A fuse that survives the first equipment startup has demonstrated one event under one set of conditions. It has not demonstrated the repeated starts, load changes, short shutdowns and automatic retries expected over the equipment's service life. Procurement needs evidence connecting the complete operating sequence to the exact fuse offered, including its installation and thermal starting state.
Direct answer: specify the current waveform through the fuse, pulse duration, interval, background current, event count and operating environment. Require the fuse manufacturer's applicable pulse or cyclic-load method, exact part-number curves and documented treatment of thermal recovery. Review single-event withstand, repeated-duty endurance and fault protection separately. A pulse below nominal melting I²t is not a cycle-life guarantee, and adding every pulse's I²t together does not calculate remaining fuse life.
This guide addresses repeatable normal-operation events in DC equipment. It complements the DC fuse coordination and I²t guide, which covers broader protection comparisons. Temperature correction and upstream selectivity remain separate assessments. The focus here is whether an offered fuse has evidence for the actual mission profile before a purchasing decision.
The calculations below organize requirements; they do not select a production fuse or prescribe a high-energy test. Engineering approval belongs to the responsible system designer using current manufacturer evidence. Nothing here establishes a SINAWATTS fuse rating, certification, test capability, manufacturing arrangement, stock level, price, minimum order quantity or delivery commitment.
Define the quantities before comparing curves
A useful RFQ gives each current and time quantity one meaning. “Surge current” can otherwise describe a measured fuse current, an equipment input specification or a semiconductor limit, even when those values answer different questions.
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Quantity | Meaning in the RFQ | Evidence boundary |
|---|---|---|
| Pulse I²t | Integral of squared fuse current over a defined event | Describes that waveform and integration window |
| Melting or pre-arcing I²t | Integral associated with melting the fuse element under specified conditions | Must distinguish nominal, minimum and applicable test conditions |
| Total-clearing I²t | Integral through both melting and arcing until interruption | Used in fault-protection assessment, not as a normal-duty pulse allowance |
| Repetition period | Start-to-start interval between events | Differs from the time spent at background current |
| Background current | Current remaining between the identified events | Determines whether an interval is unloaded or still thermally active |
| Cycle-life evidence | Manufacturer-supported endurance statement or report for defined duty | Requires product, waveform, installation and acceptance conditions |
I²t has units of ampere-squared seconds. It is an electrical integral, not energy in joules by itself. Resistive heating involves the integral of current squared multiplied by resistance; resistance and heat transfer matter when interpreting fuse temperature. Use the familiar I²t shorthand without pretending it captures the complete thermal history.
A non-repetitive surge limit means what its source defines. It must not be turned into a repetitive rating by choosing an arbitrary waiting time. Also check whose limit is being quoted: a protected semiconductor's non-repetitive surge rating does not establish the fuse's endurance.
Read official guidance within its product boundaries
The following four official manufacturer documents were checked on 10 October 2026. Their examples explain why an RFQ needs application-specific evidence. They do not qualify an unnamed replacement fuse.
Littelfuse's automotive Fuseology guide describes thermal cycling and possible mechanical fatigue from pulses. Its pulse-cycle discussion assumes sufficient cooling between events, and its illustrated endurance curve is for blade fuses. This supports asking for the relevant curve and recovery conditions. It does not authorize transferring a blade-fuse curve to a different high-power DC construction.
Eaton's Bussmann high-speed fuse application guide 10507, effective June 2016, separates basic current selection, overload influence and cyclic loading. The cyclic section considers both overall RMS loading and individual pulses relative to the melting curve, and states that general rules cannot cover every application. Use it to request the manufacturer's supported method for the offered high-speed fuse; do not copy its numerical factors into every battery-fuse specification.
Eaton's separate electronic fuse technology reference explains that pulse response depends on fuse construction and discusses pulse factors and matched comparison conditions. Its examples include small electronic fuse families. The procurement lesson is to identify the element construction and applicable documentation, not to claim that all fuses share the same pulse factor or degradation behavior.
