EE 311: Electrical Engineering Junior Lab Grounding and EMI ...

EE 311: Electrical Engineering Junior Lab
Grounding and EMI Experiment
Purpose:
The purpose of this laboratory experiment is to demonstrate the potential problems that can arise due to poor
grounding and/or shielding. In order to demonstrate these effects in a practical (and somewhat repeatable)
laboratory situation, a few liberties are taken regarding the sources- the reader is left to make the connection between
the laboratory test situation and the practical issues involved in this work.
Two types of electromagnetic interference exist-- conducted and radiated. This laboratory focuses on conducted
interference, with short demo on radiated interference as the final step.
The issues involved in these subjects are quite extensive, and the work in this experiment should be viewed only as
an introduction to the subject. For further reading in these areas, the following books are good introductions:
1. Introduction to Electromagnetic Compatibility, by Clayton R. Paul, Wiley Interscience.
2. IEEE Standard 1100-1992. Recommended Practice for Powering and Grounding Sensitive Electronic Equipment
(IEEE Emerald Book).
3. Any number of FCC regulations and reports, such as Code of Federal Regulations, Title 47 (47CFR) Part 15,
Subpart B: “Unintentional Radiators.”
4. IEEE Std. 1250-1995. IEEE Guide for Service to Equipment Sensitive to Momentary Voltage Disturbances.
5. Noise Reduction Techniques in Electronic Systems, by Henry W. Ott, Wiley Interscience, 1988. ISBN 0-471-
85068-3.
Grounding:
The grounding issues can be stated simply as 1) Maintain a “clean” ground. and 2) Avoid ground loops.
It is convenient (but erroneous) to view “ground” as a large number of points which are always at 0 volts with
respect to each other (and with respect to the outside world). In fact, “ground” at the device level consists of a
conductor which may or may not be larger than other conductors, and likely will have a longer and more circuitous
path than other conductors- therefore leading to higher impedances in the ground path, particularly at the higher
frequencies. A clean ground for a computer room, for example, will consist of conductors of suitable size
connecting the case of each piece of equipment back to a central grounding point, with no loops and a minimum
conductor length. The central grounding point then is connected to the building (or dirty) grounding system with a
single lead. The electronics in each piece of equipment is typically tied through a common point to the case ground.
Note that there can be exceptions to this in specialized equipment. Also note that the case ground is NOT tied to the
neutral conductor of the ac power supplied to the equipment.
A ground loop is a path in a grounding system through which unintended currents can flow. For example, you may
have a sensor located on a compressor in an industrial location. The sensor is connected to the compressor case for
safety purposes- and the case is connected to the “dirty” ground. The sensor output is carried back to the controller
by a pair of wires labeled “signal” and “ground”, and the “ground” lead is connected to the clean instrumentation
ground- this creates a second connection between the clean ground system and building ground. Two things can
happen- 1) A potential difference between the clean and dirty ground system will exist and the ground lead between
sensor and instrument will carry stray currents; and 2) Currents will be induced to flow around the loop due to
inductive coupling. An improved approach is to ground the CASE of the sensor, but leave the sensor itself
(thermocouple, position sensor, limit switch, etc.) floating. One of the two sensor wires then be tied to ground AT
THE instrumentation location. Another alternative is to ground the sensor at its location, and to use a differential
amplifier to avoid a ground loop.

The circuit shown below illustrates a common situation: a sensor (thermocouple, strain gauge, etc.) is located some
distance from it's amplifier. There is some ground potential difference due to stray current flow. Connecting the
common point of Vi to ground grid 1 and the polarity side of Vi to a simple amplifier would cause amplification of
both the sensor voltage and the ground potential difference. Reconnection of the Vi common to ground grid 2 could
remedy this problem- but now the sensor is floating with some potential difference to the local ground voltage, which
can be a safety problem. The third approach we will see is the use of a differential amplifier, which allows Vi to be
tied to local ground but amplifies only Vi and not the ground potential difference.
Figure 1. A sensor at one location to be connected to an amplifier at another location.
Figure 2. Proto-board 1 schematic
On your laboratory bench, you are provided with a variety of equipment. Each piece of equipment is tied to the
bench ground wire (the green wire) through the third prong of the plug. In addition, the electronics common point of
some of this equipment is tied to the case, and hence to the green wire. In particular, you can count on any standard
oscilloscope to have the common lead tied to case. Also on this bench, the counter and the signal generator are
reference to case. The measurement terminals on digital voltmeters are typically NOT tied to case, and this is true
for both DVM’s you will be using in this experiment. You will also be using a signal generator from the instrument

