Motion Lab Systems

Whisper User Manual

Distributed by Motion Lab Systems, Inc.

Roessingh
Research and
Development
Whisper User Manual
Published February, 2001
Revision 1.51
Intended AudienceThis manual is written for anyone using the Roessingh Research and Development Whisper EMG
Test Set. It is assumed that the user is familiar with the principals of the Electromyography equipment that they use, and
any associated EMG data collection system. Separate manuals for these products are available from their manufacturers.

The following are registered trademarks:

Whisper Roessingh Research and Development
Windows 98, Windows NT Microsoft Corporation
RData2 Motion Lab Systems, Inc.

Y2K compliance statement

The Whisper is Year-2000 compliant. It will not produce errors in date data related to the year change from
December 31, 1999 to January 1, 2000. It will handle leap years correctly.
All MLS products either do not use internal date codes or utilize specific, non-ambiguous representation,
handling and interpretation of centuries.

Electrical Safety
The Whisper is a battery operated electronic signal source that generates a sequence of test signals that simulate
some characteristics of an electromyographical (EMG) signal. The unit is not designed to perform any
electrical safety tests on your EMG system or replace any subject safety tests that are performed by a
biomedical engineering safety check.

The Whisper is manufactured by:

Roessingh Research and Development
Roessinghsbleekweg 33b,
7522 AH Enschede, Netherlands
Phone: +31 53-4875777
Fax: +31 53-4340849
e-mail: rrd@rrd.nl

Motion Lab Systems, Inc.

4326 Pine Park Drive • Baton Rouge, LA 70809
Phone +1 225 928-GAIT (4248) • Fax +1 225 928-0261
Email: support@motion-labs.com
http:\\www.motion-labs.com

Printed in the United States of America

© Motion Lab Systems, Inc. 2000
Contents
Introduction 1
The Whisper...............................................................................................................................1

Installation 3
Getting started ............................................................................................................................3

The Test Signal 9

A Functional description ............................................................................................................9
EMG signal specifications .......................................................................................................11

Using the Whisper 13

Standardized EMG testing .......................................................................................................13
Protocols for standardized testing ............................................................................................16
Interpreting the Test Results ....................................................................................................16
Whisper Applications...............................................................................................................18

A Whisper Tutorial 21
Introduction..............................................................................................................................21
Preparing for an EMG Test ......................................................................................................24
Gain Calibration .......................................................................................................................26
Recording the Test Signals.......................................................................................................33
Converting the data ..................................................................................................................36

The EMG Test software 43

Introduction..............................................................................................................................43
Using the EMG-Equipment-Tester software............................................................................46

File Descriptions 57
Protocol standardization...........................................................................................................57
A full Equipment Test..............................................................................................................61

Appendix A1 63
EMG-Equipment Tester Output ...............................................................................................63

Appendix A2 65
Technical Issues .......................................................................................................................65
Troubleshooting the Whisper ...................................................................................................66
Introduction

The Whisper
Congratulations on your purchase of the Whisper EMG equipment test package,
manufactured by Roessingh Research and Development and distributed in the United
States by Motion Lab Systems, Inc. From now on you can use the Whisper to test
your EMG instrumentation in an easy, fast, and reliable, standardized manner. The
Whisper can be used in seconds to perform a quick functional test or can perform a
specification check and comparison with previous test results.
A quick functional test of EMG equipment is based on the idea that a faulty EMG
system, and a normal system, will produce different EMG records when tested with a
known, reproducible EMG input signal. The inspection of a system's response to a
standardized EMG signal is therefore a good way to verify the performance of an
EMG instrument or an EMG data collection system. The Whisper generates this
standardized test signal for your EMG instrumentation and software included with
the Whisper helps you interpret the results.

Figure 1 - The Whisper is a compact, easy to use, system.

This manual provides complete information to enable you to connect the Whisper to
your EMG instrumentation and verify that the installation was successful.
The standardized test signal generated by the Whisper is described together with the
information that can be calculated from the signal. The manual describes each of the
Test Modes possible with the Whisper. Recording protocols and test parameter
calculation algorithms are discussed to allow you to determine the test parameter
values of your EMG instrumentation. After reading this manual you will be ready to
use the Whisper to obtain many useful insights into the performance of your EMG
instrumentation.

