An auxiliary light circuit that blows its fuse the instant the switch closes has a short circuit, and fitting a larger fuse converts a cheap fault into a harness fire. The safe order is to de-energise the circuit, disconnect the lamp, measure resistance through each section of the loom, then reconnect one section at a time until the fault returns. No step in that sequence requires the circuit to be live.
In most cases of an LED light bar blowing fuses, the damage is already visible once the loom is exposed: a cable crushed behind a bumper, a connector full of green copper, a housing that has been holding water, or a relay coil that has failed short. The fuse is not the fault. The fuse is the only part of the circuit that did its job.
Three numbers decide whether a repair holds: the continuous current the lamp draws, the current rating of the smallest cable in the run, and the fuse rating. They must be in that order, and the fuse must sit between the load and the cable. When those three numbers are wrong, the circuit keeps blowing fuses no matter how many lamps are fitted.
A fuse and a relay respond to completely different things, and confusing them sends diagnosis in the wrong direction on the first attempt. A fuse is a sacrificial element in series with the conductor. It opens because current through it exceeded its rating for long enough to melt the element, and once open it stays open. A relay is an electrically operated switch. It has no current sensing at all.
That distinction matters because “the relay is tripping” is almost never overcurrent protection. A relay contact opens when the coil stops being energised, or when the contact itself can no longer carry the load. When a driver reports an LED light bar blowing fuses and a clicking relay, there are usually two separate events happening, and only one of them involves the fuse.
The 12 Volt Planet relay guide describes the internal arrangement clearly: a coil around an iron core pulls an armature onto a contact, and the coil circuit is electrically isolated from the switched circuit inside the relay. Terminal numbers follow DIN 72552, with 85 and 86 on the coil, 30 as the common terminal, 87 normally open and 87a normally closed. A mini relay coil typically draws 150–200 mA at a coil resistance of roughly 60–80 Ω, while the high-current contact is normally rated between 25 A and 40 A.
| Device | What opens the circuit | Does it reset | What its operation tells you |
|---|---|---|---|
| Blade fuse | Melting of the element from sustained overcurrent | No, one-shot | Current exceeded rating: there is a fault or an overload |
| Thermal circuit breaker | Bimetallic element heating and bending | Yes, self-resetting or manual | Same overcurrent event, but it can re-close into the fault |
| Relay contact | Coil de-energised, or contact damaged | Yes, every time the coil changes state | Nothing about current unless the contact is burnt |
| Relay coil | Overheat or shorted turns | No | A shorted coil can blow its own control-side fuse |
A self-resetting breaker is the device most often misnamed as a tripping relay, because it clicks, cools and re-closes on its own. 12 Volt Planet notes that most breakers in 12 V and 24 V systems operate thermally, and can take longer to operate than a fuse. If a circuit re-arms itself after a few seconds and clicks off again, the component is a breaker and the fault has not gone anywhere.
A fuse protects the cable, not the lamp. The fusing guide published by 12 Volt Planet states the purpose directly: fuses protect a circuit’s cabling from excessive current that could lead to damage and, very often, an electrical fire. A lamp with an internal fault is not protected in any meaningful sense, because by the time the fuse acts, the lamp is already the cause.
The same guide lists three causes of excessive current: incorrect wiring from human error, damage to the circuit such as a wire working loose or insulation wearing through so that a positive cable shorts to ground, and overloading through equipment that draws more than the circuit was designed for, or cable that is too small for the intended load. All three appear in light bar installations, and all three are installation faults rather than lamp faults.
Position matters as much as rating. A fuse only sees the current that flows through it, so it must sit upstream of the switch and upstream of the point where the short occurs. The guide makes this explicit: if a section of cable shorts to ground, it is only protected if there is a fuse before the shorting point, because otherwise the fuse never experiences the excess current. It also flags the one length of cable that is effectively unprotected in almost every vehicle, the run from the battery positive terminal to the first fuse or distribution block, and recommends keeping that run as short as possible.
That is why a badly built circuit can blow its fuse and still burn. If the fuse holder was spliced in two metres down the loom, those two metres are unprotected, and a short there heats the cable without ever opening the fuse.
Four physical paths account for nearly every short in an auxiliary lighting circuit, and each leaves different evidence. Knowing which one you are chasing shortens the job considerably.
