Content uploaded by Arnold J Wilkins
Author content
All content in this area was uploaded by Arnold J Wilkins on Mar 23, 2014
Content may be subject to copyright.
LED Lighting Flicker and Potential Health Concerns:
IEEE Standard PAR1789 Update
Arnold Wilkins
Dept. of Psychology
University of Essex; UK
Jennifer Veitch
National Research Council Canada
Ottawa; Canada
Brad Lehman
Dept. Elect. & Comp. Eng
Northeastern University; USA
Abstract -- The IEEE Standards Working Group, IEEE
PAR1789 "Recommending practices for modulating current in
High Brightness LEDs for mitigating health risks to viewers"
has been formed to advise the lighting industry, ANSI/NEMA,
IEC, EnergyStar and other standards groups about the
emerging concern of flicker in LED lighting. This paper
introduces power electronic designers for LED lighting to
health concerns relating to flicker, demonstrates that existing
technologies in LED lighting sometimes provide flicker at
frequencies that may induce biological human response, and
discusses a few methods to consider when trying to mitigate
unintentional biological effects of LED lighting. The paper
represents on-going work in IEEE PAR1789 that is vital to
designing safe LED lamp drivers.
Index Terms-- LED, health risk, flicker, lighting, power
electronics, ergonomics, drivers, headache, seizure, standards
I. NOMENCLATURE
Flicker: a rapid and repeated change over time in the
brightness of light.
Modulation: a measure of light variation that is often
applied to periodic oscillations. This report refers to
modulation as variation in luminance as a proportion of the
average luminance (commonly referred to as Percent Flicker,
Peak-to-Peak Contrast, Michelson Contrast, or Depth of
Modulation). For a time-varying luminance with maximum
and minimum values:
Modulation = (Lmax - Lmin) / (Lmax + Lmin)
(Lighting Design Glossary)
Visible Flicker: Flicker that is consciously perceivable by
a human viewer.
Invisible or Imperceptible Flicker: Flicker that is not
consciously perceivable by a human viewer.
The effects of flicker can range from decreased visual
performance to non-specific malaise to the onset of some
forms of epilepsy.
II. INTRODUCTION
This paper summarizes a public report created by the IEEE
Standards PAR1789 group on LED lighting that is examining
biological effects of flicker in emerging LED lighting
technologies. (The full length version of the report can be
found at http://grouper.ieee.org/groups/1789/ ) The intention
of this document is to provide an objective summary of the
effects on human health for both visible and invisible flicker
and to draw attention to implications for the design of LED
lighting. Specifically, contributions of this paper include
making the reader aware of
1. Risks of seizures due to flicker at frequencies within the
range ~3- ~70Hz;
2. Human biological effects due to invisible flicker at
frequencies below ~165Hz;
3. The differences between “visible” flicker and “invisible”
flicker and any relation to health risks;
4. A few, typical driving approaches in LED lighting that
may produce flicker.
This report does not attempt to make recommendations on
safe flicker frequencies or modulation depths for LED
lighting. Its purpose is to describe possible health
implications of flicker. By bringing these issues to the power
electronic and lamp designers now, it will permit better
ethical discussions and decisions to be made on development
of future LED lamps as the market continues to have
explosive growth. This report endorses no technology for
driving LED lamps. Specifically, Section III of the report
gives tutorial surveys on health effects of flicker. Section IV
introduces a few typical LED driving methods that introduce
flicker in frequency ranges of interest.
III. FLICKER
The health effects of flicker can be divided into those that
are the immediate result of a few seconds’ exposure, such as
epileptic seizures, and those that are the less obvious result of
long-term exposure, such as malaise, headaches and impaired
visual performance. The former are associated with visible
flicker, typically within the range ~3- ~70Hz, and the latter
with invisible modulation of light at frequencies above those
at which flicker is perceptible (invisible flicker). Human
biological effects are a function of flicker frequency,
modulation depth, brightness, lighting application, and
several other factors.
A. Photosensitive Epilepsy
About one in 4000 individuals is recognized as having
photosensitive epilepsy. Repetitive flashing lights and static
repetitive geometric patterns may induce seizures in these
individuals, and in perhaps as many again who have not been
diagnosed and may be unaware that they are at risk.
