Electronic systems have one thing in common—they all need power supplies, power supplies that must handle a host of diverse requirements. These include wider input ranges, increased switching frequencies, higher power density and broader fault protection plus the continual demand for high-quality DC output voltages.
Fortunately, the industry has standardized power supply specifications; but the same is not true for actual test conditions, which need special attention. First, you need to ask: Do the power supply production tests simulate appropriate conditions for my applications?
Whether your role is power supply design, production test or incoming inspection, you want a flexible test system. It must have the capability and ease of use to test for multiple conditions in simulating a variety of real-world applications as described for the following power supply specifications.
Voltage Accuracy: Voltage accuracy is the deviation of the output, normally specified in percentage from its nominal value. One vendor may measure the voltage accuracy at full load while another may test at no load. Yet a third vendor’s architecture may have a 10% or 20% minimal loading required and may choose 50% loading as the test condition.
Some commonality appears for multiple power supply outputs when all outputs are at the same test-loading condition. But, does your application have the same balanced loading? What happens when the line voltage is at its lowest input, and one load is at 80% and the other at 20%?
A test system that allows you to quickly select multiple test conditions for a particular specification is welcome here. Figure 1 shows a Windows-based test-editor screen that allows you to double-click on the desired specification and quickly edit the specification’s test conditions.
Line and Load Regulation: Line regulation is the amount of output voltage deviation for a change in input, normally from a nominal voltage to a high or low line voltage. The worst case is assumed to be under full load.
Load regulation is an output voltage change for a change in load current, perhaps from a minimal load, or 50% load to full load. Bipolar or triple output supplies have all the outputs at the same loading condition for these tests.
Power-up: Power-up ensures that the power supply can start up under a full load condition (soft-start circuitry allows gradual capacitance charging, avoiding latchup) and enter into a final error band within a certain time frame. A minimal or no-load start-up test, sometimes performed at the coldest operating temperature, may also provide insights into the design.
Transient Tests (L-H and H-L): Transient testing is performed using a step-function change in the load current and monitoring the time required to settle within a specified time. Tests are often performed from 50% load to full load in both directions at nominal input line conditions. Multiple outputs are tested individually (vs cross-linked) while other outputs are kept at their 50% load levels (Figure 2).
Ripple and PARD: Ripple is the peak-to-peak amplitude of the AC disturbance on the DC output. Periodic and random deviation (PARD) is the noise on the output measured over a 20-MHz bandwidth. Ripple and PARD tests are usually performed at a nominal input voltage and under full load. Graphical presentation of these tests, combined with a capability to vary the time base down to a single-cycle inspection, is highly desirable in a test system (Figure 3a).
Ripple Frequency Measurements: Ripple frequency is the switching frequency of a DC-DC converter under nominal line input conditions and typically under full-load conditions (Figure 3b).
Efficiency: For a DC-DC converter, efficiency is the ratio of output power to input power expressed as a percentage. Efficiency tests measure the output current and then the input current under full load and nominal input voltage.
Do you need to know the efficiency at minimal loading? How about at 50% loading? How about over the input voltage range? How about a graph of efficiency over three orders of magnitude of current vs input voltage (Figure 4)?
Fault Testing
Short-Circuit Testing: Your power supply says it is short-circuit protected, but for how long? You want a test system that controls the time a short is can be applied.
Current-Limiting: If your load impedance keeps decreasing, drawing more current, your power supply must go into current-limiting to protect itself. Eventually, as you draw more power, the output voltage begins to decrease so the output power does not go into a runaway state. Figure 5 shows a graph of a current-limiting test.
Overvoltage Protection: Your power supply may be powering expensive circuitry such as telecommunications or imaging equipment. If it experiences a failure, you want the damage limited to the power supply. Overvoltage and, in some cases, undervoltage tests can ensure that no out-of-specification conditions will damage the equipment.
Summary
A test system must provide enough flexibility to allow a thorough power supply test under multiple conditions through the design stage and even into the production stage. This is especially true for parameters sensitive to design-yield issues. Also, a buyer-beware wisdom suggests incoming QA departments may also want a test system that ensures the power supply operates under its own intended application.
About the Author
Bob Leonard is the Director of Marketing and Sales at ELTEST. He has a B.S.E.E. degree from Northeastern University. ELTEST, 26 Oxford Rd., Mansfield, MA 02048-1127, (508) 339-8210.
Copyright 1996 Nelson Publishing Inc.
June 1996