BLOG Icon-ImageBy: Eeyun

LED Strip Voltage Drop: 12 V vs 24 V on Long Lightbox Runs

Two identical five-metre runs, same strip, same driver wattage: one at 12 V, one at 24 V. They do not end at the same brightness. On 12 V the far end gives up roughly 4 % of its light; on 24 V, roughly 1 %. And the gap widens with length, because the loss grows with the square of the run, not in a straight line.

The rule of thumb. With 14.4 W/m strip on 0.5 mm² copper, a single-feed run stays inside 3 % loss for 4.1 m on 12 V and 8.3 m on 24 V. Past that, feed from both ends or move to 24 V.

What is actually dropping

A constant-voltage strip is a resistor that also emits light. Its copper trace has resistance, and every LED along the run draws current through the copper that sits between it and the feed. The strip nearest the driver sees the full voltage; the strip at the far end sees the driver voltage minus everything the copper ate on the way.

Because the load is spread evenly, only half of the run’s total resistance acts on the far end — which is why the drop rises with the square of the length. Double the run length and you quadruple the loss.

Why 24 V is four times better, not twice

Two effects stack. The same wattage at double the voltage draws half the current, so the copper loses half the volts. And the supply itself is twice as tall, so the same lost volts is half the share. Half of a half: 24 V ends up about four times cleaner on the same run.

12 V

Loss at the far end

1.2 A per metre

1 m
0.18 %
2 m
0.70 %
3 m
1.6 %
5 m
4.4 %
7 m
8.6 %
10 m
17.5 %
020 %
24 V

Loss at the far end

0.6 A per metre

1 m
0.04 %
2 m
0.18 %
3 m
0.39 %
5 m
1.1 %
7 m
2.1 %
10 m
4.4 %
020 %

Brightness loss at the far end of a single-feed run

Assumptions: 24 V strip at 14.4 W/m (0.6 A/m), the 12 V version of the same strip (1.2 A/m), copper at 0.0175 Ω·mm²/m and an effective 0.5 mm² conductor. Treat it as an order of magnitude to design against, then measure your own run.

The numbers behind the chart

Run length12 V: drop and loss24 V: drop and loss
1 m0.02 V · 0.18 %0.01 V · 0.04 %
2 m0.08 V · 0.70 %0.04 V · 0.18 %
3 m0.19 V · 1.6 %0.09 V · 0.39 %
5 m0.53 V · 4.4 %0.26 V · 1.1 %
7 m1.03 V · 8.6 %0.51 V · 2.1 %
10 m2.10 V · 17.5 %1.05 V · 4.4 %

Same assumptions as the chart. Percentages are of the supply voltage; on a constant-voltage strip the light loss tracks them closely until the strip stops regulating.

Four fixes, cheapest first

1. Move to 24 V. No new hardware beyond a 24 V driver, and it buys you the four-fold margin above. It is the default for any run over 3 m in our own builds.

2. Feed from both ends. Two feeds halve the length each half has to carry, and the square law turns that into a quarter of the loss. On a 10 m 12 V run that is the difference between a visible gradient and none.

3. Heavier copper. Going from 0.5 to 1.0 mm² halves the resistance and the loss with it. Worth it on long curtain runs where you cannot add a second feed.

4. Split the run. Two 5 m sections on separate feeds lose a quarter of what one 10 m section loses. A constant-current driver sidesteps the problem differently: it holds current rather than voltage, so the far end cannot dim — but the driver must match the total load and the strip must be rated for it.

Where it shows up in a lightbox

On a long edge-lit face it rarely shows as a number; it shows as a photograph. The corner furthest from the feed sits dimmer and, on a white face, slightly cooler. Clients notice it on white, not on colour. The same 4 % that is invisible in a test rig becomes the thing they point at in a showroom.

Two habits keep you out of that conversation: measure at the far end of the run, not at the driver terminals; and if your driver has a voltage trim, resist the urge to wind it up — that lifts the near end above spec while the far end still lags.

Frequently Asked Questions

Does voltage drop change the colour temperature?

Only once the loss gets large. Below roughly 5 % nobody sees a shift; past 10 % the dimmed end reads a little cooler on a 3000 K run, and the gradient becomes visible on white. If colour consistency matters across a long face, treat 5 % as the ceiling and design to it.

Can I just raise the driver voltage?

A trim pot lifts the whole run, so the near end now runs above its rated voltage — hotter, shorter-lived — while the far end only partly catches up. It moves the problem rather than solving it.

Is 48 V worth it?

For very long runs, yes: the same maths gives you another four-fold margin over 24 V, and the copper can be thinner. The trade-off is driver choice and the fact that fewer strips are rated for it. Below 8 m of 24 V run, it is usually not needed.