Ask Dr. Bob

Group 2A Tests:
Supplementary Connector Housing Testing Options
By Max Peel, Senior Fellow, Contech Research

The test sequence Dr. Bob outlined in his last article is a supplementary sequence of tests performed at the same time as Group 2 tests. The basic difference is that the Group 2 variable measurements concentrate on measuring low-level circuit resistance (LLCR) or contact resistance (CR), and the supplementary group deals with insulation resistance (IR) and dielectric withstanding voltage (DWV).

Since IR and DWV are high-voltage tests, it’s important that these measurements are performed on contacts that are not being measured for LLCR or CR. High voltages can puncture thin films and/or oxides, which may indicate electrical degradation of the contacts. Thus, if all three variables are measured on the same contacts, problems may be “masked,” and therefore missed. As a result, separate samples are used to avoid that possibility.

For printed circuit board (PCB) connectors, it is also recommended that the IR and DWV samples are tested when they are not mounted to test boards. This is done to eliminate the influence of any test board degradation, which may affect the final test measurements.

There are two techniques used to perform IR and DWV. The first is the point-to-point technique, whereby each adjacent pair of contacts is measured. The second is to bus all of the “plus (+)” polarity contacts together, and the same for all of the “minus (-)” polarity contacts. For the latter technique, a complete connector is tested in one measurement. It is also a bit more severe than the point-to-point technique. If a failure occurs, it is necessary to break the circuit up to determine if it occurred as a result of an isolated position, or as a result of a gross defect. If connectors are tested in a mated state, they will have to be unmated to determine if the plug or the receptacle is at fault.

The “bussed” technique with mated connectors is the most common technique. It is significantly less expensive and it’s more efficient from a time point of view. For example, the point-to-point technique takes three to five minutes to perform a measurement. If a 100-position connector is tested (50 adjacent pairs), the test would take two-and-a-half to four (or more) hours per mated pair, or double that if testing is required in an unmated state. For many of today’s high-density connectors on 1.0 to 2.0mm centers, the time for test explodes (not unusual for 300 to 400 positions to exist). Since the variables are measured at least twice, and more likely three times in the sequence, it could take up to 60 hours to perform this test, contingent on the sample size and excluding sample prep time.

The bussed technique takes the same amount of time to perform. However, it measures the variable across the whole connector, as opposed to across individual pairs. Using the above example (100-position and 50 adjacent contact pairs), the time per connector is three to five minutes, as opposed to two-and-a-half to four (or more) hours. Thus, a significant decrease in cost and time can be realized.

Due to requirement levels (particularly IR levels) and the plastic used for connector housings in today’s connectors, there are very few cases for failures (U.S.- and European-manufactured products). The exception would be the use of lower grade nylon families and/or other high-moisture absorption-type plastics, especially the non-glass filled variants.

If test boards have to be used, the trace size, annular rings, and soldering test leads have to be carefully evaluated to be sure the proper spacing is maintained. It is not unusual to have failures recorded that are associated with improper spacing between traces, annular rings, or improper soldering, which decreases the distance between adjacent features that have nothing to do with a design fault. This is also related to the 1.0mm to 2.0mm center connector styles.

A secondary issue is when small connector densities are exposed to a humidity-type test. Contingent on the mass involved in combination with different temperature levels, a minute amount of condensation may occur, which will result in failures. Again, this is more commonly found with the high-density, small center-type connectors.

In conclusion, the requirements which have been established are:

1. Insulation Resistance (IR): To determine the resistance of the plastic housing (adjacent positions) and/or seals of a connector relative to leakage through or on the surface of the material.

2. Dielectric Withstanding (DWV): To establish that a given connector can operate at its rated voltage and withstand momentary overpotentials due to switching, surges, and other similar phenomena.

It is basically a spacing issue. The smaller the centers, the greater the potential for failure in both instances if the requirement levels have not been verified. Many, if not most, product specifications indicate failure levels of the above. The basic issue is that many have no idea where these requirements came from, again particularly with the newer high-density styles.

For DWV requirements, the now-canceled MIL-STD-1344 does indicate a technique to establish DWV and working voltage levels in its “purpose” statement. The EIA document (which has replaced 1344) also indicates the technique in its “normative” section.

The technique to be performed is a breakdown test on a given number of adjacent positions (10 adjacent contact pairs is the recommended sample size). The DWV requirement is established as 75 percent of the lowest breakdown voltage observed. Working voltage is established as one-third of the DWV level.

The above deratings are somewhat arbitrary, but over time (when used), they have proven to be adequate and are not, in the opinion of the writer, to be under- or overrated.

Insulation resistance requirements can be established by increasing the leakage current to a point of shorting and then establishing a derating level, to indicate the requirement levels in megohms and the test voltages to be used.

Only a few manufacturers use these techniques to establish proper criteria. Most either guess or use the same values that are in other catalogs, if available. Some specifications allow wiping surfaces clean, and in a few instances, allow a post-conditioning (50°C for 24 hours). This, for the most part, guarantees passing the requirement, which leads me to wonder, “Why bother to test?”

Send your comments and questions to AskDrBob@connectorsupplier.com.

Max Peel is a Senior Fellow at Contech Research, an independent test and research lab located in Attleboro, Massachusetts, U.S.A. For more information, visit www.contechresearch.com.