Cabinet lighting: glare, reed switches, and how much one driver actually runs

Cabinet lighting splits into two different jobs that get installed the same way and behave completely differently once finished: under-cabinet task lighting you look past, and inside-cabinet display lighting you look directly at. Treating them the same is where most installs go wrong.

Written by the people who manufacture the controller inside these fixtures. Last reviewed July 2026.

Under-cabinet: hide the source, not the light

Mounted along the bottom front edge of a wall cabinet, aimed at the counter, the strip sits close to standing eye height for anyone at the counter or across the kitchen. The complaint we hear most is glare, and it is almost always a mounting position problem rather than a brightness one: strip mounted flush on the flat underside of the cabinet is directly visible from a normal standing position, while the same strip set back a centimetre or two behind the cabinet's front lip is invisible from anywhere except directly underneath, looking straight up.

Use the lip. Almost every wall cabinet has a shallow recess at the front edge for exactly this, and setting the strip back into it rather than mounting it flat is the difference between a countertop that is lit and a kitchen where you notice individual LEDs every time you look up.

Inside-cabinet: this is where a reed switch earns its place

Glass-front display cabinets and closets want a different control model entirely — light on when the door opens, off when it closes, with no app interaction needed day to day. That is what a magnetic reed switch does: a small sensor on the frame and a magnet on the door, wired in line with the strip, with no controller logic required at all for the basic on/off behaviour. Where our controllers add value is layering brightness and colour temperature on top of that simple trigger, so the reed switch decides when and the app decides how.

Mount the reed switch on the hinge side, not the latch side. A door swinging on its hinge stays closer to the frame through more of its travel on the hinge side, which means a more consistent gap between switch and magnet and a more reliable trigger point. On the latch side the gap changes quickly as the door swings, and a switch tuned to trigger reliably there often triggers too early or not at all.

PIR or radar: the sensor decides more than the light does

A reed switch answers "is the door open". Everything else — a light that comes on as you reach the worktop, or as you walk into the pantry — needs a motion sensor, and there are two kinds in this product class. They fail in opposite directions, which is why the choice matters more than the fitting does.

PIR is passive infrared. It watches for a warm body crossing between its detection zones, so it needs a clear line of sight and it is blocked by anything solid — a cupboard door, a panel, even glass. It is poor at seeing someone walking straight towards it, because that motion does not cross zones. What it is very good at is doing almost nothing while it waits: standby current is in the microamps, which is the whole reason battery-powered cabinet lights use it.

Radar — 5.8 GHz or 24 GHz microwave, or 60 GHz mmWave — detects motion by frequency shift instead. It sees slow approach and small movements a PIR will miss, and it works through thin non-metallic material. That last property is both the selling point and the trap: a radar module will happily trigger through a cupboard door, a plasterboard wall or a worktop, so in a kitchen it can be set off by someone walking past on the other side of the units. It also draws milliamps rather than microamps.

If it runs on a battery, the standby current is the specification

Battery cabinet lights are sold on runtime and the runtime figure is almost always quoted for the light being on. That is not what drains them. A sensor light spends nearly all its life waiting, so the standby draw of the sensor and the controller sets how long a charge lasts, and a radar module in a cupboard that nobody opens for a week will flatten a cell that a PIR would have kept alive for months.

Two things worth confirming before buying, both of which reputable products state and cheap ones do not: the standby current, and whether the pack has over-discharge protection. A cell run flat and left flat does not always come back, and a cabinet light that will not recharge after a long holiday is usually this rather than a fault.

Colour temperature: task light and display light want different answers

Under-cabinet and inside-cabinet lighting are lit for different reasons, and the colour temperature that works well for one is often wrong for the other. Under-cabinet task lighting is there so you can see what you are cutting and read a label accurately — a cooler white, in the 4000–5000K range, renders colour more neutrally and is easier to work under for the few minutes at a time it is actually used for a task. Inside-cabinet display lighting is there to make glassware, crockery or a collection look good while you are not actively working, and a warmer 2700–3000K tends to flatter those surfaces the way a warm-white gallery light does, rather than giving them the slightly clinical look a cool white produces on glass and ceramic.