Whisper User Manual Introduction • 1

 The Whisper can be connected to any mono-polar or bi-polar EMG system that
generates either smoothed, rectified EMG signals or raw EMG signals and provides a
means of recording these signals to a disk file in a format that the EMG-Equipment-
Tester software (supplied with the Whisper) can read.

Figure 2 - A typical EMG test report generated using Whisper test data.

RData2 reads many common The Whisper package supplied by Motion Lab Systems includes two software tools
binary recording formats and to help the user analyze their recorded EMG signals. The EMG-Equipment-Tester
generates the ASCII files software (written by Roessingh Research and Development) which performs the
required for analysis by the EMG test signal analysis, and also generates a report of the Whisper test results.
EMG-Equipment-Tester The package also includes RData2, a file translation application from Motion Lab
Software. Systems, that generates the DST files (required by the EMG-Equipment-Tester
software) from C3D and Dataq binary data files.
The Whisper is CE-marked. The CE mark is the official marking required by the
European Community for all electronic equipment sold or put into service for the
first time, anywhere in the European community. It proves to the buyer or user that
the Whisper fulfills all essential safety and environmental requirements as they are
defined in the European Directives.

Electrical Safety
The Whisper is a battery operated electronic signal source that generates a sequence
of test signals that simulate some characteristics of an electromyographical (EMG)
signal. The unit is not designed to perform any electrical safety tests on your EMG
system or replace any subject safety tests that are performed by a biomedical
engineering safety check. Testing an EMG system with the Whisper must be in
addition to your regularly scheduled safety testing and is not a substitute for regular
biomedical equipment tests.

2 • Introduction Whisper User Manual

Installation

Getting started
The best way to avoid problems during installation of the Whisper is to read this
entire chapter before you start the installation. The chapter lists all of the items that
you should receive with the Whisper and describes the simple process of connecting
the Whisper to your EMG instrumentation.
In order to be able to install the Whisper properly, the package contains the
following items:
• The Whisper – an EMG signal generator.
• Four AA batteries.
• An EMG signal cable with LEMO connector.
• A ground reference cable.
• The EMG-Equipment-Tester software.
• The Whisper User manual (this document).
RData2 is a powerful binary In addition to the Whisper equipment and software, Motion Lab Systems includes a
to ASCII conversion utility copy of the RData2 file translation software. This software package can be used to
supplied by Motion Lab generate the DST files required by the EMG-Equipment-Tester software supplied
Systems, Inc. with the Whisper. RData2 reads C3D and Dataq data files and converts them to
CAMARC standard DST files. The items supplied with RData2 are:
• RData2 Software.
• RData2 User Manual.
• Customer Registration code.
Please check that you have all the Whisper and RData2 items before continuing the
installation. Please contact Motion Lab Systems if you can not find any of the items
listed. Our mail address, phone numbers and email address can be found at the front
of this manual.

Setting up and testing the Whisper

The Whisper is powered by four AA alkaline batteries (supplied) – these are easily
obtained at any hardware store. If you plan to make extensive tests with the Whisper
then you may want to purchase additional batteries.

Whisper User Manual Installation • 3

 Start by examining the Whisper and becoming familiar with the layout of the front
panel connections and controls.

Figure 3 - The Whisper front panel

The front panel of the Whisper is simple - from the left to right you find:
• EMG Ground (Signal Reference)
• EMG Signal Output
• Differential Mode Signal On/Off Switch
• Common Mode Signal On/Off Switch
• Start Button
• Power On/Off Switch
• LED indicator: Simulator On/Off/Active
The rear of the Whisper provides access to the battery compartment.

Figure 4 - The Whisper back panel showing the battery compartment.

Start the installation by checking that the POWER switch, on the front of the
Whisper is in the off position. The battery holder can be removed from the back of
the Whisper by pressing the two clips together in the direction shown by the two
arrows.

