Near-Field Communication is different from other wireless technologies, yet it's complex and requires production testing.
In this blog originally published in EE Times, LitePoint VP of Marketing, Curtis Schmidek, discusses the misconceptions surrounding NFC testing. LitePoint’s innovative analytical tools are specifically designed to keep pace with evolving wireless technologies, challenging device environments and pressing market demands. MCS Test Equipment are the sole distributor of LitePoint test solutions in the UK.
Near Field Communication (NFC) technology will soon become a mission-critical enabling technology in the devices and services we use on a daily basis. NFC has already proliferated into public transit systems, letting millions of travelers rapidly gain access to subways, buses, ferries, and planes.
It's being adopted to help authorize billions of dollars in financial transactions in ways that provide far greater levels of security than conventional credit cards. It's also being deployed to simplify how we enable our wireless devices to connect and interact with each other through simplified pairing techniques (sometimes referred to as "bootstrapping"). It will offer new means to streamline and protect our healthcare services through improved identity methods for patients and medications. And NFC offers new ways for advertisers to reach consumers though "smart posters" and tagged consumer items.
NFC capabilities can be found in literally hundreds of millions of smartphones, but you may be surprised to know that the method employed to test the NFC functionality in these devices is very primitive, relative to how other wireless technologies in smart devices are tested. Because the practical applications for NFC-enabled smart devices is only now beginning to take root, it will be critical to ensure the necessary NFC performance to guarantee a high-quality, consistent, and reliable user experience. While only about 30% of smartphones employed NFC technology in 2013, that number is expected to grow to 70% by 2018 -- in other words, well over 1 billion NFC-enabled smart devices.
The following misconceptions are commonly held when it comes to NFC testing:
Misconception No. 1: NFC is similar to other wireless technologies, such as WiFi and Bluetooth.
There are numerous differences between NFC and other short-range wireless communications standards, such as Wi-Fi and BT (Bluetooth). The profound differences between NFC and Wi-Fi or Bluetooth are due to each having very distinct use models and applications.
BT and Wi-Fi are intended for sustained high-speed data connections between devices such as PCs, access points, headsets, keyboards, smart phones/tablets, etc. Data rates are much higher with BT and Wi-Fi than with NFC, but NFC was not intended for continuous data transfers. Instead, NFC was created to enable short bursts of data over very short distances.
Some of the key operating parameters that differentiate NFC from Wi-Fi and Bluetooth include their operating frequencies: NFC operates at just 13 MHz, while Bluetooth spectrum is placed two orders of magnitude higher at 2.4 GHz and Wi-Fi at 2.4 GHz and 5 GHz. Bluetooth, designed for audio transmission and limited data transmission, has rates ranging from 1 Mbps to 24 Mbps. Wi-Fi, created for very high speed data communication, supports rates from 1 Mbps to nearly 7 Gbps. NFC, with a focus on the transfer of only small amounts of data, moves information at rates ranging from a mere 106 kbps to 848 kbps.
NFC's operating range is specifically designed to operate over very short ranges (1 cm to 4 cm, typically) in order to provide an additional layer of security against eavesdropping. NFC utilizes near field communication, which relies on the coupling of two highly-tuned, low-frequency, narrowband coils. These loop antennas provide strong electromagnetic mutual induction resonance that enable both the transfer of energy -- enough for an active "initiator" to power a passive "target" -- and the transfer of data. The passive NFC "target" can take simple forms with minimum circuitry, such as a plastic payment card. Wi-Fi and Bluetooth have much longer operating ranges and use far field communication, which requires signal amplification from both the transmitter and receiver. Bluetooth is designed for up to 10 m of range, but can be up to 50 m line of sight, while Wi-Fi, depending on the operating frequency and bandwidth of a given connection, can have a usable range of up to 100 m.
Misconception No. 2: NFC is a low-cost, simple radio. There is no reason to test it.
Cost aside, it's the functionality that NFC technology delivers that is important—the value of which far exceeds the cost of the components. NFC is being used across a variety of applications to enable financial transactions, entry into public transportation systems, access to businesses/public venues/homes, identification verification in healthcare, etc.
NFC is becoming a "mission critical" technology and, as such, it is crucial that its consistent performance be assured. These applications will affect users of NFC-enabled smart devices every day. Poor or inconsistent performance of NFC functionality will quickly bring on the ire of consumers who are relying on it to work. Retailers, banks, credit card companies, transportation systems and numerous other institutions are counting on NFC applications simply to be reliable. Therefore, production test of NFC should be considered mandatory. The ignition switch in a car is not the most expensive component in the vehicle—but you certainly want to be sure it was tested!
Misconception No. 3: Pass/Fail testing using a "Golden Unit" is all you need for NFC. Anything more is overkill and too complicated to implement in manufacturing.