The first is a positive conductor touching the chassis or any earthed metal. This is the classic dead short and the one that blows a fuse fastest. The second is a short inside the lamp: a pinched internal lead, a failed driver component, water bridging a potted board, or a damaged grommet where the cable enters the housing. The third is the loom itself, worn through where it passes a metal edge, rubs on a bracket, or is pinched by a body panel after a repair. The fourth is the connector, where water and salt create a conductive path across terminals that should be isolated, and where corrosion raises resistance until the joint heats.
| Path | Where it hides | Typical evidence | Fastest check |
|---|---|---|---|
| Positive to ground | Loom behind bumper, under tray, at a body seam | Shiny metal, flattened insulation, scorch mark | Resistance from fuse load side to chassis |
| Inside the lamp | Driver board, entry gland, internal lead | Water, burnt smell, cracked potting | Resistance across the lamp input |
| Chafed loom | Bracket edges, firewall grommet, suspension travel | Rubbed sheath, copper visible, intermittent on bumps | Wiggle test while watching the meter |
| Connector ingress | Low-mounted plugs, splash zones, unsealed crimps | Green or black copper, tide marks, swollen seal | Visual inspection, then pin-to-pin resistance |
Water deserves its own mention because it is the slowest and most deceptive of the four. Water reaches a joint and does not fail immediately. It wicks along the strands by capillary action, oxidises the copper, and raises the resistance of the joint. The 12 Volt Planet cable guide describes exactly this: moisture wicking up the strands, copper oxide appearing as brown or black discolouration, verdigris appearing greenish-blue where the oxide reacts with acid, and contact performance degrading until a significant length of cable has to be cut away to find shiny copper again.
The arithmetic of a short explains why some faults blow a fuse instantly and others never blow it at all. Ohm’s law gives the current for a given fault resistance, and the range is enormous.
These are illustrative calculations, not measurements. They ignore the battery’s internal resistance and the cable’s own resistance, so a real circuit reads lower than the table suggests.
| Fault resistance | Calculated current at 12 V | What a 15 A fuse does | What it means in the bay |
|---|---|---|---|
| 0.05 Ω | 240 A | Opens almost instantly | Hard short, often with a visible arc mark |
| 0.1 Ω | 120 A | Opens instantly | Classic positive-to-ground short |
| 0.5 Ω | 24 A | Opens within about a second | Wet connector or partly chafed cable |
| 1.0 Ω | 12 A | May never open | The dangerous case: heat without protection |
| 5.0 Ω | 2.4 A | Never opens | Corrosion or a partial path; shows as heat |
The 1 Ω row is the one that catches people. A fault at that resistance draws less current than the fuse rating, so the fuse holds, the cable warms, and nothing obvious happens until insulation softens or a connector discolours. A fuse is a current-responsive device with a time curve, not a thermostat and not a fire detector.
12 Volt Planet’s guide gives the time dimension in plain terms: a 10 A fuse exposed to 11 A might take many minutes to blow, while the same fuse exposed to 20 A may blow in a fraction of a second. Manufacturers publish this as a current-time chart. So a fuse that blows instantly points to a low-resistance short, and one that blows after minutes points to an overload or a fault that heats as it runs.
The safety rules for this job are short and non-negotiable, and they apply even though the system is nominally 12 V. Anyone chasing an LED light bar blowing fuses repeatedly is already working around a live battery and a fault that has demonstrated it can pass current through places it should not. HSE’s guidance on electrical injuries notes that a voltage as low as 50 V applied across the body can disturb the signals between brain and muscles, and adds the sentence that matters most here: even low voltage batteries such as those in motor vehicles can get hot and may explode if they are shorted out.
So the first rule is isolation, and isolation means more than switching off. HSE’s guidance for work on electrical equipment requires that equipment be prepared by isolating and releasing all sources of energy, and that the isolation be secure, meaning energy cannot be inadvertently re-introduced. On a vehicle that means disconnecting the battery negative lead and making sure the terminal cannot swing back into contact. Tape over the terminal, or tuck the lead where it cannot move.
The second rule is that nothing gets plugged or unplugged while the circuit is energised. A live positive blade terminal in a plastic connector is a small exposed conductor with a whole battery behind it, and LED drivers with capacitive inputs can draw a substantial inrush at the moment of connection. Make the connection, then energise, then observe. If you must move a probe, disconnect first.
