nue
Santa Barbara, CA 93109
Phone: (805) 564-1260
Fax: (805) 966-7875
tustin@equipment-reliability.com
http://www.equipment-reliability.com
http://www.vibrationandshock.com
May 2006
Equipment Reliability Institute
Volume 23
In this issue...
May 2006
Equipment Reliability Institute
Volume 23
by William H. (Bill ) Parker
An Introduction to EMI Suppression for Sensor Leads
Abstract
EMI (electromagnetic interference) control
is often a confusing aspect of modern
equipment, system, and sensor design.
After all, no one intentionally designs in the
sporadic, unpredictable upset of a perfectly
functioning system by EMI gremlins. During
my upcoming EMC (electromagnetic
compatibility) short course in Las Vegas on
July 12-15, I anticipate student questions
concerning practical fixes of EMI problems
with regard to low level sensors operating in
electromagnetically noisy environments. The
intent of this brief treatise is to explain the
essence of EMI upsets, and to address how
certain EMI corrective measures (fixes) may
be used to reduce or eliminate those upsets.
Operating Bandwidth
First of all, we should recognize that
an electrical/electronic system has
an intentional operating bandwidth.
The concept of operating bandwidth is
commonly recognized for actual radio
receivers, but is not so commonly
appreciated for non-receiver devices. For
example, the power input to a device may
have an intentional operating bandwidth
of 60 + 5 hertz. This intentional power
frequency operating range gives us an
An Introduction to EMI Suppression for Sensor Leads -
by William H. Parker
Internally Isolated Enclosure -

by Herb Lekuch
Instruction Manual Stories Contest - meet the winner!
Test Lab Musings (part 12) -
by Robert L. Renz
C
ontrol Accelerometer
Location - Fixture
design instructor (next class
at Las Vegas in November)
Steve Brenner and I have
long criticized vibration
and shock test standards
and specifications for not
instructing test labs precisely
where to put the control
accelerometer. Experienced
test personnel know that
control accelerometer
location can greatly affect
test severity and test
outcome. Rather than
criticize further, we decided
to offer guidance on this
subject. We plan to offer our
ideas to the group gathering
at Phoenix from 8-10 am on
Monday, May 8 at the Hyatt
Regency Hotel, Phoenix,
during the Institute of
Environmental Sciences &
Technology annual meeting.
The purpose of this meeting
is to further develop the
upcoming G revision to
Military Standard 810. The
F revision has been in
use since the year 2000.
edge in rejecting potential electrical
interference on the power line; we can
simply filter out electrical energy that is
outside this frequency range. If we have
a sensor placed into an electrically noisy
environment, we can also install filtering
on the output of the sensor, or at the
input of the device to which the sensor
sends its signal, to reject unintentional
signals outside the sensors intentional
operating frequency range. Such filtering
may consist of a bandpass filter, a lowpass
filter, or a highpass filter, depending upon
the specific requirements. Filters may be
rather complex circuits purchased from a
filter manufacturer, or they may be simple
capacitors or inductors.
Example: Two Wire Sensor
Let us consider an example, where a
two-wire output sensor is monitoring
some aspect of performance machine
performance, such as temperature or
vibration. The sensor is mounted directly
onto the machine being monitored, and the
two-wire output is routed some distance
through a hostile environment, with strong
electromagnetic fields, to the device
which receives the sensor output. For our
example, let us assume that the strong Equipment Reliability Institute
1520 Santa Rosa Avenue
Santa Barbara, CA 93109
Phone: (805) 564-1260
Fax: (805) 966-7875
tustin@equipment-reliability.com
http://www.equipment-reliability.com
http://www.vibrationandshock.com
May 2006
Equipment Reliability Institute
Volume 23
fields induce interference voltages and
currents on the sensor output leads which
are outside the normal output bandwidth
of the sensor, but are within the bandwidth
of the receiving device, which performs
some processing of the sensor output,
such as amplification or analog to digital
conversion. Let us assume further, for our
example, that the interference imposed
upon the sensor leads causes the receiving
device to misinterpret the sensor output. By
common sense, the solution would seem
to be to prevent the voltages and currents
induced by the hostile environment from
reaching the receiver.
Simple Filtering with Small Capacitors
For our example, let us assume in this
case that the usual output of the sensor is
a very low frequency analog signal, with
an intentional operating frequency range
of DC (0 hertz) through 100 hertz. One
potential fix of this problem is to simply
apply filtering at the input of the device
receiving the sensor signal, to exclude
all frequencies above 100 hertz. A simple
way to accomplish this would be to apply
a simple capacitor at the receiver input.
