Flicker and Stroboscopic Visibility Measure (SVM) to be introduced by the revised Ecodesign
Regulation for light sources by the end of 2019.
Evaluation and measurements of optical flicker is a new task for electronic developers who provide solutions for LED lighting applications. When new drivers, power supplies or dimming electronics is purchased or designed, engineers need to consider LED lighting performance parameters including Temporary Light Artefacts (TLA) i.e. flicker and stroboscopic effects. The challenge is that up until now there was no regulation or agreed standard describing acceptable or minimum flicker levels. Soon this situation may change because the European Commission is currently working on revision of the Ecodesign and Energy Labelling regulations on light sources including LEDs.
Published in LED Lighting Magazine
by Miko Przybyla and Pawel Czarnecki GL Optic
Updated on 4 OCT 2019 | On 1 October 2019 the European Commission adopted 10 ecodesign implementing regulations, setting out energy efficiency and other requirements for product groups including LEDs. more
Updated on 8 Sep 2020 I At the end of 2021, the new Commission Regulation EU 2019 2020 (Ecodesign) enters into force. The specific requirements for LED light sources are changing. Completely new minimum requirements for flicker and the so-called stroboscopic effect are introduced. Read more about it HERE
Future regulation will most likely include minimum requirements on flicker. Should this change be introduced by the end of 2019 and it can come into force in 2021. Member states and national lighting associations are discussing what will be the recommended level and appropriate metric to evaluate flicker. At present the Stroboscopic Visibility Measure (SVM) is introduced in the Ecodesign Regulation for light sources by the Member States (Regulatory Committee) with a limit of ≤ 0.4.
Including minimum requirements for flicker in the European standard will be the first international legal document after California Title24 to recognize flicker parameter and setting the standard level. Should a limit of ≤ 0.4 be officially agreed it could mean that major part of vailable LED mains connected lighting products will not comply with the standard.
Before these changes are introduced in Europe let’s take a closer look at basic flicker metrics like flicker index, flicker percent and frequency once again to make sure we know how to measure these values and what are the available instruments and suggested measurement methods. Additionally it is important for all lighting professionals to understand the use and the meaning the new Stroboscopic Visibility Measure (SVM) a metric developed at Philips Research. In this article we will also presents details of measurement systems and discuss ways to eliminate flicker effect.
Light flicker in LED lighting systems
In addition to the spectral power distribution of the light and its intensity, light flicker is one of the factors influencing the visual comfort in the workplace. Light flicker is defined as a rapid, periodic change in light intensity. Under these conditions, most people are unable to see a light flicker higher than 80 Hz, which is considered to be the lower limit of visible frequency.
Long-term light flicker can cause of visual discomfort at work. It can cause headaches, migraines and epileptic seizures. In an industrial environment it can provoke accidents involving people, because the stroboscopic effect caused by light flicker leads to a disturbance in the perception of the speed of rotating objects, e.g. machine parts. Flicker is highly undesirable in sports facilities, concerts, where high quality film cameras are used, often with the possibility of fast-frame recording (60 frames per second or more) when the effect of changing the intensity of light becomes visible. Peripheral vision is more sensitive to the flicker effect, which can disrupt the concentration of the vehicle’s driver and redirect attention to the light source, thus posing a risk to road safety.
The rapid development of lighting technology based on LEDs has made this issue more and more relevant due to the very fast response time of this type of a semiconductor light sources at the level of a few dozen, or several dozen ns. In incandescent bulbs, the reaction time, and thus also the heart rate, was limited by the large thermal inertia of the fibre. Therefore flicker is more likely to be visible in LED lighting installations.
Photometric devices consisting of a high performance photodetection system (photodiode with trans-impedance amplifier) supplemented by an optical V-lambda filter allowing the sensitivity curve of the system to be adjusted to the characteristics of the human eye shall be used to measure the light flicker parameters of the light source (Figure 1).
Basic flicker parameters
When characterizing light flicker, one of the basic issues is to determine the light flicker frequency of the light source.
Figure 3. a) LED E27 bulb, 12 W 230 V AC, b) multi LED module supplied with 230 V AC. The measurements were made with the GL SPECTIS 1.0T Flicker with a sampling frequency of 125 kHz.
In tungsten halogen or fluorescent lighting, the flicker frequency is usually twice the frequency of the power grid, so in the European Union it is 100 Hz, and in the USA, for example, it will be 120 Hz. Since there are often additional sources of flicker when measurements are made outside the laboratory, a fast Fourier Transform (FFT) is used in the measurement systems to analyse the signal in order to identify all the components present in the signal.
It is also important to be able to see the registered signal on a time chart, which enables the evaluation of the shape of the signal (sinusoidal, rectangular, triangular), modulation or frequency.
The most important measurements from the point of view of an engineer who evaluates the flicker of LED light are based on the measurement of signal amplitude (flicker percent) and shape (flicker index). Figure 2 shows the sinusoidal change of light intensity (photodiode current) over time. The chart shows the most important amplitudes and the areas under the curve used to determine the
Flicker Percent (FP) is the simplest measure of flicker based on maximum and minimum flicker amplitude. If the waveform changes (while maintaining the values of the flicker amplitudes), e.g. from sinusoidal to rectangular, or if a straight wave is filled in, the value of the flicker factor remains the same. The range of possible values ranges from 0% (with no flicker) to 100%. IEEE 1789-2015 recommendations suggest flicker percent values <0.08 times the flicker frequency (8% at 100 Hz). This is a low-risk level. The second, higher criterion assumes that the flicker level will be in the NOEL (no observable effect level) and will not exceed 0.0333 x flicker frequency (3% for 100 Hz).