The third rule is eye protection and no jewellery. A shorted vehicle battery can deliver enough current to melt a ring or a watch strap in contact with a terminal, and HSE notes that even a low voltage battery can produce a spark strong enough to ignite a flammable atmosphere. Work away from fuel vapour, and do not lean over a battery while making or breaking connections.
| Prerequisite | Why it matters | How to confirm it |
|---|---|---|
| Battery negative disconnected | Removes the return path so no circuit can complete | Lead physically off the post and taped or stowed |
| Circuit proven dead | Confirms no backfeed from a second battery or accessory block | Meter reads no voltage at the load side of the fuse holder |
| Eye protection fitted | Arc, molten metal, and battery electrolyte | Worn before the first probe, not after the first spark |
| Rings and metal watch removed | A short across jewellery causes deep burns | Pockets empty before work starts |
| No fuel vapour or solvent present | A low voltage spark can ignite an explosive atmosphere | Ventilated area, no spilled fuel, no aerosol |
| Second person aware | Somebody knows the battery is disconnected | Tell someone before you go under the vehicle |
One more boundary belongs here. Do not probe wiring you cannot identify. Modern vehicles route airbag, restraint and high-current body circuits through the same harnesses as lighting, and a probe in the wrong cavity can trigger a fault that costs far more than the original repair.
A digital multimeter with a resistance range and a continuity mode covers almost everything in this article. Add a fused test lead, insulation tape, the correct crimp tool and terminals, and a good light. That is the whole kit. If you are still choosing the parts of the circuit in the first place, the LED light bar wiring kit buying guide covers relay, fuse and switch selection in more depth.
The multimeter needs one property that is often overlooked: know the resistance at which its continuity mode beeps. Many meters beep below a threshold in the tens of ohms, which means a beep can indicate a joint that is far too resistive to carry 20 A. A beep is a convenience, not a verdict. Read the number.
Two tools create more faults than they find. A jumper wire or a paperclip used to bypass a fuse removes the only protection in the circuit. A probe pushed into a sealed connector from the front spreads the terminal, reduces contact pressure and creates the high-resistance joint that fails six months later. Back-probe at the terminal, or use the correct extraction tool.
An ohmmeter must only be used on a dead circuit; measuring live produces meaningless numbers at best, and at worst damages the meter.
The conditions under which the fuse fails narrow the search more than any single measurement, and they cost nothing to collect. Write them down before the first fuse is replaced, because memory is unreliable and the pattern is the diagnostic.
Record when it fails: on the first switch-on, after a few minutes of running, only when the engine is running, only after rain or a wash, only over bumps, or only when a second accessory is switched on. Record which fuse failed and its rating, and what else shares that fuse. Record how many times it has been replaced and whether the replacement was the same rating. Photograph the fuse box and the lamp connectors before you touch anything.
The pattern is genuinely diagnostic. A fuse that fails on switch-on points to a hard short downstream of the switch. A fuse that fails after minutes points to a resistance that rises with heat, often a connector or a partly chafed cable. A fuse that fails only in the wet points to ingress. A fuse that fails over bumps points to a mechanical fault, and the meter will read perfectly until the vehicle moves.
Photographs also protect you: if a connector is already discoloured, you need to know the condition existed before you worked on it. The companion piece on a light bar that is not working at all covers the no-fuse-blown version of this starting point.
The single most valuable measurement in this whole procedure is also the simplest: disconnect the lamp and see whether the fuse still blows. That one test divides the circuit into two halves and eliminates half the possible causes in under a minute.
Disconnect the lamp at its own connector, not at the battery. Fit a fuse of the correct rating. Reconnect the battery, switch the circuit on, and observe. If the fuse holds with the lamp disconnected, the fault is in the lamp, its pigtail or its internal wiring. If the fuse still blows with the lamp disconnected, the fault is in the supply side: the loom, the relay, the switch or the feed itself.

Do not skip obvious checks with the lamp off. Look into the lamp connector for water, corrosion or a bent terminal, at the pigtail where it enters the housing for a damaged gland, and at the harness end for a partly withdrawn seal. Many faults are visible here and need no meter.
With the battery negative disconnected and the lamp still unplugged, resistance measurements become meaningful. The meter now compares the circuit against a known open, and any low reading is a path that should not exist.
Put one probe on the load side of the fuse holder, the terminal that feeds the lamp rather than the terminal that comes from the battery, and the other probe on a clean chassis point. With the lamp disconnected and the switch off, a healthy circuit reads open or a very high resistance. A reading of a few ohms, or a continuity beep, means the positive side is connected to ground somewhere downstream.
Then measure across the lamp’s own input terminals, positive to negative, with the lamp disconnected from the vehicle. A reading close to zero indicates a short inside the lamp. A very high or open reading on an LED lamp is not automatically a fault, because the driver input is not a simple resistive load, but a dead short is unambiguous.