The seizures reflect the transient abnormal synchronized
activity of brain cells, affecting consciousness, body
movements and/or sensation. The onset of photosensitive
epilepsy occurs typically at around the time of puberty; in the
age group 7 to 20 years the condition is five times as common
as in the general population. Three quarters of patients remain
photosensitive for life (Harding and Jeavons, 1994; Wilkins,
1995; Fisher et al. 2005). Many factors [see Fisher et al.,
2005 for extensive reference list] may combine to affect the
likelihood of seizures including:
• Flash Frequency. Any repetitive change in a visual
stimulus within the frequency range 3 Hz to 70 Hz, is
potentially a risk and the greatest likelihood of seizures is for
frequencies in the range 15 Hz to 20 Hz, see Fig. 1. The
flashes do not have to be rhythmic.
• Brightness. Stimulation in the scotopic or low mesopic
range (below about 1 cd/m2) has a low risk and the risk
increases monotonically with log luminance in the high
mesopic and photopic range.
• Contrast with background lighting. Contrasts above
10% are potentially a risk.
• Distance between the viewer and the light source and its
location which determine
• Total area of the retina receiving stimulation. The
likelihood of seizures increases according to the
representation of the visual field within the visual cortex of
the brain. The cortical representation of central vision is
greater than that of the visual periphery, and so
• Location of stimulation within the visual field is
important: Stimuli presented in central vision pose more of a
risk than those that are viewed in the periphery, even though
flicker in the periphery may be more noticeable.
• Wavelength of the light. Deep red flicker and
alternating red and blue flashes may be particularly
hazardous.
• Open or closed eyes. Bright flicker can be more
hazardous when the eyes are closed, partly because the entire
retina is then stimulated. However, if flickering light is
prevented from reaching the retina of one eye by placing the
palm of a hand over that eye, the effects of the flicker are
very greatly reduced in most patients.
Figure 1. Percentage of patients with photosensitive epilepsy exhibiting
epileptiform EEG responses to the flicker from a xenon gas discharge lamp
shown as a function of flash frequency. After Harding and Jeavons (1994).
Figure 2. Escalator stair tread
In addition, a substantial minority of patients (usually
those who are sensitive to flicker) are sensitive also to spatial
patterns; see Fig. 2 for an example. About one third of
patients are sensitive to patterns even when there is no
flicker, and most are more sensitive to flicker if it is patterned
(Harding and Jeavons, 1994; Wilkins, 1995; Fisher et al.,
2005; Wilkins et al. 1979). The worst patterns are those of
stripes in which one cycle of the pattern (one pair of stripes)
subtends at the eye an angle of about 15 minutes of arc.
B. Imperceptible Flicker
The frequency of the alternating current electricity supply
is 60Hz in America and 50Hz in Europe; in Japan, both 50Hz
and 60Hz are used in different regions. The circuitry in older
fluorescent lamps with magnetic ballasts operate so as to
flash the lamps at twice the supply frequency (100Hz or
120Hz). However, as the lamps age, the flashes that occur
with one direction of current may not equal those that occur
with the other direction, and the lamps may emit flicker with
components at the frequency of the electricity supply. It has
been determined that photosensitive seizures should not occur
if fluorescent lamps are operating properly. However, when
the lamps malfunction giving flicker below 70Hz,
electroencephalographic recordings indicate a risk of seizures
(Binnie et al., 1979). Nevertheless some photosensitive
patients do complain of normally functioning (older)
fluorescent lighting (with magnetic ballasts).
Measurements of the electroretinogram have indicated that
modulation of light in the frequency range 100-160Hz and
even up to 200 Hz is resolved by the human retina although
the flicker is too rapid to be seen (Burns et al. 1992, Berman
et al., 1991). In an animal (cat), 100Hz and 120Hz
modulation of light from fluorescent lamps has been shown
to cause the phase-locked firing of cells in the lateral
geniculate nucleus of the thalamus, part of the brain with
short neural chains to the superior colliculus, a body that
controls eye movements (Eysel and Burandt, 1984). There are
several studies showing that the characteristics of human eye
movements across text are affected by modulation from
fluorescent lamps and cathode ray tube displays (see work of
Wilkins, 1986; Kennedy and Murray, 1991), and two studies
have shown impairment of visual performance in tasks
involving visual search as a result of flicker from fluorescent
lamps (e.g. Jaen et al., 2005). Under double-masked
conditions the 100Hz modulation of light from fluorescent
lamps has been shown to double the average incidence of
headaches in office workers, although this effect is
attributable to a minority that is particularly affected (Wilkins
et al., 1989).
