# Setting Practical Resistance Specs for Continuity Testing

The IPC\WHMA-A-620 Rev A standard states …“in the absence of agreed on test requirements between manufacturer and user the continuity spec for class 3 assemblies should be 2 ohms, or 1 ohm plus the actual resistance.” Cirris endorses this as a reasonable method for determining continuity test specs. However, when more stringent continuity test specs are dictated, the following guidelines should be followed in order to avoid setting a spec so tight that “in-specification” cables fail testing.

## How to determine the lowest “practical” test resistance threshold

When testing a cable or harness for continuity, there are three factors that must be considered:

1. Theoretical resistance of the “ideal” device under test (DUT). This includes both the wire and mating-contact resistance of the DUT.
2. In-tolerance variations in the actual resistance of the DUT.
3. Added resistance for measuring the DUT (test fixture resistance and measurement tolerance of the tester used to test the DUT).

These three factors work together to create the following equation:

(Theoretical Resistance of Ideal DUT )  +
(Worst-case variances for in-spec wire and contacts)  +
(Practical variances of tester accuracy and fixturing resistance)

=
Lowest practical resistance threshold for DUT test

In a perfect world we would determine the “exact” resistance of the assembly and then add some margin of error to determine the test threshold. In reality, many factors come into play that must be added to this “ideal” resistance in order to be able to practically test harnesses in a production environment

### 1. Theoretical resistance of the “ideal” device under test (wire and contact resistance):

Calculate the theoretical resistance based on length and gauge of wire and the connector manufacturer’s specification for the contact resistance. When all wires are of similar length, use the wire with the smallest diameter. Measuring a single sample will NOT give you a proper theoretical ideal. It should give you a resistance value that is less than the theoretical ideal plus some portion of the total in-spec variation as described below:

### 2. Variations in the actual resistance of the device under test:

#### Variance due to length:

Most drawings include some kind of tolerance in the end-to-end length of the harness. Make sure to use the LONGEST ALLOWABLE length when figuring your “ideal” theoretical resistance. Also, allow for “pig-tails” or extra wire length required in the connector.

Variances of the wire resistance:

Where a maximum resistance spec is given on stranded copper wire, it is typically 8% above the “nominal” resistance given from a table for solid copper wire due to reduced wire cross sectional area diameter of copper. If you use the nominal resistance of solid copper to theoretically determine the “ideal” resistance of your DUT you MUST factor in this resistance variation (at 500 milliohms, an 8% variation in wire resistance could result in as much as 40 milliohms of variance!).

#### Variance due to twisted wires

Twisted pair wiring has higher resistance-per-foot because each wire is longer than the end-to-end length of the cable itself (if you don’t believe it, untwist a pair and measure the exact length of the individual wires!). Length and resistance increases with the twists per foot and with larger outside wire diameters (includes insulation). If you divide the outside wire diameter into the length for one full twist, you have a ratio that can be used to compute this increase in length. Typically, twisted pair wire has between 0.5% and 3% HIGHER resistance than wires without twists.

#### Variance due to temperature:

Most wire resistance charts are based on 20 degrees C (68 degrees F). This will change by approximately 4% for every 10 degree C rise in temperature (2.2% for every 10 degree F rise in temperature)

## 3. Variations in measurement (Tester Tolerance and Fixturing Resistance):

Even the most accurate testers have some kind of measurement resolution and tolerance. Typical for most Cirris testers is +-1% with 0.1 ohm resolution in 2-wire mode, and +-2% with 0.001 ohm resolution in 4-wire mode.

To test cables in a production environment, you must plug them into an automatic cable tester. This is done through a mating fixture or “test adapter.” You can use either a “2-wire” fixture or a “4-wire Kelvin” fixture. For a detailed description of 4-wire fixturing click here. With a 2-wire fixture, the resistance of the interface wiring must be added. With a 4-wire fixture you can eliminate the resistance of the fixture wiring.

Whether you use 2-wire or 4-wire fixturing, you must include the mating contact resistance of the connectors in the DUT as part of the ideal DUT. This is a spec you should be able to get from manufacturer’s data sheets. However, you will likely soon exceed the specified mating cycles on your mating connectors of your test adapters as you test production volumes. You may be surprised by the small number of mating cycles allowed for this contact resistance number to remain in spec. By allowing for some additional increase in your resistance test spec due to this “connector wear” under high mating cycles, you may save yourself a lot of fixture maintenance/replacement costs. Usually female contacts wear out faster than male contacts because of mechanical fatigue of the spring that is part of the female contact forces.