By Doug Rathburn
Implementing a connector characterization system for connector testing is a crucial step in ensuring product reliability.
 Figure 1. Electrical equivalent of a connector, showing leakage and continuity paths. |
The importance of quality electronic and electrical connectors has increased dramatically in the past several years in direct correlation with a dynamic electronics market. Connector performance is vital in applications ranging from motor vehicles to transatlantic telecommunications systems. Implementing a connector characterization system for proper connector testing is a crucial step in ensuring overall product reliability. The two most commonly measured parameters in connector testing are isolation and continuity. Isolation measurements are usually performed between individual connector pins, or between connector pins and the outer shell of the connector. Continuity testing is performed on pins to ensure that each one provides a low-resistance electrical path through the connector (see Figure 1).
Isolation (Insulation) Resistance
Isolation measurement tests determine the degree that individual pins are insulated from each other, or from other parts of the connector. In the worst case, two connector pins might be shorted together. However, ever-shrinking circuit and connector geometries, in conjunction with higher frequency signals, require that connectors also be tested for possible crosstalk between pins. At high frequencies, this sort of radiated leakage can be as disruptive to proper circuit operation as outright short circuits.
Isolation is usually tested by applying a voltage across two pins in a connector and measuring the resulting current flow between them. The corresponding resistance from the test is compared to a predetermined threshold value. If the resistance level is too low, the connector is rejected. Common threshold levels range from 106 to 1012 Ω.
Pin Continuity
As long-term performance of connectors becomes increasingly important, so does the testing of continuity through a connector system. Typically, pin continuity is tested by sourcing a constant current through a pin and measuring the corresponding voltage drop across the pin. Virtually all connector pins are made from metal alloys, frequently with a conductive plating (gold, silver or solder tinning) to provide as low a connection resistance as possible. Therefore, the aforementioned test scenario typically results in a very low voltage reading.
 Figure 2. Diagram of a single source/measure instrument with switch matrix card. |
Using high currents for continuity testing has two advantages. First, it ensures the resulting voltage signal will be above the noise floor of the test system. The ratio of the test current to the noise current determines the signal-to-noise ratio, and the accuracy of the continuity test. Second, higher test currents can serve as a stress test for the connector. Often, connector pins are tested at a higher current level than the product's actual rated current capacity in order to verify a performance margin. If stress testing is unnecessary, a digital multimeter (DMM) containing a low-noise source for ohms measurement can be used to measure the pin resistance directly.
Single Source/Measure Instrument Test System
As connector pin counts increase, it becomes increasingly advantageous to simplify test systems with multifunction instruments, thereby reducing the overall cost and complexity of the test system. Integrating as much functionality as possible into a single source/measurement instrument can greatly reduce programming, cabling and potential error sources (see Figure 2). Precision readback of voltage or current levels is also easier to accomplish, without requiring a separate meter.
A fundamental requirement of the single-instrument test system is a bipolar power source featuring low noise and provision for constant current or constant voltage output. A complete test plan can include both continuity and isolation measurements, with and without stress testing. Measurements can encounter resistances as high as 20 GΩ, and may have to resolve resistances as low as 10 µΩ, so the dynamic range of voltage and current sourcing should be adequate to the task. Desirable output capability can range up to 1,000 V and several amps.
Connectors can contain anywhere from 1 to 100 pins or more, so many measurement channels can be required to thoroughly test a connector. The most cost-effective way of adding channels is to include switching in the test system. A switching system enables specific points or pins of the connector to be tested, while others are isolated from testing. A personal computer can control the instrument and switching system, automating the entire connector assembly test process without the need for constant operator intervention.
High-resistance Test System
For specialized connectors, test specifications can require high-resolution resistance measurements at resistance levels greater than 1 GΩ. Such tests fall into the category of sensitive measurements that can exceed the capabilities of source/measure types of instruments. When accuracy and resolution are critical at high resistances, a high-resistance meter or electrometer may be a better choice than a single-meter instrument.
Low-resistance Test System
A DMM offers a convenient, self-contained method for connector testing because it can apply a test current across the connector pin's unknown resistance, directly measure voltage drop across the pin and translate this reading to resistance. For continuity testing, however, a DMM is often incapable of forcing enough current through a pin to produce a voltage drop sufficient to deliver the desired resistance measurement resolution. The noise floor even of a high-end DMM is seldom better than about 75 to 100 nV, which limits the resolution of resistance measurements to about 500 mΩ. Actual pin resistance values may be on the order of only a few micro-ohms. In such a case, it becomes necessary to use a precision low-level voltmeter (i.e., a nanovoltmeter) in conjunction with a source (or the source output of a source/measure instrument).
Switching Options
A typical switching system consists of a switch mainframe with one or more cards containing appropriate relays, connectors and circuitry. The relays can be programmed to connect the test instruments to various points of the circuit under test. Multiple tests can be configured easily by programming and storing test sequences in the switch mainframe. Some instruments and mainframes have built-in trigger hardware that allows instruments to trigger the switch mainframe, and vice versa. More sophisticated triggering capabilities enable the instruments to execute programmed test routines without computer intervention.