Sensitivity effects due to flicker at frequencies above
perception have also been observed in normal people with
good vision and health. A substantial decrement in sensitivity
to visible flicker at 30 Hz, used as a testing condition, occurs
in normal observers when there is a prior exposure of only 2
minutes duration with flicker frequencies about 20% above
the observers critical fusion frequency (CFF) (Shady et al.
2004).
Computer monitors and backlights
When making a rapid jerk (saccade), for example when
reading, the eyes move at a velocity of about 180 degrees per
second. As a result, any intermittently lit contour is displaced
at a succession of retinal positions during the flight of the eye
and can sometimes be seen as a set of repetitive targets. The
LED rear lamps of motor vehicles can produce such an effect.
Some displays on netbook computers have LED backlights
and exhibit significant flicker at 60Hz. Their flicker also
results in the perception of multiple images during a saccade.
It is possible that this effect is responsible for the known
disturbance of ocular motor control by high frequency flicker,
a disturbance which, in its turn, may be responsible for the
known impairments in visual performance.
Modulation depth and the Fourier fundamental.
The effects of flicker depend not only on the frequency of
the flicker but also on the modulation depth. For visible
flicker, the amplitude of the Fourier fundamental predicts
flicker fusion (de Lange Dzn, 1961). For imperceptible
flicker the effects of different waveforms have not been
studied in detail. The peak-trough modulation depth of the
100-120Hz flicker from older fluorescent lamps with
magnetic ballasts varies with the component phosphors, some
of which exhibit persistence, varying the chromaticity of the
light through its cycle (Wilkins and Clark, 1990). The peak-
trough modulation depth known to induce headaches from
fluorescent lighting at 100Hz is about 35% (Wilkins et al.,
1989). The IEEE Standards PAR1789 working group is
developing new measures/definitions of flicker that rely on
Fourier series of the flicker. The present definitions for
modulation do not distinguish the difference between low
frequency and high frequency modulation. But for
sufficiently high flicker frequencies, there are limited human
biological effects.
C. Summary of Biological Effects
The obvious biological effects occur: 1) immediately from
2) flicker that is visible. The risks include seizures, and less
specific neurological symptoms including malaise and
headache. Seizures can be triggered by flicker in individuals
with no previous history or diagnosis of epilepsy.
The less obvious biological effects occur: 1) from flicker
that is invisible 2) after exposure of several minutes. Invisible
flicker health effects have been reported to include headaches
and eye-strain.
The table in the Appendix summarizes and categorizes the
types of flicker and the biological effects they cause.
(c) Simulation of current through HB LEDs.
Luminous intensity is proportional to current,
causing lamp to flicker at twice the line frequency
(shown periodic every 1/120 sec)
I_LED
LED LED LED
R1
-+
BRIDGE
1
4
3
2
LED string
AC 50-60Hz
LED
I_LED
LED LED LED
R2
AC 50-60Hz
LED
I_LED
LED
LED
LED
LED
(a) Rectify AC and send to LED string
(b) Directly power two LED strings with opposite
Anode/Cathode connections
Or a
capacitor
Figure 3. Two methods to drive LEDs at twice line frequency: (a) Full bridge
rectification, (b) Opposite connected parallel strings, and (c) Current/Luminous Output in
D. A Few General Implications for Practice
Visual flicker is an undesirable attribute of any lighting
system. The Appendix Table summarizes research suggesting
that, for both visible and invisible flicker (in the mentioned
frequency ranges), there may be a special at-risk population
for which flicker is more than just annoying and can be a
potential health hazard. Any hazard will, however, depend
on modulation depth, ergonomics, flicker parameters and
their relation to perception and the ability to
measure/determine the influence of these parameters with
human diagnostics. These topics are beyond the scope of this
paper and will be covered in future IEEE PAR1789
documents. However, it is possible to make general
comments about the research cited in the Appendix Table:
1. Frequency. Normally functioning fluorescent lighting
controlled by magnetic ballasts has been associated with
headaches due to the flicker produced. LEDs driven so that
they flicker at a frequency twice that of the AC supply may
have a depth of modulation greater than that from most
fluorescent lamps. The effects of the flicker are therefore
likely to be more pronounced in these cases.
2. Field of view. Point sources of light are less likely to
induce seizures and headaches than a diffuse source of light
that covers most of a person’s field of vision. Flicker from
LEDs used for general lighting is therefore more likely to be
a health hazard than that from LEDs used in instrument
panels.
