A sealed connector can arrive with an IP claim, a correctly latched housing and an intact rear seal yet still leave a buyer unable to answer a practical question: did this exact terminated harness remain leak-tight when assembled with the quoted wire, cavity plugs, accessories, routing and process? A family brochure cannot answer that question by itself. Neither can one unexplained pressure number or a photograph of a connector under water.
For procurement, the useful deliverable is a controlled evidence chain. It identifies the connector halves, terminals, individual wire seals or mat seal, unused-cavity plugs, backshell, wire construction and production tooling. It then defines the pressure or vacuum boundary, fixture, fill and stabilization periods, measurement interval, calibrated leak standard, environmental conditioning, acceptance rule and raw result for every serialized specimen.
This guide explains how to request that evidence without turning one laboratory method into a universal product rating. It complements the IP67 versus IP68 procurement guide, the DEUTSCH DT sealed-connector BOM guide, the bulkhead pass-through and grommet guide and the cable strain-relief and bend-radius guide. Those articles address rating scope, component identity and mechanical installation. This one concentrates on leak-test configuration and evidence.
Nothing in this article verifies a SINAWATTS connector, harness, seal, process, test capability, certification, material, pressure limit, leak-rate limit, stock position, price, MOQ, sample timing or lead time. The connector manufacturer, applicable standard, equipment designer, qualified laboratory and project approval authority must define the controlling requirement for the exact application.
Direct answer: what belongs in the RFQ?
Require the bidder to return a leak-test plan and report that state, at minimum:
- exact connector manufacturer, family, housing part numbers, terminal part numbers and drawing revisions;
- mating state, mechanical-assist or latch position, CPA or secondary-lock state and any shipping or protective cap;
- seal architecture: individual wire seal, mat seal, interface seal, radial seal, gland, overmould or potting;
- wire manufacturer and article, conductor size and construction, insulation material, finished outside-diameter tolerance and surface condition;
- unused-cavity plug part number, permitted cavity range and installed orientation;
- backshell, dress cover, conduit, tape, heat-shrink, breakout and first-support position;
- production stripping, crimping, seal loading, terminal insertion, plug insertion and inspection instructions;
- test boundary and diagram showing every pressurized volume, port, tube, adapter, closure and potential fixture leak;
- positive pressure, vacuum or differential method, including units and gauge versus absolute reference;
- fill, equalization, stabilization, measurement and exhaust times;
- part and ambient temperature control, preconditioning and permitted drift;
- complete effective test volume, including connector cavities, tubing, manifolds and adapters;
- transducer range, resolution, calibration status, acquisition rate and software revision;
- traceable calibrated leak or master used to challenge the complete station;
- gross-leak check, fine-leak limit, pass/fail calculation, guard band and retest rule;
- sample count, lots, worst-case wires, cavity fills, orientations and destructive witness plan;
- conditioning sequence and whether the same samples are tested before and after exposure;
- raw pressure-versus-time data, calculated result, photographs, fixture ID and operator record; and
- change-control triggers for every component, wire, tool, setting, fixture, method and acceptance limit.
Do not accept “pressure tested,” “waterproof,” “IP67” or “zero leakage” as a complete return. Each phrase omits the test boundary, sensitivity and relationship to the offered harness.
Separate an enclosure rating from a production leak screen
The current IEC page for IEC 60529, checked on 2026-10-03, describes the classification of degrees of protection provided by enclosures. An IP code communicates protection under the defined standard conditions and scope. It does not, by itself, prescribe the supplier’s pressure-decay fixture for every terminated cable harness, nor does it convert a pressure-decay result into an IP rating.
A production pressure-decay screen answers a different question: does the assembled unit behave within an established pneumatic limit under one controlled station method? It can be fast, dry and non-destructive, but its sensitivity depends on test volume, pressure, temperature, stabilization, fixture leakage and measurement time. Qualification may use submersion, spray, thermal cycling, pressure/vacuum conditioning or other methods. Production may use a correlated dry-air screen. The RFQ must state how those decisions connect.
