Battery Disconnect Switch Contact Resistance, Voltage Drop and Temperature Rise: Continuous and Cranking Duty RFQ Evidence

Two battery disconnect switches can carry the same catalogue ampere label and produce different system voltage drop, heat and cold-start behaviour. The difference may come from internal contact geometry, contact pressure, stud material, plating, terminal hardware, cable size, lug preparation, mounting, enclosure temperature or the test method itself. A single milliohm number without measurement boundaries cannot separate those effects.

For procurement, the direct answer is: require the supplier to define where voltage is sensed, how current is applied, which exact switch and terminal configuration is tested, and how the result changes from initial condition through thermal and mechanical conditioning. Use low-current resistance, high-current voltage drop and temperature rise as related but distinct evidence. Do not convert one into a universal rating or acceptance limit.

This guide focuses on the resistive and thermal path of a manual or remotely operated low-voltage battery disconnect. It complements the broader battery disconnect switch specification guide, the battery switch cranking-duty guide and the battery cable voltage-drop guide. Those articles address topology, duty and system conductors. Here the purchasing question is narrower: what evidence shows that the offered switch path has acceptable loss and heat in the buyer's declared installation?

This article does not set a universal milliohm, millivolt or temperature-rise limit. It does not verify a SINAWATTS product, factory, laboratory, certification, current rating, material, inventory, price, MOQ, lead time or customer result. The responsible design and compliance parties must define acceptance for the exact system and market.

Direct answer: what should the RFQ require?

Require one controlled evidence package that identifies:

• switch manufacturer, full part number, suffix, revision, pole arrangement and actuator configuration;
• rated voltage and the separate continuous, intermittent, cranking and live-switching claims returned for that exact model;
• contact, stud and terminal materials or finishes only to the extent documented by the responsible manufacturer;
• cable conductor size and construction, cable length, lug part number, stud stack, washer/nut arrangement and tightening instruction;
• mounting orientation, cover, enclosure, nearby heat sources, ambient definition and airflow;
• four-wire contact-resistance method, instrument, range, test current, polarity, zero/offset treatment, sense points and stabilization time;
• high-current voltage-drop waveform, current waveform, sampling rate, sense points and duration;
• temperature sensors, locations, local ambient sensors, data interval, test endpoint and shutdown rules;
• initial values by serialized sample, not only an average;
• preconditioning, operating cycles, thermal cycles, vibration or environmental sequence where applicable;
• post-condition values, individual deltas and acceptance decision;
• calibration status, raw data, photographs, drawing/BOM traceability and deviations;
• production controls and sample frequency appropriate to project risk; and
• change-notification rules for every feature that can alter the current path or heat flow.

The buyer should be able to reproduce which resistance belongs to the closed switch alone, which voltage drop includes external joints, and which temperature belongs to the switch body, stud, lug, cable or ambient. If those boundaries are missing, apparently precise numbers are not comparable.

Separate four different electrical boundaries

“Switch resistance” can describe several measurement loops. Put the boundary on the RFQ drawing before requesting a value.

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

These boundaries answer different questions. A manufacturer development report may use internal points that purchasing cannot access. A finished-equipment test may intentionally include lugs and cables because it evaluates the installed system. Both can be valid if labelled correctly; they are not interchangeable.

For stud-to-stud or lug-to-lug tests, show the sense-probe locations in a photograph or drawing. Moving a sense point across a bolted interface adds or removes that interface from the result. Likewise, clamping a probe to a threaded end can produce a different reading from a defined smooth stud or lug-palm location.

Map the duty before comparing numbers

Contact loss creates voltage drop and heat whenever current flows, but the consequence depends on time. Separate at least four duties:

1. Continuous duty covers steady loads such as distribution, charging or an inverter operating for long periods. Thermal equilibrium, enclosure ambient and cable heat sinking matter.
2. Intermittent duty covers a defined current for a defined on-time, off-time and number of repetitions. The switch may not reach thermal equilibrium, but accumulated heating can still matter.
3. Cranking duty covers a high, changing starter current over seconds. Voltage available at the starter and peak/short-term heating matter; the waveform is not a flat current rectangle.
4. Making or breaking duty concerns closing or opening current. Arc energy, contact erosion and utilization category are separate from the ability to carry current while already closed.

