A well-engineered electronic load is not designed for exclusive use in the laboratory. It should be carefully engineered and built to test real power systems. This is true for portable systems for military applications, such as fuel cells, which must be tested under the real-world conditions in which they will be deployed. To meet this challenge, electronic loads can be manufactured as portable units, capable of running from a battery pack for use in the field. They can also be packaged as fully encapsulated units for which harsh conditions and other environmental hazards, such as sand, dust or silt, present no operational problem. There are also no maintenance or calibration problems to take into consideration with these units (Figure 1).

However, before running tests on any power system with an electronic load, it must be determined which type of electronic load is the most suitable for recreating the conditions that will be encountered in its operating environment. Because the variety of load types is broad, understanding the strengths and weaknesses of each type can make selecting the appropriate load nearly as challenging as the actual power-system testing process itself.

Broadly defined, there are several categories of electronic loads. For example, there are the older standard transistor types, the newer electronic FET high-power types, and then there are the latest, most reliable types of electronic loads. These latter types are of the low-voltage, low-on-resistance FET variety. And they work well for testing portable power systems based on batteries and fuel cells.

Two additional load types should be mentioned — as well as their inherent disadvantages. One type is based on the dated technique of using an array of power resistors. While relatively low in cost, this technique suffers from limited accuracy. The other type is what is called the switching electronic load, and this type should not be used for load-testing a power system as it can cause serious problems for the device being tested. However, there are several alternatives to this specific load type, as well as several crucial factors that must be considered when making the final selection.

Essentially, an electronic load functions as a constant-current power resistor, with an adjustable range of current controlled by some external control signal, such as an analog voltage. This feature removes the uncertainty of the power-system's output current. The electronic load can, therefore, maintain a constant current during testing, something that is impossible to do using only a power resistor. This simplifies the calculation for the power-system's output power, which essentially becomes a function of the power system's output voltage. This aspect of operation is favorable to the pass/fail testing used in high-volume production of batteries. This application, however, has the additional requirement that the load be able to quickly ramp up to the desired current (Figure 2).

Selecting the proper electronic load for a specific application can be an involved process that should be as streamlined as possible. In a fuel cell system, for example, the first requirement is to know the cell voltage or stack voltage of the fuel cell. Second, the current of the cell (or the entire fuel cell stack) must be determined. From these two parameters, the output power of the fuel cell system — or any other portable power system — is easily calculated. Once the power capacity of the unit under test (UUT) is known, the range of electronic loads suitable for testing can be restricted to those rated for this power level.

Nearly all electronic loads are rated in Watts at room temperature (25 °C). If testing at other than room temperature is required, then the electronic load's specifications must be carefully reviewed, because most loads will require de-rating at higher temperatures.

The next factor to consider when selecting a load is not as simple as it might first seem: both the minimum and maximum voltages supported by the electronic load must be discovered. As the voltage applied to a load decreases, the load's ability to accurately control the current also decreases. Many electronic loads only work down to about 0.7 V (or even only down to 3 V). Below that, the load may function improperly, or even fail completely.

Wiring is an important issue in electrical testing, and it should be considered together with the load selection. Unfortunately, many engineers fail to consider it when testing high-power systems, such as fuel cells that can generate from 100 A to 250 A. Yet, a few milliOhms of resistance in the wiring and connections can have a tremendous impact on the final voltage drop felt across an electronic load (Figure 3).

Current shunts represent a common and effective way to measure high currents, but they must be used with great care. For example, to measure 100 A, a 0.001 V/A shunt would produce 0.1 V. However, if the power system under test is a single fuel cell that can generate only 0.3 V under load, then at best only 0.2 V can be applied across the load, and resistive losses in the wiring alone could fully consume this voltage.

Therefore, especially in the case of fuel-cell-based power systems, it is desirable to have as much voltage at the electronic load as possible in order to accurately test a power system at high currents.

Because the difficulties associated with low-voltage operation in electronic loads can be severe in fuel-cell systems, one method used to deal with this drawback is to use a second power supply. This power supply is placed in series with the power system under test (and the electronic load). This boosts the voltage applied to the load, restoring it to the range needed for proper operation.

However, this practice introduces drawbacks of its own, including voltage instability of the second power supply, instability of the electronic load, and increased electrical noise (contributed from the second power supply). Other drawbacks include the following: overheating of the UUT, invisible high-frequency components in the output current during testing, dynamic instability in the UUT, and reduced slew-rate in the UUT. All of these problems can sabotage the effectiveness of using a second power supply in the test platform.

For example, the instability of the power supply line and load regulation can be several hundred millivolts (and, in some cases, even greater than the voltage of a single cell of a fuel cell stack). If a second power supply is used to boost the voltage, then electrical measurements must be taken frequently, and at multiple points in the output-current circuit path, in order to correctly gauge the performance of the UUT.

If the practice of using a second power supply is unavoidable, then switching power supplies should never be used, as they have problems concerning noise and current spikes that may be undetectable on standard test equipment, such as a multimeter.

When testing the UUT, such as a portable power system based on fuel cells, with an electronic load in conjunction with a second power supply, the line and load regulation of the power supply are the two most important considerations.

Before use, the specifications of a power supply intended for placement in series with an electronic load should be independently measured. This is due to the unfortunate reality that data-sheet specifications can make a power supply's performance appear greater than it would be under actual working conditions.

There can be yet another issue associated with the use of a second supply. Incorrectly connecting the second supply's sense leads. The original purpose of these sense leads is to bypass the voltage drops caused by high currents through the supply's output wires. Therefore, the traditional practice is to connect them directly to the interface terminals of the application being powered (or at the ends of the power supply output wires).

However, in the case of a test-platform configuration in which a second power supply is used in conjunction with an electronic load, the sense lines should be connected directly to the output terminals of the supply, not the ends of the supply output wires (Figure 4). This will help stabilize the power supply and allow it to give optimal performance when testing fuel cells.

Attaching the second power supply's sense leads to any other nodes in the main output current loop will invariably lead to false readings in the power-system measurements. Of course, the sense leads should be kept as short as possible, and sufficiently far away from any obvious sources of electrical noise.

Nevertheless, despite the limited effectiveness of these measures, the use of a power supply in series with an electronic load should only be used as a last resort. Furthermore, a power supply engineer should be consulted when using this method.

1 2 Next

RSS    Save to Del.icio.us  Digg This