Figure 5 – The Whisper Battery pack

4 • Installation Whisper User Manual

 Remove the battery holder from the Whisper and insert the four AA batteries, paying
particular attention to the polarity of the batteries.
Once the battery pack is inserted firmly into the back of the Whisper you are ready to
generate EMG signals. The Whisper can be checked by turning on the POWER
switch at the right side of the front of the unit. The LED will glow green if the
batteries have been inserted correctly. If the LED does not light up then check that
the batteries are inserted correctly and that the battery pack has been firmly inserted
into the Whisper
If the batteries are inserted into the Whisper correctly and the LED lights when the
power switch is turn on then you can test that the unit is functioning by turning the
POWER switch on and pressing the START button, located to the left of the
POWER switch. The LED will change from green to orange-red as the Whisper
starts to generate an EMG test sequence.
The Whisper is now operational.
Two cables are supplied to enable you to connect the Whisper to your EMG system –
you will need to connect both of them to you EMG system. These are the EMG
signal cable and a ground reference cable.

Figure 6 – The EMG signal cable with LEMO connector.

LEMO p.n. FGG OB 302 The EMG signal cable has a LEMO connector at one end that connects to the EMG
CLA D45 Z w/ GMA OB 045 OUT connector on the left side of the Whisper front panel. The two output leads of
D6 the EMG signal cable have spade connectors.
The ground reference cable has a single connector that mates with the black ground
reference jack on the left of the EMG OUT connector, while the other end has a
standard safety DIN connector.

Figure 7 - The Whisper ground reference cable.

The ground reference cable should be connected to the to the grounded input of your
EMG measurement chain – normally the EMG ground or indifferent reference

Whisper User Manual Installation • 5

 connector. The EMG signal cable is connected to the input of your EMG
measurement chain. The EMG signal cable is supplied with two simple spade
connectors that are suitable for most common EMG surface preamplifiers. You may
need to modify this cable if your particular EMG input device does not interface to
these connections.
The standard Whisper test The EMG signal connector has a positive and negative lead. The cable should be
signal starts with 5 leading connected in such a way that the red lead on the EMG signal cable is connected to
pulses that check the the positive input of the EMG system. If you don't know which of the inputs of your
recording system polarity. EMG instrument is the positive one, then connect the cable in the most convenient
manner. A test will be described later that will check that the EMG signal cable has
been connected correctly.

Figure 8 – A typical EMG test setup

It's a good idea at this point to connect the Whisper to your EMG recording system
in order to test that the Whisper installation has been successful. EMG systems and
recording configuration vary considerably so these instructions are quite general.
You may need to read the instruction manual for your EMG system and recording
system in order to complete this test.
Prepare your EMG recording instrumentation to record a signal during at least 22
seconds. Choose the gain of your EMG instrumentation channel such that your EMG
instrumentation will not saturate when the standard Whisper test signal is applied.
Your EMG instrumentation should expect to see a maximum EMG test signal
amplitude of 1 mV.
If your EMG system has high and low pass filters then they should be set to allow
the maximum bandwidth signal to pass with minimum attenuation.
Set both the DIF. MODE (Differential Mode Signal) and COM. MODE (Common
Mode Signal) switches in the on position if you have bi-polar EMG instrumentation.
If your EMG instrumentation is mono-polar then set the COM. MODE switch off.
When the Whisper LED is Now the Whisper is ready for use - you can switch on the Whisper using the
green the unit is ready to start POWER switch on the right side of the front panel. The LED above the switch
generating an EMG test should light up green. At this point the Whisper is ready to generate a signal at
signal. surface EMG levels when you press the start button.
The Whisper will start generating an EMG test sequence as soon as you push the
START button, located to the left of the POWER switch. You will need to start your
EMG recording system at the same time as you press the Whisper START button.
It's a good idea to turn off the With the Whisper turned on, and the power LED green, press the START button on
Whisper when you have the Whisper and start you EMG data collection recording system. The LED above
finished an EMG test to the Whisper power switch will turn orange / red and your EMG recording instrument
preserve the batteries. will record a standard EMG test signal from the Whisper for the next twenty two
seconds. At the end of the EMG test sequence the Whisper LED turns green again
and you can stop your recording device and store your data.

6 • Installation Whisper User Manual

 Figure 9 - The standard test signal recorded by raw EMG instrumentation

Once you have completed an EMG test recording you can view the recorded EMG
data using your normal methods. If your EMG instrumentation records raw EMG
then your recording should look like the signal that is shown in Figure 9.
If your EMG instrumentation records a rectified and filtered EMG signal (also called
"envelope" EMG) then your recording should look like the signal that is shown in
Figure 10. Both illustrations show highly compressed EMG signals to illustrate the
entire EMG test sequence – each illustration covers a period of about twenty-two
seconds.

