Engineering resource

Asset tracker PCBA: radio modules and battery connections

An asset tracker can work on a development bench and reset when its radio transmits from the actual battery pack. The board, battery cable, connector, firmware and enclosure form one operating system. A turnkey assembly order should identify that system clearly enough that the supplier can reproduce the approved configuration and distinguish an assembly fault from a network or power-source limitation.

This article covers the manufacturing handoff for a tracker whose radio architecture has already been selected. It does not recommend a battery chemistry or establish a radio approval. Use the IoT assembly requirements to define the board deliverable, then document the power connections and test conditions that make that deliverable meaningful.

Freeze the complete module identity

Record the full module ordering code, hardware revision where applicable, modem firmware and application firmware. A shortened family name is not enough for procurement or test configuration. Document the antenna part, cable assembly and any regional build variant in the same release. The person approving substitutions needs to understand which combination was evaluated, not just whether a replacement fits the footprint.

Keep identifiers assigned to individual devices separate from generic firmware settings. Define which identifier is read from the module, which is programmed by production, and which is associated with a customer account later. The manufacturing record should reveal a duplicate or missing identity without exposing unnecessary account credentials to operators or including them in ordinary build reports.

Prepare an approved recovery path for a module that fails programming or identification. Record its original state and the recovery action. Repeatedly replacing firmware until communication succeeds can leave the batch with mixed versions unless the final verification explicitly reads back the intended configuration.

Review voltage at the radio supply pins

Nordic’s nRF9160 supply guidance identifies transmit-related voltage drops as a design concern and gives device-specific supply and decoupling requirements. It also distinguishes the conditions associated with RF specifications from the device’s lower operating boundary. For a different radio, obtain the corresponding limits from its own documentation rather than transferring the nRF9160 values.

Measure at the point relevant to the specification. A stable bench-supply display does not demonstrate a stable voltage at the module after the fixture cable, contacts, protection circuit and board traces. Provide accessible measurement locations and a documented probing method during design validation. Include the ground reference so different teams do not compare measurements made across different parts of the power path.

Use the actual operating sequence in the investigation. Startup, network registration, message transmission and location acquisition may impose different demands. Record which event caused a minimum voltage, and correlate it with the current trace and reset indication when available. A single number without its event context is difficult to use for corrective action.

A simple resistance calculation reveals a hidden margin

Consider an illustrative power path with a combined series resistance of 0.40 ohm and a transient current of 0.35 ampere. The resistive voltage drop is 0.14 volt. If a different cable and connector raise the total resistance to 0.70 ohm, the same current produces a 0.245 volt drop. The change consumes another 0.105 volt of supply margin before other dynamic effects are considered.

These are invented engineering-review values, not tracker specifications. The calculation treats the path as a simple resistor and does not model battery dynamics, regulator response, inductance or capacitor behavior. It is useful for showing why a connector or cable substitution deserves review. It cannot establish the complete minimum operating voltage or guarantee that a product will survive a transmit event.

During qualification, replace the illustrative numbers with measured or justified values for the actual configuration and conditions. Compare the result with the selected module’s requirements and the design margin approved by engineering. If the trace shows a failure, preserve it before changing capacitance or wiring. Otherwise the team may lose the evidence needed to identify which change resolved the problem.

Specify the battery harness as an assembly

Give the harness its own controlled drawing and part identity. Show connector manufacturer and ordering code, contact positions, wire specification, length, strain relief and the view from which pin numbering is defined. Make polarity explicit at both ends. A drawing viewed from the mating face can be misread when an operator is holding the wire-entry side.

Do not use wire color as the sole acceptance criterion. Define an electrical continuity and polarity check that the assembler can execute before connecting the harness to the board. If the product includes a temperature-sense or identification connection, include it in the pinout and test record. A mechanically compatible two-wire harness is not equivalent to a qualified multiwire battery interface.

Record how the cable is routed and retained in the enclosure. Show the permitted bend path and the features that prevent the cable from loading its solder joints or connector. Have the mechanical owner specify any retention verification. Avoid improvising a pull force or bend radius from an unrelated connector family.