3. Visual task. The invisible flicker described in the
Appendix Table is more likely to cause problems when the
visual task demands precise positioning of the eyes, as when
reading.
4. Spatial distribution of point sources of light. Spatial
arrays of continuously illuminated point sources of light have
the potential to induce seizures in patients with photosensitive
epilepsy when the field of view is large and when the arrays
provide a spatial frequency close to 3 cycles/degree (e.g.
large LED public display boards viewed from close
proximity).
IV. TYPICAL LED DRIVING METHODS IN LOW FLICKER
FREQUENCY RANGE
There are several common methods that are used to drive
LEDs that can operate with frequency of modulation in the
ranges discussed in the above table (below 120Hz, including
frequencies in the vicinity of 15Hz.) For example,
commercially available LED lamps have been reported (Rand
et al., 2007; Rand, 2005) to malfunction and produce visual
flicker in the 15Hz range when connected to a conventional
residential dimmer.
Below, we present only a few driving approaches that
modulate in frequency ranges from zero to 120Hz. The list is
not exhaustive, and the discussions are only meant to
demonstrate typical driving LED currents with frequencies in
this range.
A. LED Driving Current Frequencies in Range: ~100Hz–
120Hz
(1) Full Wave Rectifier Connected to LED String
In this approach, the AC input source is sent into a full
wave rectifier, causing the (approximate) absolute value of
the input voltage to be sent to the load. In this case, the
current through the LEDs has a waveform shape similar to a
scaled absolute value of a sine wave. That is, the rectified
sine wave may be of the form, where Vp is
wave and ω is the
the amplitude of the sine
angular frequency in radians ω = 2*π*f. In this case, the LED
current is of similar shape, as Fig. 3 shows. In a first
approximation, the LED current is equal to a scaled rectified
voltage, with the additional deadtime (zero current) caused by
the LED bias voltage. Thus, when properly functioning, the
direct full wave rectifier driving approach modulates the
LEDs at twice the line frequency, which in North America
leads to 120Hz modulation and in Europe leads to 100 Hz
modulation. As Fig. 3(a) shows, often a resistor is added in
series with the LED string for current limiting protection.
(2) Directly Drive Two Parallel LED Strings with
Opposite Anode/Cathode Connections
A second LED driving method that doubles line frequency
is shown in Fig. 3(b). Two strings of LEDs are powered in
parallel, with anode of one paralleled string connected to the
cathode of the other parallel string. When the AC line voltage
is positive, energy drives one of the LED strings. When the
AC line voltage is negative, the other paralleled LED string is
driven. At most, one of the LED strings has current through
it. The net effect is that the effective LED driving current is
modulating at 120 Hz in North America or 100 Hz in Europe.
Thus, for both driving methods illustrated in Fig. 3, the
LED current modulates at twice the line frequency. Since the
intensity of the LEDs is (ideally) proportional to the current
through the LEDs, this causes the LEDs to flicker at
frequency equal to twice the AC line frequency, i.e.
100Hz~120Hz. There are many variations of the approach in
Fig. 3 that are not shown here.
(3) Simple Dimming Pulse Width Modulated (PWM)
Circuits
It is common to dim LEDs by pulsing the current through
them intentionally.The luminous intensity of the LED can be
adjusted by varying the length of time that the LED current is
High or Low,. Thus, PWM dimming circuits may be designed
to operate at any frequency, whether the input is DC or AC.
(It should be noted that it is not uncommon for LED drivers
using AC residential phase modulated dimmer circuits to
attempt to “emulate” PWM type signals with frequency
120Hz. That is, when the AC dimmer shuts off, no current is
sent to the LEDs.)
It should be mentioned that there are alternative
approaches to dimming, such as amplitude dimming, in
which the current through the LED is continuous and not
|)sin(| tVp
ω
pulsing. By reducing the value of this continuous current
(amplitude), the brightness is dimmed. This approach does
not use flicker to adjust brightness and therefore, should not
induce flicker related health risks.
(4) Power Factor Correction Circuitry
Even when sophisticated high frequency switching power
supplies with power factor correction circuits are used to
drive LEDs from AC mains, there is commonly a frequency
component in the current (and luminous intensity) of the
LEDs at twice the line frequency. Depending on the design of
the circuitry, the harmonic content of this flicker may vary
from being small and unnoticeable to being significant in
magnitude.