Figure 10 - The standard test signal recorded by "envelope" EMG instrumentation.

The Whisper is working and has been installed correctly if an EMG recording from a
known working EMG systems is comparable to the EMG signals illustrated in either
Figure 9 or Figure 10. Please consult the troubleshooting guide in Appendix A2 at
the end of this manual if you do not see these signals.

Whisper User Manual Installation • 7

8 • Installation Whisper User Manual
The Test Signal

A Functional description
Roessingh Research and Development have developed the Whisper as a tool for
quality assurance and routine checking of the performance of EMG instrumentation
in a biomechanics laboratory. Its primary goal is to detect degradation in the
performance or malfunctioning of the equipment in order to alert the users to any
problems with their system at an early stage. Information generated by the Whisper
allows the laboratory engineers and technicians to schedule regular maintenance to
avoid making invalid EMG measurements and /or recordings.
The basic idea of an equipment test is that a standardized EMG test signal will be
applied to the EMG instrumentation on a regular basis. The output of the EMG
instrumentation will be recorded and test parameter values will be calculated from
the recorded signal in order to obtain a record of the performance of the EMG
instrumentation and recording system.
By comparing the test results from a Whisper test with the results of earlier tests it is
possible to check that the EMG system and EMG recording channels for any signs of
degradation. A faulty EMG system will produce a different EMG record from a
functioning EMG system when both systems are presented with a standard EMG test
signal such as generated by the Whisper.
The standardized test signal is stored in EPROM's inside the Whisper. Each Whisper
contains an identical set of EPROM's and thus every Whisper generates an identical
EMG signal for each test. The signal consists of two components – these are:
• A differential mode component.
• A common mode component.
The differential component consists of five, positive, leading pulses that are followed
by a digitized EMG signal. The common mode component consists of a 40 Hz sine
wave. Each of the two components can be generated and applied to the EMG output
signal independently via the two DIF. and COM. mode switches on the front of the
Whisper.
Both signals are generated as soon as the START button is pushed and released to
allow the Whisper to produce a range of test signals under control of the user. The
ability to combine and separate the differential and common mode components of
the EMG test signal is invaluable in fault finding and testing EMG systems and
components.

Whisper User Manual The Test Signal • 9

 The standard test signal
The standard Whisper test signal consists of a differential mode signal with a
common mode component superimposed on the signal. This allows a number of
different EMG instrumentation parameters to be observed in a single test signal.

The differential mode signal

The differential mode signal consists of five leading pulses, followed by fifteen
seconds of standard EMG generated from a signal stored in the Whisper EPROM's.
The pulses are spaced 1 second apart. Their duration is one millisecond, their
amplitude is 500 μV. There is one second between the release of the START button
and the first pulse and one second between the fifth leading pulse and the stationary
EMG signal. The stationary EMG signal is fifteen seconds long. The Whisper stops
one second after the stored EMG signal has finished. Figure 11 shows a sample of
the differential mode EMG signal while Figure 9 illustrates the entire fifteen-second
sequence.

Figure 11 - Differential mode EMG signal.

The differential mode EMG signal, that is stored within each Whisper, is a simulated
signal so that it has a precise amplitude and known frequency components. This
avoids the problems inherent in recording a long, continuous burst of EMG from a
human subject. In Frequency parameters of the myoelectric signal as a measure of
muscle conduction velocity (IEEE Trans. Biomed. Eng. Vol. 28 (1981), p. 515-523),
Stulen, F.B. and De Luca, C.J. show, that a stationary EMG signal can be obtained
by band-pass filtering a white noise random sequence. The fifteen seconds of EMG
used by the Whisper were obtained by low-pass filtering and then high-pass filtering
fifteen seconds of white noise. The low-pass filter was a 2nd order, 80 Hz filter and
the high-pass filter was a 1st order, 40 Hz filter. The digital filter used a 1024 Hz
sample rate.
The resulting EMG test sequence has been scaled such that the standard deviation of
the amplitude of the sequence equals 200. The overall z-transfer function of the
digital band-pass filter is:

0.035719( z 3 + 1.0022( z 2 - z) - 1 )
H(z) = 3 2
z - 1.9792 z + 1.2948z - 0.2804
The simulated EMG data is provided in the file EPROM_e1.DST on the diskettes
supplied with the Whisper. This file is a standard ASCII file and can be read by
many common spreadsheet programs such as Lotus, Quattro, and Excel etc). If the
DST data is read into any of these products and plotted then you should get a display
that is similar to that illustrated in Figure 9. The y-scale values correspond to the
amplitude of the differential mode signal at the output of Whisper (in microvolts).
Thus a value of 345 in the data file is converted into an output of 345 μV at the EMG

10 • The Test Signal Whisper User Manual

 signal output. The file EPROM_e1.DST thus provides you an exact description of the
standard differential mode test signal.

The common mode signal

The common mode signal generated by the Whisper consists of three parts. During
the first six seconds, the common mode signal remains 0.0 Volt. As soon as the
differential mode signal starts to output the stored EMG signal, the common mode
signal becomes a sinusoidal signal with an amplitude of 1.00 Volt at a frequency of
40 Hz. This common mode signal remains present for fifteen seconds and then
returns to 0.0 Volt during the last second of each simulation.
The common mode signal is stored in the same way as the differential mode signal
and is also converted on a one-to-one basis to each differential output of the
Whisper.

EMG signal specifications

The standard EMG test signal, used by the Whisper to test the performance of your
EMG recording instrumentation has the following characteristics. These can be
verified from the file EPROM_e1.DST that is installed in the data subdirectory to the
EMG-Equipment-Tester software. A full description of the format of this file is
given in File Descriptions later in this manual.

Test parameter Value Unit

EMG signal – RMS Value 200 [+/- 5 %] μV
EMG signal – Mean Value 0.1 [+/- 5 %] μV
EMG signal – Fmed 59 [+/- 1 %] Hz
EMG signal – Fmode 31 [+/- 1 %] Hz
Leading Pulse Width 1 [+/- 1 %] ms
Leading Pulse Amplitude 500 [+/- 2 %] μV
Interpulse time 1.0 [+/- 1 %] seconds
Common Mode Amplitude 1.0 [+/- 5 %] Volt
Common Mode Frequency 40 [+/- 1 %] Hz
Table 1 - Standard Test signal values

• RMS-value indicates the Root Mean Square value of the fifteen

seconds of digitized EMG signal.
• Mean is the mean value of the 15 seconds of digitized EMG signal.
• Fmed is the median frequency of the power spectrum of the first two
seconds of the digitized EMG signal.
• Fmode indicates the frequency at which the power spectrum of the first
two seconds of the digitized EMG signal is at its maximum.
• Pulse width and Amplitude refer to the 5 leading pulses. The interpulse
time is the time between 2 successive leading pulses.
• Common Mode Amplitude is the amplitude of the sine wave.
• Common Mode Frequency is the frequency of the sine wave.

Whisper User Manual The Test Signal • 11

12 • The Test Signal Whisper User Manual
Using the Whisper

Standardized EMG testing

The Whisper has been developed to provide a well-defined, reproducible test signal
for standardized EMG equipment testing. Standardized testing can be performed in a
number of separate steps. First the EMG test signal from the Whisper is applied to
the input of the EMG instrumentation under test. The output of the Whisper is then
recorded through the EMG instrumentation. The recording is then evaluated and
parameters calculated from the recorded data. Finally, the calculated values can be
either compared to the known input values or compared with the results of prior tests
in order to be able to draw conclusions about the performance of the EMG recording
system and EMG instrument under test.
In order to be able to draw reliable conclusions about the performance of the EMG
instrumentation, it is important that each of these steps is performed in exactly the
same standardized way each time the equipment is tested. This chapter will discuss
how all this can be achieved.
Since the differential mode and common mode component of the Whisper EMG test
signal can be selected independently, the standard test signal can take any one of four
different configurations. Each of the four configurations can reveal different aspects
of your EMG instrumentation.

Standard EMG test modes

The standard EMG test signal, stored within the Whisper, has been chosen so that
several EMG instrumentation features can be quickly tested. Since both the
differential and common mode components of the EMG test signal can be controlled
independently, it is possible to define four operation modes.