For example, a supplier may qualify a connector system under a defined environmental sequence and then screen each finished harness with pressure decay. That approach is reviewable only when the supplier shows that the production method detects the defect size or failure modes relevant to the qualified design. A dry test that can find a missing cavity plug but cannot distinguish a damaged wire seal is not automatically an adequate screen for both.
Likewise, do not infer that a connector marketed with an ingress rating retains that rating with every cable. TE Connectivity’s official Heavy Duty Sealed Connector Series FAQ, checked on 2026-10-03, ties sealing statements to named configurations and directs users to the relevant product and application specifications. That illustrates the evidence boundary: family-level performance still depends on the specified components and assembly instructions.
Freeze the sealed assembly before choosing the test
The part under test is more than two plastic housings. Freeze the complete sealed bill of materials and the process that creates it.
Housing and interface seal
Record both mating housings, keying, cavity count, latch and any connector-position assurance device. Identify the interface seal and its controlled revision. Confirm whether the test is performed fully mated, unmated with a manufacturer-approved cap, or in both states. A mated test may exercise the interface seal while hiding an unmated-service vulnerability; an unmated test with an improvised flat plate may load the seal unlike the intended cap.
Terminals, wire seals and cavity plugs
Identify terminal and seal pairings by part number. Wire outside diameter is a sealing input, not a cosmetic attribute. Two wires with the same conductor area can have different insulation diameter, ovality, surface texture or hardness. Request the manufacturer’s approved insulation range and compare the complete measured tolerance band, not only the nominal diameter.
Unused cavities need the specified seal plug. A similar-looking plug may have a different rib geometry or insertion depth. Record plug cavity locations in the pinout and inspect their final position. The connector terminal-retention guide explains why terminal lock and secondary-lock state should be verified separately; a terminal that is not fully seated can also disturb a rear sealing system.
Rear accessories and harness routing
Backshells, dress covers, conduits and tape can redirect cable forces. Freeze the first-support distance, bend direction and bundle diameter. If the qualification or leak method bends the wire bundle, document where the bend begins and which accessories are installed. A tightly taped test bundle can load an individual wire seal differently from a free bundle.
The official USCAR change letter for USCAR-2 Revision 2-9-1, Revision 1, checked on 2026-10-03, is valuable because it makes the configuration visible. Its pressure/vacuum procedure discusses introducing air through tubes or suitable actual wires, notes that tubing dimensions, wire flow and connector volume affect the time to reach the target, and adds wire-bundle bending instructions during the submerged observation. Those details belong to that specification and approval context; they are not universal pressure values for every connector. They demonstrate why an RFQ must capture the port path, bundle condition and monitoring point.
Draw the pneumatic boundary
Require a one-page schematic before accepting a test result. Shade the volume that is filled or evacuated. Label:
- supply and regulator;
- fill valve and isolation valve;
- transducer and reference volume, if used;
- calibrated leak connection;
- manifold, hoses and quick connectors;
- custom adapter or seal head;
- connector cavity or harness passage under test;
- blocked outlets, open outlets and vent paths; and
- exhaust or vacuum path.
The diagram should make it possible to ask, “If this O-ring at the fixture leaks, does the instrument report a connector failure?” It should also show whether air can escape along stranded conductors or between insulation layers. A result taken at the instrument inlet may include far more volume than the connector itself.
Run a fixture-only check with a verified blank. Then run a known-good assembly and a controlled leak challenge. Record each configuration. If the fixture’s baseline approaches the product limit, redesign the fixture or revise the measurement method; subtracting an unstable fixture leak from every unit is a weak control.
