Resistor Network Inspection: Catch a Terminal Swap Before Calibration

Measurement purpose

Verify accessible resistor-network terminal identity before calibration.

Specimens and conditions

Circuit state

Declared passive topology and external connections removed or modeled.

Orientation

Defined face, asymmetric landmark and fixture channel map.

Equipment and records required

• Resistance measurement: Known excitation, suitable range and controlled lead/contact effects.
• Fixture verification: Independent channel-to-contact continuity reference.

Method sequence

1. Predict

   Calculate correct and candidate-swapped pair values.

   Record: Expected fingerprint matrix.

2. Measure

   Acquire discriminating pairs before calibration.

   Record: Raw values and mapping decision.

3. Resolve identity

   Use independent adapter checks or physical landmarks for ambiguous signatures.

   Record: Accepted terminal map and unresolved symmetric nodes.

Decision and uncertainty

Accept identity only when the measured signature distinguishes the permitted map from relevant alternatives.

Overlapping windows and symmetric terminals remain unresolved electrically.

Network drawing owner and test-method owner.

Traceable outputs

Start with a physical terminal map

Number terminals from a defined viewing side and an asymmetric landmark. Show whether the drawing is viewed from the printed face, the back or the mating connector. Photograph the actual orientation in the fixture. A correct electrical table attached to a mirrored photograph can produce exactly the terminal swap the inspection is intended to catch.

Separate the product terminal name from the fixture channel name. The mapping should state, for example, that product terminal B connects to fixture channel 7 in a particular nest. Preserve that relation through adapter cables and switch matrices. A label on the fixture proves only its label; an independent continuity check of the empty adapter establishes which physical contact the channel actually reaches.

Define the unpowered network being measured

Identify every resistor and internal connection between the accessible terminals. Disconnect external circuitry only through the approved safe procedure and document the resulting state. Protection diodes, amplifiers and switches can make an ohmmeter reading depend on polarity or range. A passive printed resistor network must not be evaluated using an isolated-resistor equation while it remains connected to those paths.

Record the stimulus and range used for the fingerprint. Keep test power low enough for the agreed method so heating does not move resistance during the sequence. Where lead resistance is material, use appropriate sensing connections. These controls support repeatable identity checks; they do not by themselves establish absolute resistance accuracy or a final precision ratio.

Calculate the expected pairwise fingerprint

For each selected terminal pair, calculate the equivalent resistance with the remaining terminals in their stated condition. In an isolated series string, a pair spans the sum of the intervening elements. In a bridge or ladder, current can travel along several paths, so use the complete network equations. Do not fill a matrix by adding the resistor values that happen to look closest on the layout.

A network with n accessible terminals has n(n−1)/2 distinct unordered pairs. Four terminals therefore provide six pairs. This is the size of a complete pair list, not a requirement to measure every pair. First enumerate the plausible assembly swaps, then select measurements whose predicted results differ. Keep a complete expected matrix available to investigate a failed shortened check.

See why one correct total does not prove the pinout

Assume a hypothetical series string A–B–C–D containing 1 kΩ, 2 kΩ and 4 kΩ in that order. The expected A–D reading is 7 kΩ. If the labels or fixture contacts for B and C are exchanged, A–D is still 7 kΩ and B–C is still 2 kΩ. Those two apparently correct results do not detect the swap.

The expected A–B value is 1 kΩ but the swapped mapping returns 3 kΩ. A–C changes from 3 kΩ to 1 kΩ, while B–D and C–D exchange 6 kΩ and 4 kΩ. These complementary changes provide a much stronger identity signature than a single failed value. The example assumes an isolated, linear series network and ideal contacts; its resistance values are not company product specifications.

Separate identity windows from precision calibration

Select identity limits that account for the expected element variation and measurement uncertainty while remaining separated for the candidate maps. In the example, differences of several kilohms may be far larger than normal reading uncertainty, but that separation must be evaluated for the actual drawing. Do not copy a universal percentage into every fingerprint.

If two candidate mappings have overlapping predicted ranges, the measurement cannot reliably choose between them. Add a different pair, change the accessible circuit state where permitted, or use a physical landmark. Do not improve the apparent separation by calibrating the same readings first. Calibration coefficients should be applied only after the identity test has established that each observation belongs to its intended node.

Recognize terminals that resistance cannot distinguish

Electrical symmetry can make a terminal exchange invisible to every passive pairwise resistance measurement. For example, a perfectly symmetric network may have two branches whose terminals can be exchanged without changing the entire matrix. More repetitions of the same electrical test cannot establish a physical distinction that the topology does not contain.

Use orientation marks, asymmetric geometry, keyed tooling or another traceable physical discriminator when the drawing requires those terminals to remain separately identified. A powered functional state can sometimes break the symmetry, but only if the actual circuit introduces a known difference. Record the limitation explicitly. Passing a resistance fingerprint means the measured map is electrically consistent with the model, not that every hidden marking or construction detail has been verified.

Distinguish a swap from contact or resistor faults

A terminal swap tends to permute a coherent group of expected readings. A poor contact can raise or destabilize several readings involving one terminal without producing the complementary exchange expected from a swap. A changed resistor affects the pairs whose current paths include that element. Compare these patterns before replacing the product or editing the test program.

Use an independently verified adapter or direct accessible contacts to separate product and fixture mapping. Preserve the failing channel assignment and raw data before any cable is moved. If a fault follows one fixture nest across different specimens, investigate the fixture. If the same physical product terminal remains anomalous in an independent connection, the product or its terminal condition deserves closer examination.

Keep identity verification ahead of adjustment

The production sequence should retain a terminal-map result before trim, ratio calibration or system compensation. Re-run affected checks after an adapter, nest assignment, terminal drawing or test-program mapping changes. Store the raw pair values and method revision so a later engineer can see why the mapping was accepted rather than seeing only a final pass flag.

For a custom thick-film resistor network enquiry, send the circuit graph, accessible terminals, viewing convention and the functions controlled by each node. ChipSimple can review the drawing-specific passive network and inspection access. The identity fingerprint complements absolute and ratio testing; it does not replace thermal, loaded-circuit or environmental verification. Its practical benefit is preventing a calibration process from making the wrong connection look temporarily correct.