Mode Dif. Com. Test options recording

mode mode length
0 On On Bi-polar EMG: EMG and timing parameters. > 22 sec
1 On Off Mono-polar EMG: EMG and timing parameters. > 22 sec
2 Off On Bi-polar EMG: Common Mode Rejection Ratio > 15 sec
3 Off Off Bi/Mono-polar EMG Noise and offset parameters > 10 sec
Table 2 - Operation Modes and Test Features.

Whisper User Manual Using the Whisper • 13

 It is worth noting at this point that all the Whisper tests produce results that measure
not only the EMG system but also the data recording and measurement system that is
used to access the EMG data. The results reported from a series of Whisper tests
may provide as much information about the performance of the analog EMG
recording system as the EMG instrumentation that is used to acquire the EMG
signal. It is important that the Whisper user considers all possible error sources
before evaluating the results of a series of Whisper tests.

Mode 0: EMG and Common Mode

In mode 0 the EMG parameters, the timing parameters and the dynamics of any
bipolar EMG instrument can be tested – this includes almost all common surface
EMG preamplifiers. Mode 0 applies a differential EMG signal to the EMG
instrument under test and superimposes a common mode component over the EMG
signal. This tests the response of your EMG instrumentation to a differential mode
signal in the presence of a large common mode signal. Modern EMG preamplifiers
should reject the common mode component and amplify only the differential EMG
signal.

Mode 1: EMG only

Mono-polar EMG instrumentation (often used with traditional needle electrodes –
not to be confused with fine wire electrodes) can not reject common mode
disturbances. Therefore this mode does not include a common mode signal and only
applies the differential EMG component during tests to determine the EMG
parameters, the timing parameters and the dynamical properties of the EMG
instrumentation.

Mode 2: Common Mode

Mode 2 allows testing of the Common Mode Rejection Ratio (CMRR) by applying a
large common mode signal to both EMG signal lead simultaneously. Since mono-
polar EMG instruments can not reject common mode disturbances, this mode is only
of interest for bipolar EMG systems.

Mode 3: Noise and Offset

Mode 3 allows testing of the noise and offset properties and can be applied to any
EMG recording instrument. It does not produce any output signal other than the
background noise level of the Whisper, which is present in all other tests. This
allows us to calculate the signal to noise ratio of the EMG instrumentation.

Testing EMG instrumentation

The most important task of an EMG recording instrument is to measure and process
EMG signals in a reliable and reproducible manner. Using the Whisper Mode 0 and
Mode 1 EMG test signals it is possible to check that your EMG instrumentation is
working correctly, since in these modes the differential mode component will be
applied to the EMG instrument under test. From the fifteen seconds of recorded
EMG a number of parameters can be determined. The most important parameters
are:
• The RMS value of the recorded EMG signal.
• The mean value of the recorded EMG signal.
• Median and mode frequency of the power spectrum of the recorded
EMG signal.

14 • Using the Whisper Whisper User Manual

 Since the precise values of the EMG signal, stored in the Whisper, are known, we
can calculate the expected parameter values and compare them to the values that we
measure. A regular EMG instrumentation test using the Whisper allows verification
of the gain, the offset and the frequency characteristics of any EMG instrumentation.

Timing and Dynamics

The timing and the dynamics of the EMG instrumentation and recording system can
be tested using the Whisper Mode 0 and Mode 1 EMG test signals by studying the
five leading pulses in the differential EMG signal. Since the shape and the time
between the leading pulses is exactly known, they can be used to check the:
• The sample rate of the EMG recording system.
• The dynamic response of the EMG recording system.