B. LED Driving Current Frequencies in Range: 3Hz~70Hz
(1) Failures in rectification or LED strings: 50Hz ~ 60 Hz
Modulation
In either of the two methods of Fig. 3, there is risk of
failure that can cause LED current modulation at AC line
frequency, thereby entering the range of frequencies that may
induce photosensitive epilepsy. For example, if one of the
legs of the full wave rectifier bridge fails, then it is common
that the leg becomes an open circuit. Open circuits prevent
current flow, and therefore, the LED modulation frequency
may change. This single diode failure in the rectifier will
cause the output voltage for the full wave rectifier to become
the input voltage for half the AC line cycle, and then 0 volts
for the remaining half line cycle. This means that if the AC
Mains line frequency is f and the period is T=1/f, then non-
zero voltage is applied to the LEDs for 0.5*T seconds and
then is zero for the next 0.5*T seconds, causing the LED
current to modulate at line frequency.
Similarly, when the two strings of LEDs are connected in
parallel with opposite anodes and cathodes in each string, a
failure in one string of the LEDs may cause an open circuit to
occur in that string. The net effect is the same as before: the
current is modulating at line frequency, i.e. 50Hz ~ 60Hz.
This low frequency driving current leads to brightness flicker
in the LEDs at 50Hz~60Hz, since the current in the LEDs is
proportional to their intensity. This is in a range of
frequencies that are at risk of causing photosensitive epilepsy.
(2) Residential Dimmer Switches Can Cause Low
Frequency Flicker (~3Hz – 70Hz)
Residential dimmers for incandescent bulbs primarily
utilize phase modulating dimming through triac switches to
control the power sent to the bulb. These dimmers control the
RMS voltage applied to the bulb by suppressing part of the
AC line voltage using a triac. The effect is a chopped sine
wave as shown in Fig.4. Thus, as the dimmer switch is
manually adjusted, the value of the off-time, α (often referred
to as the phase delay) changes. As α is increased in Fig. 4,
less power goes to the incandescent bulb and brightness is
reduced
Some LED lamps and their associated drivers may not
perform properly with residential phase modulated dimmers.
Often on the LED bulb application notes or on the lamp’s
manufacturer web sites there are warnings to the user that
their bulbs may not work properly when used with residential
dimmer switches. Rand’s work (Rand et al., 2007; Rand,
2005) explains the causes of these failures and shows that low
frequency flicker may occur.
Dimmer Voltage
α
Figure 4. Residential dimmer and its output voltage sent to
the driver (Rand et al., 2007).
Fig. 5 illustrates how one type of commercially available
LED lamp flickers in the noticeable visual range when
connected to a dimmer switch. The particular lamp involved
has a common LED driver configuration (further discussed
below) of a full bridge rectifier with capacitor filter within
their Edison socket, described in more detail by Rand et al.,
2007; 2005). The results presented in the figure may be
typical of similar driving configurations. The circuit will
continuously peak charge the filter capacitor to the peak
voltage of the input waveform, i.e. 169Vdc for standard
120Vac line voltage. This high level DC voltage may then be
fed into a large string of LEDs in series. For example, some
typical lamps may have parallel strings of many Red, Blue,
Green LEDs, in series attached through a current limiting
resistor to the high level DC voltage. The particular lamp
tested utilized a combination of 64 Red, Green and Blue
LEDs to produce white light.
Figure 5. Commercial LED lamp flickers at 3.15Hz when
connected to typical residential dimmer switch.
The experimental data in Fig. 5 represents the voltage of a
photo-sensor placed directly underneath the LED lamp.
Specifically, a photoresistor circuit is used to generate a
voltage proportional to the light intensity shining on it. As the
experimental voltage shows, the bulb malfunctions when
connected to (phase modulated) residential dimmer switch. It
produces a noticeable visual flicker. In particular, the flicker
varies between around 3.0Hz and 3.3Hz, with average over
many cycles of 3.153Hz. This frequency is in the range that
has been shown to be a risk for causing photosensitive
epileptic seizures.
The flicker illustrated in Fig. 5 is typical of a few LED
lamps on the market when connected to a dimmer. However,
the precise flicker frequency is hard to predict, as it may
either be higher or lower depending on various factors such
as number of lamps on the dimmer, position of the dimmer
switch (the value of desired off-time α), and/or internal
characteristics of the lamp. However, as the experimental
oscilloscope plot shows, the flicker frequency may be in the
range that induces photosensitive seizures.