The Blue Sea Systems 9003e page, checked on 2026-10-03, is a useful example of why these rows must remain separate. For that named model, the official page publishes a 350 A continuous rating, 600 A intermittent rating for five minutes, a 1,200 A 30-second starting rating, a 25 A switching rating, a 4/0 AWG cable basis, terminal size and terminal torque. Those values belong to 9003e and its stated conditions; they are not generic battery-switch limits and do not establish a resistance or temperature-rise limit for another model. Blue Sea Systems 9003e official product record.

Littelfuse's official battery-disconnect selection guide similarly tells readers to compare alternator output with continuous rating, starting demand with intermittent duty, and short-circuit current/time with the switch's stated capability. It also says cable gauge affects switch performance because larger cables extract heat, and that a rating is based on a minimum wire size. This supports an RFQ requirement to record the conductor and duty with every thermal result. It does not authorize transferring a Littelfuse value to another manufacturer's switch. Littelfuse battery disconnect switch selection guide.

If the offered device is claimed as a switch, disconnector or switch-disconnector under IEC 60947-3, obtain exact-model evidence and the applicable utilization and test conditions. The IEC catalogue states that IEC 60947-3:2020 with Amendment 1:2025 applies to specified switches, disconnectors, switch-disconnectors and fuse-combination units up to 1,000 V AC or 1,500 V DC; it also notes DC critical-load-current tests and power-loss measurement in the current edition. The public catalogue scope does not prove that an automotive, RV or marine rotary switch is within that standard or has passed it.

Use four-wire resistance measurement for a controlled baseline

Low resistances are easily obscured by test-lead and fixture resistance. A two-wire ohmmeter sends current and senses voltage through the same leads, so lead and connection resistance become part of the reading. A four-wire, or Kelvin, arrangement forces current through one pair and senses voltage through a separate high-impedance pair close to the defined device boundary.

The official IEC 60512-2-2:2003 record, checked on 2026-10-03, describes a specified-test-current method for contact resistance across mated contacts or a contact with a measuring gauge. IEC 60512 is written for connectors; its method does not automatically become the governing product test for a battery switch. It is still a useful reference vocabulary when a detail specification deliberately applies a comparable method and states its deviations.

NI's current DMM manual explains the measurement principle directly: force current through source terminals and measure voltage through separate high-impedance sense terminals so source-lead voltage drop is removed from the measured path. NI four-wire resistance measurement documentation. Instrument documentation supports the method principle, but the switch manufacturer or governing specification must still define test current, limits and sample conditioning.

An RFQ method sheet should freeze:

• serialized sample and switch state;
• ambient and sample stabilization time;
• operating history before measurement;
• exact force and sense points;
• current magnitude, duration and direction;
• current reversal or offset-compensation method, if used;
• instrument model, range, resolution, uncertainty and calibration status;
• fixture conductor size, connection pressure and cleaning rules;
• number and timing of readings;
• treatment of thermoelectric offset and self-heating;
• raw voltage and current as well as calculated resistance; and
• retest rules when a reading is unstable.

Do not use a test current high enough to change contact temperature materially unless that effect is part of the written method. Do not use a very low-current reading to claim high-current thermal performance. The low-current value is a traceable baseline and change indicator; the loaded voltage and thermal tests answer different questions.

Calculate voltage drop and power without inventing acceptance limits

For an approximately resistive closed path over a defined interval:

Voltage drop: Vdrop = I × R

and

Resistive power: P = I² × R = Vdrop × I

These equations help check units and compare the consequences of an agreed resistance. They do not predict the full transient thermal response, contact behaviour, current sharing or arc performance.

Worked calculation: a teaching example only

Assume a buyer's draft test plan uses a hypothetical stud-to-stud resistance of 0.25 mΩ. This is not a SINAWATTS value, a supplier result or a recommended limit.

At 200 A continuous test current:

• resistance = 0.25 mΩ = 0.00025 Ω;
• voltage drop = 200 A × 0.00025 Ω = 0.050 V, or 50 mV; and
• resistive power = 200² × 0.00025 = 10 W.

At an illustrative 900 A cranking point using the same resistance:

• voltage drop = 900 A × 0.00025 Ω = 0.225 V; and
• instantaneous resistive power = 900² × 0.00025 = 202.5 W.