Understand the pressure-decay sequence
Pressure-decay systems normally use a sequence rather than one reading:
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Stage | Purpose | Evidence to retain | Frequent error |
|---|---|---|---|
| Fill | Bring the complete circuit toward the target pressure or vacuum | target, actual pressure, fill time, valve state | judging a restricted fill path as a leak |
| Equalize or settle | Allow pressure to distribute through cavities and ports | time and end-pressure band | assuming the connector is at instrument pressure |
| Stabilize | Reduce thermal and mechanical transient effects | part/air temperature and drift | shortening time until good parts merely appear to pass |
| Isolate | Separate the test volume from the source | valve timing and pressure at isolation | including regulator behavior in the decay |
| Measure | Record pressure change over a fixed interval | raw time series, interval and calculation | reporting only a rounded pass/fail flag |
| Exhaust | Return the part safely to ambient | exhaust time and interlock | disturbing the seal before the record is saved |
COSMO Instruments’ official explanation of the air leak tester pressure-decay principle, checked on 2026-10-03, separates charging, equalization, stabilization and detection. Cincinnati Test Systems similarly describes pressure-decay leak testing and emphasizes controlling ambient-temperature effects and minimizing circuit volume. These are instrument-maker explanations, not connector acceptance standards. They support the method questions a buyer should ask.
Temperature can imitate a leak
Compressed gas warms during filling and cools afterward. A warm part moved from a process station can cool during measurement. Flexible seals, tubing or housings can creep. Each effect can change pressure without a physical through-leak. Conversely, a rising part temperature can hide pressure loss.
Control part soak, ambient band, fill method, stabilization time and operator handling. Log temperature near the test volume when the required sensitivity makes it relevant. Challenge the cycle after planned line stops, at the start of shift and across expected room conditions. Do not establish a cycle using only repeated tests of one already-conditioned master; repetition can make stability look better than normal production.
Volume controls sensitivity
For a fixed leak and observation period, a larger effective volume produces a smaller pressure change. Hoses, adapters and unused manifold branches count. Minimize and document them. If the product variant changes cavity count or harness volume, show whether the existing recipe remains sensitive enough.
An approximate isothermal relationship can help purchasing review the order of magnitude:
volumetric loss at test conditions ≈ (effective volume / absolute test pressure) × (pressure change / time)
This simplified relationship is not a substitute for the instrument manufacturer’s calibrated algorithm, reference-condition conversion or leak-standard correlation. It assumes a rigid, fixed-volume system with stable temperature and small changes.
Worked example: why the fixture and temperature matter
Consider a fictional dry-air setup. The connector and adapter contain 35 cm³, while the manifold and hose add 50 cm³, so the effective isolated volume is 85 cm³. The initial absolute pressure is 150 kPa. After stabilization, pressure decreases by 0.80 kPa over 10 seconds.
Using the simplified screening relationship:
loss ≈ (85 cm³ / 150 kPa) × (0.80 kPa / 10 s)
loss ≈ 0.045 cm³/s ≈ 2.7 cm³/min at the test condition
This is an invented calculation, not an acceptance limit or measured product result. It shows three procurement points.
First, if the buyer ignored the 50 cm³ fixture volume and used only the 35 cm³ connector volume, the estimated result would be understated. Second, a temperature-driven pressure change can be interpreted as leakage. Third, a displayed leak-rate number is not reviewable unless the effective volume, absolute pressure, time basis and calibration model are known.
Now suppose an otherwise identical fixture uses a 20 cm³ hose/manifold volume instead of 50 cm³. The total volume becomes 55 cm³. For the same actual gas loss, the pressure response is larger and easier to resolve. Reducing test volume can improve separation, but the station still needs a leak-standard challenge and a capability study around the approved limit.
Correlate the station with controlled leaks
A calibration certificate for the pressure transducer does not prove the complete station can detect the required leak. The fixture, valves, volume, timing, software and part all participate.
Request a traceable leak standard or another approved master that introduces a known flow under defined conditions. Verify it at the station’s test pressure and connection point. Challenge at least near the decision boundary and, where the quality plan requires it, at more than one point. Document directionality: a standard calibrated for one gas, pressure differential or flow direction may need conversion or may not be suitable for another.