| Reading | Where you see it | What it means | Next action |
|---|---|---|---|
| Open or overload | Fuse load side to chassis | No path to ground, which is correct | Move to the loaded voltage test |
| Under about 1 Ω | Fuse load side to chassis | Hard short downstream | Segment the loom, Step 4 |
| A few ohms | Fuse load side to chassis | Partial short, often water or chafe | Inspect connectors and loom routing |
| Near 0 Ω across lamp input | Lamp terminals | Internal short in lamp or driver | Replace or return the lamp |
| Reading changes when you wiggle the loom | Any segment | Mechanical fault, intermittent contact | Repair that segment and re-test |
| Reading drifts while you watch | Corroded joint | Thermal or chemical instability | Cut back to bright copper and re-terminate |
A loom is not one circuit, it is a chain of segments, and the fault sits in exactly one of them. Dividing the chain turns a hunt into a process of elimination. The natural dividing points are the ones that already exist physically: the fuse holder, the relay socket, the switch, any inline connector, and the lamp plug.
Work from the supply end. Measure from the fuse holder’s load terminal to chassis, then move to the relay socket terminal 87 with the relay removed, then to the inline connector, then to the lamp plug. At each point you are asking the same question: is this point still connected to ground through a low resistance when it should be open? The first point where the reading goes low tells you which segment contains the fault.
The segments worth separating on a typical installation are the feed from the battery to the fuse, the fuse to the relay contact, the relay contact to the lamp, the coil feed and coil ground, and the lamp’s own pigtail. Do not forget the coil side. A shorted relay coil is a genuine short circuit, it lives on a different fuse in many installations, and it behaves nothing like a lamp fault.
A wiggle test belongs in this step. With the meter connected across the segment, move the loom by hand along its length, flex it at each mount, and press on each connector. A reading that jumps while the loom moves has found the fault, even if it reads perfectly when still. Do this with the battery disconnected and the lamp unplugged, so a sudden short cannot damage anything.
Reassembly is the proof. Reconnect in reverse order, energising with a correctly rated fuse at each stage, until the fault returns. The segment you just added is the segment with the fault, and you now know something a meter alone could not tell you.
Start with the lamp disconnected and the supply side fully assembled. Fit the correct fuse and switch on. If the fuse holds, the supply side is sound and the fault is in the lamp or its pigtail. If the fuse blows, remove the load-side connection at the relay and repeat. Working backwards through the joints one at a time is slower than guessing, and far faster than replacing parts at random.
Two rules make it safe. Use only a fuse of the correct rating for every test, and never bridge the holder with wire to see what happens. If a step blows a fuse, stop and go back to Step 4 for that segment rather than continuing, because continuing means repeatedly energising a known short.
Fuse selection is arithmetic, not opinion, and there are only three numbers involved. The continuous current the load draws, the current rating of the smallest cable in the circuit, and the fuse rating, which must sit between them.
12 Volt Planet’s fusing guide states the rule and gives the worked example: the fuse rating should lie between the current rating of the smallest cable in the circuit and the normal expected current draw, so that normal operation is possible but an overload opens the fuse. Its example is a 10 A normal draw on 25 A cable, where a 15 A fuse is appropriate. The same guide advises against running a fuse above 75% of its continuous rating, to leave headroom for momentary surges that would otherwise fatigue or nuisance-blow the element.
That 75% figure is why an LED lamp needs headroom rather than a fuse matched exactly to its running current. A lamp that draws 14 A continuously should not be on a 15 A fuse, because it is at 93% of rating before any surge. A 20 A fuse on 25 A cable gives the lamp room and still protects the cable.
The fuses FAQ on the same site puts the same rule another way: fuse at around 70–80% of the cable’s rating, and place the fuse as close to the power source as possible. Both formulations describe the same band on a number line, with the load at the low end and the cable’s capacity at the high end.
| Family | Common names | Typical rating range | Typical use on a light bar circuit |
|---|---|---|---|
| Standard blade | APR, ATC, ATO, ATS | 1 A to 40 A | Main and branch fusing in a fuse block |
| Mini blade | APM, ATM | 2 A to 30 A | Tight fuse blocks, factory auxiliary circuits |
| Low-profile mini | APS, ATT | 2 A to 30 A | Compact aftermarket holders |
| Maxi blade | APX | 20 A to 120 A | Battery-side feed to a distribution block |
The families are not interchangeable, and the physical sizes explain why. As documented in the Wikipedia article on automotive fuses, the regular blade fuse measures about 19.1 × 5.1 × 18.5 mm, the mini about 10.9 × 3.6 × 16.3 mm and the Maxi about 29.2 × 8.5 × 34.3 mm. The same article notes that blade fuses are generally rated for circuits no higher than 32 V DC, with some types rated to 42 V, and that the regular family is built to ISO 8820-3 and SAE J1284. A 12 V auxiliary lighting circuit sits comfortably inside that voltage window, which is one reason the fuse itself is rarely the weak point.
