IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 15, NO.

1, JANUARY 2000 307

Magnetic Field Effects on CRT Computer Monitors

Balazs Banfai, George G. Karady, Fellow, IEEE, Charles J. Kim, Member, IEEE, and
Kate Brown Maracas, Senior Member, IEEE

Abstract—This paper discusses the effect of external low TCO Certification [8] recommends the monitor immunity level
frequency magnetic field interference on cathode ray tube (CRT) as 12.6 mG and the jitter threshold as 0.1 mm. There is no immu-
computer monitors. The paper describes a new test facility nity level defined by US standards and generally the available
and presents a quantitative measuring method which has been
developed to characterize the field effects. A total of 21 monitors CRT monitors do not limit jitter to the specified limits.
from major manufacturers were tested. It was found that larger Recently, Edison Electric Institute completed a comprehen-
monitors are more sensitive and that the relationship between the sive guide that defines flicker and jitter as monitor disturbances
magnitude of jitter and the magnetic flux density is linear and and also suggests numerous mitigation techniques [10].
independent of the refresh rate and the frequency of the inter- Due to the rapid improvement of the technology in the com-
fering magnetic field. The monitors are most sensitive to magnetic
fields parallel to the screen. Monitor sensitivity is specified and puter industry, a systematic investigation is needed to determine
presented for the tested monitors. the effect of the size of the CRT, resolution, refresh rate and
A statistical survey was carried out to determine the human per- harmonics on the tolerable magnetic field strength. This paper
ceptibility level of jitter. It was found that 12 mG may cause de- presents quantitative analyzes of image distortion on CRT com-
tectable jitter for the common 14″ monitor. This value drops to puter monitor screens.
around 7 mG for a 21″ monitor.
Index Terms—CRT, EMI, flicker, jitter, perceptibility, refresh II. EXPERIMENT
rate, sensitivity, VDU.
A. Test Setup

I. INTRODUCTION For testing monitors with a homogenous magnetic field, Ari-

zona State University built a coil system, consisting of three par-

U TILITIES and computer monitor manufacturers have re-

ceived numerous complaints about monitor disturbances.
Measurements and case studies have shown that the sources of
allel 2 m × 2 m coils. The coils have 12 turns of 12 gauge wire
placed in a 1″ plastic tube. Each coil can be supplied indepen-
dently and the distances between the parallel coils can be ad-
the disturbances are magnetic fields. These magnetic fields exist justed. A signal generator supplies current to the coils through
in all working environments and originate from electrical instal- an audio-frequency amplifier. The current varies between 0–30
lations, motors, switchboards, transformers or cables carrying A, which produces a magnetic flux density of 0–2000 mG in
large currents [1]–[3], [9]. The interference caused by the AC the 30–2500 Hz frequency range. It was found that the devia-
magnetic field can result in jitter, a positional instability of the tion of the magnetic field generated by each of the coils does
image on the screen, or in flicker, a periodic luminance vari- not exceed 3% in the test area, which is located in the middle
ation. The presence of a high DC magnetic field can produce of the coil system. The coil system and the stand of the monitor
static color patches on the screen. This paper deals with jitter have been built on a wooden frame to avoid magnetic distortion.
only because customer complaints most often focus on this type Fig. 1 shows the experimental setup.
of disturbance.
In 1987 Baishiki and Deno published a paper [4] which dis- B. Measuring Method
cussed the magnetic field effects on different monitors. The During the measurement, the monitor is placed in the middle
paper presented quantitative results and mitigation techniques. of the coil system and the magnetic field is varied. A single
The authors were able to realize a one character shift on the lens reflex camera with large magnifying lens was focused on
monitor screen by application of a 250 mG magnetic field. Sand- a pixel on the screen, as shown in Fig. 2. The magnification of
strom et al. [5] carried out studies which investigated the effects the camera is set to 20× to meet the test method requirements
of the cathode ray tube refresh frequency on human percep- recommended by an international standard [8]. The jitter of the
tion. The ECMA defined jitter based on IEC and CISPR stan- pixel due to the magnetic field resulted in an elongated path on
dards [7]. The standard defines the monitor immunity levels of the screen, which has been photographed with a long exposure
5, 30, and 100 mG depending on the working environment. The time, as shown in Fig. 3. The length of the path, as a function of
the applied magnetic flux density, has been measured.
The magnetic flux density was measured using an Emdex-C
Manuscript received May 26, 1998; revised September 10, 1998. This study magnetometer. The magnetometer is placed touching the under-
was sponsored by Salt River Project, Southern California Edison Company, and
the American Public Power Association. side of the monitor. To measure only the magnetic flux density
B. Banfai and G. G. Karady are with the Department of Electrical Engi- generated by the testing coil system, without the field emitted
neering, Arizona State University. by the monitor itself, the monitor was turned off temporarily.
C. J. Kim is with Southern California Edison.
K. Maracas is with Salt River Project. The background magnetic flux density was less than 3 mG in
Publisher Item Identifier S 0885-8977(00)00546-X. the direction of the external applied magnetic field.
0885–8977/00$10.00 © 2000 IEEE
308 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 15, NO. 1, JANUARY 2000