The reasons that the dimmer switch may fail when
connected to LED lamp bulbs are given in (Rand et al., 2007;
2005).
(3) Uneven Brightness in Different LED Strings When
Connected as in Fig. 3(b)- With Strings Having Opposite
Anode/Cathode Connections
Consider the circuit in Fig. 3(b). Notice that each LED
must have the same dynamic characteristics (forward voltage
and dynamic resistance) in order for the current to be
perfectly balanced in each alternating illuminated string. If
for some reason this does not occur (aging, temperature
gradients, poor design), then the current through the strings
will not be identical each cycle.
Figure 6. Unbalanced LED Current in Each String of
LEDs Using Driving Method in Fig. 3(b). The unbalanced
driving will cause uneven luminous output in the lamp and
low frequency flicker.
For example, suppose over time, aging causes degradation
of one of the two strings in Fig. 3(b) such that its string
resistance increases by 50%. This could also be caused by
improper design of each string in Fig. 3(b) so that the current
in each string is not balanced. This is quite possible since
LEDs are binned by different voltages, and further, each
string may be composed of different color LEDs that have
different nominal voltage drops for the same current. Then,
the effective LED current through the bulb will look as in
Fig.6.
For example, the effective DC LED current in the
numerical simulation of Fig. 6 has average value of around
233mA. However, the Fourier component at 60 Hz (taking
FFT) is 80mA and the Fourier component at 120Hz is nearly
240mA. Thus, in this example the low frequency component
of 60Hz represents over 33% of the DC component, while the
120 Hz component represents 100% of the DC current.
Higher frequency components of the LED current in the
above figure are also present in multiples of 60Hz. However,
the typical analysis above indicates that LED lamps may
demonstrate flicker frequency at line frequency, similar to
older fluorescent lamps (previously discussed) that aged
unevenly: the flashes/brightness with one direction of line
current may not equal those that occur in the other direction.
The above example also illustrates that it is possible for
flicker in a lamp to have harmonics with multiple low
frequency components, here at both 60Hz and 120Hz.
V. CONCLUSIONS
The purpose of this paper is to make lamp and power
electronic designers aware of biological effects of flicker and
to introduce the reader to a few LED driving methods that
will have flicker. The LED driving approaches described in
this paper are not exhaustive and are only meant to introduce
the reader to a few common approaches. Other
approaches/applications of LED lighting that may also have
flicker include, but are not limited to, pulse amplitude
modulation driving, triangle wave currents through LEDs,
using LED flicker for wireless communication (see IEEE
Standard 802), beat frequencies created through the
interaction of different lamp flicker frequencies, etc.
This paper assigns no health risk to the biological effects
of flicker in the various LED lamps. The hope is that by
discussing the issue of flicker within the power electronic
community, it will be possible to decide as a community
whether or not standards or recommended practices are
necessary. We do not attempt to do so here. However, we do
offer simple suggestions as to what should be considered
when designing lamps, such as flicker frequency, angle of
viewing, task being performed, spatial distribution, AC
dimmer flicker, etc. Further, it is not difficult to create shut-
down or other safety prevention circuits that prevent
flickering in the 3Hz-70Hz range when the lamp is in failure
mode. This is the flicker range that has risk of photosensitive
epilepsy for small minority of the population.
ACKNOWLEDGMENT
This paper was assembled with input from many members
of IEEE Standards PAR1789 who at the time of writing
consisted of (including observers):
Maurizio Acosta, Ian Ashdown, Rolf Bergman, Sam
Berman, Anindita Bhattacharya, Subramanyam Chamarti,
Clint Chaplin, Yvonne de Kort, Montu Doshi, Kevin
Dowling, John Halliwell, Mark Halpin, Steve Hayes, Chris
Horton, Wijnand IJsselsteijn, Michael Jennings, Jiao
Jianzhong, David Keeser, Faisal Khan, Thorbjorn Laike,
Brad Lehman, Ihor Lys, Theron Makley, Naomi Miller,
Lesley Murawski, Brandon Oakes, Yoshi Ohno, Steve
Paolini, Radu Pitigoi Aron, Michael Poplawski, Conor Quin,
Eric Richman, Anatoly Shteynberg, Michael Shur, Roger
Shuttleworth, David Sliney, Jennifer Veitch, Joachim
Walewski, Arnold Wilkins, Howard Wolfman, Wei Yan,
Regan Zane.
