A step-by-step diagnosis for a camera system that only fails in wet weather — what to check, in what order, and which faults are workmanship rather than weather.

Rain does not damage a camera by falling on it. Outdoor cameras are sold with a weather rating for the housing, and the housing is almost never the thing that fails. What rain does is find every joint, gland and screw hole that someone made after the camera left the factory, and it finds them in a very specific way: not by pouring in from above, but by wicking along the outside of a cable, sitting in the bottom of a junction box, and creeping into a connector by capillary action over weeks. That is why the fault is intermittent, why it gets worse each monsoon, and why it clears up for a few dry days and comes back.
There is a second mechanism that is easy to miss. Wet cable and damp connectors lower the resistance between the conductors inside the cable. On a Power over Ethernet run that carries both power and data, a small leakage current between pairs is enough to corrupt the data long before anything looks visibly wet, so the camera stays powered up and the recorder still shows the channel as offline. A technician looking only for a puddle will walk past that fault.
And a third: rain in the Klang Valley arrives with lightning, and a nearby strike puts a surge onto the mains and onto every long metal run in the building. That damage is permanent and it looks nothing like a water fault, which is why it is worth ruling in or out early. It is covered in full in CCTV lightning and surge protection.
This single question splits the diagnosis in half and costs nothing to answer, so answer it before anything else. Watch the recorder or the app during the next shower and note exactly what goes.
Write the answer down with the date and the weather. Two observations across two storms are worth more than an hour of guessing, and any competent technician will ask you for exactly this.
| What you see | Most likely cause | Check this first |
|---|---|---|
| One camera offline in rain, back after a dry day | Water in the cable joint or the connector at the camera end | The junction box or connector directly behind that camera |
| One camera reboots repeatedly at night, worse in rain | Voltage drop on a long or undersized run, made worse by damp | Cable length and cable type on that run, then the recorder’s PoE port |
| Camera image milky, hazy or fogged from inside | Moisture inside the housing past a perished dome seal or an unsealed screw hole | The dome gasket and the cable entry on the camera body |
| All cameras offline, recorder screen also dead | Power — a tripped circuit, the recorder’s adaptor, or surge damage | The socket and the adaptor before anything on the roof |
| Cameras record locally but the phone app is dead | Internet or router, not the camera system | Whether the recorder is still recording; if it is, call your ISP |
| Channel says “no video” but the camera’s infrared is still glowing | Data fault on a powered run — corroded or wet connector, damaged pair | The connector at both ends, then the cable itself |
| One camera dead permanently after a storm with thunder | Surge along the cable, not water | Whether the recorder’s port for that channel also stopped working |
| Everything worked for two years, now fails every wet season | An outdoor joint that was never properly sealed, slowly wicking | Every joint made outside the building envelope |
Send us a photo of the job on WhatsApp — we reply with an instant quote.
💬 Get Your Instant QuoteThe gland is the fitting where the cable enters the camera or a junction box, and it is the most common failure point on a Malaysian install for a simple reason: it is the one part of the assembly the installer makes on site, in the heat, at the top of a ladder, at the end of the day. Three things go wrong with it.
The gland is not tightened onto the cable jacket, so it grips nothing and the cable can move; water follows the cable straight in. The box is mounted with the entry pointing up, which turns it into a cup — rainwater running down the wall collects at the entry and sits there. And there is no drip loop: the cable runs downhill directly into the fitting, so every drop that touches the cable is delivered into the joint. A drip loop is nothing more than a deliberate sag in the cable below the entry point, so water runs to the bottom of the loop and falls off. It costs nothing, it takes seconds, and its absence is visible from the ground.
The related myth worth killing: a camera marked with a weather rating is rated for its own housing, as it left the factory. It says nothing about the joint you made in its cable, the box you screwed to the wall, or the hole you drilled through the render. Those are your installer’s work and they carry no rating at all unless the parts used were chosen for outdoor use.
On most jobs the camera arrives with a short pigtail lead that has to be joined to the cable running back to the recorder. That joint is the single most abused detail in the trade. Done properly it lives inside a weatherproof box that is sealed, mounted with the entry downwards, and positioned where a person can reach it again. Done badly it is two connectors pushed together and wrapped in ordinary electrical tape, taped to the underside of an eave, and completely inaccessible.
Ordinary PVC tape is not a seal. It unwinds in the heat, its adhesive creeps, and within a couple of wet seasons it is holding water against the connector rather than keeping it out. The materials that do work — a proper weatherproof enclosure, a gel-filled or otherwise water-blocked connector, self-amalgamating tape that fuses into a single mass — cost very little more, and the difference only becomes visible in year three, which is exactly why cheap jobs skip them.
Position matters as much as materials. A joint that has to be reached from a ladder over a driveway, in the rain, is a joint that will not be inspected until it fails twice. Ask where every outdoor joint is before the crew leaves, and photograph them.
Power over Ethernet sends the camera’s power down the same cable that carries its video. Over a long run, through thin conductors, some of that voltage is lost as heat in the cable itself, and the camera at the far end receives less than it should. On a dry afternoon it copes. At night, when the infrared illuminators switch on and the camera’s current draw goes up, it may not — and the camera reboots, comes back, and reboots again.
Rain enters the story in two ways. Damp in a connector adds resistance, which subtracts a little more voltage from an already marginal run. And rain arrives with darkness and cloud, which is exactly when the infrared comes on. So a marginal PoE run behaves like a water fault: it fails during storms and works in the sun. The tell is the pattern — a water fault stays down until things dry out, whereas a voltage-drop fault cycles: down, up, down, up, minute after minute.
