AC Power Testing Presents Many Challenges and Concerns

Testing products designed to operate from ac power is not as simple as one might think. Consider a factory with the straight-forward need to manufacture and test the heating element for a toaster. Even for this factory, operating on an unreliable power grid, the job was not considered a major challenge. The manufacturer, however, discovered that was anything but true.

Accurate testing proved impossible. Small but unknown variations in voltage resulted in apparent deviations in the impedance of the heating element. The nominal impedance for the heating-element design was 55 Ω, with an acceptable range of 53.5 to 54.5 Ω. At 120 V, this translates to a current range of 2.2 to 2.24 A. Fig. 1 plots the expected current's minimum, nominal, and maximum impedances over the acceptable current range against a range of voltages. As you can see, you simply can't test the impedance of a toaster heating element by measuring the current if the test voltage keeps changing.

Whether the product will be used on the ground or on an airframe, if it's running on ac power someone somewhere is going to have to test it. When a new product is designed, there are a number of concerns that will eventually need to be quantified. Among these concerns are: the product's performance over a range of power conditions, the power required for operation, and negative effects the product may have on the ac line providing input power. A designer may use a wall plug at the outset; but before he is done, a full safety and compliance profile will be required. This is a task best performed using a programmable ac power source.

A closer look reveals the true complexity and the varied challenges of testing. Even though ac power technology hasn't changed significantly since the introduction of switch-mode supplies in the early 1990s, the expectation that the power source operate as a test instrument continues to grow unabated. Having an ac power source that functions as an arbitrary waveform generator is no longer just a feature, but an expectation. Users also expect voltage, current, and power-factor metering to levels of precision beyond that of many handheld instruments.

Now consider any product moving through the typical product development cycle. First, you design it. Then you test it. From there it heads to production. On the production floor, the conscientious manufacturer will run limit or production tests on every device to ensure it meets the specifications that the design and test engineers identified. The nature of the business is one of complexity — the complexity of the product itself, the complexity of the software used in testing and more. The assistance and support required by users varies as much as the products they manufacture.

TAKING THINGS ON FAITH

Assistance actually starts with the design engineer, whose challenge is to anticipate on paper the product's power requirements. Here is a hypothetical example: a widget that shoots laser beams from desktops. The problem is clear. The design engineer knows widgets. He knows laser beams. But he doesn't know about ac power. So he has to take on faith when he calls the ac power expert and says, “All I know is that I'm going to have a full-bridge rectifier on the input to my product, and I need 15 A at 120 V.” The design engineer is relying on the ac power-source applications staff for the appropriate direction. He simply wants to turn on the power to see how the product works. We now come to the test engineer, who must be able to quantify and certify all facets of this new product, which includes the power quality. He has to report whether this new device tears up the utility lines. “If I plug the new widget-laser into the same outlet as the PC,” he asks, “will I induce line harmonics that will fry the PC?”

This relationship between the ac power source and the unit under test can be critical. You can divide power sources into two basic categories: linear and switch-mode. Both technologies generate a sine wave, which is amplified to provide the necessary output. Generation of a waveform independent of the input power allows precise control of the frequency, amplitude, and other characteristics of the output. A linear source provides a faithful reproduction of the waveform using traditional amplification techniques whereas a switch-mode source utilizes pulse-width modulation to provide more efficient amplification. Table 1 provides an overview of the tradeoffs between them.

Here is a real-life example in the world of small ac-to-dc chargers for laptop computers. The original design engineer used a textbook switching-type ac supply on the input of the converter. During test and development, he used a switch-mode-type ac power source to supply power to the design. The ac power source in the lab was not of a particularly new or high-quality design; consequently, it had some excessive yet natural characteristics: high output impedance and slow response time.

When driven by this type of ac power source, he thought the new ac-to-dc power supply was just perfect. The new power-supply design was then sent off to the compliance engineer for certification. Using a different power source, the compliance engineer watched the design start to fail. To the design engineer's dismay, the new power supply “popped” as soon as input power was applied. Further investigation revealed that the compliance engineer was using a linear power source. What they both discovered is that the dc power-supply design was inappropriately dependent upon the impedance of the switch-mode power source to work properly. When connected to a linear power source with its low output impedance, high-peak current capability and fast transient response time (all characteristics of modern utility power grids), the input stage to the power supply simply demanded too much peak inrush current. So, the challenge constantly faced by the design engineer: Is the power source appropriate for the application?

FROM TEST TO PRODUCTION FLOOR

Another challenge typically occurs when the test engineer crosses paths with the facility engineer on the production floor. There is a difference between the test engineer and the production supervisor. The test engineer wants a fully programmable, transient-capable, super-fast, dropout, sag, surge, do-it-all model; the production-floor supervisor wants a start button and a stop button.

Not surprisingly, both parties often discover there is a disconnect in terms of magnitude. The test engineer fashions a test sequence that runs a single product through all its design limits. A simple event when one unit under test needs only 5 A; but now the production engineer needs to test 10 at a time. The ac-power-source expert, therefore, must be able to scale the appropriate amount of power for the particular customer. A motor design, for example, uses a 6-kVA source; now the customer wants to test 10 motors. Is it safe to assume you simply need a 60-kVA source? What about the inrush current? Are you going to start all the motors at the same time? Will you measure inrush current? Can you tolerate variations in input voltage regulation? Only an experienced applications engineer with the expertise to ask the right questions will be able to guide the production engineer through this maze and avoid many of the common pitfalls.

The customer must come to the power-source provider to explain his application, and the two must work together. Once again, the technology of the source plays an important role in the selection process as presented in Table 2. The good news is that translation of increasing or scaling up in power can be simplified for the customer. But if the customer can master the smaller ac power source, controller, and software he can master the higher-powered unit.

Additional concerns arise when the product makes it to the real world, where line power is not always as expected. A host of problems, including variations in voltage and frequency instability, can substantially impact the performance of a product. Less obvious problems such as line losses and harmonic content should not be overlooked. Careful selection of a modern ac power source can result in a powerful tool for product development and test, but there is much to consider in the process.

THINGS TO REMEMBER

For both the engineers and the power supply expert, there are some basic rules to keep in mind when working together.

1. Everyone needs to understand the purpose of an ac power source. Many manufacturers simply don't understand power. Those who deal in ac power technology need to ensure that the customer understands that the ac power source doesn't create power, but rather modifies the voltage, the current and the waveform. In other words, there must be enough power available before a source can recondition it. If the product demands 16 A of single-phase power, the engineer can't plug into a 15-A plug in the wall and expect more power.

2. Have everyone agree on what is really expected during testing. We select the ac power source that's right for the application. Typically, I'll ask the test engineer to describe the nature of the unit under test. Is it a full-wave bridge rectifier, or is it power-factor-corrected ac-to-dc converter? Many customers don't know the answer. Others hamper the process by telling me what they want to do, but not telling me what they're testing. This lack of information makes it difficult to help them engineer the best solution for their application. So the key questions are these: What am I testing, and what are the expected results?