| Lamp power at 12 V nominal | Continuous current | Minimum cable to consider | Reasonable fuse | Check |
|---|---|---|---|---|
| 60 W | 5 A | 1.0 mm² at 16.5 A | 7.5 A | Comfortable margin both ways |
| 120 W | 10 A | 1.5 mm² at 21 A | 15 A | Load at 67% of fuse rating |
| 180 W | 15 A | 2.5 mm² at 29 A | 20 A | Load at 75% of fuse rating |
| 240 W | 20 A | 2.5 mm² at 29 A | 25 A | Load at 80%, fuse at 86% of cable |
| 300 W | 25 A | 6.0 mm² at 50 A | 30 A | Long runs need a voltage check |
Currents above are power divided by nominal voltage. Cable ratings are the manufacturer’s published figures for single-core thin-wall cable: 11 A at 0.5 mm², 16.5 A at 1.0 mm², 21 A at 1.5 mm², 29 A at 2.5 mm² and 50 A at 6.0 mm². Real current differs, because running voltage is not exactly 12.0 V.
If your installation is specified in AWG rather than square millimetres, the conversion is straightforward, but it converts cross-sectional area, not current rating.
| AWG | Cross-sectional area | Nearest metric equivalent |
|---|---|---|
| 18 AWG | 0.82 mm² | 1.00 mm² |
| 16 AWG | 1.31 mm² | 1.50 mm² |
| 14 AWG | 2.08 mm² | 2.00 mm² |
| 12 AWG | 3.31 mm² | 3.00 mm² |
| 10 AWG | 5.26 mm² | 6.00 mm² |
| 8 AWG | 8.36 mm² | 8.50 mm² |
Length is the second half of the calculation, and it is the half that gets skipped. 12 Volt Planet’s cable guide works the example of a 50 W lamp drawing 4.17 A on 0.5 mm² cable with a resistance of 0.0371 Ω per metre: over a 5 m one-way run, counted as 10 m of conductor, the drop is 0.774 V, which is 6.45% and above the 3% target. Moving to 1.5 mm² at 0.0127 Ω per metre gives 0.265 V, or 2.21%, which is acceptable. The guide states that around 3–4% is the generally acceptable voltage drop for DC circuits.
Voltage drop is not just a brightness issue. A circuit dropping volts in the cable is dissipating that energy as heat in the cable, and heat is what eventually destroys insulation and terminals.

Fitting a larger fuse is the most common shortcut in this job and the most expensive one, because it removes the only protection the circuit has while leaving the fault in place. It is not a repair and it is not even a useful diagnostic. Every secondhand account of an LED light bar blowing fuses repeatedly ends the same way when the fuse was upsized: the fuse stopped blowing and something else started failing.
12 Volt Planet is unusually direct about this: if a fuse continues to blow then there must be a fault with the circuit, and a higher rating fuse should never be fitted to overcome this, even temporarily, because doing so creates a high risk of component failure and electrical fire. That sentence is the whole argument.
The engineering reason is that a fuse is designed to be the weakest link. It is specified so that it opens before any other part of the circuit is damaged. Raise the fuse rating and the weakest link moves, usually to the cable or to a connector. Cable does not fail gracefully. It heats, the insulation softens, and at some point the conductor itself becomes the fuse element, at a temperature that ignites anything nearby.
Physical fit matters too. A Maxi fuse does not belong in a standard blade holder, and forcing a larger fuse into a holder that was not designed for it reduces contact area and adds a new heat source exactly where you were trying to protect. Fuse holders are rated for a family and a current, and that rating is part of the circuit.
If a circuit nuisance-blows with the correct fuse, the correct response is to find out why. Either the load genuinely draws more than the circuit was designed for, in which case cable and fuse both need resizing, or there is a fault that only manifests under load, in which case the fault still needs finding. Neither case is solved by a larger fuse.
A relay that clicks and a switch that feels fine can both cause an intermittent failure, and their symptoms imitate a short closely enough to mislead. Contacts erode every time they make and break a load, and an inductive load accelerates it.