The causes are cable length, cable quality and joints. Copper-clad aluminium cable sold as network cable has noticeably higher resistance than solid copper and is the commonest reason a run that should have been fine is not. That is a decision made at quoting time, which is why cable specification belongs in the written quote — see CCTV wiring, concealed versus surface.
| Observation | Points to a power fault | Points to a network or data fault |
|---|---|---|
| Camera’s infrared glow at night | Goes dark when the channel drops | Still glowing while the channel says no video |
| Behaviour of the drop | Cycles — off, on, off, on | Stays off until conditions change |
| Other cameras on the same recorder | May also flicker if the supply is shared | Unaffected |
| Moving the cable to a different recorder port | Same fault follows the camera | Fault may clear, which indicts the port |
| Plugging the camera in with a short lead at the recorder | Camera works, so the fault is in the long run | Camera works, so the fault is in the long run |
| Time of day | Worse at night, when infrared draws most | No particular pattern by time |
The last row of that table is the most useful test on the list, and it is the one nobody does: take the camera down, carry it to the recorder, and connect it with a short patch lead. If it works perfectly there, the camera is fine and the problem is somewhere in the run — cable, joints or connectors. If it misbehaves there too, the camera is the problem and no amount of re-taping a joint will help.
Steps one to five need no tools and no ladder. Most homeowners can complete them in ten minutes and hand a technician a diagnosis instead of a symptom, which usually turns two visits into one.
Water faults recover. Surge faults do not. If a channel has been dead since a thunderstorm and has never come back, if the recorder’s network port for that channel is also dead, or if a network switch in the chain has stopped passing traffic entirely, the cause is electrical rather than wet. The camera cable is a long conductor running outside the building, and a nearby strike induces a voltage on it that arrives at both ends. Sealing the joint will not bring it back.
The distinction matters commercially too: a water fault is a workmanship issue and often a warranty conversation, while surge damage is an act of weather that proper protection reduces but never eliminates. The protection options, and their honest limits, are set out in CCTV lightning and surge protection. Whole-premises protection and earthing sit with an electrician — see lightning and surge arrester installation.
Wireless cameras fail in rain for entirely different reasons, and swapping a wired camera for a wireless one to “solve” a rain fault usually trades one problem for another. Heavy rain attenuates a radio signal, wet foliage between the camera and the router absorbs more of it, and a camera that was marginal in the dry becomes unusable in a storm. Battery cameras also charge less in overcast weather if they are solar-assisted, so a week of rain can flatten one that survives the dry season easily.
The consolation is that a wireless camera has no long outdoor cable to carry a surge, which genuinely reduces one failure mode. The comparison is set out in Wi-Fi cameras versus professional CCTV, and the mains-free and broadband-free options are covered in CCTV without Wi-Fi, using 4G or solar.
A water-ingress repair is not a re-tape. Done properly it means opening the joint, cutting back the cable past the wet and corroded section — corrosion travels up the conductors, so the visibly green part is not the whole of it — remaking the termination on clean copper, and re-housing it in a weatherproof enclosure mounted with the entry downwards and a drip loop below it. Where the cable itself has taken water along its length, the honest answer is a new cable, because a wicked run will keep failing however well the ends are made off.
If the camera has water inside the housing, the dome gasket and the body seals get inspected, and a camera that has been full of water more than once is usually replaced rather than dried out. And if a channel is dead from surge, the repair is replacement of whatever the surge destroyed, plus a conversation about protecting the replacement.
These are market figures published by others rather than by ClickBina (market estimate 2026, indicative Klang Valley). Read them carefully: the maintenance rows come from one published Klang Valley rate card, and the two per-camera rows from a second single operator’s card, so every figure here is one installer’s price rather than a surveyed market rate.
| Work | Market range (2026, Klang Valley) | Notes |
|---|---|---|
| Ad-hoc service call | RM150 – RM250 per visit | diagnosis, realignment and configuration; replacement parts are charged on top |
| Maintenance, 1–5 cameras | RM150 per visit | one scheduled service a year; parts are charged separately |
| Maintenance contract, 6–30 cameras | RM240 per month | twelve services a year, labour only — parts are charged separately |
| Install a camera you already own | RM80 – RM150 per camera | mount, aim and configure only; the rate drops once you pass four cameras |
| Add one camera to an existing system | RM300 – RM400 per camera | cabling and installation for one more point; the camera itself is on top |
Two things move the bill more than the visit itself. The first is whether the cable has to be re-pulled, which turns a joint repair into a cabling job at the per-point rates in the wiring guide. The second is access — a joint under a two-storey eave over a driveway needs equipment, and that is a separate line. Whole-system pricing, and what a new install should itemise, is on the CCTV installation cost guide. We do not publish a package price for this work; we survey, then quote.
ClickBina services and repairs camera systems across Kuala Lumpur and Selangor, and rain faults are among the most common calls we take. We diagnose before we quote: we want to know whether one camera dropped or all of them, whether the infrared was still on, and what the weather was doing, because those three answers usually locate the fault before anyone climbs anything. We will tell you plainly when a joint can be remade and when a cable has to be re-pulled — and we do not publish a package price for security work, so every job is quoted from what the survey finds. Send photos of the camera, the joint and the recorder, or a floor plan, and we will quote after a site survey. The mains side sits with our electrical and plumbing team.
Tell us what you need — we reply within the hour.