Distinguish these activities:
- instrument calibration confirms selected measurement channels against traceable references;
- station verification challenges the installed tester, plumbing and fixture;
- method validation shows the recipe distinguishes acceptable and unacceptable assemblies for the defined application;
- daily or shift checks demonstrate the station has not drifted; and
- product qualification evaluates the connector assembly under the required design and environmental plan.
None replaces the others. Define who owns each record, its interval and the reaction to failure. If a shift check fails, the control plan should bound product since the last known-good check and define containment.
Build a representative sample matrix
Select samples to exercise sealing risks, not merely convenient stock. Consider:
- minimum, nominal and maximum approved wire outside diameter;
- smallest and largest conductor construction that affect air flow through the wire;
- every approved wire-seal or mat-seal cavity range;
- fully populated, partly populated and plug-heavy cavity patterns;
- edge and corner cavities when geometry differs;
- maximum connector cavity count or internal volume;
- minimum and maximum harness branch load at the rear seal;
- straight and permitted worst-case exit directions;
- mated, capped or unmated service states;
- each mould cavity, seal lot or assembly line where risk requires it; and
- post-conditioning specimens from the same serialized set.
Do not test a hand-selected “golden” sample with polished wires and then transfer the result to normal stripping and insertion. Samples should be built with released production tools and settings. Record any intentional worst case and why it bounds the offered population.
The latest public USCAR material found during the 2026-10-03 source check, USCAR-2 Revision 2-9-2, includes separate connector-validation schedules and identifies pressure/vacuum, submersion, high-pressure spray, seal-retention and conditioning sequences for relevant sealed-connector changes. It also notes that changes in applied cable-seal design, cable or seal plugs matter to the schedule. Use the exact customer-approved edition and clauses; the public revision material is evidence that cable and seal changes cannot be treated as cosmetic.
Sequence leak checks around environmental exposure
A useful qualification record often includes an initial leak check, defined conditioning and a post-conditioning leak check on the same serialized assemblies. Conditioning may include temperature/humidity, high-temperature exposure, vibration, mechanical shock, fluid exposure, wire flexing, mating cycles or spray/submersion, depending on the governing requirement.
Write the order down. A passing post-test result can be misleading if a connector was dried, reseated or rebuilt after conditioning. State whether samples may be disconnected, cleaned, baked, vented or repaired. Photograph the as-conditioned state before disturbing it. Preserve failures for root-cause work.
Do not combine incompatible methods casually. Positive pressure can temporarily improve one seal contact while vacuum loads it in the opposite direction. Submersion with surfactant may leave residue. High-pressure spray introduces directional mechanical loading. The responsible engineer must decide which states and sequence represent the application.
Define acceptance without inventing a universal limit
The acceptance source hierarchy should be explicit:
- governing regulation or customer specification;
- applicable connector or equipment standard and exact edition;
- connector manufacturer’s product and application specification;
- project-specific drawing and validated test plan; and
- approved production-screen correlation.
State the limit in the instrument’s controlled quantity, such as pressure change over time or calibrated leak rate. Define units, reference conditions, rounding, measurement uncertainty and guard band. Specify whether every sample must pass or whether a stated statistical rule applies. Define gross-leak behavior: some large leaks prevent the part from reaching pressure rather than producing a normal decay trace.
“No bubbles” is not the same as “zero leakage,” and “0.00” on a display is limited by resolution and rounding. Avoid absolute claims. Report “below the approved detection or rejection threshold under the stated method” when that is what the evidence supports.