Mersen's electronic fuse application information presents a pulse-count example for a radial-lead micro fuse and attaches a ten-second heat-dissipation condition to its pulse table. That condition belongs to the published example. It is not a universal recovery time for a bolt-down DC fuse, a loaded enclosure or a tightly spaced restart sequence.
These documents support a common purchasing question: what evidence applies to this part in this duty? A supplier may use a different valid method. Ask it to identify the method, its limits and the relationship to the submitted current history rather than requiring one attractive chart from a general handbook.
Build a mission profile that includes the inconvenient events
Start with current through the actual fuse location. Equipment output current or battery current measured elsewhere may differ because branches, auxiliary loads or alternative current paths exist. Identify the measurement point on a controlled system drawing and assign each current trace a file name, revision and operating condition.
Separate normal startup, routine load cycling, commanded shutdown and automatic restart. Include commissioning, service and diagnostic modes if they consume meaningful events. A controller that retries a failed start several times changes the sequence even if the daily count of successful starts remains unchanged. Record retry behavior as an operating requirement for review, not as a suggested control strategy.
Use a compact event register:
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Event field | Required description | Why purchasing needs it |
|---|---|---|
| Shape and duration | Measured trace or defined envelope, including tails | Peak current alone cannot describe pulse duty |
| Initial condition | Prior load, shutdown interval and local temperature | A cold start and a hot restart need separate treatment |
| Repetition | Minimum interval, burst structure and typical schedule | Daily averages can conceal clustered events |
| Background load | Current between pulses and between bursts | A pause need not be a cooling interval |
| Count | Events per operating mode and required service exposure | Supplier endurance evidence needs a comparable count |
| Configuration | Firmware, load option, fuse location and assembly revision | A changed operating sequence can invalidate the comparison |
Distinguish the representative operating profile from the acceptance envelope. A representative trace might show ordinary usage; an envelope should describe the combinations engineering requires the supplier to address. Do not silently combine unrelated maxima into an impossible waveform, and do not discard a credible severe combination because it is uncommon. Have engineering state which combinations are applicable.
Where current reverses, preserve that information in the trace even though squaring removes the sign from the I²t calculation. Current direction can matter elsewhere in the system, and the underlying record should not be reduced to a single unsigned number. Identify measurement uncertainty, sampling limitations and any unobserved operating modes.
Include a trace-quality review in the evidence handover: measurement location, sensor range, time resolution, integration boundaries and any clipping or missing samples. Preserve the raw trace alongside the processed integral. Ask the responsible engineer to explain whether the recording resolves the shortest relevant event and whether omitted tails materially change the result. This is a data-quality question, not an instruction to perform a hazardous live-circuit measurement or a universal bandwidth specification.
Translate service assumptions into counts with a transparent calculation. State whether the requirement is based on operating hours, calendar years, vehicle trips or machine cycles. Mark estimates as estimates. A bidder should be able to reproduce the count without guessing how many days per year the equipment runs.
Select the right evidence for the timescale
First establish the candidate fuse's identity and basic DC suitability. Record the manufacturer, full ordering code, current rating, voltage rating, interruption conditions and protection category. Pulse endurance is relevant only after the candidate fits the circuit's protection requirements. A larger fuse chosen to survive starts can change the protection available to conductors or equipment.
Then ask which manufacturer method covers the event duration and repetition. A short pulse assessed with an applicable I²t approach is not automatically equivalent to a long load step where heat transfer materially affects temperature. Request the permitted duration range and starting conditions. If the submitted duty crosses regimes, the supplier should show how it assessed each portion and their interaction.
For a time-current curve, identify whether it shows average melting, a tolerance band, maximum clearing or another defined quantity. Ask for the exact rating's curve and its revision. Reading an average curve as a guaranteed minimum melting boundary can create an unsupported conclusion before repetitive duty is even considered.
