# Light pipe design for LED status indicators

> Good light pipe design makes status LEDs bright, even and free of crosstalk. Coupling, geometry, materials, tolerances and sealing explained.

- Author: Milo Perović, Founder & lead engineer
- Published: 2026-11-01
- Topic: Design

A status LED is often the only thing a user ever reads on a connected device, yet light pipe design is usually left until the enclosure is nearly finished. The result is familiar: a dim, blotchy dot, green light bleeding into the red one next to it, and an indicator that looks different on every unit off the line.

This post is for founders, product managers and engineers who want their indicators to look deliberate. It covers how light pipes actually carry light, the geometry and materials that work, and the tolerance and sealing problems that catch teams out at the first moulded parts.

## How a light pipe moves light

A light pipe is a solid piece of clear plastic that guides light from an LED on the PCB to a window in the housing. It works by total internal reflection: light hitting the side wall at a shallow enough angle stays inside the pipe instead of escaping. For PMMA (acrylic, refractive index around 1.49) the critical angle is roughly 42° from the surface normal; for polycarbonate (around 1.58) it is closer to 39°.

Two consequences follow. First, any light that hits the wall more steeply than that leaks out, so tight bends, scratches and sink marks all cost brightness. Second, light that never enters the pipe is simply lost. The coupling between LED and pipe face usually matters more than the length of the pipe itself.

## Getting light into the pipe

Most of the losses we find on the bench happen in the first millimetre. A typical SMD LED emits over a wide cone, often a viewing angle of 120° or more, and a pipe input face only captures part of it. A few rules keep coupling efficient:

- **Keep the air gap small.** Aim for a nominal gap of a few tenths of a millimetre between LED and pipe. Larger gaps lose light quickly because the cone spreads past the pipe face.

- **Size the input face to the LED.** The pipe should be at least as wide as the emitting area, and usually a little wider to tolerate misalignment. For 0603 and 0805 LEDs, pipes in the 2–3 mm range are common.

- **Match the LED to the path.** A top-firing LED suits a straight vertical pipe. If the window is on the side of the product, a side-firing (right-angle) LED or an edge-mounted part often beats a pipe with a 90° bend.

- **Never touch the LED.** Design the nominal gap so that the worst-case stack-up still leaves clearance. A pipe pressing on an LED will eventually crack a solder joint or the lens.

## Geometry and materials for light pipe design

Straight and short is always best. Every bend lets some light escape, and the loss rises sharply once the bend radius gets small relative to the pipe diameter. As a rule of thumb we keep the inside bend radius at two to three times the pipe diameter or more, and we avoid more than one bend per pipe if we can.

Surface finish does more than most people expect. The side walls should be highly polished in the tool so that total internal reflection works; an SPI A-1 or A-2 finish is typical for optical surfaces. The exit face is a different decision: polished gives a bright, sharp point, while a light texture or frosting spreads the light into an even glow that hides the LED hotspot.

PropertyPMMA (acrylic)Polycarbonate

Light transmissionSlightly higher, around 92%Around 88–90%

ToughnessBrittle, can crack at snap featuresMuch tougher, tolerates press-fits

Heat resistanceLowerHigher

Scratch resistanceBetterSofter surface

For most small connected devices we default to optical-grade polycarbonate because pipes get pressed, clipped and handled during assembly. PMMA earns its place when the pipe is exposed on the outside of the product and scratch resistance matters more than impact.

### Moulding details that affect light

Place the gate away from optical surfaces, ideally on a mounting tab, because gate vestige and flow lines scatter light. Keep wall sections even to avoid sink marks on the sides of the pipe. Ejector pins on optical faces leave marks you will see when lit. These are the same disciplines we describe in [designing for injection moulding before you have a mould budget](https://miloperovic.com/journal/designing-for-injection-moulding-before-you-have-a-mould-budget/), applied to a part where cosmetic flaws become optical ones.

## Stopping crosstalk between indicators

Two LEDs a few millimetres apart will light each other's pipes unless you stop them. Light escapes from the sides of the LED package, reflects off the PCB and finds its way into the neighbouring pipe. A charging indicator that glows faint green while showing red reads as a fault to the user.

The fixes are mechanical. Opaque ribs or a black shroud moulded into the housing around each LED, a black foam gasket between PCB and pipe carrier, or a two-shot part with an opaque body and clear pipes all work. Black soldermask under the LEDs also helps, because white or green soldermask reflects stray light. Test this in a dark room with only one LED on at a time; crosstalk that is invisible on a lit bench is obvious at night.

## Tolerances, assembly and sealing

The gap between LED and pipe is the end of a tolerance chain: PCB thickness, standoff heights, LED placement, housing shrinkage and the pipe itself. If the pipe locates on the housing and the LED sits on a PCB that floats on screw bosses, the gap can vary by several tenths of a millimetre from unit to unit, which shows up as visible brightness variation.

There are two common strategies. Locating the pipe on the PCB, with off-the-shelf press-fit or SMT light pipes, fixes the coupling and moves the tolerance to the exit end, where it is less visible. Locating the pipe in the housing gives a cleaner exterior but needs a proper stack-up analysis.

Sealed products add another constraint. A pipe that passes through the wall is a leak path, so it needs an O-ring, an overmoulded seal, or a design where the pipe stays inside and light passes through a thin translucent window in the housing. We cover the wider sealing picture in [IP67 enclosure design: seals, vents and leak testing](https://miloperovic.com/journal/ip67-enclosure-design-seals-vents-and-leak-testing/). "Hidden until lit" indicators, where a locally thinned wall in a light-coloured translucent material glows through, avoid the pipe entirely but depend heavily on material colour and wall thickness control.

## Brightness, colour and meaning

An indicator has to be readable in daylight and not annoying on a bedside table. That usually means driving the LED with PWM from firmware rather than fixing brightness with a resistor, so that it can be dimmed after a few seconds or adjusted for ambient light. Give firmware the hooks early; retrofitting a dimming scheme onto an LED wired straight to a GPIO with a large resistor is a board respin.

Colour needs care too. RGB LEDs need some pipe length or a diffusing exit to mix into one colour instead of three visible dots. And because a significant share of users have some form of colour vision deficiency, do not rely on red versus green alone. Pair colour with blink patterns, and write those patterns down in one table that firmware, the app and the manual all share. For industrial products, IEC 60073 sets out coding principles for indicator colours and is worth checking before you assign meanings.

## What we would do before freezing the enclosure

- Pick LEDs and their orientation together with the window position, not after.

- Run a tolerance stack for the LED-to-pipe gap and confirm worst-case clearance and worst-case brightness.

- Prototype the pipe in clear resin or machined acrylic and photograph it at a fixed camera exposure next to the target finish.

- Check every indicator one at a time in a dark room for crosstalk.

- Specify polished side walls, the exit texture and gate location explicitly on the drawing.

- Agree a written table of colours, blink patterns and meanings, with PWM brightness control in firmware.

None of this is expensive if it happens during detailed design. All of it is expensive once the tool is cut.
