Flicker and IEEE PAR1789

Flicker and IEEE PAR1789 Recommended Practices of Modulating Current in High Brightness s for Mitigating Health Risks to Viewers Chair IEEE PAR1789 Brad Lehman (speaker) Northeastern University lehman@ece.neu.edu Vice Chair IEEE PAR1789 Michael Poplawski Pacific Northwest National Lab Htttp://grouper.ieee.org/groups/1789/

Disclaimer The views and opinions of this presentation do not represent views of the IEEE or IEEE PAR 1789 working group, but only reflect opinions of the presenter 3/25/2014 2

Presentation outline IEEE PAR1789 - purpose Introduction / Motivation drivers and flicker Ongoing results and working documents Conclusions 3/25/2014 3

IEEE PAR 1789 - PURPOSE Vision: Bring together a community of lighting environmental psychologists, medical researchers, lamp designers, driver designers, and lamp users to openly discuss concerns for lighting. There is a need to create a community where experts among the above different fields can communicate. Suggest a recommended practice, not a standard. Representation on IEEE P1789 from ENERGY STAR, CIE and NEMA may later incorporate findings into standards if deemed necessary. IEEE Standards Association has a unique open process that MUST involve all interest groups including academics, national labs, industry, customers (current membership is ~50 with around 25% academics, 25% government labs, 50% industry/consulting) International participation from members and from standards groups

IEEE PAR 1789 - PURPOSE Describe some possible health risks, such as headaches, eye strain and epileptic seizure, associated with low frequency modulation of High Brightness s in different applications Provide recommended practices to aid design of driving systems to modulate at safe frequencies for their particular applications in order to protect against the described health risks. http://grouper.ieee.org/groups/1789/

Introduction and Motivation Flicker, flutter, shimmer Repetitive change in magnitude over time, or modulation, of the luminous flux or luminance of a light source Light output modulation Visible vs. invisible, sensation vs. perception Visible flicker = Light output modulation is sensed and perceived Invisible flicker = Light output modulation is sensed, but not perceived Sensation: external conditions are detected and neurons respond Perception: the brain detects AND the mouth can report it sees

Light source modulation 60W A19 T12 A19 CFL A-lamp/G-lamp R30/PAR30 R38/PAR38

Vision: The Brain Source: The IESNA Lighting Handbook: Reference & Application (9th Ed.), 2000, p. 3-2 There is good evidence that fluctuations in the light signal are detected by the nervous system up to perhaps 200 Hz: Neurons in the lateral geniculate nucleus fire in phase with the fluctuations in the signal and retinal responses in the range of 120-200 Hz have been observed with varying methods. Source: http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/veitch_flicker_philly2010.pdf

Definitions Sensation Perception Critical flicker fusion Critical flicker frequency] CFF Source: The IESNA Lighting Handbook: Reference & Application (9th Ed.), 2000, p. 3-20 Health - World Health Organization (1947): a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity Source: http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/veitch_flicker_philly2010.pdf

Eye Saccade Eye in motion from (e.g. left to right) more sensitive to flicker Experiment: CRT with flickering bar (vs.) constant illuminating bar Above what frequency is image same? Implication to tail-lights (worst case scenario here) Experiment designed by Jane Roberts and A. Wilkins, Univ. Essex, gives worst case upper bound of perceptible flicker

Experimental Setup Which do you see? Participants view flickering line in the dark and make eye saccade

Effects 1 Photosensitive epilepsy Short exposure to 3 70 Hz flicker (i.e., visible modulation) may cause seizures in sensitive people Also static repetitive geometric patterns 1 in ~10,000 people Onset around puberty; 75% remain sensitive for life Source: http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/veitch_flicker_philly2010.pdf

Effects 2 Malaise: headache and eyestrain Slower onset, to frequencies in range 100-120 Hz have been demonstrated Exact population frequency isn't known; not everyone is affected Source: http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/veitch_flicker_philly2010.pdf

Effects 3 Visual performance Longer exposures to 100-120 Hz modulation, (i.e., not perceived as flicker) have been shown to reduce group average performance on visual tasks, both when viewed on paper and on CRT screens. Source: Veitch, J. A., & Newsham, G. R. (1998). Lighting quality and energy-efficiency effects on task performance, mood, health, satisfaction and comfort. Journal of the Illuminating Engineering Society, 27(1), 107-129. Source: http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/veitch_flicker_philly2010.pdf

Drivers and Flicker What makes s different? Why the concern about flicker?