 Figure 3. Multiplexer configuration. |
The multiplexer and the matrix are the two most common switching topologies. Multiplexer (see Figure 3) switch cards are generally used when one instrument is to be connected to many discrete devices-under-test (DUTs). A device can be connected to the meter inputs and tested using a single relay switch closure.
The matrix configuration (see Figure 2) provides greater flexibility by enabling any instrument or DUT terminal in the system to be connected to any other terminal in the system.
Figure 2 shows the matrix configuration used with a single source/measure instrument, but the matrix is also ideal for multi-instrument test systems. Two channels must be closed to perform a measurement, but the matrix layout enables instruments to test any possible combination of isolation or continuity.
Guarding
Guarding is a measurement technique that reduces measurement errors by eliminating leakage paths between insulated conductors on a printed circuit board (PCB) or in a cable or connector. A meter with a guard output is required to perform guarded measurements (see Figure 4). In the case of isolation testing between adjacent connector pins, the resistance measurement can be affected by leakage between the pins of interest and surrounding pins, or between these pins and ground. By connecting the guard output from the meter to the surrounding pins, undesirable leakage between the pins of interest and other parts of the connector can be eliminated. When testing multipin connectors, it is important not only to guard the system cabling, but also to guard the pins that are not being tested.
Sources of Error
Noise can come from many sources in the production environment. When electrically charged objects are brought near an uncharged object (i.e., the DUT), small, unwanted voltages may be generated in the DUT. To minimize the effects of this electrostatic interference, all system cabling must be properly shielded. Whether the system cabling is single- or multi-conductor, only one shield should be used around the wire bundle.
 Figure 4. Guarded connections. |
All shields should be connected to a single ground point in the system to eliminate the possibility of ground loops. Ground loops can occur when multiple ground points exist in a measurement system, and voltage gradients exist across these ground points.
Current flows in the ground system and can affect the accuracy of measurements or even damage equipment.
Leakage currents in cables and fixtures can be a source of error when extremely low currents are measured. To minimize these problems, insulation in test fixtures and cables must be made of materials having impedances much higher than the impedances being tested. If proper care is not taken, a significant portion of the test current can flow through any low-impedance path to a variety of circuit points and affect measurement results.
The amount of capacitance in test system cabling determines the settling time required before an accurate reading can be made. Settling time is determined by a system's RC (resistance X capacitance) time constant. A coaxial cable, for example, can be modeled as a combination of a resistance between the signal source and instrument input and a capacitance across the signal. A high resistance value can impose a long settling time, even when the equivalent capacitance is small; higher capacitance values can increase the settling time dramatically. For best accuracy, four to five time constants should elapse before taking the measurement, at which time the reading will have reached over 99 percent of its final value. Settling times can be reduced by minimizing cable length, by guarding the system and by using the source-voltage/measure-current method for making high-resistance measurements.
A common source of voltage or resistance measurement error is the series resistance of test leads running from the source and measurement instruments to the DUT. Ordinary two-wire measurement techniques tend to see this resistance as part of the DUT. This effect is particularly detrimental when long connecting cables and high current are used because the voltage drop across the leads becomes a significant part of the measurement.
The four-wire remote sensing method can be used to overcome the limitations of the two-wire technique. With the four-wire method, a current is forced through the DUT using current sourcing leads (in this case, the Output HI/LO leads of a combined source/measure instrument), and the voltage across the DUT is measured using separate Sense HI/LO leads. Virtually no current flows in the sense leads, so the voltage measurement consists of just the voltage drop across the device.
Thermoelectric electromotive forces (EMFs; voltages) can cause measurement problems, especially for low-impedance measurements. The voltage drop across low-impedance devices, such as connector pins, is typically very small. Thermoelectric EMFs are caused by the Seebeck effect, in which the mere contact of dissimilar metals generates small voltages that vary with temperature. These voltages can have sufficient magnitude to interfere with the test signal. Many instruments can be programmed to cancel the effects of thermoelectric offset.
 Figure 5 |
Resistance measurements using the delta method of thermoelectric offset compensation involves taking two readings. One voltage reading (V1) is made with current flowing (I1) through the DUT in one direction. The second reading (V2) is performed with the polarity of the test current reversed (I2). These two measurements are then subtracted from each other and the resistance is calculated as shown in Figure 5.
Conclusion
The testing of electronic and electrical connectors, like many other active and passive components, continues to evolve. Current trends are toward greater automation of instruments that have feature sets and measurement capabilities designed for specific test applications. Highly capable test systems can now be created using instruments that integrate source and measurement functions. In the case of connector testing, such instruments, combined with an electrometer and switching matrix, can perform virtually any isolation or continuity test on connectors having up to hundreds of pins.
DOUG RATHBURN is a senior applications engineer, Keithley Instruments Inc., 28775 Aurora Road, Cleveland, OH 44139-1891; (216) 248-0400; Fax: (216) 248-6168; Web site: www.keithley.com.
SPEC SHEET
End Applications: Testing of electronic/electrical connectors
Related Products: Digital multimeters, electrometers, micro-ohmeters, cable, PCBs
Main Point: The testing of electronic and electrical connectors continues to evolve. Highly capable test systems can be created using instruments that integrate source and measurement functions. Such instruments, combined with an electrometer and switching matrix, can perform virtually any isolation or continuity test on connectors having up to hundreds of pins.