2) If the applied magnetic field is perpendicular to the

bottom and top of the monitor, the pixel shift is purely
horizontal on the screen, resulting in a “waving vibration
pattern,” as seen in Fig. 5. The magnitude of the jitter
increases when the pixels are closer to the middle of the
screen. The test pixel must be placed in the middle of the
screen for the worst case.
3) Fig. 6 shows the third case, when the monitor’s screen is
perpendicular to the applied magnetic field generated by
the coils. In this case, the pattern of the jitter is a rotational
twist of the picture.
The pixel in the middle of the screen is stable and the mag-
nitude of the jitter increases when the pixels are closer to the
Fig. 1. Experimental setup. corners of the screen. In this case the test pixel must be placed
at the corner of the screen for worst case analysis.

B. Jitter Magnitude
Jitter magnitude is defined as the displacement of the test
pixel and can be calculated as the difference between the length
of the elongated path when a magnetic field is applied (see
Fig. 3) and the size of the pixel when no magnetic field is ap-
plied (see Fig. 2).
Fig. 7 shows the relationship between the magnetic flux den-
sity in milligauss (rms) and the jitter in the three different ori-
entations of the monitors for 5 monitor sizes.
Fig. 2. Image of one pixel without magnetic field applied.
The solid curves show the amplitude of jitter when the
magnetic field is parallel to screen, normal to the sides or
the top-bottom of the monitor (Figs. 4 and 5). The dotted
curves represent the case when the magnetic field direction
is perpendicular to the screen (Fig. 6). It can be seen that the
monitors are most sensitive to the magnetic field when the field
is perpendicular to the sides or the top of the monitors. The
relationship between the amplitude of jitter and magnetic flux
density is linear. The jitter caused by the background magnetic
field is smaller in magnitude than the accuracy of the measuring
method, thus the fitted lines intercept zero.
The linear relationship between the jitter and the magnetic
Fig. 3. Vibration path of one pixel with an applied field of 213 mG.
flux density permits the monitor sensitivity to the applied mag-
netic field to be described by the slope of the curves in Fig. 7.
III. EXPERIMENTAL RESULTS Monitors with a smaller slope are less sensitive to magnetic
Twenty one monitors from major manufacturers were tested. fields than monitors with a higher slope.
To determine the effect of the magnetic field direction, the mon- Fig. 8 shows the sensitivity of all tested monitors as a function
itor was rotated relative to the coil system. This investigation of their size. The monitors were selected from major manufac-
showed that the pixel movement depends on the pixel location. turers. The sensitivity rapidly increases with the size of the CRT.
The worst location was selected for the measurement of monitor No effects were seen on non CRT type displays (e.g., laptop
sensitivity as described below. computers).