If that power were constant for ten seconds, the simple electrical energy term would be 2,025 J. A real crank waveform changes with time, resistance changes with temperature and contact state, and heat spreads into studs, cables and housing. Therefore record synchronized current and voltage traces and integrate from measured data if energy is relevant. Do not use the teaching calculation as a pass/fail threshold.

This example also shows why quoting only “0.25 mΩ maximum” is incomplete. The buyer needs to know whether the value is initial or aged, at what temperature, at which boundary, at what current, and whether it is a maximum individual value or an average.

Design a high-current voltage-drop test around the waveform

A high-current test should measure voltage at the same defined sense points used in the requirement. Record current and voltage on a common time base. For a continuous test, the data should show warm-up and the stable or specified endpoint. For a crank test, capture the initial inrush, compression peaks, decay and recovery with enough bandwidth and sample rate for the waveform being evaluated.

Specify:

• power source or load type and its control mode;
• target current profile, tolerance and duration;
• starting temperature and permitted preheating;
• cables, lengths, lugs, hardware and torque;
• switch operating sequence and dwell after closure;
• oscilloscope, recorder, shunt or current-probe identity and uncertainty;
• differential voltage connection and common-mode capability;
• trigger point and sampling parameters;
• number of repetitions and rest period;
• raw file format and channel scaling; and
• abort limits for voltage, temperature, smoke, deformation or unstable connection.

Avoid subtracting two large node voltages with unrelated handheld meters to infer a small millivolt drop. Use an appropriate differential or isolated measurement arrangement designed and reviewed for the energy and common-mode environment. Test safety, conductor protection and laboratory competence remain the responsibility of the qualified test party.

For the rest of the path, use the four-wire battery-cable resistance guide to define cable-assembly measurements and the battery-cable lug geometry guide to freeze the mechanical interface. This prevents a cable or misfitted lug from being incorrectly attributed to the internal switch.

Make temperature-rise evidence reproducible

Temperature rise is the measured temperature at a defined point minus a defined local ambient or reference temperature. Absolute temperature and rise answer different questions. A low rise in a hot compartment can still produce a high absolute temperature; a high rise measured against the wrong “ambient” can be misleading.

The IEC 60512-5-2:2002 official record, checked on 2026-10-03, describes a current-temperature derating method for electromechanical components, essentially connectors, at elevated ambient temperature. The IEC 60512-9-2:2011 record describes an electrical-load and elevated-temperature endurance method when required by a detail specification. These catalogue descriptions reinforce the need to link current, ambient and component detail. They do not define a universal battery-switch setup or acceptance value.

Require the thermal report to show:

• chamber or room arrangement and how local ambient is measured;
• enclosure, cover, mounting panel and orientation;
• airflow and distance from other heat sources;
• cable conductor size, insulation, length and routing on both sides;
• lug identity, plating, stud stack and tightening method;
• current waveform and measurement uncertainty;
• sensor type, attachment, insulation and exact location;
• switch body, each stud, each lug and adjacent cable temperatures;
• data interval and plot from energization to endpoint;
• the criterion for equilibrium or the specified test duration;
• absolute temperatures and rises for every recorded location; and
• post-test inspection and resistance/voltage-drop comparison.

Cable length matters because conductors remove heat from terminals. A short, massive busbar and a long insulated cable can create different gradients. A cover can change convection. Mounting the switch to a metal panel can change heat spreading. Therefore reject a report that lists only current and a final surface temperature.

Infrared images can help locate hot regions, but emissivity, reflected temperature, viewing angle and hidden joints affect interpretation. Use contact sensors or another validated method at the controlled points required by the test plan. If thermal imaging is supplied, request the instrument setup and a visible-light image that identifies the viewed features.

Compare initial, conditioned and production evidence

An attractive initial reading does not show retention after operation or environment. Build a sequence that reflects project risk without inventing tests solely to create paperwork.

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

Use the same serialized units and the same measurement boundary before and after conditioning. Report every sample rather than hiding spread in an average. An increase can come from contact wear, contamination, relaxed pressure, plating change, loosened hardware or a test-fixture problem; the investigation should distinguish them before disposition.

Follow a practical procurement decision process

Step 1: define the circuit and operating sequence

Draw battery, alternator, charger, inverter, starter, protection, switch and downstream loads. State which currents pass through the switch, whether it is normally operated de-energized, and what abnormal operation must be considered.