FP = maximum – minimum x 100%
maximum + minimum
• maximum – maximum signal amplitude,
• minimum – minimum signal amplitude.
Flicker Index (FI) is a measure of flicker based on the determination of integral under the curve. In contrast to the previous measure, the flicker index takes into account not only the amplitude of the waveform, but also its shape and filling. The value of the measurement may range from 0 (in case of lack of flicker) to 1. The American organization IES recommends that light sources should not exceed the value of flicker index=0.1.
FI = A
A + B
– A – integral under the curve to the level of the mean value,
– B – integral under the curve from the level of the mean value to zero.
Putting all flicker parameters together
Since flicker can be described in many different ways with the use of specific parameters we might want to know which lighting product has better optical performance parameters and which ones are worse. In this manner a new SVM metric can become helpful because it is considering frequency, modulation type and the modulation depth.
Stroboscopic Visibility Measure (SVM) is a new measure of the probability of stroboscopic effect. It was developed by a group of scientists from Philips Research taking into account the visual perception of different frequencies of the light source flicker. A value equal to or greater than 1 means that the stroboscopic effect will be visible to the observer. Unlike flicker index and flicker percent measures, SVM takes into account the different sensitivity of the human imaging system (eye and brain) with respect to flicker frequency.
The SVM parameter corresponds to the sum of the Fourier Cm transform components divided by Tm, understood as the detection threshold of the corresponding frequency for sinusoidal modulation flicker.
The measures discussed in the article refer to the actual optical measurements of the intensity of the tested light source. It should be noted that the so-called flicker meters built on the basis of the IEC 6100-4-15 standard, i.e. devices for short-term measurements (Pst index – 10 minutes) and long-term measurements (Plt index – 2 hours) of fluctuations in power network voltage (120 V or 230 V, 50 or 60 Hz) caused by time-varying reactive power (Blindleistung) of disturbing receivers, are referring electromagnetic compatibility test and not the optical flicker performance. This standard assumes measurement of network voltage rather than directly measuring the flicker of light source intensity,
simulation of the eye and brain system and the use of incandescent light sources.
How to measure?
Most often manufacturers want to specify the flicker parameters of LED modules and lamps in laboratory conditions in order to include these parameters in the data sheet. Such a test should be performed in a dark-room, where there are no other light sources that could interfere with the measurement. A measuring instrument directed towards the light source should be mounted on an optical bench or tripod, because hand vibrations can cause additional low frequencies in the signal. The PN semiconductor connector of an LED has a negative temperature coefficient, so the current flowing through the LED will increase as it heats up, so you should wait a few, several minutes to obtain stable temperature conditions of the module, which will allow you to complimentarily determine the flicker parameters of the light source.
Another method to measure flicker would be to install flicker measuring device in an integrating sphere which might simplify the measurement process for engineers who not always have the possibility to use laboratory to make fast evaluations when checking components and newly designed circuits. An integrating sphere with appropriate measurement instruments and software can create a table top measurement station.
On site measurement of several light sources installed in the building for instance is a much more difficult task because of the superposition (mutual overlapping of waves) of changes in light intensity, which occurs then. A periodic audits of the installation can, however, reassure us about the actual lamp performance, and increased flicker may suggest damage to the power supply systems or changed thermal conditions of the LED module inside.
What are the available instruments?
Light flicker meters vary in measurement time range, sampling frequencies, maximum transmission frequency, and frequency resolution depending on the size of the FFT. Not all manufacturers follow the guidelines of light flicker testing organizations. High-end devices sample a signal with a high oversampling rate of 10x higher than the highest expected frequency of the measured signal, which is important for accurate reproduction of the flicker signal shape. When designing lamps for specific applications in production halls or sports facilities, we may also be interested in the stroboscopic effect, so the presence of this parameter in the calculations will be important in this case.
The United States Department of Energy (DOE), a US government agency that deals with energy efficiency and renewables, has reviewed equipment to measure the flicker of light sources. In its report it compared 7 portable measuring devices plus one application for smart phones. The devices were compared with a reference stationary system.
For this purpose, 12 different light sources were used, each with different flicker characteristics. The devices were compared both in terms of accuracy of measurement of the basic metrics of flicker: flicker index and flicker percent, as well as the correctness of detection of the dominant frequency in the signal. Some sources were measured in maximum brightness mode and in 30% brightness mode.
Only two European instruments including Gigahertz and the GL SPECTIS 1.0T Flicker from GL Optic were included in the comparison and were able to measure the SVM (stroboscopic visibility measure) The correctness of SVM measurement was also compared with a reference device.
We have to remember that the only component that has direct influence on flicker performance is the driver. The higher driver capacity the better filtering effect to eliminate the influence of power supplied form the mains. This results in higher complication of the design and higher costs. In order to select better components and optimize the product we need to consider flicker performance during the entire process of the development and we can use available metrics and instrumentation to help us deliver better LED lighting products. Soon we will need to be able to comply with the European standards I this field.