Compare supplier evidence packages
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Supplier return | Assembly identity | Test configuration | Measurement control | Procurement assessment |
|---|---|---|---|---|
| A | Connector family and “IP67” | “Air tested” | Pass certificate only | Insufficient: no wire, seal, boundary, method or limit |
| B | Housing and wire part numbers | Pressure, time and fixture photo | Instrument calibration | Partial: no effective volume, stabilization, controlled leak or post-conditioning link |
| C | Complete BOM, process and serialized samples | Boundary schematic, states, volume, recipe and conditioning sequence | Raw traces, leak-standard challenge, capability and change control | Reviewable: engineering can compare the exact offer against the project requirement |
Package C is not automatically acceptable. The method may still have the wrong pressure, sensitivity, sample coverage or acceptance source. It is simply the package that permits a defensible review.
Use an RFQ evidence matrix
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| RFQ field | Evidence to request | Release condition | Hold point |
|---|---|---|---|
| Connector identity | Housing, terminal, seal, plug, accessory and drawing revisions | Exact offered BOM is frozen | Family name only |
| Wire boundary | Manufacturer/article, conductor, insulation and OD tolerance | Full OD band fits approved seal scope | Nominal gauge only |
| Assembly process | Strip, crimp, seal loading, insertion, plug and lock instructions | Released tools/settings and inspection | Hand-built untraceable sample |
| Harness geometry | Bundle, bend, first support, backshell and orientation | Test represents allowed installation | Free straight wires only |
| Pneumatic circuit | Boundary schematic and effective volume | Product and fixture volumes identified | Instrument screen photo only |
| Recipe | Fill, settle, stabilize, measure and exhaust settings | Each timer and pressure has an approved basis | “Default program” |
| Temperature | Part/air soak, ambient band and drift | Thermal transient bounded | No temperature record |
| Calibration | Instrument channels and traceable records | In calibration for test date | Sticker without scope |
| Station challenge | Known leak/master at installed fixture | Detects boundary near the limit | Transducer calibration alone |
| Samples | Lots, cavity patterns, wire OD and orientations | Worst cases justified | Convenient golden sample |
| Conditioning | Exact sequence and same-sample traceability | Initial and final results correlate | Separate unidentified samples |
| Result | Raw trace, calculation, pass/fail and deviations | Every serialized sample reviewable | Rounded pass flag only |
| Production control | Check frequency, master storage and reaction plan | Drift and suspect product are contained | No failed-check response |
| Change control | Components, wire, tooling, fixture, recipe and software | Advance review before shipment | Silent substitution |
Follow a five-step procurement decision
1. Freeze the application and failure consequence
Define water, dust, condensation, fluid, wash, immersion, altitude and temperature exposure. Identify mated and unmated states, service access and harness routing. State what a leak can damage and whether the test is design qualification, production screening or both.
2. Freeze the exact connector and harness BOM
Approve housings, terminals, seals, plugs, wires, accessories and drawings. Confirm the wire OD tolerance and cavity map. Link every component to the assembly instruction and traceable lot.
3. Approve the qualification and correlation plan
Choose the governing standard and customer requirement. Define sample matrix, environmental sequence, leak method, limit and controlled-leak correlation. Resolve deviations before testing.
4. Validate the production station
Review volume, temperature, timing, fixture baseline, calibration, leak-standard challenge, repeatability and operator controls. Test known-good and deliberately challenged assemblies under controlled authorization. Do not create defects in saleable product.
5. Release routine control and change rules
Set station checks, sample or 100% screen rules, traceability, data retention, containment and requalification triggers. Audit the exact recipe and software revision used in production.
Control changes that can invalidate the leak evidence
Require advance notice for changes to:
- housing, seal, plug, terminal, cavity count, mould cavity or material;
- wire supplier, insulation compound, nominal size, OD tolerance, strand construction or lubricant;
- stripping blade, crimp tool, applicator, seal-loading tool or insertion fixture;
- cavity plug tool, backshell, tape, conduit, heat-shrink or potting;
- harness bundle, bend direction, first support or branch breakout;
- test port, adapter, hose, valve, manifold, reference volume or seal head;
- pressure/vacuum target, timers, temperature band, calculation or guard band;
- transducer, leak standard, calibration interval, software or recipe access control;
- environmental sequence, sample selection or acceptance source; and
- assembly or test location.
