"Zero"
solution for shielding and electrical protection in a unit
for
electroencephalographic recording.
"Zero"
solution for shielding and electrical protection in a unit
for
electroencephalographic recording.
Abstract --The conception and solutions used for the
electrical protection and shielding of an EEG (Epilepsy) unit are
presented. Solutions are based on the "zero" method. Principle of
this method consists of eliminating all stresses in the areas of
measurement. Original solutions for the conception as well as
maintenance procedures are presented. First EEG recordings gave
satisfactory results.
I. INTRODUCTION
Because of the tiny amplitude of electrical potential at the surface
of the head, electrical problems, quality of the power supply,
protection against electrical and magnetic perturbations, as well as
safety for the equipment and the patient are major concerns for
people designing laboratories for electroencephalography (EEG). One
of the most common solution applied is the Faraday cage [1]
which gives partial isolation of low frequency electrical fields.
However it can be demonstrated that this solution is often not
sufficient to protect against perturbation like the well know 50 (or
60) Hz signal that invades the EEG. This occurs despite the high
quality bipolar amplifiers that are available now. The following
paper presents solutions that were applied to a new epilepsy unit
that includes 5 long and short term EEG rooms, 2 research and
clinical laboratories that are mainly devoted to evoked potentials
recordings and various offices and meeting rooms.
II. GENERAL PRINCIPLES
A. The "zero" method
The method developped here consisted of approaching the
"zero" condition that is "zero" current in the connections and "zero"
potential between the reference points. Limits are given by the
practical and physical considerations. i.e. 1) cost of the zero
operation, 2) constraints given by the architecture of the unit and
the neighborhood, 3) know-how of the company in charge of the
electrical installation and 4) self perturbations by the systems
inside the shielded environment (Personal Computers for
instance).
B. Measurement of the perturbations
A Perturbometer voltammeter (to be commercially available
soon) was used for the measurement of the perturbations. It measures
the actual voltages in peak value between 10 KHz and 10 MHz, in order
to reproduce the effects obtained by traditional wirings. It has been
substantiated that they are insensitive to low frequencies up to a
few KHz at which a lot of energy is needed to excite them. They are
also insensitive to very high frequencies where they are not able to
render the energy by current circulation. This device detects V
perturbation levels which correspond to an exceptional zero
level.
III. POWER SUPPLY
A. Electrical power of the unit
The entire unit has been supplied by reinforced-shielded
transformers equipped with capacitors for filtering and phasing the
cos Phi. The circuits are of two types [2], the sensitive
power supplies (SPS) for the laboratories equipments and the regular
power supply (RPS; offices, light, ventilation, etc.). SPS
transformers are monophasic with a common middle point connected to
the sensitive ground (SG; fig.1). The RPS is traditional with an
earthed neutral.
B. Electrical distribution
As much as possible, sensitive and regular power supplies
have separate tracks. Middle point wiring gives symmetrization of the
sensitive sources, thus avoiding propagation of electromagnetic
fields. The electrical outlets for the SPS are designated by a
different color and are useable only with standard plugs equipped
with a furnished accessory safeguards against connections of
unintentionned devices which could be responsible for perturbations.
Even in the non sensitive areas and for the RPS, all known sources of
perturbations have been prohibited. For instance lighting in the unit
is provided by incandescent lamps instead of fluorescent lamps, and
appliances are tested by the MES 904 before being used.
C. Ground circuitry
The building ground (BG) is a conventional electrical
earthed network and is used for the RPS. The SG network [3]
is used for the SPS. It is directly connected to the middle point of
the transformers (fig. 1) and connected to the hospital ground
through separated inductance filters (fig. 2). In this way the SG is
fictitiously disconnected from the BG but will still insure patient
and staff safety. In each room, near the patient location, the floor
is equipped with a screw inserted into a metal plate pressing the
shielding grid and is connected to the nearest outlet to the EEG
machine ground (Gnd 2).
Fig. 1 Schematic organization of the 4 main sources in the unit.
IV. ROOM SHIELDING
The 7 dedicated EEG rooms are surrounded by metal grids to produce
electrostatic shielding. They are inserted into the walls, between
two faces, and into the floor. Protection in the ceiling is insured
by metallic plates isolate from the building by insulated fixations,
and connected together for the continuity. After initial tests of the
amplitude and frequency of the perturbations, and considering cost
constraints we have chose to use regular sized grids (1 cm2). Window
panes have also been covered by metallic screen (armed glass or
mosquitos netting type) adapted on a frame connected to the grid
inside the wall and isolated from the window. To preserve the zero
condition, the heater inside the EEG rooms were isolated from the
building ground by plastic sleeves. Each panel were isolated and
connected together at the same point in a control box. This allows
for control of the isolation between the BG and the different panels
as well as the ability to localize defaults. These Faraday cages
constitute main part of the SG (fig. 1).
V. RESULTS
Overall attenuations of the perturbations in the unit are from a
factor of 5 to 500. Tensions given by the Perturbometer are 10 mV max
between every phase and ground, or phase and neutral. Electromagnetic
fields at 50 Hz and closed frequencies give less than 0.1 mV/m2 for
the magnetic component and less than 0.1 mV/m for the electric
component (fig. 2). For EEG recordings, none of the usual
perturbations, 50Hz contamination for instance has been
detected.
Fig. 2. Schematic diagram of EEG measurment and zero test.
VI. SELF PERTURBATIONS
Attention needs to be focused on using shielded and filtered
equipment, avoiding common switching power supplies, and using
shielding frame on video screens. A single typical PC/AT with
accessories created perturbation 10 times above the residual noise of
the treated environment. Permanent level of the perturbations can be
checked by a MES 904 connected to a spy looped line. This meter will
not indicate more than 0,5 V or will require localization of the
perturbation source. Annual tests will be performed to insure SPS
insulation and power line quality level.
VII. CONCLUSIONS
Virtual elimination of radioelectrical perturbations is the garanty
of reliability and confidence of the EEG signal, and is acheivable at
low cost and in a simple way, provided the influence generation and
transmission issues be dealt with at the design stage of a system.
The zero method is an original procedure for "perturbation busting",
and turns out to be simple and obvious if all operators follow
strictly predifined guide lines .
REFERENCES
[1] V. Gobin & G. Labaune. "Calcul et mesure de
l'efficacité de blindage des matériaux
composites." Ann. Télécomun., 43 , "n
11-12 - 1988.
[2] F. Vaillant. "La Compatibilité
Electromagnétique", Cahier technique Merlin-Gérin
n 149 ". - 1991.
[3] E. Montandon "Influence de la mise à la terre et
du câblage pour la CEM."Compatibilité
Electromagnétique - Presses Polytechniques Romandes , 2
Edition 1985, pp 315 to 323.
Communication presented from "Palais des
Congrès, Porte Maillot" in Paris (France), on October 31th
1992, text n 92 CH 3207-8 pp 1134 & 1135, for
"Proceeding of the 14th Annual International Conference of the IEEE
Engineering
in Medicine and Biology Society. Innovations in Biomedical
Engineering in the Year of the European Unified Market"
Accueil
E-Mail
