The short answer
Start with a complete distribution drawing and the verified current, cable, connector and fault limitations of every branch. A competent designer must then select a protective device and holder with the correct DC voltage rating, interrupting capacity, time-current behaviour and environmental rating for the prospective fault current and the equipment being protected. The power supply’s own protection does not prove that every smaller outgoing cable or branch is adequately protected.
1. Draw the source, distribution point and every branch
Record the exact power-supply model and suffix, output voltage, rated output current and the manufacturer’s overload and short-circuit behaviour. Add every distribution terminal, controller, dimmer, connector, cable and LED load. Give each outgoing circuit a stable branch ID that also appears on the physical wiring.
Show where conductors split and where their size, insulation, installation method or connector rating changes. A single 24V output can feed several parallel branches, but that does not make the branches electrically identical. Route length, load, conductor, connector and temperature can differ.
Include every possible source of energy. A second power supply, battery, charged capacitor, controller output or incorrectly linked common can change the fault path. Do not assume that the power supply’s rated load current is automatically the maximum current that a protective device may need to interrupt.
2. Define what the branch device is intended to protect
Write a precise protection objective before selecting a part: for example, to limit damage to a 0.75mm² outgoing cable after the distribution point, or to isolate one removable sign section after a fault. A device selected for one objective may not protect a smaller downstream lead, connector or controller channel.
Identify the weakest verified current-carrying component in the complete branch and the conditions behind its rating. Check conductor material and area, insulation temperature, grouping, enclosure temperature, terminal range, connector current and the LED manufacturer’s connection limits. A larger upstream supply does not increase any of those ratings.
Separate normal operating current from abnormal current. LED supplies and controllers can have start-up, charging or control behaviour that affects nuisance operation, while a high-resistance joint may overheat without producing the large current expected from a short circuit. Protection design and connection-quality control therefore solve different problems.
3. Verify the exact device for DC voltage and fault current
Use the manufacturer’s data for the complete part number, not the family name or an AC marking. E-T-A explains that DC arcs are harder to extinguish because current does not pass through a natural zero in the same way as AC; DC voltage ratings for fuses and breakers are therefore often lower than their AC ratings.
Littelfuse defines breaking or interrupting capacity as the maximum approved current a fuse can safely interrupt at its rated voltage. Confirm both values for the proposed device and confirm the holder, terminals and any disconnecting function separately. A 32V DC automotive fuse, a mains MCB and a photovoltaic fuse are not interchangeable categories.
Obtain a defensible prospective fault-current value for the actual source and wiring. That assessment may need the power-supply manufacturer’s data, source impedance, parallel sources and stored energy. If the value or the device’s DC interrupting rating is unknown, the selection is not verified.
4. Complete the branch-protection design-input table
Use one row per branch and attach the supporting datasheet revision. “Same as branch A” is acceptable only when the cable, route, load and connection hardware really are the same.
| Design input | Record for each branch | Why it matters | Evidence required |
|---|---|---|---|
| Source and fault path | Exact supply, output, other sources and common connections | Defines available energy and possible backfeed | Manufacturer data and system drawing |
| Normal load | Steady current, start-up behaviour and controller state | Separates expected operation from overload | Load schedule and measured commissioning values |
| Cable and route | Conductor material/area, length, insulation, grouping and ambient | Identifies the circuit that needs protection | Cable data and installation method |
| Weakest component | Terminal, connector, controller channel or smaller lead | The branch is not stronger than its lowest verified rating | Exact part-number specification |
| Protective device | Part number, current, DC voltage, interrupting rating and time-current data | Establishes suitability for the DC fault and load profile | Manufacturer datasheet and curve |
| Holder and enclosure | Holder rating, terminals, access, temperature and IP conditions | A suitable fuse in an unsuitable holder is not a verified assembly | Holder/enclosure instructions |
| Coordination result | What opens first for the assessed faults and what remains energised | Tests whether one fault is actually limited to the intended section | Designer’s calculation and test plan |
5. Worked example: branch arithmetic is not a fuse-size answer
Consider an illustrative 24V sign with three measured steady branch currents of 3.0A, 2.0A and 1.5A. Their total is 6.5A, or 156W at 24V. That is useful for checking the supply and distribution schedule, but it does not select 4A, 3A or 2A protective devices.
For the 3.0A branch, the designer still needs its start-up profile, cable and connector ratings, enclosure temperature, the source’s prospective fault current, the proposed device’s DC voltage and interrupting ratings, and its time-current curve. A device that carries 3.0A in normal service may still fail the cable-protection objective, operate during normal start-up or be unable to interrupt the assessed fault safely.
The correct result of this example is therefore a completed evidence set, not a universal multiplier. If any required rating is unavailable, label the branch design unverified and obtain the exact manufacturer data or competent electrical design before construction.
6. Check selectivity, isolation and replacement control
A branch device can only limit the effect of a fault if the source and upstream protection allow it to operate as intended. Review the relevant time-current information and the power supply’s current-limiting, hiccup, latch-off or shutdown behaviour. Do not promise discrimination from current labels alone.
Document what remains live after one branch opens and how the affected section is isolated for service. A dim or dark branch is not proof of safe isolation. Mains and DC work must follow a safe system of work and be carried out by a competent person where required.
Label the branch ID and the exact replacement device next to the holder and in the drawing. E-T-A warns that an unspecified replacement can remove the intended protection; record technology, rating, speed or curve and approved part reference rather than only an amp number.
7. Commission and retain the protection record
Before energising, inspect polarity, conductor preparation, terminal occupancy, holder installation, branch labels and enclosure conditions. Confirm that the fitted part numbers match the approved schedule. HSE advises checking equipment is suitable for the work and supply, and stopping use where damage, burn marks or overheating raise a safety concern.
Under the competent person’s test plan, record input conditions, steady current per branch, start-up behaviour and the functional isolation result. Do not create a short circuit as an improvised site test. Any fault-injection or trip verification must use suitable equipment and a controlled method defined by the designer.
Keep the final drawing, calculations, datasheets, device schedule and commissioning results with the sign. After a trip, investigate the cause before reset or replacement and confirm that the replacement is the documented type.
- Exact supply, controller, protective device and holder part numbers recorded.
- DC voltage and interrupting ratings verified against the assessed fault.
- Cable, connector and terminal limits checked for every branch.
- Labels, isolation method and approved replacement reference present.
- Commissioning values and any deviations signed off by the responsible competent person.
Frequently asked questions
Can I choose a DC branch fuse by multiplying the normal LED current?
No universal multiplier proves suitability. The design also needs the cable and component limits, normal start-up behaviour, prospective fault current, DC voltage and interrupting ratings, time-current behaviour and environmental conditions.
Does a power supply with short-circuit protection protect every outgoing LED branch?
Not automatically. Its protection may limit or restart the whole output without protecting a smaller branch cable, connector or controller channel to the required objective. Review the exact supply behaviour and branch design together.
Can an AC-rated breaker be used on a 12V or 24V DC LED circuit?
Only when the exact device manufacturer provides a suitable DC rating and the remaining selection checks pass. A low circuit voltage alone does not establish DC interrupting or switching capability.
What should be labelled at a branch protection point?
Record the branch ID, circuit voltage, approved device technology and full replacement reference, plus the drawing or schedule location. Do not label only an amp value.
Technical references
Use the current datasheet for the exact product you install. These manufacturer and standards resources explain the general principles.
Review the branch-protection evidence before construction
Send the distribution drawing, exact supply, branch loads, cables, connectors and proposed protective-device data so Glowfy can help organise the documented checks.