and the like. In many product classes, the NRTLs provide collections of standards and testing that result in equivalent certifications, but with different marks. In some cases, the standards have been harmonized between different NRTLs. NRTLs can also have satellite lab facilities that are approved to collect data as inputs for the certification process. Companies can have these facilities on site to speed product development.
By contrast, a product’s RF behavior is controlled by the Federal Communications Commission (FCC). Depending on the type of device, it may fall under Verification, the more stringent Declaration of Conformity, or the most stringent Product Certification. Verification can be performed by any one of almost 3,000 listed test facilities. Declaration of Conformity or Product Certification must be performed by an accredited test lab. Labs may achieve accreditation through NVLAP, A2LA, or as designated by the FCC under “a negotiated Mutual Recognition Agreement.”
Confused yet? From the paragraphs above we see that certifications are quite the mishmash of public and private entities. Navigating the EU for the CE mark is more complicated still. Then there are other processes for Australia, NOM for Mexico and so on. Rather than untangling this knot, let’s instead turn to discussing some of the techniques used in certifying products.
Testing Techniques
There are a great number of tests that products can be subjected to and a huge range of standards to match. Many of those tests fall in the realm of certification, but some go beyond to certify performance metrics like UV, water, corrosion or abrasion resistance. Some tests are applied to representative units before manufacturing and some are applied at the point of assembly. The standards dictate what tests are applied when, as well as what results constitute passing or failure.
A ubiquitous and often thorny test that products undergo is the “radiated emissions” test. Electronic products that are “unintentional radiators” must go through a battery of tests to look at the frequency response of the RF energy that leaks out of them. These tests are undertaken in anechoic chambers which function in the same way as an acoustic anechoic chamber, but for RF energy. Depending on the class of product and where it is going to be used, the RF leakage from the unit must fall below a specified threshold over a specified bandwidth. Hand and hand with radiated emissions testing is “conducted emissions” testing. The idea is the same, but instead of measuring RF in the air, the test looks for leakage out of the cables emerging from the unit (e.g., the power cord). Again, there are threshold levels for leakage in different frequency ranges.
A common electrical test is the “dielectric withstand test.” This test is also called a “hipot” test and characterizes the behavior of electrical insulation to protect against current flow. Here, a voltage is applied and the testing equipment looks for leakage current through insulation. Another electrical test is the “electrostatic discharge” (ESD) test. Here a short, high-voltage stimulus is applied to the product being certified as the testing lab looks for failures or changes in state. ESD testing can cause LEDs to light erroneously, touch pads to trigger, electronic inputs to fail and other such anomalies.
There are also electrical tests associated with line voltages, circuit continuity and fault behavior. Some of the tests are physical, like making sure a cord is properly strain-relieved, or looking at temperature behavior of wiring while the product is operating or making sure the wire is of a type approved for the product. There are electrical tests that ensure that current is not leaking off surfaces of the device while in operation. Other tests include checking the continuity of grounding conductors and/or the current carrying ability of grounding conductors.
Electronic components within products can go through their own battery of tests, and those tests result in the components being “recognized.” The recognized components are then placed in an overall product, and the final product is “listed.” In the world of industrial electronics, companies that build custom panels can become “listed panel shops” so that they are certified to apply the NRTL mark to custom products without having to go through the conventional certification process.
Each new configuration and/or design change for a product must be submitted to the NRTL, who then make a judgment about whether that change is going to require specific re-testing to maintain certification. Finally, the NRTLs have procedures for holding manufacturing accountable to producing products that are consistent with what was tested at the laboratory. Managing the relationships between the NRTL and manufacturing is part of the certification process.
What’s a Manufacturer to Do?
Engineering design is an interlocking mix of feature requests, build of materials (BOM) costing, codes and/or standards, documentation and processes to ensure consistency of the product when it is built. Conception of a product is a tiny but critical part of the overall engineering process. There are many hurdles downstream for the product, including certifications. When products emerge from design, testing and certification as ready for market, often a substantial amount of time has passed. Ensuring the product is still viable for the marketplace by the time manufacturing has been ramped up is an art form.
“Feature creep” in a product not only can delay its development, but it can also complicate or slow the certification process. As a hypothetical example, imagine that you are a powered loudspeaker manufacturer that would like to add the ability to stream Bluetooth to your product. Do you choose to integrate an older, larger and more expensive Bluetooth module in your product that has already completed the FCC Product Certification, or do you wait for the smaller, cheaper, improved new module that still has to finish design and navigate the Certification? If you pick the new part and design your circuit board layout around it, then you run the possibility that your product release cycle will be delayed by the Bluetooth module vendor. Is that worth the risk? If you decide it is not worth the risk and utilize the older, more expensive module, does Bluetooth then add enough value to the product at point of sale to justify its cost within your product? And what if the older module is incompatible with newer phones? And what happens when the older part gets “end of lifed” (EOL)? And do the majority of customers really need this feature, or merely a vocal minority?
Choices like the above are the nuts and bolts of engineering practice, and the cost and time of product certifications plays a role in these decisions. A clean pass through one basic certification regimen typically starts in the low five figures and can increase quickly if additional troubleshooting follows. Manufacturers also have to consider whether it is worth certifying their products for all potential markets, or only the most major? Even the major market certifications have different timelines and hassle factor. And then beyond this, certain emerging markets have their own certification processes for specific product segments, and those certifications add money and time.
Products remain vaporware for many reasons, and (failed) testing is one of them. Established manufacturers have received the school of hard knocks education for getting products through certification. Even then, when manufacturers branch out into new product categories they run the risk of growing pains around certification. Manufacturers can also run afoul when switching component vendors or contract manufacturing (CM) vendors.
Conclusion
My hope for this article is to bring a spotlight on the concept that engineering a product and bringing it to market is more than placing some speakers in a box or slapping some electronic components on a PCB in a metal box. One of the central additional activities is ensuring that the product can pass a battery of tests developed to help keep end users safe. Another is ensuring that products don’t interfere with each other electrically.
The captured cost of a product is more than just the parts and the design. It is also the design verification process, the third-party certifications, creating documentation, developing the manufacturing process, implementing and sustaining manufacturing, providing support, analyzing failed products, and ensuring enough working capital to do it all again on new models next year. When you are facing the price of your next major pro audio purchase, remember all the engineering pain that has already been expended to build something worthy of your money.