For a pulse-count curve, require the definitions of both axes. “Relative I²t” needs an identified denominator, including whether it is nominal, minimum or measured melting I²t and under which conditions. Ask whether the plotted count represents a design recommendation, observed results, a statistical bound or another basis. Do not select a convenient curve line without identifying its meaning.
For a cyclic-load method, request the calculation inputs, applicable correction factors and the source of each factor. The response should address the load between events and the individual pulse conditions. Multiplying several unrelated handbook factors is not a documented method. If supplier engineering introduces a project-specific adjustment, require its explanation and approval scope.
Finally, request a signed or otherwise controlled statement connecting the selected part and installation to the required profile. It should identify unresolved conditions explicitly. “Suitable for inrush” and “high I²t” are descriptions, not an endurance acceptance record.
Send your fuse pulse profile and evidence requirements with the proposed part number, waveform files, event-count basis and questions that remain open. Ask for a scoped quotation response identifying the available documents and any additional information needed.
Use arithmetic to expose missing evidence, not invent life
Consider an entirely hypothetical DC duty for comparing RFQ inputs. Assume the current through one fuse is exactly 120 A for 10 milliseconds and exactly 40 A for the rest of each period. There are no other transitions, tails or ripple in this simplified calculation. The installation, voltage, interruption duty and fuse selection remain unspecified.
Compare Profile A, with a 60-second start-to-start period, with Profile B, with a one-second period. Both profiles run for an assumed 10,000 operating hours. These values are illustrative requirements, not measured equipment data or recommended operating conditions.
For the defined 10-millisecond pulse window:
Pulse I²t = 120² × 0.010 = 144 A²s.
This uses total current through the fuse during that window. Subtracting the 40 A baseline contribution would give an incremental integral of (120² − 40²) × 0.010 = 128 A²s. Those are different quantities. Do not substitute the incremental value into a manufacturer's method unless that method explicitly requires it.
The full-period RMS calculation is the square root of the period's current-squared integral divided by the period:
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Calculation | Profile A | Profile B |
|---|---|---|
| Period | 60 seconds | 1 second |
| Time at 40 A | 59.990 seconds | 0.990 seconds |
| Period integral | 144 + 40² × 59.990 = 96,128 A²s | 144 + 40² × 0.990 = 1,728 A²s |
| Period RMS current | Approximately 40.03 A | Approximately 41.57 A |
| Events in 10,000 hours | 600,000 | 36,000,000 |
The pulse integral is identical, while event count and time between pulses differ substantially. RMS current helps describe the complete waveform but does not, on its own, establish the local element temperature or endurance. The 40 A background also means neither interval is an unloaded cooldown. The supplier must determine whether either profile meets the recovery assumptions of its method.
Suppose a fictional datasheet listed nominal melting I²t of 1,200 A²s. Dividing 144 by 1,200 gives 0.12, or 12%. This ratio alone approves neither profile: the applicable pulse-count relation, integral conditions, thermal starting state and statistical basis remain unknown. There is no universal acceptable percentage established by this example.
Likewise, multiplying 144 by 600,000 would produce an arithmetic sum, not a life-consumption model. Heat can leave the element between events, while construction-specific irreversible changes may follow a different relationship. A sum ignores those mechanisms and the sequence. Any cumulative-damage model used for procurement must have a documented manufacturer basis and a stated validity range.
The purchasing result is therefore a better evidence request: submit both profiles, their complete periods and counts, and ask the manufacturer to explain which is supported. The calculation deliberately ends before selecting an ampere rating or claiming years of life.
Make thermal recovery an explicit acceptance question
“Fully cooled” needs a definition tied to the method. It might refer to an initially unenergized condition, a specified background load or a verified repeatable starting temperature. A long-looking interval does not prove recovery when the fuse remains loaded or its surroundings continue to warm.