A. Jitter Pattern C. Frequency Response

Depending on the orientation of the monitor to the magnetic The frequency of the applied magnetic field was varied from
field, three jitter patterns can be distinguished. 15 Hz to 15 kHz. It was found that the magnitude of the jitter is
1) If the applied magnetic field is perpendicular to the sides independent of the frequencies in this range.
of the monitor, the displacement of the pixels on the The magnetic field generated by electrical appliances con-
screen is purely vertical, resulting in a vertical oscillating tains harmonics that are superimposed on the power frequency
motion, as shown in Fig. 4. The magnitude of the jitter magnetic field. We added third or fifth harmonics to the power
increases when the pixels are closer to the middle of the frequency magnetic field. The magnitude of the harmonic was
screen. For the worst case, the test pixel must be placed 30% of the fundamental. The amplitude of the jitter was found
in the middle of the screen. to be proportional to the peak value of the magnetic field. A
BANFAI et al.: MAGNETIC FIELD EFFECTS ON CRT COMPUTER MONITORS 309

Each horizontal line is divided into pixels. Pixels are simple

picture elements that carry the actual information of the picture.
The number of resolvable pixels on the horizontal and vertical
lines is referred as the resolution [10].
Although the detailed explanation of the monitor’s operation
is not the subject of the paper, it is important to briefly explain
some basic principles. The image on the screen is a result of a
shifted electron beam, which constantly scans the whole surface
of the screen. The beam starts from the top-left edge, rapidly
moves horizontally to the end of the line, jumps back to begin-
ning of the next line and repeats this sequence until it has traced
Fig. 4. Jitter pattern due to magnetic field oriented perpendicular to the side
of the monitor. the whole screen. The number of times that this process occurs
in a second is called the refresh rate (also known as the vertical
scan rate), and is usually 50–120 Hz. Computer manufacturers
like to increase the refresh rate above 60 Hz to avoid flicker.
Each horizontal line is divided into pixels. Pixels are simple
picture elements that carry the actual information of the picture.
The number of resolvable pixels on the horizontal and vertical
lines is referred as the resolution [10].

B. Jitter Frequency
In addition to the magnitude, jitter frequency also affects the
perceptible jitter, as discussed in Section V.
Fig. 5. Jitter pattern due to magnetic field oriented perpendicular to the Fig. 9 shows an example of vertical pixel displacement when
top-bottom of the monitor. the direction of the interfering magnetic field is perpendicular
to the side of the monitor.
Due to the difference between refresh frequency and the inter-
fering magnetic field frequency, the electron beam is displaced
at the “difference” frequency. The dotted curve is the magnetic
field waveform generated by the vertical deflection coil. This
signal drives the beam from the top to the bottom of the screen at
the vertical refresh rate of 72 Hz in this example. An interfering
magnetic field (60 Hz—solid line) is superimposed, resulting
in an error in the position of the pixels on the screen. For ex-
ample, a beam positioned in the vertical center of the screen ex-
periences a resultant magnetic field that is changing with time.
Fig. 6. Jitter pattern due to magnetic field oriented perpendicular to screen of This magnetic field creates a linearly proportional vertical pixel
the monitor. displacement on the screen. The arrows indicate the vertical dis-
placement as a function of time. The frequency of the envelope
of the pixel displacement is the difference between the refresh
special case exists when the refresh frequency equals the funda-
frequency and the magnetic field frequency, 72 − 60 = 12 Hz.
mental frequency of the applied magnetic field. The frequency
The investigation also shows that the amplitude of the jitter is
of the jitter is 0 Hz, resulting in a stable image.
independent of the screen refresh frequency. Similar simulations
also show that applying a 120 Hz refresh frequency will NOT
IV. MONITOR PRINCIPLES eliminate the pixel displacement, however the jitter might not
A. Principle of CRT Operation be perceptible due to its high frequency (60 Hz) as discussed in
Section V.
Although the detailed explanation of the monitor’s operation
is not the subject of the paper, it is important to briefly explain
V. HUMAN PERCEPTION
some basic principles. The image on the screen is a result of a
shifted electron beam, which constantly scans the whole surface The other aspect of monitor interference is the human re-
of the screen. The beam starts from the top-left edge, rapidly sponse to the jitter. Earlier studies have shown that jitter is most
moves horizontally to the end of the line, jumps back to begin- perceptive when the external magnetic field is in the vertical di-
ning of the next line and repeats this sequence until it has traced rection [5], [6]. These studies also concluded that the human
the whole screen. The number of times that this process occurs perception level is the lowest when the frequency of the jitter is
in a second is called the refresh rate (also known as the vertical around 4–15 Hz. If the jitter frequency is barely above 0 Hz, the
scan rate), and is usually 50–120 Hz. Computer manufacturers jitter becomes a slow waving motion and is not as perceptible
like to increase the refresh rate above 60 Hz to avoid flicker. as when the frequency is a few hertz. At jitter frequencies in the
310 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 15, NO. 1, JANUARY 2000