oom that has three terminals - the ground terminal is connected directly to case and to the green wire through the
third prong of the plug. To reference the signal to ground, a grounding strap is provided which will tie one side of
the output to ground. A similar situation exists with the portable dc power supply on your bench, which you will use
for a ground source of dc current. The bench power supply, which has a ground terminals tied directly into the green
wire circuit.
In this experiment, connect your electronics to the bench green wire ground through a single point- the black lead on
one of the oscilloscope channels is a reasonable choice.
Lab Procedure:
1. On proto-board 1, set up the Hpxxxx oscillator as shown in Figure 2. Verify that the oscillator ground stud is
connected to the third prong of the power plug. The voltage divider is used to provide better control of the
voltage in the milli-volt range that we will be using. The voltage divider will be the only thing located on
the first proto-board that you have been issued. The middle terminal of the oscillator will be the common
point of the voltage divider. In the first set of experiments, the ground strap connecting the ground terminal
and the middle terminal will be on. Connect the output of this divider to a pair of studs on the proto-board.
This voltage will be referred to as V i in the following steps.
2. On proto-board 2, set up the op amp as an inverting amplifier with gain of 100. The circuit diagram is shown in
Figure 3. Take some care to make the layout clean and neat. Do some quick measurements to make sure
the amplifier is working. At the end of this step, you should have a working amplifier with channels 1 and 2
of the oscilloscope connected to V i and V o , respectively, and DVM's at both locations. There should be a
single ground on each proto-board through the oscillator ground strap on proto-board 1 and through either
the scope ground or power supply ground on proto-board 2.
Figure 3. Amplifier schematic—proto-board 2.
3. Connect v i on proto-board 1 to v i on proto-board 2 by a single lead from the respective hot sides. Put the ground
strap in place on the oscillator. Set the oscillator at 1000 hertz. Vary v i from its minimum value to 100
milli-volts, and measure the amplifier gain over this range. Plot the results. Also watch the oscilloscope
traces as you do this and note anything interesting. (Comment: You are taking a sinusoidal input and
amplifying it with a linear amplifier- what should the output be like? We are interested in the gain, which
you are measuring with an rms instrument. We are also interested in the wave-shape of the output).
4. Connect the portable dc power supply so that 0.6 amps or so is circulating around the bench ground system. This
simulates a ground loop with this much current flow- a very practical situation. You can complete this
connection by strapping the ground stud on the dc power supply to one of the active terminals of the dc
supply. Connect a lead from the other active terminal to the ground stud of the oscillator. Put the power
supply meter to measure 1.8 amps full scale, and adjust the power supply output. Note the effect of the dc
current on the two scope traces- make sure you have the scope set on the dc mode on both channels.
Comment: If you don’t observe anything different, talk to the lab instructor at this point. Repeat the
measurements of Step 3, plot the results, and note observations of the scope traces.