Classify the effect before deciding on document review, station study, correlation check, partial requalification or full requalification. A wire color change can still require review if it changes the controlled insulation compound or diameter. A shorter hose changes test volume even when the connector is unchanged.
Send a complete sealed-connector leak-test RFQ
Provide the application environment, connector states, circuit and cavity map, exact wire schedule, harness routing, support geometry, required ingress or customer standard, qualification sequence, production-screen objective, data-retention requirement and approval authority.
Ask bidders to return the exact BOM and revisions, manufacturer instructions, pressure-boundary drawing, recipe, effective volume, temperature controls, calibration and controlled-leak records, sample matrix, initial and post-conditioning traces, deviations, production checks and change-notification list. Request current commercial terms separately in writing; do not treat a technical report as a price, MOQ or lead-time commitment.
Send a sealed cable-harness connector leak-test RFQ
Buyer FAQ
Does an IP67 or IP68 connector rating eliminate the need for a harness leak test?
No. The rating has a defined product configuration and standard scope. The finished harness adds actual wires, seals, plugs, accessories, routing and assembly process. Determine whether qualification evidence covers that exact configuration and whether a correlated production screen is required.
Is pressure decay the same as an immersion test?
No. Pressure decay measures gas-pressure behavior in a defined pneumatic circuit. Immersion evaluates liquid ingress or escaping bubbles under its own conditions. They can complement each other, but one result should not be relabelled as the other.
Why must the fixture volume be included?
Because the pressure response depends on the complete isolated gas volume. Extra hose and manifold volume reduce the pressure change created by a given leak and can alter the displayed calculated rate.
Can a transducer calibration certificate prove the leak station is capable?
No. It verifies the covered instrument channel. The installed fixture, valves, plumbing, volume, timing, software and temperature behavior also affect the decision. Challenge the complete station with an approved controlled leak or master.
Why is stabilization time important?
Filling changes gas temperature, while parts, seals and tubes can expand or settle. Measuring too early can create false decay or hide a leak. Establish stabilization with representative parts and temperatures, then lock the approved recipe.
Should positive pressure and vacuum give the same answer?
Not necessarily. They load seals in different directions and can change interface contact. Use the method and direction required by the controlling specification and application; do not substitute one without approval.
Can air be introduced through the harness wires?
Some official procedures permit suitable actual wires or dedicated tubes under defined conditions, but conductor construction, length and tubing dimensions affect airflow and time to pressure. Document the route and confirm it does not compromise the part or method.
What wire dimensions matter to sealing?
Use the exact insulation outside-diameter tolerance, ovality or other manufacturer-controlled characteristics, not only AWG or mm² conductor area. Surface condition and insulation construction may also matter within the connector manufacturer’s approved range.
Is a display reading of zero proof of zero leakage?
No. Every method has resolution, uncertainty, rounding and a detection boundary. Report the approved threshold, method and actual trace rather than claiming absolute zero.
Should the same specimens be checked before and after conditioning?
Usually that provides the clearest trend when the approved plan allows it. Preserve serialized identity and do not reseat, clean or repair samples between stages unless the procedure explicitly requires and records that action.
How often should a production station be challenged?
Set the frequency through the quality plan using risk, stability, shift pattern and method capability. Define start-up, changeover, post-maintenance and failed-check reactions. There is no universal interval for every station.
What is the most important change-control trigger?
Any change that affects the sealed BOM, wire OD, assembly process, harness load, pneumatic volume, fixture, recipe or decision algorithm deserves review. The approval authority should determine the required level of revalidation before shipment.
Does this guide confirm that SINAWATTS can perform pressure-decay leak testing?
No. It is a procurement framework. Ask for current written capability, equipment, calibration, scope and project-specific evidence for the exact offered assembly.