Fig. 7. Jitter versus magnetic flux density.

Fig. 8. Summary of monitor jitter sensitivity.

range of 30 Hz and above, the human eye can not respond and rates of 56 and 72 Hz were used. Fifty observers participated,
the jitter becomes a blurring disturbance resulting in a higher of which 8 were female and 42 were male. The average age
perception level. of the observers was 28 yr and they all use computer moni-
Tests were conducted in our laboratory to statistically deter- tors more than 2 h per day. The external magnetic field strength
mine the minimum perception level of jitter. The tests were all was increased until the subject observed jitter. The jitter level
conducted on a 17-in monitor with an 800 × 600 screen reso- was recorded and the field was then further increased. The field
lution. The viewing distance was approximately 24 in. Refresh strength was then reduced until the subject no longer observed
BANFAI et al.: MAGNETIC FIELD EFFECTS ON CRT COMPUTER MONITORS 311

Fig. 9. Pixel displacement as a function of time. Fig. 10. Threshold of observed jitter at 56 and 72 Hz refresh rate.

jitter. The corresponding jitter level was recorded. The average

of the two levels is considered the threshold of observed jitter.
The histograms of Fig. 10 statistically show the threshold of ob-
served jitter for the 50 observers, at refresh rates of 56 and 72
Hz.
Due to the large sample size, the mean can be approximated
with normal distribution (central limit theorem). The mean
threshold of observed jitter is found to be 0.127 mm ± 13% for
a 56 Hz refresh rate and 0.142 mm ± 11% for a 72 Hz refresh
rate for a 95% confidence interval.
These results have been combined with Fig. 7 to prepare
Fig. 11, which shows the expected magnetic flux density
causing perceptible jitter, for external fields vertical and
parallel to the monitor screen and at a refresh rate of 72 Hz. It Fig. 11. Magnetic flux density causing perceptible jitter for parallel external
fields at 72 Hz refresh rate.
was assumed that monitor size and resolution do not affect the
perception of jitter.
ACKNOWLEDGMENT