Common Mode Rejection Ratio

By applying only the common mode component (a Mode 2 test) to an EMG
instrumentation system, it is possible to evaluate:
• The Common Mode Rejection Ratio of the EMG instrument.
The CMRR test (Mode 2) can During a Mode 2 test the common mode signal component is applied to a bipolar
be very informative if you are EMG instrument. Since this signal is present at each of the differential outputs of the
able to view the live output of Whisper, the EMG instrument should reject this signal (a 40 Hz sine wave). It is
the EMG system under test. known that the amplitude of the sinusoidal signal from the Whisper equals 1.0 Volt.
Poor CMRR performance is Thus a simple measurement of the amplitude of the sinusoidal present in a Mode 2
easy to spot as it produces a recording enables us to measure the extent to which the Common Mode signal has
large signal output during been suppressed. This is called the Common Mode Rejection Ratio and is important
this test and instantly because it is the primary method by which bi-polar EMG recording systems
identifies a problem EMG eliminate line frequency signals (50 or 60 Hz depending on your location) from the
channel. EMG signal present on the subject.
It is worth noting that there is also a common mode component to the standard bi-
polar EMG test signal (Mode 0) and that any EMG instrument that performs poorly
with a Mode 2 test will probably need to be evaluated with Mode 1 EMG test signals
instead of Mode 0 test signals.

Noise and Offset

It is possible to test:
• The noise and any DC offset in the EMG instrumentation
This is done by applying neither a common mode nor a differential mode signal to
the input of the EMG instrument (Mode 3). Although it seems unusual to use the
Whisper for this, it is very useful since the input of your EMG instrumentation will
be forced to 0.0 Volt (ground) if you switch off both signals.
The signal from the Whisper will, however, still contain the same noise and
frequency components that may be present during the other EMG tests. Thus, by
recording the null output signal from the Whisper, you can measure any noise or
other electrical disturbances, which are present in your EMG measuring and
recording chain and are superimposed on your EMG recordings.
You can evaluate the noise contribution of the Whisper by comparing two Mode 3
tests – one test sessions being recorded with power applied to the Whisper and the
second test session with the Whisper turned off.

Whisper User Manual Using the Whisper • 15

Protocols for standardized testing
A standardized measurement protocol is of major importance for reproducible
equipment testing. It is therefore recommended to use the following protocol every
time you execute an equipment test:
1. Connect the Whisper to your EMG instrumentation as described earlier.
2. Select the test mode. Set the differential mode and common mode switches
in the appropriate positions.
3. Prepare your recording instrumentation such that it can record the standard
test signal without saturating for at least twenty-two seconds.
4. Switch the Whisper on.
5. Push the Start button, and commence recording the EMG signal, release the
Start button within one second and check that the LED turns orange / red.
6. Wait until the LED turns green again.
7. Store the recorded data for future analysis.
8. Switch the Whisper off.
This protocol can be used in every test mode. In order to be able to extract test
parameter values from a recording, it is necessary to verify whether the recording has
been made in an appropriate way. For the following test modes the recording should
therefore satisfy the following criteria:
Mode 0 / mode 1:
• Be sure that you have recorded each of the five leading pulses.
• Be sure that you have recorded fifteen seconds of digitized EMG.
Mode 2:
• Be sure that you can see a sine wave in the recording after six or seven
seconds.
If your recording satisfies the particular criteria it is possible to determine the values
of the test parameters. If your recording does not satisfy the particular criteria you
need to execute another recording following the protocol described above.

Interpreting the Test Results

The results of the Whisper tests contain a lot of information about the performance
of the EMG instrumentation under test. In this section we will explain how the
various results can be evaluated.

Sample Rate
The sample rate of the recording device can be determined by counting the average
number of samples between any two consecutive leading pulses produced at the start
of a Mode 0 or Mode 1 recording. Since the five leading pulses repeat at a fixed rate
the average of the number of EMG data samples between the five pulses must equal
the sample rate of the recording device.
The measured sample rate of a Whisper test should equal the sample rate of your
EMG data collection system.

16 • Using the Whisper Whisper User Manual

 Amplitude parameters
Since the Whisper generates exactly the same EMG test signal during every Mode 1
and Mode 2 test you can easily measure the gain and any DC offset in your EMG
instrumentation and data collection system. The RMS and the Mean values of the
input EMG signal are known – these are 200 μV RMS with a Mean of 0 μV. By
measuring these values from the recorded EMG data, after it has passed through the
EMG instrumentation and data recording system, you can learn a great deal about the
performance of the entire EMG instrumentation and recording system.
In order to measure the EMG amplitude parameters reliably it is important that the
RMS and Mean values are measured correctly from the EMG portion of the recorded
signal. If the five leading pulses and the recording system sample rate have been
determined, it is simple to locate the start of the EMG signal since it starts exactly
one second after the last leading pulse and is exactly fifteen seconds long.
By calculating the RMS and Mean values of the fifteen seconds of standard EMG
test signal in a Whisper Mode 0 or Mode 1 recording you can determine the gain and
DC offset parameters of the EMG instrumentation and data recording system under
test.