Control the point at which the product becomes powered

State whether production ships a board, a board with an unconnected battery, or a complete activated tracker. Identify the operation that connects or enables the power source and the checks required before it. This is a manufacturing sequence decision as well as a product decision: once powered, the unit may begin drawing energy, joining a network or recording data.

Define the intended state after programming and test. A successful functional test can leave firmware in a diagnostic mode or a modem repeatedly trying to connect. Verify the approved shipping or storage state using the product’s documented procedure, and retain the evidence required by the release owner. Do not assume that unplugging the programming cable changes the application state.

Keep battery handling and charging instructions tied to the approved pack and product design. An assembler should not infer that a pack can be charged because its connector fits the fixture. If charging is included in the scope, identify the approved equipment, settings, termination criteria and exception handling in the controlled process documentation.

Measure the whole product over a defined cycle

Nordic’s historical current-measurement guide explains the measurement challenge presented by a large difference between idle current and active peaks. The page is marked as unmaintained. Its general measurement discussion is useful context; use current product and instrument documentation for settings, limits and firmware instructions.

Define whether the measurement includes the radio alone or the complete tracker. Sensors, indicators, regulators and connected accessories can consume energy outside the measured module path. Document every external connection, including programming and debug cables. A test arrangement that supplies part of the circuit through another cable can make the battery input appear better than the actual standalone product.

Select a measurement interval that captures the events relevant to the application. Label startup separately from repeated reporting cycles and prolonged inactivity. Retain the waveform or event record when it is needed to explain the result. An average value is useful only when its interval and operating state are known.

Use charge accounting without promising battery life

For a deliberately simplified example, suppose one hourly reporting cycle consumes 3 coulombs during its active portion. Dividing by 3.6 converts that amount to approximately 0.833 milliampere-hour per cycle. Twenty-four identical cycles would consume approximately 20 milliampere-hours per day from those active events alone. The example demonstrates units and accounting rather than predicting a real product’s endurance.

Add the remaining intervals instead of assuming they cost nothing. Include idle consumption, unsuccessful communication attempts, location searches and any scheduled maintenance activity in the test model where relevant. Keep the source of each value visible: measured data, a bounded assumption or an unresolved requirement. This makes the estimate reviewable when field behavior differs.

Do not turn nominal battery capacity divided by an optimistic average into a customer guarantee. Usable capacity and delivered voltage depend on the selected source and operating conditions. Ask the product team to define the evidence and margin supporting its life claim. Manufacturing can then verify the configuration and screens that the approved claim depends on.

Separate radio tests from uncontrolled network conditions

A failed message does not by itself locate a manufacturing defect. Preserve the test state, device identity, firmware, supply observations and available connection diagnostics. Distinguish a reproducible local electrical failure from a service or coverage issue. The production procedure should define when to retry, when to hold the unit and when to escalate a station or network problem.

Use a controlled test setup when engineering requires repeatable RF measurements. A desk test with an unspecified antenna location and changing surroundings is weak evidence for comparing boards. If routine production uses a simpler communication check, identify its detection purpose and its limits. Do not label that check as complete RF characterization.

Verify the assembled enclosure separately where its antenna placement or cable routing affects the qualified configuration. The box build scope should name the antenna installation, battery retention and final checks included in the order. Approval of a bare board does not automatically approve every mechanical arrangement around it.

Close the release with reproducible records

For the pilot batch, review one complete unit record from incoming module identification through programming, power checks, communication testing and final state verification. Confirm that another operator can follow the same instructions and produce interpretable evidence. Correct ambiguous pin views, missing firmware identifiers and undocumented fixture connections before increasing volume.

When preparing to request a project review, provide the board files, module and antenna identities, harness drawing, enclosure instructions and test scope together. Declare which parts are customer supplied and who approves changes. The resulting quotation can separate board assembly from harness work, configuration and final integration without losing the relationship between those operations.

For repeat orders, compare the proposed supply and process with that approved record. A new cable, changed firmware or alternate battery pack may require a targeted review even when the PCB revision is unchanged. Preserve the reason for each accepted change and the checks repeated afterward. That discipline makes field behavior easier to investigate and keeps an apparently identical tracker from becoming an undocumented variant.

Your next build,
clearly scoped.
Get a Quote ↗