Drivers and Flicker: The Concern? AC Powered Lighting System SOURCE AC source DRIVER AC-DC Converter LIGHT STRINGS OF S LIGHT FIXTURE 1. AC-DC converters often have 120Hz harmonics (flicker) in their current. How much is acceptable? (120Hz = twice the line frequency, which would be 100Hz in Europe.) 2. It is possible to eliminate AC-DC converter using a few special techniques: Reduce costs, eliminate capacitors, smaller size, increased lifetime. But this gives 100Hz/120Hz flicker.

Typical Dual-Stage driver

PFC Architecture Trade-Off Dual-stage PFC Near perfect PFC possible With proper control and component design, flicker can also be kept minimal BUT: Boost stage adds components and cost Single-Stage with PFC Modulating the input impedance improves PF BUT: Flicker at twice line frequency (100%) usually remains (Some have proposed combining passive pfc with single stage power converter to impact both pf and flicker. ) It Comes Down to Flicker Performance 3/25/2014 18

Several Different Alternative Approaches with Low Frequency Modulated Current (Described from publicly available documents) 1. Rectify AC and send to string I_ 1 AC 50-60Hz string 2 - + 4 R1 BRIDGE 3 2. Directly power two strings with opposite Anode/Cathode connections I_ AC 50-60Hz I_ R2 Or a capacitor Luminous Flux (periodic every 1/120 sec) is proportional to current

3 1 Failures may cause 60 Hz flicker: Open circuit in rectifier or in string (a) Rectify AC and send to string I_ AC 50-60Hz 2 - + 4 string R1 BRIDGE (b) Directly power two strings with opposite Anode/Cathode connections AC 50-60Hz I_ I_ R2 Or a capacitor (c) Simulation of current through HB s. Luminance is proportional to current, causing lamp to flicker at the AC mains line frequency (shown periodic every 1/60 sec) IEEE PAR1789 members are suggesting to consider simple shutdown alarms circuits so that the lamp never flickers at 60 Hz

PWM Dimming Used with either AC Mains or DC power as source Can increase or induce flicker Vdimmer I_ Driver string (1/T=f=100~120Hz)

PWM Dimming Example 3/25/2014 Switch 100% dimmer ~75% dimmer ~50% dimmer ~25% dimmer ~0% dimmer 22

Status-Timelines MAJOR Developments : 3 rd Revision Hazard Analysis Draft Report (30 pages). Has been released to general PAR1789 membership Gives low-risk levels for flicker Report development led by same authors that developed the European Union Commission s policy on consumer product recall Writing draft recommended practice report (~ 50 pages written) Portions of recommended practices have been proposed by members already Visible flicker to be avoided No effects levels understood Low risk levels proposed (controversial) Discussions on the 70Hz 1000Hz range for flicker recommended practices still to be formalized in writing Another Step Soon: Opening of Ballot 23

Risk Assessment Table 2 Risk Levels Risk Level Low Color code Medium Serious High Figure 1 Risk Matrix by Hazard. Greater opacity corresponds to greater certainty G, Ryder, R. Altkorn, X. Chen, JA Veitch, M. Poplawski, Safety 2012, the 11th World Conference on Injury Prevention and Safety Promotion

Discussing Possibility of New Flicker Metric Higher frequencies are filtered out by the LPF effects of the eye; Why not use Fourier Series and define measurements of flicker based of frequencies of interest below some f threshold? x( t) = Xavg + m= 1 c m cos( mωt + φ ) Xtrunc( t) = Xavg + c cos( ω t + φ ) + c2 cos(2ωt + φ2) +... + c cos( nωt + φ ) 1 1 n n m Consider the truncated Fourier Series representation of x(t) represented by Xtrunct(t) with n terms (n*f < f threshold, where f = 1/T is the frequency of signal x(t)). See IEEE ECCE paper (2010, Lehman et al).

Conclusions IEEE PAR1789 Committee intends to provide a recommended practice for how to apply this information Designers need to understand and apply the new flicker metrics and risk matrix Third-party testing may be needed for interaction of dimmer and products

Recent Flicker Measurement Work Methods No standard measurement procedure Metrics Percent Flicker (does not account for duty cycle, frequency) Flicker Index (does not account for frequency) TBD (accounts for modulation depth, duty cycle, frequency) IEEE PAR1789 Flicker Measurement Subcommittee Looking for others interested in collaborating on the development of, reviewing, and/or sharing test procedures, data Contact Michael Poplawski (michael.poplawski@pnnl.gov) Flickering Light Source Photodiode Transimpedance Amplifier Oscilloscope

PNNL Flicker Measurement Yokogawa WT500 Power Meter Test Chamber UDT TRAMP Amplifier Tektronix DPO 2014 Dimmer Computer Running LabVIEW Chroma 61502 Power Supply UDT Photo Detector