REFERENCES
Primary References for This Report are below. (Please refer to the IEEE
PAR1789 website for additional references
http://grouper.ieee.org/groups/1789/ that were also used. The web site will
continuously be updated.)
Berman, S.M., Greenhouse, D.S., Bailey, I.L., Clear, R.D., and Raasch,
T.W. (1991) Human electroretinogram responses to video displays,
fluorescent lighting, and other high frequency sources. Optom Vis Sci.,
68(8),645-62.
Binnie, C.D., de Korte, R.A., and Wisman, T. (1979) Fluorescent
lighting and epilepsy. Epilepsia,20,725-7.
Burns, S.A., Elsner, A.E., and Kreitz, M.R. (1992) Analysis of
nonlinearities in the flicker ERG. Optom Vis Sci.,69(2), 95-105.
de Lange Dzn, H. (1961) Eye's Response at Flicker Fusion to Square-
Wave Modulation of a Test Field Surrounded by a Large Steady Field of
Equal Mean Luminance. Journal of the Optical Society of America, 51(4),
415.
Evans J.E., Cuthill I.C., and Bennett A.T.D. (2006) Animal Behaviour,
72(2),393-400.
Eysel, U.T., Burandt, U. (1984) Fluorescent tube light evokes flicker
responses in visual neurons. Vision Res., 24(9),943-8.
Fisher, R., Harding, G.F.A., Erba, G., Barkley, G.L., and Wilkins, A,J.
(2005) Photic- and pattern-induced seizures: a review for the Epilepsy
Foundation of America Working Group. Epilepsia, 46(9),1426-41.
http://www.essex.ac.uk/psychology/overlays/2005-168.pdf
Funatsuka, M., Fujita, M., Shirakawa, S., Oguni, H., and Osawa, M.
(2003) Pediatr Neurol, 28(1),28-36.
Halpin, S.M., Bergeron, R., Blooming, T.M., Burch, R.F., Conrad, L.E.
and Key, T.S. (2003) Voltage and lamp flicerk issues: should the IEEE adopt
the IEC approach, IEEE Transactions on Power Delivery, 18(3), 1088-1097.
Harding, G.F.A. and Harding, P.F. (2008) Photosensitive epilepsy and
image safety. Appl Ergon, Oct 16.
Harding, G.F.A. and Jeavons, P. (1994) Photosensitive Epilepsy. Mac
Keith Press.
Hazell, J. and Wilkins, A.J. (1990) Psychological Medicine, 20,591-96.
Ishiguro, Y., Takada, H., Wantanabe, K., Okumura, A., Aso, K., and
Ishikawa, T. (2004) Follow-up survey on seizures induced by animated
cartoon TV program "Pocket Monster", Epilepsia, 45(4) 377-383.
Okumura, A., Aso, K., Ishikawa, T A (2004) Follow-up survey on
seizures induced by animated cartoon TV program "Pocket Monster",
Epilepsia, 45(4) 377-383.
Jaen, M., Sandoval, J., Colombo, E., and Troscianko, T. (2005) Office
workers visual performance and temporal modulation of fluorescent lighting,
Leukos, 1(4), 27-46.
Kelly, D.H. (1969) Diffusion model of linear flicker responses, Journal
of the Optical Society of America, 59(12), 1665-1670.
Kelly, D.H. (1971) Theory of flicker and transient responses, I. Uniform
Fields, Journal of the Optical Society of America, 61(4), 537-546.
Kennedy, A., Brysbaert, M., and Murray, W.S. (1998) Q J Exp Psychol
A., 51(1), 135-51.
Kennedy, A. and Murray, W.S. (1991) The effects of flicker on eye
movement control. Q J Exp Psychol A., 43(1),79-99.
Lighting Design Glossary
http://www.schorsch.com/kbase/glossary/contrast.html
Maddocks, S.A., Goldsmith, A.R., and Cuthill, I.C. (2001) General and
Comparative Endocrinology, 124(3), 315-20.
Rand, D. (2005) Off Line Dimming for High Brightness LEDs. MS
Project, Northeastern University, Boston, MA.
Rand, D., Lehman, B. , and Shteynberg, A. (2007) Issues, Models and
Solutions for Triac Modulated Phase Dimming of LED Lamps, Proc. IEEE
Power Electronics Specialists Conference.
Rea, M.s. and Ouellette, M.J. (1988) Table-tennis under high intensity
discharge (HID) lighting, Journal of the Illuminating Engineering Society,
Vol. 17, No.1, 29-35.