Step 2: build the duty table

List maximum voltage, continuous current and duration, intermittent loads with on/off cycles, measured or justified crank waveform, expected number of operations, ambient range and installation zone. Do not select a test point from the switch headline alone.

Step 3: freeze the physical current path

Specify switch, cables, lugs, studs, washer/nut stack, torque, cover, mounting and enclosure. Include cable clearance and support. The battery-busbar terminal-stack guide provides a related evidence structure for bolted high-current joints.

Step 4: define measurement boundaries and methods

Put force and sense points on the drawing. Separate low-current baseline resistance, high-current voltage drop, continuous temperature rise and crank-transient capture. Assign acceptance and data-retention rules to each.

Step 5: request exact-model evidence

Ask for controlled drawings, current ratings with conditions, applicable reports, raw or summary data at the agreed level, certificate scope where required and a signed deviations list. Family brochures may support screening but do not close exact-model gaps.

Step 6: compare evidence before price

Normalize each bid into the same matrix. A supplier that states a different boundary or test condition has not necessarily failed, but the deviation must be reviewed. Do not rank a lower resistance number above a complete thermal report unless the data are actually comparable.

Step 7: approve samples against written criteria

Link samples to drawing/BOM revision and evidence. Record installation and test setup. Quarantine unexplained outliers rather than averaging them away.

Step 8: release with change control

Define production checks, lot traceability, retained records, nonconformance response and change notification. Requalification depth should follow the affected risk.

Worked bid comparison: evidence, not a customer case

The following is an invented purchasing exercise. It is not a customer project, supplier endorsement, laboratory result or SINAWATTS performance claim.

An OEM requests an ON–OFF switch for a system with a declared continuous load and a measured cold-crank waveform. The RFQ defines lug-palm sense points, a four-wire baseline method, a continuous loaded thermal setup and a synchronized cranking voltage-drop trace.

Bid A returns a catalogue current rating and “very low resistance.” It provides no boundary, test current, individual results, cable or ambient. Its price is lowest, but the resistive and thermal evidence is not comparable.

Bid B supplies a milliohm table measured stud-to-stud at room temperature on three samples. It identifies the instrument but not the terminal torque, cable arrangement, operating history or temperature-rise test. The values may support preliminary screening; they do not close the installed thermal requirement.

Bid C identifies the exact switch revision, stud and lug stack, cable, torque and mounting. It returns serialized initial four-wire results, continuous current/voltage/temperature plots, a crank waveform, post-cycle results, calibration references and deviations. Bid C gives the buyer the strongest evidence path, even if its nominal resistance number is not the smallest.

The decision is not “choose Bid C automatically.” Engineering still checks the specified limits, safety, compliance and system fit, while purchasing checks commercial terms. The exercise shows why a complete boundary and method are more valuable than an untraceable number.

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.

Plan first-article and lot controls

A first article should verify identity and the interfaces that make the development evidence relevant. Check full part number, markings, dimensions, terminal material/finish only where specified and verifiable, stud size, usable length, hardware, handle states, mounting, cable clearance and controlled assembly instructions.

For electrical checks, use the approved method and limits. A low-current four-wire check can be useful for trending when its fixture is capable and stable. It does not replace the high-current qualification. If production uses a voltage-drop test at a defined current, control fixture contacts, cable temperature, dwell and sense location. Record serialized or lot-linked results and apply a reaction plan for drift, instability or an outlier.

Incoming buyers should not loosen and retighten a critical power joint merely to “check torque” unless the approved procedure calls for it. Audit assembly records, witness the controlled tightening process where appropriate, or use a defined verification method that does not damage the joint. The battery terminal clamp torque and post-damage guide explains why torque values and interfaces must be kept distinct.

Control changes that affect loss or heat

Require advance notice and buyer review for changes to:

• contact alloy, geometry, surface finish, plating thickness or supplier;
• spring, cam, actuator travel, lubricant or contact-force setting;
• stud material, plating, insert method, thread, usable length or retention;
• nut, washer, cover or supplied terminal hardware;
• internal conductor, rivet, weld, braze or fastening process;
• housing resin, insert, seal or mounting feature that influences pressure or heat flow;
• cable, lug or terminal instructions used to substantiate ratings;
• assembly tooling, critical setup, inspection method or limits;
• manufacturing or test location;
• resistance, voltage-drop or thermal test fixture/software;
• rating, certificate or report scope; and
• drawing, BOM, label or packaging revision that can affect identity or handling.