Each operation draws a small arc as the contact closes or opens. Over thousands of cycles the contact surface pits, the contact area shrinks, and resistance rises. A joint of even 0.1 Ω in a 20 A circuit dissipates 40 W at the contact, which is enough to discolour a socket, soften plastic and eventually weld or burn the contact. The visible result is a relay whose plastic body has browned, or a socket with melted plastic around terminal 30.
The coil side fails differently. A coil cooked by heat can develop shorted turns and drop from its normal 60–80 Ω to a few ohms, which raises coil current from 150–200 mA to several amps and can blow the coil-side fuse without any lamp fault at all. An open coil reads infinite resistance and simply never operates. Both are measurable across terminals 85 and 86 with the relay removed.
| Symptom | Likely cause | Safe check | Action |
|---|---|---|---|
| Relay clicks rapidly, lamp flickers | Coil supply dropping out, or bad coil ground | Measure coil voltage at 85 and 86 while it clicks | Repair the coil feed or ground |
| Relay does not click, lamp dead | Open coil | Resistance across 85 and 86 | Replace relay |
| Coil-side fuse blows | Shorted coil or reversed diode relay | Coil resistance, relay part number | Fit the correct relay, wired correctly |
| Fuse blows only when relay closes | Short downstream of terminal 87 | Disconnect 87 and re-test | Trace the load-side loom |
| Relay body brown, socket melted | Contact erosion from sustained load | Visual, plus voltage drop across 30 and 87 | Replace relay and socket |
| Switch warm or scorched | Switch rating below load current | Compare switch rating to measured current | Fit a relay and a rated switch |
The diode detail catches people out. A relay with an internal suppression diode is polarity sensitive, and the 12 Volt Planet guide warns that the diode is installed with its stripe towards terminal 86, so positive must go to 86 and ground to 85, or the diode can be damaged. A damaged suppression diode can short the coil feed directly, which presents as a fuse that blows the moment the circuit is switched on with no lamp connected. A relay without a diode is not automatically the right answer either: the guide notes that coil de-energisation produces a reverse voltage spike of several hundred volts, which is exactly what the diode or resistor suppresses.
Water and abrasion develop over months, pass a resistance test when dry, and fail in conditions the owner describes as random. They are the two most often missed, because the vehicle behaves perfectly in a dry workshop.
Water enters through a damaged gland, an unsealed crimp, a connector pushed together with the seal folded, or a splice that was taped rather than sealed. Once inside, it does not simply short the circuit. It wicks along the strands, oxidises the copper and, in the presence of salt, creates a conductive path across the connector face. The evidence is distinctive: green or blue-green deposits on the copper, black discolouration of the strands under the insulation, tide marks on the connector body, and a seal that has swollen or hardened. The condensation and water ingress diagnosis guide covers how to tell normal venting fog from a real leak before you condemn the lamp.
Abrasion is mechanical. A loom that passes a metal edge without a grommet, rests on a bracket, or is pulled tight around a corner will eventually wear through its insulation. The evidence is a rubbed or flattened sheath, visible copper, a polished metal edge where the loom has been moving against it, and a fault that appears when the suspension moves.
The repair for both is the same principle: remove the damaged material, do not cover it. Tape over green copper seals the corrosion in. A splice over a chafed section leaves the chafed section in place. Cut back to bright copper, crimp a proper terminal, seal the joint with adhesive-lined heat shrink, and re-route the loom so it is not touching whatever damaged it.
Routing changes are part of the repair, not an extra. Add a service loop so the loom is not under tension, a drip loop so water runs off, and a grommet wherever the loom passes metal. If the installation does not change, the fault returns at the same place.
A lamp can be the short, and the common internal causes are worth knowing even though most are not owner-repairable. Water bridging a driver board, a pinched internal lead at the entry gland, a failed input capacitor, and a damaged grommet are the usual four.
The meter can identify a dead short across the lamp’s input, and a short from either input pin to the metal housing. It cannot test a potted driver board, see a cracked solder joint, or reproduce a fault that only appears when the housing heats. A meter that reads open does not prove a lamp is good.
Multi-function bars complicate the picture. A bar with a separate daytime running light feed and a main beam feed has two circuits inside one housing, and a fault in the DRL module can blow the DRL fuse while the main beam works perfectly. The Jungle Series 40 inch slim straight bar, which combines a dynamic start-up sequence with an amber and white DRL, is a working example of that arrangement. If the installation uses more than one feed, test each feed separately rather than treating the lamp as a single load.
Do not open a sealed housing to investigate. The seal is part of the product’s ingress protection, and once disturbed the lamp will take water. If a lamp is inside its warranty period, a resistance reading across its input plus photographs of the installation are usually what a supplier needs.
