Finding No Faults

Universal Synaptics offers a special method for locating intermittent and potentially dangerous faults in aircraft wiring.

By Ed Maher

Many do not want to believe that avionics wear out, but the electro-mechanical components of avionics systems are wearing out proportionally to mechanical failures. This includes wiring, connectors, crimps, splices, relays, switches, solder connections, circuit breakers, sensors, line replaceable units, and edge connectors.

For some reason, test engineers have thrown the solid-state devices that have low failure rates in with electromechanical components and the connections that mate the harnesses.

Existing testing methods consider both of these areas the same, so if the system can pass a functional test with no hard failures, then it is considered operational and safe to fly. The downside of this type of testing is it does not test for inter-connectivity wear that may result in intermittencies.

What this approach to maintenance and testing neglects is the aging problem, where pilots experience avionics failures, yet when ground tested, the failure cannot be duplicated. The degree of this problem is reflected in high no fault found rates on older avionics systems. These rates often exceed 50 percent or more of all avionics failures, on older systems.

Unlike mechanical systems, no direct testing or measurement operation presently verifies or enhances the inherent reliability of the repaired unit. The physical wear and tear that aircraft wiring and avionics components receive over time is not being tested. It is that simple.

But there is a solution. Just as you can monitor the noise levels in engine bearings, so, too, can you monitor the "noise" from electrical interconnections. You just need to use test equipment that is capable of detecting such degradation. A product called IFD-3000 made by Universal Synaptics introduces a new testing paradigm called reliability testing, which directly tests for intermittency, the predominant result of aging avionics.

The Intermittent Fault Detector or IFD-3000 began as the IFD-2000, but evolved due to increased testing demands. Research into aging wiring suggests that any intermittent, either opening or shorting, but at low levels, is likely ignored by the aircraft computers and especially the high-tech test sets designed specifically to check continuity. Because the underlying digital measurement technology used in virtually all test sets operates basically as a sampling system that can miss intermittent faults, a wide spectrum of age-related intermittencies might not be seen for what they really are, accidents waiting to happen. These unseen defects comprise the bulk of the no fault found problems found in flight, and they cannot be duplicated on the ground using digital-based testing systems.

A digital meter simply won't "see" real-world defects. These are ignored by averaging test devices, such as a high quality digital meter.

Digital averaging was simulated at the Universal Synaptics laboratory by connecting high quality meters to a pulse generator that is similar to those used both on the bench and in the field for troubleshooting. The generator delivered a steady 4-volt signal and was set to drop the signal to 0 volts when a single pulse was injected into the test harness. This simulates a short or open lasting from 1 to 100 milliseconds, depending on test parameters. At various "glitch" periods, Universal Synaptics sampled the meter via its computer and the latest IFD software, which plotted what the meter should be measuring, sort of like a history over time. Even at 100 milliseconds, the meter missed a couple of glitches with the rest of the injected pulses showing a 4-volt change at about 1 to 2 volts.

The same test was conducted at 100, 10, and 1 milliseconds. At 1 millisecond, the meter displayed no glitching at all. At 10 milliseconds, the meter presented only a 0.2-volt change. If the computer hadn't been recording the readings, a technician's eyeballs could easily have missed the meter's display momentarily blink. Early detection of these intermittents, both shorts and opens, is like a look into the future of a totally failing system.

Interestingly, the more accurate the meter, the more the errors were averaged out with a no-fault-found display. Some of the best meters available, the Keithley 2100 as well as an old Simpson 260, were integral to the test. The less accurate meter indicated a slight bump, but the more accurate meter displayed no change at all. This is not to say that the Keithley 2100 isn't a good meter, quite the contrary, it is a great meter and is part of the IFD-3000 test system partnership.

The Universal Synaptics tests demonstrated that the suspected problems were real, but more was needed to clearly demonstrate the gravity of the problem. A relay circuit board from a failed system was introduced into the test. To make sure the relay's operation could be monitored, the case was removed. A naked eye inspection revealed nothing of the defects, but after a blacklight test, it was immediately apparent that there were more than 19 almost invisible circuit cracks on the board.

Blacklight-sensitive Zyglo was applied to the board with remarkable results. The light revealed that more than 90 percent of the board's connection-to-edge connectors were defective. Admittedly, this board had gone around the block several times and many of the failures could be from installation and removal, but the results were dramatic.

As electronic and electrical systems age, they become more difficult to test and diagnose because of the random and often infrequent nature of faults. Intermittent systems that do not display this random failure during testing and diagnostics will most likely end up as a no fault found. Because a specific fault cannot be tied to these problems, it is often said that the no fault found problem is probably gremlins or if it goes away, it is left to fail later, ending up as another no fault found on the pilot's squawk sheet. If the problem can't be found but continues to occur, parts start flying off the shelves in a shotgun approach, which usually compounds the problem even further.

Finding this type of problem is readily accomplished using the added features of the latest version of the IFD-3000, with its new software, interface, and accompanying options. The latest addition to Universal Synaptics's arsenal is the Keithley 2100 for complying with ohmic and capacitive testing requirements and the Eclypse ESP test set for finding definitive short or open failures and their exact location in any type of cable or wiring. They form a harmonious relationship to target a failure and its location.

Finding the Fault

Eclypse International's ESP fault-location meter is a unique test unit that can be used on a wide range of wiring and coaxial cabling, such as triaxial, multistranded, and even twisted pair. A spin-off from NASA's reflectometer prototype design, the ESP standing wave reflectometer is a handheld, battery-operated test set with the capability of testing up to 1,000 feet from the test unit. Priced at $5,500 dollars, it is probably too pricey for many small shops, but not for the heavy metal or commercial maintenance. This is a relatively small price to pay for slashing the cost of maintenance.

The ESP features a liquid-crystal display that provides system status, menu items, cable type, and the "ready for test" display. There are five buttons, one turns the ESP on and off while yet another provides a menu and the third is used to enter changes. To scroll up and down, there are buttons to search though the menu. The next one is really important, the test button, which starts the automatic test function and the display backlight. The last two are the decimal point and the numerical keys for changing impedance and velocity.

Although, it works more efficiently using a known good wire or cable, the ESP is capable of reaching a reasonably accurate reading with only the onboard programmed algorithms. Turn it on, select the cable to be tested and attach it, then depress the test button; that is about it for testing for shorts or opens with an accuracy of just a few inches. That plays out to an extremely accurate reading when looking at a harness that may be as long as 20 or 30 feet. Once you've established the velocity and impedance for a given coaxial cable, the data can be entered into a list that can be used over and over without reconfiguring.

During a recent test at a general aviation facility, the ESP showed its capability of checking the status of the bonding on an antenna system, plus the connecting cable quality. These features could help pay for the ESP via antenna tests during scheduled inspections.

On the right side of the ESP there is a serial data port that allows it to connect to the RS232 port on a laptop. With the software installed and opened to the display screen and the computer tied to the ESP, the user simply has to depress the "3" key for about two seconds; the screen will automatically display "RS232 port" with its status displayed as either "enabled or disabled." When the test set's test button is engaged, the data starts to flow into the ESP with a graphical display showing the condition of the cable or wire. These results tell the user if the cable is kinked, too sharply bent, or crimped by being wedged between two racks. This data screen can then be saved to a text file for placement into a library for future reference against other installations, and Eclypse will add this data to an established library for access by users all over the world. Customers can access Eclypse's website and obtain the latest library additions.

Contact:

Universal Synaptics corporation
Ogden, Utah
Phone: 801-731-8508
http://www.usynaptics.com/

Eclypse International
Corona, California
Phone: 909-371-8008
http://www.eclypse.org/