5. Turn off the dc ground current. You have been given a special light bulb circuit that takes current from the hot
wire of the plug through the light bulb and puts it onto the third prong rather than the normal neutral
connection. Plug this bulb into the far slot on the power strip on the rear of the bench. You are now
injecting about an amp of 60 hertz current into the bench ground system. Repeat the observations and
measurements of Item 4. Change the oscillator frequency to approximately 60 hertz, and observe variations
in waveform with small changes in frequency. What difference do you observe in the "noise"? Comment:
If no effect is noticed, change the location where the light bulb/scope/oscillator are plugged in until you get
a noticeable change. Double check that you have no ground loops in your setup.
6. Put the oscillator back to 1000 hertz, remove the light bulb, and put the dc supply back as it was in step 4. An
improvement on the previous results can be made by tying the v i common into the amplifier ground rather
than into the ground on the oscillator. Remove the ground strap from the oscillator, leaving the dc supply
on the oscillator ground stud. Make a connection from v i common on proto-board 1 to the common bus on
proto-board 2. Repeat the measurements of step 4. Note that this step provides a "clean ground" to protoboard
1 by putting in a common connection between proto-boards through which no circulating ground
current can flow. The drawback here is that Vi is not referenced to the local ground- this can create
problems.
7. A better solution yet is to connect the amplifier as a differential amplifier. Rewire the amplifier for the
configuration shown in Figure 4. Replace the ground strap on the oscillator. Repeat the measurements of
step 4.
Figure 4. Differential amplifier schematic.
8. In this type of work, the desired signal voltage is referred to as the differential mode voltage, while the voltage
between one of the oscillator leads and amplifier common (as measured by the scope here) is termed the
common mode voltage- common because both signal leads are floating above “ground” by this many volts.
Figure 5 shows an equivalent circuit with the ground loop potential difference shown here as the common
mode voltage above “true ground”. In this experiment, "true ground" is the ground on proto-board 2. The
common mode voltage Vcm is induced in the circuit by either the dc supply or the light bulb. The
differential mode voltage Vdm is the voltage supplied by the oscillator through the voltage divider. Figure
5 shows these voltages with the differential amplifier.
The goodness of an amplifier configuration is often judged by the common mode rejection ratio; see your
circuits book for a definition of this (see pre-lab step 4). Measure the common mode gain and the
differential mode gain, and determine the common mode rejection ratio for this circuit.
9. Summarize your findings.
10. Electromagnetic interference (EMI) takes two forms- conducted and radiated. We saw an example of conducted
interference in the previous section. In this section, we will look at radiated electromagnetic interference.
KEEP YOUR PREVIOUS SETUP INTACT. You have two leads of about 4-5 feet each carrying a signal

from Vi to the amplifier. (If they are shorter, arrange for leads of this length). These leads may form a loop
with a significant “window” area (if not, rearrange them). You know from previous work that magnetic flux
passing through this window will induce a voltage around this loop. This voltage will be seen as a
differential mode voltage by the amplifier, and create errors. Ask the instructor to provide some radiated
interference, and observe the resulting wave-shape on the oscilloscope. How can this form of interference
be reduced? Provide specific wiring examples, i.e., wiring type and connection topology.
Figure 5. Differential amplifier equivalent circuit showing common mode and differential mode voltages.
Pre-lab Assignment:
1. Determine op amp pin assignments for the LM741 op amp, dip package, and prepare a layout for proto-board 2.
2. Study the differential amplifier of Figure 4, and determine the gain of the amplifier shown.
3. Prepare a PSpice file for Figure 4 circuit and verify the gain of step 2. Include verification in the pre-lab report.
4. Find definitions of common mode and differential mode voltages, as well as common mode rejection ratio.
Determine a method to experimentally measure the common mode rejection ratio of the differential amplifier. What
is the common mode rejection ratio (CMRR) of an amplifier that amplifies the common mode and differential mode
signals with the same ratio?
5. For the circuit illustrated below, what restriction must be placed on the value of R IN in order to keep the ground
noise coupled into the differential amplifier to less then 0.1% of the signal voltage V S ? What is the CMRR of the
circuit expressed in dB for the calculated value of R IN ? Use manufacturer's datasheets to compare the calculated
value of R IN with the input impedance of a general-purpose LM741 series op amp, an AD624 precision
instrumentation amplifier, and an AD215 isolation amplifier. State, if any, the general relationship between an
amplifier's CMRR and its input impedance?

6. Determine the optimum cabling and grounding arrangement for the circuit illustrated below. The circuit consists
of a grounded, low-level low-frequency signal at location A; a differential amplifier at location B; and a grounded
load at location C. Do not use any transformers and assume that the source at A and the load at C must remain
grounded for safety purposes. Describe the benefits of your topology over other possible configurations.
Equipment/Parts List:
Equipment/Parts:
Bench Equipment Used:
1. HP oscillator
2. 2 proto-boards
3. 741 op amp oscilloscope
4. Resistors: 3- 1Kohm and bench power supply, +12 volts.
2- 100 kohm, 1- 50ohm
5. Special light bulb variable power supply
6. leads 2 dvm’s
Last revised: 8-25-02/jjc