VI. CONCLUSION The authors would like to thank Prof. R. G. Farmer of Arizona
State University for his technical help.
A quantitative technique was developed to measure jitter on
the screen. A total of 21 monitors of different sizes from major REFERENCES
manufacturers were tested. With the help of 50 observers, a sta-
[1] M. Silva, N. Hummon, and D. Rutter, “Power frequency magnetic fields
tistical survey was performed to determine the jitter percepti- in the home,” , Feb. 1988, IEEE-PES, Paper 88 WM 101-8.
bility level. The following results were obtained: [2] W. T. Kaune and R. G. Stevens, “Residential magnetic and electric field
measured over 24H period,” New York State Power Lines, June 1986.
• The monitor is most sensitive to the magnetic field when [3] J. R. Gauger, “Household appliance magnetic field survey,” IEEE Trans-
the field is perpendicular to the sides or the top of the actions on Power Apparatus and Systems, vol. PAS-104, pp. 2436–2444,
monitor. 1985.
[4] R. S. Baishiki and D. W. Deno, “Interference from 60Hz electric and
• The screen refresh rate influences only the frequency of magnetic fields on personal computers,” IEEE Trans. Power Delivery,
the jitter. The frequency of the jitter is the difference be- vol. PWRD-2, no. 2, pp. 558–563, Apr. 1987.
tween the refresh rate and the interfering magnetic field [5] M. Sandstrom, K. H. Mild, and A. Berglund, “External power frequency
magnetic field-induced jitter on computer monitors,” Behaviour and In-
frequency. When the refresh rate equals the magnetic field formation Technology, vol. 12, no. 6, pp. 359–363, 1993.
frequency, the frequency of the jitter is 0 Hz (constant dis- [6] “Work with visual display units (VDUs),” Swedish National Board of
placement of the image) and no distortion occurs. Occupational Safety and Health, S-171, 84 Solna, Sweden, Ref: AFS
1992-14, June 1992.
• The screen resolution does not affect the magnitude of the [7] Immunity of VDUs to Power Frequency Magnetic Fields, European
jitter. Computer Manufacturers Association, Standard ECMA-199.
• The relationship between the jitter magnitude and the [8] “Requirements for environmental labelling of personal computers,” The
Swedish Confederation of Professional Employees, Report no. 1-2, 3rd
magnetic flux density is linear for a given monitor. ed., Mar. 1996.
• The human eye can detect about 0.13 mm jitter. [9] “Magnetic field mitigation to reduce VDU interference,” Electricity
• For a 14″ monitor, a 12 mG field may cause a detectable Supply Association of Australia Limited, CAN 052 416 083, July 1996.
[10] “Computer CRT monitor performance in the presence of power-fre-
disturbance. This value drops to around 10 mG for a 17″ quency magnetic fields,” Edison Electric Institute EMF Working Group,
monitor and to 7 mG for a 21″ monitor. 701 Pennsylvania Avenue, NW, Washington, DC 20004, Dec. 1997.
312 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 15, NO. 1, JANUARY 2000

Balazs Banfai holds the M.S. degree in electrical engineering from the Tech- Charles J. Kim (M’93) was born in Korea in 1967. He received the B.S.E.E.
nical University of Budapest. Currently he is working toward the Ph.D. degree degree from University of California, Irvine in 1992. He is working toward the
at Arizona State University. His research interests include power electronics, M.S. degree in electrical engineering. Currently, Mr. Kim is a Project Man-
high voltage testing, and electromagnetic interference problems. ager and a Research Engineer in R&D for Southern California Edison Com-
pany, where he is directly responsible for EMF engineering research. His areas
of interest include transmission and distribution system power system analysis,
power quality, and communication. He is a member of the IEEE AC/DC Field
Effects Working Group, Bionmagnetics Society, and the Phi Kappa Phi National
Honor Society.

George G. Karady received the M.S. and Ph.D. in electrical engineering from
Technical University of Budapest in 1952 and 1960, respectively. Dr. Karady Kate Brown Maracas was born in Pittsburgh. Until recently, she worked as
was appointed to Salt River Chair Professor at Arizona State University in a manager of the Environmental Affairs Division of Salt River Project. Ms.
1986. Previously he was with EBASCO Services where he served as Chief Maracas received the B.S.E.E. degree from Arizona State University in 1986,
Consulting Electrical Engineer, Manager of Electrical Systems, and Chief and the Certificate of International Business Management from the American
Engineer of Computer Technology. He was Electrical Task Supervisor for the Graduate School of International Management in 1997. During her tenure at
Tokomak Fusion Test reactor project in Princeton. Dr. Karady is a registered SRP, Ms. Maracas worked extensively with both technical and policy issues re-
Professional Engineer in New York, New Jersey, and Quebec. He is the author lated to electric and magnetic fields (EMF’s), and in 1992 was appointed by the
of more than 100 technical papers. Secretary of Energy to act as advisor to the National EMF Research Program.