If the kitchen only has one driver output and you want both behaviours, tunable-white strip is the fix — it lets the same run sit cool while you are prepping food and warm once the cabinets are just for ambience in the evening, without running two different strips. Fixed-colour strip forces a single compromise temperature for both jobs, which is usually why a fixed 3500K install feels slightly wrong in both roles rather than clearly right in either.

How much one driver actually supports

Cabinet runs are usually short individually — one run per cabinet — but kitchens commonly daisy chain several cabinets off a single driver to avoid a transformer behind every door. This is where the same arithmetic from sizing the supply applies at a smaller scale: each metre added to the chain increases the current the driver and the first metre of cable have to carry, and the cabinet furthest from the driver is the one that goes dim first, not the one nearest it. If the last cabinet in a run looks dimmer or warmer than the first, that is voltage drop along the chain, not a bad strip — either inject power partway along the run or split it into two shorter chains from the same driver.

A worked example makes the arithmetic concrete. Take a common 12V strip drawing roughly 4.8W per metre and a driver rated at 60W with a sensible 80% headroom applied, which leaves about 48W of usable output — call it 10 metres of that strip before the driver itself is the limit. That sounds like plenty for a run of wall cabinets, until the wiring is counted separately from the strip's own draw: thin daisy-chain cable between cabinets adds resistance that the strip's own copper does not have, and every metre of that connecting cable between the driver and the far cabinet drops a small amount of voltage before the LEDs ever see it. On a 12V system that drop is proportionally much larger than the same voltage lost on a 24V run, which is the practical reason longer kitchen chains are usually specified in 24V rather than 12V — the same power delivered at double the voltage halves the current, and drop scales with current, not power. If a chain is running to six or more cabinets, ask whether the kit is 12V or 24V before assuming the driver's wattage rating alone tells you how far it will reach evenly.

Corners: the detail that separates a tidy run from a visible kink

Cabinet runs rarely stay in a straight line — an L-shaped counter or a cabinet run that turns into a peninsula both need the strip to change direction, and how you handle that corner is more visible than almost anything else in the install. A strip simply bent flat around an outside corner puts stress on the copper traces at the exact point they are already thinnest, and repeated flexing there is a common cause of a dead section starting right at a corner months after installation. Two better options: a purpose-made corner connector piece, which most of our tunable-white and single-colour strip kits include for exactly this, or cutting the strip at the marked cut point before the corner and rejoining it with a short jumper cable bent to the angle instead of bending the strip itself. Either keeps the actual LED strip straight and puts the flex in a joint designed for it.

Common questions

Can I mix under-cabinet and inside-cabinet lighting on one driver?

Yes, as long as the combined length stays within the driver's rated output — treat it as one run for sizing purposes even if it serves two different cabinets.

Does the reed switch work with dimming?

Yes — the reed switch controls on/off, and brightness or colour temperature set in the app applies whenever the switch turns the strip on. You are not choosing between the two.

Can I cut the strip to fit a specific cabinet?

Yes, at the marked cut points, the same as any of our single-colour or tunable-white strip.

Does colour rendering matter for a kitchen strip?

Yes, more than most buyers expect. A cheap strip's colour rendering index can be low enough that food and produce look slightly off under it even at a "correct" colour temperature — the light is the right white on a colour temperature scale but renders individual colours poorly. Ours are specified for food-preparation use; if you are pricing a competitor's strip against ours, ask for its CRI figure specifically, not just its colour temperature.

Glass cabinet doors reflect the strip — can I stop that?

Set the strip further back behind the door lip so the light source itself is out of the reflection angle for someone standing at a normal viewing distance, the same fix as the under-cabinet glare problem above. A frosted or diffused strip channel also softens the reflected point sources into a wash rather than visible dots.

Which app runs this?

Lotus Lantern. Full mapping here.

Contact us if a chained run is dimming unevenly and you want the driver sizing checked.