The End of Flicker
If it’s not visible, it’s not a problem, right? Non-visible flicker is a common byproduct of LED circuit designs, and it can be hazardous to health, safety, and work performance.
Lighting manufacturers should be aware of some upcoming standards changes—including from the DLC—that could impact their ability to sell products with flicker in the near future.
Flicker is not a new problem. At the 1972 Olympic Games in Munich, many photographers noticed that their pictures were vastly underexposed or pitch black. This is because all the HID stadium lights were connected to the same AC power phase, and they briefly turned off 50 times per second.
Single-phase sports lighting could also cause fast-moving objects to appear as stationary or slow-moving. As a result, HID stadium lighting is now often staggered to all three AC phases, which temporally offsets flicker such that there are no periods of pitch-black light.
Fluorescent lighting was also prone to migraine-inducing flicker, but this issue was largely resolved with the advent of electronic ballasts. These new ballasts operate at very high frequencies, which are considered both imperceptible and safe.
LEDs present a new challenge regarding flicker, as trade-offs are made between cost and performance.
Why do LED lights flicker?
An easy way for LED lighting manufacturers to reduce the size and cost of products is to modulate the incoming AC signal as little as is necessary to drive the LEDs. As a result, many low-cost lighting products flash on and off at 100–120 Hz, twice the AC line frequency (50–60 Hz).
In addition, many drivers cause LEDs to flicker when dimmed. This is either due to poor compatibility with an AC line dimmer or due to the method of dimming (such as PWM). When dimmed, these products could flicker in the range of 100 Hz to over 3 kHz. Note that in this range, flicker is often called the stroboscopic effect.
Purely constant current drivers, which modulate current and voltage in order to dim, are immune to this problem. However, be aware that a number of constant current drivers implement a “hybrid” strategy and still cause flicker under certain conditions.
Is all flicker problematic?
Most lighting products are outside the flicker range that can induce photosensitive epileptic seizures (up to 70 Hz), but this does not automatically mean they are not hazardous.
Stroboscopic effects above 100 Hz are not directly visible to most people. However, rapid head or eye movements, along with rapid movement of objects, can cause the conscious or unconscious perception of flicker up to about 2 kHz. Beyond the fact that it is annoying, high-frequency flicker (100 Hz to 2 kHz) can cause eyestrain, fatigue, and headaches.
Perhaps the most hazardous aspect of flicker is that it could cause fast-moving objects to appear at a standstill or slower than the actual speed of travel.
Keep in mind that many lighting products flicker in the range of 100 Hz to 2 kHz.
Perhaps the most hazardous aspect of flicker is that it could cause fast-moving objects to appear at a standstill or slower than the actual speed of travel. This could be dangerous in environments with rotating machinery, which might appear safe to touch even at high speed, leading to injuries. It could also be problematic on roadways, where is is harder to judge the actual speed of travel of external objects.
As the flicker frequency increases, it becomes less of a problem. Lighting products with 100% flicker at 1.25 kHz or higher are considered low-risk, and those with 100% flicker at 3 kHz or higher are currently regarded as completely safe.
New flicker guidelines
In late 2015, IEEE published the first widely-accepted lighting flicker recommendations, largely based on the work of Arnold J Wilkins of the University of Essex. IEEE 1789–2015 is aimed at quantifying lighting flicker and the associated health and safety risks.
I won’t dive into the specifics of IEEE 1789 in this article. However, you should know that it is not only concerned with flicker frequency, but also flicker index and modulation. The IEEE recommendations define high-risk, low-risk, and risk-free flicker characteristics based on these three terms.
Before IEEE 1789, some forward-thinking manufacturers and scientists relied on the 2012 LRC ASSIST flicker recommendations. It’s hard to say why there was a general lack of awareness of these guidelines, but one can guess they got lost in the LED hype of the time.
In 2017, NEMA published its own guidelines. NEMA 77–2017 employs two different metrics, one for flicker below 80 Hz and the other for stroboscopic effects above 80 Hz. The latter is claimed to closely model human eye sensitivity to stroboscopic effects.
NEMA 77 provides different flicker thresholds for indoor and outdoor lighting. In addition, it is concerned with the flicker effects of pairing LED lighting and dimmers. Lamps and dimmers that meet certain requirements can display a compatibility logo on the packaging.
Flicker’s final days: Title 24 and the DLC
California is historically a progressive state in terms of energy regulation. But 2016’s Title 24 was unprecedented—it placed a higher emphasis on quality of light than on lighting energy savings.
Most are aware that the California Energy Commission’s Title 24 requires many light sources to have a CRI of 90 or higher. But another important requirement is flicker. Based on the requirements laid out in Joint Appendix 10, high efficacy lighting must have less than 30% flicker at frequencies below 200 Hz. There are other requirements for frequencies up to 400 Hz.
In addition, California Title 20 requires light sources to be tested for flicker after pairing with controls.
The DLC is currently researching a flicker requirement for inclusion in an upcoming version of the standard.
ENERGY STAR published specific flicker measurement recommendations in 2015. Flicker index and percentage should be measured for all products, and the testing must extend to five different dimmers for dimmable lamps. Although this is a recommended practice and not a requirement at this time, a future ENERGY STAR revision will likely require flicker testing for a wider array of lighting products.
Finally, the DLC is currently researching a flicker requirement for inclusion in an upcoming version of the standard. Lighting OEMs should take note because there is a lot of DLC-compliant lighting today that flickers, especially products targeted at cost-sensitive markets (like ESCOs). For the health and safety reasons listed previously, I am strongly in favor of a DLC flicker requirement.
Your head start opportunity
Although California is the only place where a flicker requirement is currently written in to law, it is likely that the DLC will have a flicker requirement within the next 12–18 months.
Considering many manufacturers rely on DLC compliance to drive sales, if your products produce unacceptable flicker, it would be wise to get a head start and work toward reduction before the new standard lands.
While the DLC’s specific requirements are not yet finalized, I recommend to adhere to the flicker guidelines of IEEE 1789–2015 and California Title 24 JA10. Any product that complies with these recommendations should meet or exceed what the DLC comes out with.
I am very passionate about the human-centric lighting space and the work toward reducing unacceptable flicker.