All LED downlights are 100% efficient

Independent lighting consultant Alan Tulla explains how easy it is to be misled by measurements of the light output of LED luminaires

Obviously, not all LED downlights are 100 per cent efficient. Some produce more light output from an LED chip than others.

Despite this, both a well engineered, optically efficient LED luminaire and a poor one will demonstrate an ‘efficiency’ — measured as light output ratio (LOR) – of 100 per cent.

So, you might ask, how is it possible for a luminaire to have an LOR of 100 per cent? The answer goes back to the way the light output of luminaires is measured. The starting point is the lamp and knowing how much light it emits.

This is how we have measured the output of luminaires for the past 75 years. Without going too much into the details, this is what you do.

There are lamps available from the National Physical Laboratory that have a known and precise light output, measured in lumens.

Measure of a lamp

The bare lamp is put inside what is known as an integrating sphere. This is a large globe, typically two or three metres in diameter painted matt white inside.

The light from the lamp, burning in free air, inter-reflects a zillion times from the interior and by measuring the lux value on the inside of the sphere, you can relate the lux value to the known value of lumens emitted from a calibrated lamp.

When you switch on the test lamp, you must wait until the light output stabilises and is constant. This may take as long as as half an hour for low-wattage lamps. Measurements are taken at an ambient air temperature of 25ºC.

The next stage is that you put the lamp, whose output is known, in the luminaire you want to test and switch it on. Once the output has stabilised, you again measure the lux level on the inside of the sphere.

Since the lamp is inside a luminaire, there are some losses, and the lux level on the inside of the integrating sphere is less. The difference between the two tells you what percentage of light is lost inside the luminaire and hence gives you the LOR.

That’s all LOR is. This method works for all conventional lamps such as filament, HID, fluorescent and SOX. Experience has shown that varying the wattage of the lamp makes little difference to the LOR.

This is because lamps that are significantly different need a larger luminaire in which to fit. A 1kW SON/T floodlight is larger than a 100W unit of the same design.

The big difference with LEDs is that it is almost impossible to achieve a stable output from a bare chip in free air for any practical length of time.

High-power ones, especially, will simply get hotter and hotter and maybe even burn out. A heatsink is required to keep the temperature within operational limits. Data sheets often refer to the critical ‘junction temperature’.

This brings us to the second issue with measuring LED luminaires. As you know, as an LED heats up, its light output drops. This reduction in lumen output may be as much as 30 per cent. So the lumens emitted from the chip will depend on the construction of the luminaire and the way the chip is mounted inside.

It is important to realise that the light emitted from the luminaire depends not only on the optics and reflectors but also the thermal management and working temperature of the LED chip. Simply changing the heatsink will affect the quantity of light – lumens – emitted.

In effect, the luminaire and LED chip combined become the light source. This is why LED downlights have a LOR of 100 per cent. It cannot be less because, as explained earlier, you cannot measure the output from a bare chip.

It is worth noting that as a result of their small size, a 12W LED luminaire may not be any larger than a 1W unit. This is why thermal management is so important.

Do LEDs affect your calculations?

Not at all. Many readers of Lux magazine will be familiar with lighting software packages such as Relux, Dialux, AGI32 and Radiance. With conventional luminaires, you insert the lamp lumens into the luminaire data.

The software (taking account of the LOR) converts this into an intensity distribution for the luminaire and, hence, calculates the illuminance values in the space.

An LED luminaire is no different except no LOR conversion is necessary because losses have already been taken in to account. That said, what you must never do is plug in your own lumen value for a different LED chip and assume it will translate into an equivalent output from the luminaire.

The term LOR does not have any real meaning when applied to LED luminaires. The only way to calculate the output from an LED luminaire is to use the photometric data for the exact LED and fitting you wish to use.

Do not be tempted to substitute another LED chip or smaller style of luminaire and expect a similar result.


Ceramic metal halide (CMH) lamps combine high output with great colour rendering and good efficiency. They produce light by passing an electric arc through a mixture of gases. In a metal halide lamp, the compact arc tube contains a high-pressure mixture of argon, mercury, and a variety of metal halides.

The doctor will see you now…

Lux magazine’s resident technical boffin takes the fear out of lighting technology by answering the questions you are afraid to ask