Pass/fail testing based on a Golden Unit approach (e.g., use of an ideal example of a device against which all later devices are tested, yielding a simple pass/fail result), whether for NFC or any other wireless technology, is a risky solution that doesn't provide the quality assurance or production control that a factory manager needs to properly run an assembly line. In general, employing a golden unit approach will result in high variability in device performance because the performance of each Golden Unit is unknown and each unit will perform differently from the other. In addition, there is no way to extract parametric performance data from the devices being tested, so there is no way to detect if device performance is running at the edge of acceptable performance limits -- this can lead to shipping large quantities of marginal devices, with no way to even know when the problem started. If there are failures, a golden device test approach gives no indication as to the source of the problem. Additionally, the Golden Unit approach won't let you monitor production quality shifts that are often caused by changes in component tolerances and/or manufacturing process. It is critical to identify issues early and take actions to avoid production down time and ensure product performance.
Misconception No. 4: Pass/Fail with a Golden Unit is the shortest way to test; anything other than pass/fail would decrease my factory throughput.
There are numerous standards and modes of operation related to NFC. Doing a single Golden Unit type test (for example using a Suica payment card as the golden unit) would only indicate that the DUT (device under test) was able to operate in that one particular way (in this case, NFC-F as a Reader/Initiator). However, to get verification that the DUT is able to properly function over all standards and modes, one would need at least three different passive cards and a multi-mode reader. This would mean four different insertions! Even if it only took 5 seconds for each insertion and test, this would equate to easily 20 seconds of test time. Compare this to a multi-standard NFC tester utilizing a Universal Test Head that can perform parametric test on all NFC standards and mode with only a single insertion—typically less than 5 seconds or about 20% the test time of using Golden Unit test and at far greater factory test throughput.
Misconception No. 5: Sample testing is fine for NFC devices.
Sample testing, especially if parametric in nature, is certainly better than no testing. In situations where test times can be extremely long -- for example a final signal-based test that may be used for cellular technologies at the "end of line" for a fully assembled smart device -- this may be the most practical method. But to ensure consistent quality and deliver high quality products that consumers and service providers expect, 100% testing of devices in production is the best route. As described in Misconception 4, with the very short test times made possible with an optimized, multi-standard, multi-mode test system, there is little reason not to perform 100% test.
Misconception No. 6: The NFC standard used in my market is all I need to implement from a test perspective.
While it's true that certain NFC standards have become dominant in some countries (for example, NFC-F in Japan and NFC-A in Taiwan), the reality is that each "flavor" of NFC brings a different set of capabilities and advantages, so application developers and device manufacturers will design products implementing a variety of NFC standards. Nearly all chipsets used in smart devices today have the capability to operate in numerous modes and virtually all the top smart device manufacturers have international distribution. This means that testing all the standards and modes of NFC is basically a necessity.
Misconception No. 7: I have decided to test. Now I need to ensure both the NFC digital protocol is working for each unit, as well as the analog.
The primary purpose of production test is to screen out malfunctioning and marginal devices while monitoring performance changes that may be caused by defective components or manufacturing process drifts. It is critical to carry out RF parametric testing of analog functionality because it's the only way to provide quantitative quality control in production.
Testing at the protocol level may not be a good use of factory test capacity because software won't change from unit-to-unit. The digital operation of a device is rarely, if ever, affected by changes in manufacturing tolerances. Earlier phases in the manufacturing flow -- including ICT (in-circuit test) and verification of software download -- have already determined that the digital/software functionality will work as expected, but the analog performance can vary from unit to unit. Factors affecting assembly and chip defects include ESD damage, bad components, and changes in component tolerances and supply voltages. Thus, the RF/analog operation must be measured. By focusing only on NFC's analog performance, factory throughput is optimized becuase test time is minimized. Note that this method of "non-signaling" (i.e., physical layer only) test is universally accepted in the test of cellular, Wi-Fi and Bluetooth technologies in today's smart devices.
As applications using NFC technology become widely adopted, it is even more important to carry out thorough parametric testing and quality control on the production line to ensure the consistent performance of NFC-enabled devices. Reliable performance is the key for delivering a good user experience and increasing brand loyalty for the device manufacturer. The primary purpose of production testing is to screen out malfunctioning and marginal devices while monitoring performance changes that may be caused by defective components or manufacturing process drifts. Therefore, it is critical to carry out RF parametric testing (as opposed to functional "golden unit" testing), becuase this is the only way to provide quantitative quality control in production.
The LitePoint IQnfc is a simple-to-setup and easy-to-use system that allows one-click operation to quickly characterize the NFC physical layer in both R&D and Production.
You can email us for more information on the IQnfc or the full range of LitePoint products or give us a ring on +44 (0)4853 62 63 65.