FMS Publications

ELF Magnetic Field Interference With Computer Monitors: Field Characteristics, Choices And Costs Of Remediation

INTRODUCTION:

Work on threshold values of ELF magnetic fields which cause computer monitor interference have been studied with conflicting results. Early work (Baishiki 1987) concluded that interference occurred with an increase in the difference between 60 Hz and the monitor vertical scan frequency. More recent work (Sandstrom 1992) argues that sensitivity increases as the difference between the frequency decreases. Sandstrom also discloses that individual users have significantly different tolerances for monitor distortion.

There are, generally, four solutions to the problem of computer monitor interference:

1. increase the monitor-source distance
2. shield the monitor
3. shield the source, and
4. modify the vertical scan frequency of the monitor

Typically, increasing distance is the least expensive and source shielding is the most expensive. Changing the refresh rate is sometimes technically feasible and inexpensive, but often creates a new "flicker" problem. Monitor shields are widely available at different costs and effectiveness.

OBJECTIVE:

The objective of this study is (1) to further investigate at what range of ELF magnetic field levels, monitor distortion is apparent on modern computer monitors, (2) to characterize the ELF magnetic field strengths and dominate field direction which cause interference, and (3) to gauge the effectiveness and relative costs of available monitor shields.

Of particular interest is the question of what, if any, effect the most recent developments in monitors and application or operating system software technology has had on computer monitor sensitivity to external magnetic fields. In pursuit of that objective, this study departs from the use of traditional laboratory Helmholtz coils as a magnetic field source and to substitute a more realistic source - a commercial service entrance panel - which produces a complex range of fields. Lastly, this study seeks to examine cost vs. performance values for the use of commercial and custom computer monitor shields.

METHOD:

In laboratory facilities of Electro-test Inc., a Nationally Recognized Testing Laboratory (NRTL) accredited by Federal OSHA, a test commercial electric service panel, rated at 800 amps was connected to a three-phase power network shown in Figure 1. Three single-phase variable transformers were fed by a 4-wire, three-phase 208 volt source. The variable transformers were capable of supplying changeable voltages from 0 to 140 volts to a set of three loading transformers which where configured for 120 volt input to 5 volt output at up to 1000 amps. The output of each loading transformer was connected to the test service panel. When energized by the variable transformers, the asymmetrical conductors within the power panel created a test magnetic field in the area adjacent to the service panel which could be adjusted both for magnetic field strength and dominant field direction. This source provided a "real world" test environment in which a variety of different monitors and monitor shielding products were subjected to controlled magnetic field conditions.

After characterizing the magnetic fields produced by the test service panel, a test location was selected adjacent to the test panel, approximately one meter from the floor and one meter from the vertical centerline of the test panel. Progressive values of current were circulated through the test panel and the resulting magnetic field values noted for each incremental increase.

A variety of computer monitors where then placed, one at a time, at the test location and connected to either a 220 CDS laptop computer running Windows 95 and displaying a Microsoft Word 97 text document or a Macintosh 5300 laptop computer displaying a Microsoft Word 6.0 Document. A diverse group of volunteers were asked to first identify the threshold of interference for each of the test monitors and then the level of interference which would be objectionable for use. Each volunteer was requested to assume a "normal" viewing position in front of the test monitor and their eye to screen distance was noted. The test service panel was then energized at a low current value and with each setting change, the volunteer was asked if "jitter" was present on the test monitor and their response noted. Magnetic field values were varied in a random manner to minimize the perception of a linear increase of field strength during each test session.

Commercially available monitor shields were then evaluated utilizing the same test location adjacent to the service panel. After positioning a subject monitor shield in the test location, an Emdex-C magnetic field gaussmeter was placed in the interior at the center of the subject test shield. Currents were then circulated in the service panel at levels calibrated to create magnetic field levels ranging from 5 mG to 50 mG at the test location. At each setting, magnetic field data was collected with the Emdex-C located in the test shield. This data was utilized to evaluate shielding factor performance for each of the test shields at each of the various magnetic field levels.

Monitor shields tested included: (1) standard JitterBox; 2) JitterBox enhanced with Mu-metal; (3) custom-manufactured mu-metal enclosures, and (4) shields constructed of other ferromagnetic and conductive shielding materials.

RESULTS:

As in the Sandstrom study, a large variability existed in the threshold level of magnetic field strength at which individuals noted monitor distortion. However, as shown in Table I, monitor distortion or "jitter" was noticeable to all of the volunteers at ELF magnetic field levels of between 5 and 10 mG for monitors with vertical refresh rates or 75 Hz and somewhat higher at 10 to 20 mG for a monitor with a vertical refresh rate of 60 Hz. Monitor distortion appeared to be unacceptable to nearly all individuals with monitors in a 10 to 20 mG magnetic field.

Results of the distortion perception testing suggest that computer monitor shields must be capable of reducing external magnetic field strengths to a level of approximately 5 mG in the interior of the shield to control monitor distortion. Analysis of the test data also indicates ELF magnetic fields which were predominately horizontal or vertical tended to have a greater effect on the stability of test monitor images in unshielded test cases. However, magnetic fields which were perpendicular to the face of the test monitor had a disproportionate effect on the stability of monitors in the shielded cases, regardless of the shield used.

As expected, the custom manufactured Mu-metal and other high performance shield enclosures were capable of satisfactorily reducing the interference pattern in all cases. However, in the majority of cases tested, the less expensive, standard monitor shielding products also produced satisfactory results. In virtually every test case, it appears likely that the "custom" enclosures will continue to provide adequate protection above 100 mG and in one case, above 150 mG.

Within general categories of construction materials, enclosures which were smaller, generally enjoyed an advantage in shielding effectiveness. This advantage was amplified by the fact that larger monitors tended to be more sensitive to an external magnetic field.

CONCLUSION:

The ETI laboratory test results suggest that external ELF magnetic field levels greater than 5 to 10 mG will cause distortion or "jitter" on most modern computer monitors. Although there is great variability in the actual threshold level experienced from person-to-person, the data from this study finds than an average perception threshold of 9 mG across all monitors and computer combinations. These tests also found that magnetic field levels of 10 to 20 mG will likely produce objectionable interference on nearly all computer monitors.

The test data also indicates that ELF magnetic field shielding effectiveness provided by standard monitor shielding products and custom enclosures is highly related to cost. Inasmuch as even the relatively inexpensive standard monitor shielding products approach the cost of a new monitor, the options of increased distance and scan rate reset should be investigated prior to shielding. If those choices are either unavailable or impractical, the results of this study strongly suggest the need to scale the shield effectiveness to the particulars of the magnetic field, both strength and direction.