Testing for intermittent faults in high value Wiring Harnesses

If a wired assembly has been “tested” does it mean that it is definitely good? What about that bugbear of electrical assemblies – The matter of intermittent faults manifesting in electrical cable assemblies and wire harnesses?

Like most electronic assemblies, cable and wire harnesses are subject to manufacturing defects and it is standard procedure to electrically test them. Hard, (repetitive) failures are relatively easy to detect, and testing is an effective means to eliminate them. Soft, (intermittent) failures can be much more difficult to find. However, it is possible to improve product through testing by understanding the source of intermittents and attacking their root cause.

Since intermittent problems are just that, intermittent, they are much more difficult to find than solid defects and the best techniques for finding them are often overlooked.  Sadly, most cable harness testers in use in the aerospace industry use old test point designs that require relays to switch from point to point and this technology cannot be expected to find fast intermittent connection problems.

A test engineer has to ask;

1. Will the harness tester, and the interface fixturing, support the detection of intermittent problems?
2. Can test data be utilised for feedback, and can appropriate action be taken to remove root causes, rather than just futilely attempting to “test quality into” assemblies.

What is an intermittent fault?

There are three basic manufacturing problems that occur in wire harnesses. Opens, shorts and mis-wires. Since wires generally don’t change location by themselves, intermittent shorts and opens are the problem. Intermittents are usually caused by some mechanical change; temperature, mechanical vibration, or physical flexing, which changes (temporarily) the electrical characteristics of the device under test.

Intermittent Shorts: Failures in insulation between conductors that should be isolated.

In most cases intermittent shorts are caused by some compromise of the insulating material intended to keep wires isolated from each other. When insulation is breached, by whatever means, there is now only an “air-gap” isolating wires from each other for some portion of the cable. When some force (bending, flexing, vibration, aging, etc.) causes the conductors isolated only by an air-gap to come into physical contact with each other, a short-circuit occurs. This short may become a hard failure, or remain an intermittent failure, depending on the nature of the force, however it is almost certainly to become a future field failure. Many intermittent shorts may be found by high speed continual-scan flex-testing in production, however a more thorough test is using non-destructive (low current) high voltage tests (Dielectric Withstand Voltage – DWV).

Intermittent Opens/High Resistance Connections: Failures in the continuity of a conductor.

An intermittent open, or high-resistance connection, occurs when there is temporarily no electrical connection, or weak electrical connection (high resistance) from one end of a wire to the other. This can occur in the contact-to-contact mating at either end, the contact-to-wire bond at either end, or the wire itself.

Intermittent Contact-to-Contact Problems:

1. Micro-fretting. Some metals, when they come into contact with each other, form surface oxides over time. This is common for tin/lead contacts and is often recognised as micro fretting. Tin contacts are especially susceptible to this problem unless they have high mating forces (good wiping action) or the application uses more than 10 volts. Micro-fretting is not a problem for gold plated contacts as long as the plating remains intact. Often this problem does not show up in test. As soon as connectors are re-mated it goes away. A common field solution of re-mating connectors can cause this problem to disappear only to later reappear. See Tyco’s technical report “The Tin Commandments: Guidelines for the Use of Tin on Connector Contacts” for more information on problems with micro-fretting.
2. Improper mating forces. Intermittent opens can occur when contacts do not mate with the proper force. This can be caused by:
   1. Wear that weakens the spring forces in female contacts or reduces the size of male contacts.
   2. Contacts that are not fully mated. (crimped pins aren’t fully locked into the connector housing and push back when mated)
   3. Connector shells/mating hardware that interferes with full engagement of the mating connectors.
   4. Mechanical damage/ deformation of contacts.
3. Contamination on contacts. This can be caused by flux left from soldering, dirt, etc.

Intermittent Contact-to-Wire Problems – Bond is not Gas Tight

1. In crimped connections intermittent opens (or high resistance connections) are often caused by poor crimps due to improper mechanical adjustment or misuse of crimping tools. Consider electrical crimp testing in millionths of ohms or crimp force monitoring.
2. Cold solder joints and poor hand soldering can cause intermittent problems.
3. IDC connectors can have intermittent problems due to bent tines, cut strands, misalignment of planar cable, including variation in spacing of wires in ribbon cable, and improper use of application tools.

Intermittent Problems in the Wire Itself

Apart from problems caused at the point of wire-to-contact termination as discussed above, intermittent problems in the wire itself usually occur after the assembly has been put into use. This can be caused by flexing, stretching, or bending a cable more than expected. These types of problems are usually not solved by improving testing, rather by design changes such as using more strands of wire in the conductors, limiting the flexing and stress experienced by the assembly with improved strain reliefs, etc., and by protecting the assembly with better jacketing.