The supplier should return a marked comparison, reason, affected lots, proposed validation and implementation date. A cosmetic change may be low risk; a contact-plating or spring change can affect resistance and endurance. The responsible team chooses proportionate revalidation rather than assuming every change needs the same test or that no test is needed.

Buyer FAQ

Is contact resistance the same as voltage drop?

No. Resistance is calculated from voltage and current under a defined method and boundary. Voltage drop is the measured result at a stated current and time. A resistance value can help compare baselines, while a loaded voltage waveform shows performance in a particular duty. Always state sense points and conditions.

Can a handheld two-wire meter verify a battery switch?

It may show gross continuity, but lead and probe resistance can dominate a milliohm-level path. Use an approved four-wire method and suitable instrument for quantitative low-resistance evidence. Test safety and instrument capability must match the circuit.

What is an acceptable millivolt drop?

There is no universal value for every switch and system. The responsible designer must allocate the system voltage-drop budget across cables, lugs, joints, protection and switch, then define a test at the required current, temperature and life state.

Can we calculate temperature rise from contact resistance alone?

No. The I²R calculation estimates electrical dissipation at a moment, but temperature also depends on contact distribution, materials, heat capacity, cables, mounting, enclosure, ambient, airflow and time. Require a representative thermal test or validated analysis as the project requires.

Should cranking voltage drop be measured at a constant current?

A constant-current point can support comparison, but a real starter current changes during the event. If crank performance matters, obtain synchronized current and voltage waveforms under a defined representative profile and starting condition.

Does a high continuous ampere rating prove low resistance?

No. A rating without its test conditions does not provide the measured resistance, voltage drop or installation temperature. Ask for the exact-model rating basis and the project-specific evidence rows separately.

Should the sense leads touch the studs or the installed lugs?

That depends on the requirement. Stud-to-stud emphasizes the switch path; lug-to-lug includes bolted interfaces. Mark the chosen points on the drawing and use the same boundary for comparisons and before/after measurements.

How many samples are required?

There is no universal sample count. Base it on governing requirements, design novelty, evidence maturity, process capability, consequence of failure and project risk. Record individual results and the rationale instead of quoting an unsupported sample-size rule.

Does a post-cycle resistance increase automatically mean contact failure?

It is a signal for investigation, not a diagnosis by itself. Check fixture stability, temperature, sense points, hardware, contamination and instrument offsets before examining contact wear or pressure. Apply the approved limit and disposition process.

Can supplier test data replace finished-system validation?

Supplier data can support component selection when the exact model and conditions match. The finished system can introduce different cables, lugs, enclosure temperature, installation and duty, so the responsible OEM must decide what system-level validation remains necessary.

What commercial details belong beside the evidence request?

State sample, pilot and production quantities; destination; packaging; records; requested quotation validity; and any customization. Ask the supplier to return price, MOQ and timing assumptions in writing. Do not infer availability or lead time from technical qualification data.

Prepare the inquiry package

Send one controlled package containing the circuit, duty table, switch and terminal drawing, environmental and installation description, measurement boundaries, qualification sequence, acceptance criteria, production controls, change rules and commercial quantities. Require every bidder to list exceptions rather than silently changing the setup.

Send your battery disconnect switch RFQ evidence package to SINAWATTS. Include the current waveform, cable and lug details, sense-point drawing, ambient/enclosure condition and required records. SINAWATTS must review current written evidence for the exact project; this article does not establish product performance, compliance, certification, price, MOQ, inventory or lead time.

Source and review record

The official IEC, Blue Sea Systems, Littelfuse and NI pages linked above were accessible and checked on 2026-10-03. IEC catalogue pages provide titles, scopes and lifecycle information, not the licensed procedures or proof that any offered switch complies. Manufacturer values are quoted only for the named products and documents. Recheck all editions, URLs and exact-model documents on the actual sourcing date.

Ten-step review: 97/100 — topic positioning 10, content quality/E-E-A-T 10, search intent 10, high-conversion long-tail coverage 9, AI readability 10, structure/technical SEO 10, conversion path 10, evergreen reuse 9, authoritative sources 10, user experience/readability 9. The remaining deductions reflect that commercial terms and product-specific acceptance limits must stay project-defined and that licensed standard text is not reproduced.