The table below condenses everything above into the order a fault actually presents. Use it to choose the next check, not to skip the checks. A symptom can have more than one cause, and the only way to be sure is to measure.
This is also where most LED light bar blowing fuses cases resolve: the symptom pattern points to one of four paths, and the isolation steps confirm which.
| Symptom | Likely cause | Safe check | Action |
|---|---|---|---|
| Fuse blows the instant the switch closes | Hard short on the load side | Disconnect lamp, re-test | Segment the loom, Step 4 |
| Fuse blows a few seconds after switch-on | Capacitive inrush or a rising-resistance fault | Measure with lamp disconnected | Check lamp input and connector heat |
| Fuse blows after several minutes | Connector or cable heating under load | Feel connectors after a run, then measure | Cut back and re-terminate |
| Fuse blows only in rain or after washing | Water ingress at connector or gland | Inspect for verdigris and tide marks | Reseal, re-route, replace the section |
| Fuse blows only over bumps | Chafed loom or loose terminal | Wiggle test with meter connected | Repair the segment and re-route |
| Relay clicks, fuse intact, no light | Coil circuit or contact failure | Coil resistance, contact voltage drop | Replace relay or repair coil feed |
| Fuse intact, cable warm, lamp dim | High-resistance fault below fuse rating | Voltage drop along the run | Fix the joint; the fuse will never find it |
| Fuse holder melted, fuse intact | Poor contact at the fuse clips | Visual and voltage drop across the holder | Replace the holder, not just the fuse |
| New lamp blows fuses immediately | Lamp internal short, or wrong voltage rating | Resistance across lamp input | Return lamp; verify 12 V versus 24 V rating |
Not every fuse failure starts at the lamp. Supply-side conditions change the current a circuit draws and the voltage its relay coil sees, and both can produce a failure that looks like a short.
A constant-power load draws more current as supply voltage falls, because the driver holds its output constant and the input power has to come from somewhere. A circuit whose fuse is already at 80% of its rating at nominal voltage has very little margin on a weak battery, and a marginal fuse can nuisance-blow for a reason that has nothing to do with a short. That is an argument for the 75% rule rather than against it.
Shared grounds cause a different class of problem. A corroded chassis ground raises the resistance of the return path, which reduces the voltage the relay coil actually sees and can make a relay chatter, drop out under load, or fail to hold. Measure the voltage between the relay’s coil ground terminal and the battery negative post while the circuit is loaded; a significant reading there is a ground fault.
Cranking is the other upstream event worth knowing. The supply voltage dips, and any auxiliary circuit live at that moment sees the dip. A relay whose coil feed drops out during crank will release and re-make, producing a click that sounds like a fault. If the symptom only occurs while starting, suspect the coil feed.
A repair is not finished when the lamp lights, but when the circuit has survived the conditions that caused the failure. Five checks cover that ground.
First, fit the correct fuse and run the circuit continuously for at least twenty minutes with the lamp on. Second, feel every connector and the fuse holder immediately after switching off; anything noticeably warm is a joint that will fail. Third, measure the voltage at the lamp with the circuit loaded and compare it with the voltage at the battery. Fourth, if a clamp meter is available, measure the running current and compare it with the fuse rating to confirm the margin. Fifth, drive the vehicle over the surface that originally triggered the fault, and if the fault was water related, apply water to the repaired area deliberately.
| Check | Pass condition | If it fails |
|---|---|---|
| Twenty-minute continuous run | Fuse holds, output stable | Re-isolate and return to Step 2 |
| Connector and holder temperature | No connector noticeably warmer than ambient | Re-terminate or replace the joint |
| Loaded voltage at the lamp | Close to battery voltage, within a few percent | Upsize cable or repair the joint |
| Measured running current | Comfortably below fuse rating | Re-check cable and fuse selection |
| Road or water test | No repeat failure under the original trigger | The original cause was not found |
Write the result down. A record of fuse rating, measured current, loaded voltage and conditions tested turns a repair into evidence, and is what a supplier asks for on a warranty claim.