Ask the supplier to address cold startup, restart after a short interruption, clustered starts and operation after extended background load. Keep the conditions associated with each trace. A single ambient-temperature entry cannot describe all of these histories.
The DC fuse ambient-temperature correction guide covers local temperature and correction evidence. For repetitive duty, add the time sequence: when the environment reaches the specified condition, what current has already flowed, and what happens before the next event? This connects a static installation description to the mission profile.
Record the approved holder or mounting arrangement, conductor or busbar interface, enclosure condition and cooling arrangement. Ask which changes require reassessment. Procurement does not need to derive a thermal model, but it needs to prevent a test performed with one heat-transfer arrangement from being accepted for another without explanation.
A report should distinguish measured quantities from calculated internal temperatures. An external temperature measurement does not directly measure every part of the fuse element. Where a model supplies the missing relationship, ask for its application scope and validation basis, without demanding disclosure of proprietary construction details.
Review endurance reports and competing quotations consistently
Require a report index that connects the tested specimens to the proposed production part. Record sample count, lots if provided, waveform tolerances, event spacing, preconditioning, installation, achieved count, failures and acceptance criteria. A statement that samples remained continuous is useful but does not automatically cover every requested end-of-duty characteristic.
Have the responsible engineer define what evidence is needed after exposure. Depending on the application, that may include documented examination, electrical characteristics or subsequent protection behavior under an approved validation plan. The article does not prescribe destructive tests, apparatus or acceptance limits. These must come from the responsible laboratory, manufacturer and project requirements.
A test ending at the requested count without failures establishes the reported outcome for those samples and conditions. It does not reveal their eventual failure count, establish unlimited life or create a statistical reliability claim without the associated analysis. Ask how the supplier connects sample evidence to the purchasing requirement.
Compare bids using evidence status rather than an unsupported numerical ranking:
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Supplier response | Purchasing interpretation | Next action |
|---|---|---|
| Exact part, applicable method and matching profile | Potentially complete technical submission | Review conditions and approve through engineering |
| Exact part, but isolated pulse only | Single-event evidence available | Request repetitive-duty and recovery assessment |
| Related family with similar ampere rating | Applicability not established | Request written basis or exact-part evidence |
| Generic electronic-fuse pulse table | Product scope may not match | Require method for the offered DC construction |
| Claimed cycle life without report conditions | Claim cannot yet be compared | Request count, profile, installation and acceptance basis |
Keep the fault-protection review active. If the proposed solution changes the fuse rating, category or characteristic, revisit the protection study, including DC breaker-to-fuse selective coordination where relevant. Endurance acceptance cannot substitute for interruption or selectivity evidence.
Put the approved duty into the purchase record
Issue one controlled RFQ package containing the system requirement, event register, current files, calculation sheet, installation drawing and evidence checklist. Assign a revision to the package so a supplier cannot unknowingly quote against an obsolete restart schedule.
Ask each bidder to identify compliance, deviations and missing information by requirement. Separate the fuse price from any proposed evaluation, documentation or sample charges. Request availability, MOQ and lead time for the exact configuration and quantity, with the quotation's validity period. Those commercial facts require a current supplier response.
Before order release, connect engineering acceptance to the quotation line item. Preserve the approved manufacturer and ordering code, duty-profile revision, evidence references, installation conditions and permitted substitutions. A procurement note such as “equivalent fuse allowed” can undo the specific endurance assessment unless equivalence includes the same required evidence.
Control later changes to firmware retry rules, peak load, event interval, cooling, mounting and substitute parts. The relevant trigger is a change to the assessed duty or configuration, even when nominal operating current stays the same. Assign responsibility for deciding whether the existing approval remains applicable.
Buyer FAQ
1. Can two DC fuses with the same current rating have different pulse life?
Yes. A matching ampere rating does not establish matching construction, time-current response, pulse withstand or endurance. Compare complete part numbers and their applicable documentation. Ask both suppliers to address the same waveform, installation and count requirement. A dimensional cross-reference alone does not show that a replacement meets repetitive-duty requirements.
