|
Design Impact of IEC 60601, Second Edition
By William D. Kimmel, P.E.
and Daryl D. Gerke, P.E.
We try to stay away from any articles on EMC regulations – our specialty
is EMC design and troubleshooting and there are better sources
of information on regulations . But the emergence of new requirements
inevitably impacts the design people, and that brings us into
the picture. A case in point is the adoption of IEC 60601-1-2,
second edition.
Medical device EMC (and safety) is covered by IEC 60601-1-2,
which is recognized by the FDA (in the U.S.), in the European
Union, and many other countries (under various names). The
original edition, dated 1993, will soon be superseded by the
second edition, dated 2001. The transition time, where use
of the first edition was still permitted, comes to an end
in November.
|
|
| Figure 1. Protecting isolated
circuits from RFI |
The first edition had numerous deficiencies, including references
to outdated standards and unclear pass/fail criteria. The
second edition is greatly expanded, answering a lot of questions
raised by the first edition (and, of course, raising more
questions). We’re not going to read the requirements
to you, but we will highlight a few of the changes that significantly
affect designers.
Medical electronics encompasses a wide range of equipment
and environments. There are cases where IEC 60601 will not
be adequate, and other cases where the specified levels are
not achievable. It will take a specialist to make a determination
of what the requirements are for your particular equipment
and how to design for them, but here are some basic guidelines.
Essential Performance
The biggest single aspect is the incorporation of “essential
performance,” a definition of failure. In the first edition,
the interpretation was the equipment could “fail safe”
– as long as the equipment did not become unsafe, it
was deemed to have passed the test. This, of course, resulted
in interrupted operation, and was generally unsatisfactory.
The second edition specifies that the equipment must meet
its intended function, but the requirements are not very well
worded. We like the viewpoint given by the FDA(1), which is
pretty explicit, and is given below:
The following degradations are not allowed:
Component Failures
Changes in programmable parameters
Reset to factor defaults (manufacturer’s presets); change
of operating mode
False alarms
Cessation of any intended operation, even if accompanied by
an alarm
Initiation of any unintended operation, including unintended
or uncontrolled motion, even if accompanied by an alarm
Noise on a waveform in which the noise is indistinguishable
from physiologically-produced signals or the noise interferes
with interpretation of physiologically-produced signals
Artifact or distortion in an image in which the artifact is
indistinguishable from physiologically-produced signals or
the distortion interferes with interpretation of physiologically-produced
signals
Error of a displayed numerical value sufficiently large to
affect diagnosis, therapy or treatment
Failure of an automatic diagnosis or treatment equipment and/or
systems to diagnose or treat, even if accompanied by an alarm
We think that pretty well covers it – it won’t
give the lawyers much room to wiggle, will it?
ESD
The ESD requirements have been increased from 3kV to 6kV
(contact), which will have a definite impact on design –
lots of equipment that passed the 3kV will crash at 6kV.
RFI
The RFI immunity level is increased from 3V/m to 10V/m and
conducted immunity is increased from 3V to 10v (in the ISM
bands, approximately 6.7, 13.5, 27 and 41MHz) for life supporting
equipment. At 3V/m, the impact was largely on the sensitive
analog circuits, but, with the increase, we expect to see
more problems showing up with digital electronics, as well.
There are a number of new requirements, including magnetic
fields, flicker and harmonics, voltage dips and interruptions.
In the EU, these requirements are generally applied over all
electronics, not just medical electronics. The requirements
have a moderate effect on the design, depending on the equipment.
Here are some observations:
Harmonics on power lines arise from non-linear loads in your
power supply and, of course, do not apply to battery-powered
devices. Voltage dips and interruptions also involve the power
supply.
Flicker only occurs with significant load changes in the equipment
– not an issue with most electronic equipment. Magnetic
fields are generally of interest only with a few low impedance
devices, notably the CRT.
Systems Aspect
The second edition includes a discussion of system aspects,
as well as equipment aspects. This is in recognition of the
fact that equipment sensitivities and needs vary widely. Some
equipment will need to be operated in a shielded room, either
because it is too sensitive to achieve the immunity levels
spelled out, or because it is too noisy. Other equipment will
not need such protection, but will need to be operated some
distance away from radio sources – so a calculation of
recommended separation is required.
How is Design Affected?
As this is a design column, we would like to provide a glimpse
as to how these changes impact the design.
You first need to look at how “essential performance”
impacts your design philosophy. If you have already made the
interpretation that the equipment needs to work during interference,
you won’t find a difference. Otherwise, you will be confronted
with some significant changes.
Having decided that, your biggest problems are likely to be
RFI and ESD.
RFI
Regardless of what level of RFI you need to meet, the big
problem will remain the sensitive analog input devices. These
amplifiers, especially the first few sensitive ones, may be
sensitive to signals in the millivolt or even the microvolt
range – even 3V/m may well be too much for your particular
application.
|
|
| Figure 2. ESD path through
seem in plastic. |
Add to this the case where the equipment is to be patient-connected.
Physiological signal levels are quite low, some down in the
low microvolt range, which means that you will have sensitive
input amplifiers connected to patient cables that can’t
be effectively shielded – so filtering is the only real
option. As we are faced with isolation requirements, the filters
need to be referenced to isolated ground (Figure 1) –
then you need to boost the signal to the point where you can
bridge the isolation barrier.
ESD
ESD poses a problem, especially with portable plastic enclosures.
If you have a plastic enclosure, ESD can only enter at seams
and openings – there you look for possible discharge
paths to a metal element inside the enclosure (Figure 2).
All too often we find a box has been metallized to contain
RF emissions or immunity, only to find new ESD points at the
seams (Figure 2 lower). We find the best way to work ESD is to recess
the metal members or protect them with a dielectric. If you
can avoid a direct discharge path, your ESD problem is reduced
to an indirect coupling path, a much lesser problem. If you
do find you need a metallized enclosure and can’t isolate
the contact points, you will need to do a very good job of
shielding.
Summary
The second edition of IEC 60601-1-2 incorporates some significant
changes – a major improvement, in our opinion. If you
haven’t been minding the store, you are running out of
time – your new designs will have to be improved.
The big change is the interpretation of pass/fail – you’ll
have to make the equipment work. RFI always poses a problem
to sensitive analog input devices – this problem will
increase when systems are exposed to the higher levels of
RFI and ESD test levels will also increase, posing a problem
for perhaps the first time.
References
1) The Draft Second Edition of IEC 60601-1-2: A further
Update, Jeffrey L. Silberberg, IEEE International Symposium
on EMC, August, 1999.
2) IEC 60601-1-2 Medical Electrical Equipment Part 1-2: General
Requirements for Safety – Collateral Standard - Electromagnetic
Compatibility – Requirements for Tests.
|