Almost every outcome of this procedure falls into one of three categories, and deciding which applies is the last job before reassembly. Repairing a cable is cheap and permanent if the routing is changed. Replacing a relay or a fuse holder is cheap and permanent if the rating was correct. Replacing a lamp is expensive and temporary if the installation caused the failure in the first place.
| Finding | Repair | Replace | Escalate |
|---|---|---|---|
| Chafed cable at a bracket | Cut back, re-terminate, add grommet | Cable section if damage is long | No |
| Water in a connector | Dry, clean, reseal, re-route with drip loop | Connector and pigtail if pins are corroded | No |
| Failed relay | No | Relay, and socket if discoloured | No |
| Melted fuse holder | No | Holder and fuse | If surrounding wiring is heat damaged |
| Shorted lamp input | No | Lamp, after correcting the installation | If the lamp is sealed and in warranty |
| Fault in a factory loom | No | No | Yes |
| Repeated failure after five steps | No | No | Yes |
When the lamp is condemned and the installation corrected, the replacement decision is about the electrical load you actually need. The LED light bar range lists bar sizes and beam options; for fixed work areas the LED work light range covers lower-power lamps that are easier to fuse. Matching the new lamp’s current draw to the existing cable and fuse is cheaper than rebuilding the circuit.
There is a point where continuing stops being diagnosis and becomes guesswork, and the cost of guesswork on a vehicle electrical system is high. HSE’s guidance is a useful line in the sand: people working on electrical equipment must be competent, competence depends on training, comparable experience and regular reassessment, and anyone who cannot demonstrate competence should not work unsupervised.
The practical triggers are easier to apply. Stop if you cannot identify a wire’s function from the vehicle’s documentation. Stop if the fault is inside a factory harness, behind a dashboard, or in a circuit shared with restraint, braking or engine management systems. Stop if a fuse box or a connector has melted, because heat damage usually extends further than the visible char. Stop if the correct fuse blows again after the five steps have been completed honestly.
A good auto electrician will ask three questions: what fuse, what rating, and what were the conditions. If you have followed this procedure, you can answer all three, and the diagnosis will take minutes rather than hours. That record is the most valuable thing you can hand over.
Running voltage is higher than the voltage of a battery at rest, so the circuit carries more current and any marginal joint carries more load. It is also the point at which the whole electrical system is at its highest normal voltage. A fuse sized exactly at the lamp’s nominal current can fail here. Measure the running current and confirm the fuse has real margin.
A bad ground on its own does not normally blow a fuse, because it raises circuit resistance and reduces current. What it does do is make relays chatter, cause intermittent operation, and force current to find another return path, which can overload a different circuit. If a fuse blows and the ground is suspect, fix the ground first and re-test, because the symptom may change.
No. Bridging a fuse holder removes the only overcurrent protection in the circuit and converts a diagnostic step into a fire risk, with a battery behind it that can deliver hundreds of amps. Use a fuse of the correct rating for every test.
The fuse is at fault only if it has opened; it is a one-shot device and it does not fail intermittently. If the fuse is intact and the lamp does not work, the fault is elsewhere, and the relay is a reasonable next suspect. Test the relay by measuring resistance across terminals 85 and 86, which should read roughly 60–80 Ω on a typical mini relay, and by checking voltage drop across terminals 30 and 87 under load.
Sometimes, and with caveats. A thermal breaker resets, which is convenient for a load that is expected to stall, but 12 Volt Planet notes that a breaker can take longer to operate than a fuse and that a breaker should never be used where sensitive electronic equipment is involved. For an LED light bar, a fuse is the simpler and more predictable choice.
Because the fault resistance was high enough that the current stayed below the fuse rating. A fault of about 1 Ω on a 12 V circuit draws roughly 12 A, which a 15 A fuse will pass indefinitely while the cable dissipates the energy as heat. This is the case where a fuse gives no protection at all, and it is why voltage drop and joint temperature matter as much as current.
At a nominal 12 V a 120 W lamp draws about 10 A. On 1.5 mm² cable rated at 21 A, a 15 A fuse sits comfortably between the load and the cable rating, and the lamp runs at about 67% of the fuse rating, which leaves headroom for inrush. Confirm against the lamp’s own published current figure before fitting anything.
It may stop the fuse blowing, and it will not stop the fault. A larger fuse raises the current at which the circuit is protected, which means the cable or a connector becomes the weakest link instead. If a circuit nuisance-blows with the correct fuse, the cause is either an undersized circuit for the load or a fault that only appears under load, and both need diagnosis rather than a larger fuse.
Written and reviewed by the SAEC Lighting Engineering Team, Foshan. This article describes a diagnostic order for 12 V and 24 V auxiliary lighting circuits and does not replace the vehicle manufacturer’s service information, which governs fuse locations, relay assignments and any circuit shared with safety systems. Current ratings, cable ratings and resistance values quoted here are taken from the sources listed below or are labelled as illustrative calculations. Last updated: 2026-10-17.