Shady, S., MacLeod , D.I.A. and Fisher , H.S. (2004) Adaptation from
invisible flicker. Proceedings of the National Academy of Sciences U.S.A.,
101, 5170-5173.
Veitch, J.A. and McColl, S.L. (1995) Lighting Res. Tech., 27(4),243-
256.
Wilkins, A.J. (1986) Intermittent illumination from visual display units
and fluorescent lighting affects movements of the eyes across text. Human
Factors, 28(1),75-81.
Wilkins, A.J. (1995) Visual Stress. Oxford University Press.
http://www.essex.ac.uk/psychology/overlays/book1.pdf
Wilkins AJ, Clark C. (1990) Modulation of light from fluorescent lamps.
Lighting Research and Technology, 22(2):103-109.
Wilkins, A.J., Darby, C.E. and Binnie, C.D. (1979) Neurophysiological
aspects of pattern-sensitive epilepsy. Brain 1979; 102:1-25.
http://www.essex.ac.uk/psychology/overlays/1979-15.pdf
Wilkins, A.J., Nimmo-Smith, I.M., Slater, A. and Bedocs, L. (1989)
Fluorescent lighting, headaches and eye-strain. Lighting Research and
Technology, 21(1), 11-18.
J. F. Fuller, E. F. Fuchs, and K. J. Roesler, "Influence of harmonics on
power distribution system protection," IEEE Trans. Power Delivery, vol. 3,
pp. 549-557, Apr. 1988.
arnol
d
Digitally signed by
arnold
DN: cn=arnold
Date: 2010.07.12
14:39:21 +01'00'
APPENDIX
TABLE I
FLICKER AND BIOLOGIC AL EFFECTS
Source of flicker Fre
q
uenc
y
ran
g
eBiolo
g
ical effect Evidence
Sunlight through roadside trees
or reflected from waves
Various Seizures Clinical histories (Harding and
Jeavons, 1994)
Xenon gas discharge photo-
stimulator
3-60Hz Epileptiform EEG in
patients with
photosensitive epilepsy
Many clinical EEG studies e.g
(Harding and Jeavons, 1994)
Malfunctioning
fluorescent lighting
Large 50Hz
component
Epileptiform EEG in
patients with
photosensitive epilepsy
(Binnie et al., 1979)
Television 50Hz and 60Hz
(discounting 25Hz
component)
Epileptiform EEG in
patients with
photosensitive epilepsy
Many studies eg (Harding and
Harding, 2008; Funatsuka et al.,
2003)
Flashing televised cartoo
n
~10H
z
Seizures in children
with no previous
diagnosis of epilepsy
Major incident (Okumura et al,
2004)
Normally functioning
fluorescent lighting (50Hz
ballast)
100Hz (small
50Hz component)
Headache and eye strain Many anecdotes.
Normally functioning
fluorescent lighting (50Hz
ballast)
100Hz (small
50Hz component)
Headache and eye strain Double-masked study (Wilkins et
al 1989)
Normally functioning
fluorescent lighting (50Hz
ballast)
32% modulatio
n
Reduced speed of visual
search
Two masked studies (Jaen et al.,
2005)
Normally functioning
fluorescent lighting (60Hz
ballast)
120H
z
Reduced visual
performance
(Veitch and McColl, 1995)
Normally functioning
fluorescent lighting (50Hz
ballast)
100Hz (minimal
50Hz component)
Increased heart rate in
agoraphobic individuals
(Hazell and Wilkins, 1990)
Normally functioning
fluorescent lighting (50Hz
ballast)
100H
z
Enlarged saccades over
text
(Wilkins, 1986)
Visual display terminals 70-110Hz raste
r
Changes in saccade size (Kennedy et al., 1998)
Visual display terminals ~70H
z
Raster
Many anecdotal reports of
prolonged photophobia
Normally functioning
fluorescent lighting
100Hz and 120Hz Phase-locked firing of
LGN neurons in cats
(Eysel and Burandt, 1984)
Various Up to 162H
z
Human
electroretinogram
signals at light
frequency
(Berman et al.,1991; Burns et al
1992)
Normally functioning
fluorescent lighting (50Hz
ballast)
100H
z
Inconsistent changes in
plasma corticosterone
levels in captive
starlings
(Maddocks et al., 2001)
Normally functioning
fluorescent lighting (50Hz
ballast)
100H
z
Mate choice in captive
starlings
(Evans et al., 2006)