Expensive Solutions for Intermittent Opens/High Resistance

High-speed event detectors (discontinuity testers that monitor resistance simultaneously on each conductor) may be used as part of a rigorous set of tests under controlled conditions, including temperature cycling and vibration over a range of frequencies on a shake table. This solution, while effective for catching intermittent problems, lacks popularity as a production test solution because:

1. Production processes must be streamlined for productivity. It takes time and expensive equipment (temperature chamber, shake table, event detector) to accurately perform these kinds of tests; hence they are usually reserved for connector testing qualification testing by connector manufacturers.
2. Concern of degradation of the assembly because of the rigours of the test.
3. Difficulty of replicating an event that causes real problems.
4. Lack of “complete” test (discontinuity testers don’t test for shorts or insulation problems).
5. False failures due to ESD (was it a real fault or just an ESD event in the vicinity of the test equipment?)

Inexpensive Solutions for Intermittent Opens/High Resistance

In order to solve intermittent problems in your cables and harnesses, an awareness of the test systems capability is first needed.  Can it enable high speed, (>256 tests per second), testing – not just high-speed relay switch closing etc.

Some cable test systems only perform a “single test” each time the test is invoked. If the tester is used in the normal “single test” mode, there is only a minimal chance of recognising an intermittent at the one brief moment when the wire with the intermittent is truly open.

Since the operator often has no feedback that they are experiencing an intermittent, the cable may be inspected and hand tested with no problem found. The cable may then be re-tested with the likelihood that this time again the intermittent is missed.

Some systems operate in a “test-until-good” mode, sometimes called “wiggle until good or “wiggle and jiggle.” If your process allows the operator to “wiggle until good”, you are blind to intermittent problems, even if the tester is capable of finding them. An operator just wiggles the cable until it passes. This type of testing has been popular in automotive harness testing where the assemblies tested are “not expected” to have intermittents because manufacturing processes have eliminated any chance of them occurring. It also allows production testing using test fixtures with intermittents.

Ultimately what is needed is a system that can capture what the intermittent problem was, locating exactly which connections were involved, even if you can’t re-create the intermittent fault. In some areas this is known as RTS (real time scanning) and it is something that the unique hybrid architecture of Cirris testers can provide as standard.  This is why F1 teams and their suppliers exclusively use Cirris to re-test each race car harness, (continuously) after every race, thereby finding the faults that conventional testers can’t see!

Once the relevant test system is installed and in use, the key question is, “how can one guarantee that the intermittent fault is in the test fixturing, and not in the DUT – device under test?” Test fixtures (interface cards, adapter cables, test blocks, etc.) have all the same characteristics of the actual harness being tested; connectors, wires and terminations, along with one big problem in that the test connectors see many more mating cycles than the assembly being tested. Ignoring intermittents (wiggle until good) saves on fixture costs and maintenance, at the expense of being blind to real intermittent problems in wired assemblies.

Upgrading Test Fixture Quality

Before intermittents can be eliminated in the product under test, intermittents in the test fixturing need to be resolved. Typically, there are two types of test fixtures employed in testing cable/harness assemblies; actual mating connectors, or test blocks using spring-loaded (often gold), contacts that imitate the actual mating connector.

Actual Mating Connectors

In lower-volume, long-life connectors, (typically computer, medical, military, communications) it is popular to just use the mating connector in the test fixture. In this case be sure to use gold plated versions of the mating connector. In high volume testing “gold flash” plating will not hold up. Refer to manufacturer’s data sheets to insure adequate gold plating, and for information on expected mating cycles. In the event of very high volumes even gold-plated connectors will wear out. For these applications you may want to choose a “replaceable” style adapter, where the connector can be easily replaced on the same test fixture.

Test Blocks Using Spring-Loaded Contacts

High-volume, low mating-cycle connectors (typically automotive and appliance) generally use tin plating and the actual mating connectors are unsatisfactory for high volume test requirements. In this case milled, or moulded test blocks using spring-loaded contacts are used. These types of fixtures can be problematic where intermittents are concerned due to the pogo pin connection in lieu of an actual mating connector. In these instances, use spring-loaded pins with higher forces and designs that minimise the resistance in the spring-loaded contact, such as Everett Charles’ bias ball. Barrel to shaft style pogo pins often result in higher resistance. Pogo pins with sharp points help deal with tin oxides on the surface of pins in the harness being tested. The drawback: possible damage to connector contacts and test that pass despite contacts that are not “locked in”, damaged, or that have severe contamination on the surface. (Pushback test blocks can be used to test for pins not locked in).

Maintaining Test Fixture Quality

Test fixturing must be absolutely free from intermittents. To assure ongoing intermittent-free test, any fixturing requires a consistent program of test fixture verification and maintenance. A “shorting block” adapter can be created for each different test connector. This shorting block should be plugged in at regular intervals to “test the test connector” for intermittent problems. When discovered the test fixture should be repaired/replaced immediately.

Recommended Action Plan

Assess if intermittent problems can escape your current production process.  If Yes:

1. Update test equipment to that capable of catching very short-duration intermittent events (fast continuity scanning/ high-voltage testing)
2. Upgrade fixturing, eliminating the possibility of intermittents on your test fixtures.
3. Change your test process to call intermittent problems BAD, not “wiggle until good”.
4. Remove the root causes of intermittents. Testing for intermittents is far from 100% effective. After making your intermittents visible, attack the causes of these intermittents at the source.