Actually, three capacitors will be required;
one capacitor from each of the two sensor
leads at the receiver to chassis ground, to
suppress common mode (line to ground)
high frequency interference, and one
capacitor across the two leads, to suppress
differential mode (line to line) interference.
The intent of each capacitor is to act as
a high frequency short circuit. At the low
frequencies (DC to 100 hertz) intentionally
used by the sensor, the capacitor values
need to be chosen such that the capacitor
0.1 uF Capacitor with 1 Inch Leads
0.1 uF Capacitor with 0.1 Inch Leads
New Distance Learning
Program 2006
Over the past 5 or 6
years, some readers have
studied vibration and shock
technology with me via
Distance Learning. Weve
sent you a CD carrying
thousands of PowerPoint
slides, text, photographs,
animations and video clips,
organized in 31 lessons.
31 sets of review problems
and questions have gone
back and forth between us
(sometimes more than once
per lesson, if I wasnt certain
that you understood a point).
Most of you have eventually
received your Certificates.
Webmaster Cris Barzellay
forced me to help her (once
my 2005 random vibration
text was completed) rewrite
and update the Distance
Learning Program. She
and I want to announce
that the 33 Distance
Learning 2006 lessons
are now ready for you.
For detailed information
about this Program and how
to purchase it, visit our site.
Were also releasing a
package with both Distance
Learning CD and my
Random Vibration and
Shock Testing book.
Click here for more info. Equipment Reliability Institute
1520 Santa Rosa Avenue
Santa Barbara, CA 93109
Phone: (805) 564-1260
Fax: (805) 966-7875
tustin@equipment-reliability.com
http://www.equipment-reliability.com
http://www.vibrationandshock.com
May 2006
Equipment Reliability Institute
Volume 23
William H. (Bill) Parker is a registered electrical engineer in California, with 32 years
of experience in defining, testing, analyzing, and solving electromagnetic interference
problems. He has performed and supervised many EMI tests in laboratory and field
environments, and has taught EMC seminars throughout the USA as well as overseas
since 1982. For more information about this ERI specialist, please visit http://www.
equipment-reliability.com/consultants/parker.htm.
Bill Parker will be teaching a
Four Day Practical EMI (Electromagnetic Interference)
Test and Fix Seminar on June 12-15, 2006, in Las Vegas, Nevada. Please visit http://
www.equipment-reliability.com/course10.htm for detailed information about this course.
Enroll by May 11th to take advantage of a $100 discount. If you and two more people from
your organization enroll before May 11th, each person can take another $100 discount.
Visit http://www.equipment-reliability.com/regist_form.htm to register.
Vibration and Shock
courses coming up
Wayne will teach short
courses on vibration testing,
shock testing, measurement,
analysis, calibration, HALT,
ESS and HASS at the
following locations:
May 17-19, 2006,
El Segundo (Los
Angeles), California
July 18-20, 2006,
Hillsboro, Oregon
August 22-24, 2006, Santa
Barbara, California
September 19-21, 2006,
Montreal, Canada
October 16-18, 2006,
Las Vegas, Nevada
December 5-7, 2006,
Detroit, Michigan
reactance is so high (nominal open
circuit) that proper circuit function is not
affected. Manufacturer design notes for the
sensor and receiver may provide insight for
suitable capacitor values.
As a general rule for many applications,
suppression capacitors in the range of 0.01
microfarad to 0.1 microfarad have been found
to be adequate. With capacitive reactance
Xc = 1/2fC ohms, the capacitive reactance
at 100 hertz for 0.01 microfarad is 159
kilohms, and for 0.1 microfarad the capacitive
reactance is 15.9 kilohms. For many sensor
circuits, these values may be high enough
to be considered open circuits which will not
adversely affect sensor circuit operation. For
an interference frequency of 100 kilohertz,
the capacitive reactance of the two capacitor
values are 159 ohms (for 0.01 microfarad) or
15.9 ohms (for 0.1 microfarad).
Capacitor Design and Installation Details
To achieve optimum suppression of higher
interference frequencies, the capacitor
design and installation are critical. The
capacitors should have inherently low
inductance. The self resonant frequency of
the capacitors (determined by the capacitor
value and equivalent series inductance,
where fresonance = 1/[2LC]) should
be higher than the highest interference
frequency of concern. The inductance
of the lead lengths of the suppression
capacitors add dire