Programmable power supplies have been a staple of the compliance testing market for years. The complexities of simulating various line conditions present in dynamic environments such as an automotive or avionics power bus, however, have provided a challenge to test, design, and compliance engineers.
Configuring systems to execute this type of testing can consume many hours of valuable time. Traditional power supplies, both AC and DC, cannot perform a high percentage of the required simulations as stand-alone units.
Systems can be configured using arbitrary waveform generators, linear amplifiers, surge networks, and power sources. Some of the problems in setting up systems such as these include obtaining repeatable results, documenting the test setup, and validating test setups to the desired criteria.
The advent of low-cost, higher-performance microprocessors and improved switch-mode amplifiers coupled with the introduction of programmable arbitrary waveform power sources mean that the tasks of subjecting products to sequences of both DC and AC transients and waveform anomalies can be performed with one product simulating real-world power conditions. By using this integrated capability, cost savings for hardware and software can be realized and the setup reduced to a fraction of the time compared to in-house-configured systems for compliance testing.
A recent example of the application of this technology involves Garwood Laboratories, a full-service calibration and testing company located in Pico Rivera, CA. The company, which performs a variety of EMC compliance tests for the telecommunications, military, and commercial aerospace industries, began using a programmable arbitrary waveform power source, the Elgar SW5250A Series SmartWave™, in 1997.
One area of testing where the power source can save time over other methods is for Radio Technical Commission For Aeronautics DO-160 Section 16 tests. The D0-160 is a commercial aerospace test specification covering many areas of avionics. Section 16 addresses equipment with input power of 115 VAC or 230 VAC at 400 Hz or the new variable-frequency 360-Hz to 800-Hz ranges as well as 14 VDC and 28 VDC.
Many of the Section 16 tests subject the equipment under test (EUT) to a variety of voltages with various rates of change. The test criteria depend on the category of the EUT and the type of input power that it receives.
The EUT must function to the required standards defined by the manufacturer. For example, a fluorescent lamp may flicker during a momentary drop to 0-V input but not suffer any damage. On the other hand, a computer may be expected to operate without fault under the same conditions.
DO-160 Sections 16.5.1.4 and 16.5.2.3 require 13 test conditions for a category A product using six different levels for the voltage drop and varied voltage decay and rise times. These test conditions for category A apply to 115-VAC, 230-VAC, 28-VDC, and 14-VDC products.
To test a product that uses all of the possible input voltages, 52 tests are required. This is a fraction of the possible tests. In some instances, a customer may require alterations of the DO-160 requirements that could add to or change the required tests.
By using the SmartWave source, the standard test requirements can be stored and recalled later. If the client’s test requirements define a change to the standard tests, a simple change to the stored values for voltage and time can be implemented.
Garwood also performs testing for category Z equipment under Section 16.5.2.3 Requirement for Equipment With Digital Circuits. Table 1 is a modified form of DO-160 Table 16-1.
This portion of DO-160 requires the applied bus voltage to be reduced to various levels for selected time periods with specified rates of change. Figure 1 shows the basic profile of the voltage applied to the EUT.
T1 is defined as the time from the start of the interrupt to the start of the recovery. T2 designates the time for the decaying voltage to reach 0 V if allowed to do so. T3 denotes the time to rise from 0 V to Vnom.
Since many of the interrupts do not reach 0 V and the programming of SmartWave sources typically requires the true start and stop points for time and voltage, the slew rate for the voltage decay and rise must be calculated. When these rates are known, the time period needed at the reduced voltage can be calculated. The time at reduced voltage (% of Vnom) is shown as T4a in Table 1.
Test Condition 1 requires a voltage ramp-down and ramp-up in under 1 ms. A typical DC supply with its large filter capacitance has a great amount of stored energy and could not be varied at such rates.
In the past, one method of testing used a linear power amplifier driven by an arbitrary waveform generator programmed by a PC. The EUT then was subjected to each of the test conditions for category Z a minimum of two times.
Using the SmartWave power source, Garwood Laboratories also tests products at 115 VAC, 400 Hz for customers. The tests required for Section 16.5.1.5.1, group 1 equipment are as follows: operate at 115 VAC for 5 min. Repeat the following three times: 160 VAC for 30 ms, 115 VAC for 5 s, 70 VAC for 30 ms, and 115 VAC for 5 s.
An alternative method for meeting these surge tests is shown in Figure 2. This older approach created some problems:
- • Poor load regulation when a variable transformer is used.
- • Transients caused by switch bounce.
- • Contact erosion caused by switching under load.
- • Loss of precise timing due to relay physics.
- • Momentary loss of power to EUT when the relay is switched.
A Fault-Verification Example
Most products today, including your car fuel-injection system, the video display on the airline seat in front of you, or a fighter jet’s laser target designator system, must be subjected to EMC testing to assure they are not affected by normal power bus conditions or do not create any conditions on the power bus outside of a defined envelope.
For example, consider a failure of an actuator on the Presidential yacht. It is removed and sent back to the factory for failure analysis. Using a storage scope, the waveform anomalies on the ship’s power bus can be captured and uploaded to a programmable arbitrary waveform power source. The source recreates the suspect-event waveform and sends it via e-mail to the actuator manufacturer to verify that the seemingly minor transient could be the cause of the failed actuator.
Once the actuator manufacturer has the waveform file uploaded into the power source, verification of the fault can be performed. The failure analysis reveals the motor-controller module latched up when subjected to the power transients.
The waveform file is e-mailed to the actuator motor controller manufacturer. This subcontractor duplicates the condition as it occurred in the field. The controller manufacturer’s design staff can accurately simulate the real-world bus phenomena and validate the design changes implemented to improve the controller’s immunity to this type of disturbance.
The existing specification defining the required compliance testing now can be revised to prevent the same failures from occurring in future products. The data can be forwarded to the committee that maintains the standards for compliance for review and possible inclusion in the next revision of the compliance standards.
The results: No ambiguities regarding the true cause for the failure and a clear means to evaluate the proposed solution. Each supplier tier can exchange the same data for evaluation. All parties have a clear roadmap to resolve the issue at hand quickly and prevent similar occurrences in the future.
About the Author
Grady Keeton is an application engineer at Elgar Electronics. He has been an engineer for more than 20 years with the last 10 involved in power conversion. Elgar Electronics, 9250 Brown Deer Rd., San Diego, CA 92121, 800-733-5427, ext. 2251,
e-mail: [email protected]
FOR MORE INFORMATION
on DO-160D tests enter this rsleads URL
www.rsleads.com/304ee-195
Return to EE Home Page
Published by EE-Evaluation Engineering
All contents © 2003 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher.
April 2003