2. Is a pulse below the time-current curve automatically acceptable?
No. Identify what the curve represents, its tolerances and initial conditions. A comparison with average melting behavior does not by itself establish repeated-event survival. The supplier must apply a supported pulse or cyclic-load method, address thermal recovery and show that the offered part remains suitable for the required protection function.
3. Can I divide melting I²t by pulse I²t to calculate cycle life?
No. That quotient is a ratio between two electrical integrals, not a count of available operations. It omits the manufacturer's endurance relationship, pulse spacing and thermal history. Use a ratio only as the applicable method defines it. Do not interpret a fuse as a container with a fixed lifetime I²t allowance.
4. How much cooling time should I specify between pulses?
Specify the equipment's actual minimum interval and background load, then ask whether the offered fuse's method covers them. There is no universal cooling time established here. If a manufacturer curve assumes a particular interval or starting state, the submission must show that the application meets it or provide an approved alternative assessment.
5. Does passing a million-pulse test guarantee a million field operations?
Only the documented test outcome is established directly. Review waveform, thermal conditions, sample population and acceptance criteria against the field requirement. The meaning of any reliability claim needs its own statistical basis. A test under a convenient isolated-pulse sequence may not represent closely spaced retries or a continuously loaded installation.
6. Can an ambient-temperature correction replace the pulse review?
No. Temperature correction addresses the conditions covered by the particular current-rating method. Repetitive duty adds event sequence, preload, heat transfer and endurance questions. Keep the temperature calculation and pulse assessment connected to the same installation and part number, while preserving their separate assumptions and acceptance criteria.
7. Can I use an electronic-fuse pulse table for a high-power DC fuse?
Only with an explicit applicability basis from the responsible manufacturer. Small electronic-fuse examples do not establish a universal relationship for high-power DC constructions. Also distinguish a conventional fusible link from an electronic protection device marketed as an eFuse. Their operating principles and evidence are different; the shared word does not make their endurance data interchangeable.
8. Should I increase the fuse rating if repeated startup opens it?
Do not approve a substitution from that symptom alone. Request assessment of the actual duty, installation and protection requirements. A higher current rating may change conductor or equipment protection and coordination. A suitable resolution needs engineering acceptance of the complete system, followed by documentation of the approved part and any operating changes.
9. What quality or certification documents should a supplier provide?
Request traceable product identification, applicable declarations or certificates for the exact part, and the relevant endurance evidence. A certification claim must be checked for scope; it does not automatically certify the project's specific pulse history. Ask for the report reference, controlled revision and explanation of any difference between tested specimens and the quoted product.
10. Must we disclose confidential equipment details to obtain an assessment?
Provide the electrical and installation information needed for a valid assessment using controlled, appropriately shared documents. A simplified system drawing and an agreed waveform envelope may be sufficient if engineering confirms their completeness. Ask whether an NDA is available when necessary. Removing background load, retries or thermal conditions merely to simplify disclosure can invalidate the answer.
11. What determines price, MOQ and lead time for this inquiry?
The quoted part, quantity, documentation, samples and any proposed additional evaluation all belong in the commercial response. Do not infer availability or delivery from a general catalog. Ask the supplier to separate confirmed terms from estimates and to state what technical decisions remain open before the quotation can be finalized.
12. What should be ready before requesting a repetitive-duty quotation?
Prepare the complete fuse identity if one is proposed, current traces, event-count calculation, shortest intervals, background load and installation conditions. Include the DC protection requirements and a list of evidence needed for acceptance. If data is incomplete, identify the gaps clearly so the response can be scoped without disguising assumptions as approved requirements.
Send your DC fuse repetitive-duty RFQ with the controlled mission profile, candidate part number and required evidence list. Request a response that identifies applicable curves, remaining technical questions and current commercial terms for the exact proposed configuration.