Frequency parameters
An EMG recording instrument can be regarded as a chain of filter and amplifier
stages. In general the settings of these stages can be changed (gain, cutoff
frequencies, bandwidths, etc.). Each of these stages has its particular task and
together they ensure that an EMG signal will be recorded without distortion, with
minimum noise or disturbances and maximum common mode rejection.
The power spectrum of the standard test signal produced by the Whisper is known
and consequently the value various frequency parameters of the power spectrum
(Fmed, Fmode etc.) are known. In order to be able to record the EMG signal without
distortion, the bandwidth of the EMG power spectrum has be equal to or preferably
less than the bandwidth of the whole recording chain. By calculating the frequency
parameters of the recorded EMG signal produced by the Whisper, it is easy to check
that the various filter stages of the EMG instrumentation have been set correctly.
The frequency parameters are calculated from the first two seconds of digitized
EMG signal this is generated in any Mode 0 or Mode 1 recording. Once they have
been measured the smoothed Power spectrum can be calculated. Smoothing yields
multiplication of the power spectrum with a Papoulis window. The median and the
mode frequency can be determined from the power spectrum.

Dynamic response
In theory, the impulse, or dynamic response, of any electronic system contains all the
information about the dynamics of a system. Any changes in the dynamic response
of the system will change the shape of the impulse response. The impulse response is
therefore a useful tool to evaluate the dynamic response of any data recording
system.
You must obey the Nyquist In order to measure the impulse response of your EMG instrumentation you must
sampling theorem when make either a Mode 0 or a Mode 1 recording. The response of your data recording
recording data from the system and EMG instrumentation to each of the five leading pulses reveals the
Whisper to test the impulse impulse response of the entire EMG instrumentation and recording system to a one
response of any system. millisecond, 500 μV pulse.

Whisper User Manual Using the Whisper • 17

 Common Mode Rejection Ratio
The Common Mode Rejection Ratio (CMRR) indicates the extent to which any
common mode components of a recorded signal will be rejected by a differential
amplifier. The differential mode component will be amplified by the EMG
instrumentation measuring chain and can be used to calculate the overall system
gain. The Common Mode Rejection Ratio can therefore be determined by dividing
the amplitude of the Whisper's common mode sine wave (40 Hz, amplitude 1.0 Volt)
by the amplitude of the sine wave recorded in a Whisper Mode 2 recording. The
Common Mode Rejection Ratio of your EMG instrument is the 20-log value of this
ratio. In general most EMG systems should have a Common Mode Rejection Ratio
of greater than 80 dB.
The Whisper generates a test signal for Common Mode Rejection Ratio calculation
in Mode 2. The first five seconds of a Mode 2 signal consists of a quiet baseline
followed by a low-level sine wave. The amplitude of this sine wave will depend on
the ability of your EMG amplifier to reject Common Mode signals.

Noise and offset parameters

A Whisper Mode 3 recording enables you to measure the noise and DC offset
parameters of your EMG instrumentation. In this mode the output of the output of
the Whisper is 0.0 volts and contains only the internal noise that is inherent in any
signal system. By measuring the RMS and Mean values of the recorded EMG signal
in Mode 3, it is possible to calculate the amount of noise and the offset of the EMG
instrumentation and data recording system.
While the internal noise generated by the Whisper is very small it is not zero. This is
possible to measure the internal noise of the Whisper by comparing the results of a
Mode 3 recording with the results of a similar recording made when the Whisper is
switched off.

Whisper Applications
The Whisper can be used to perform a number of different tests within any clinical
gait or biomechanics laboratory:
• Multi channel equipment test
• System performance test
• System specifications test
• System comparability test
Each of these tests will be described shortly in the following sections.

The multi-channel equipment test

The goal of a multi-channel equipment test is to compare the performance of a
number of EMG channels in a measurement chain with each other. This allows you
to determine if the performance of any particular channel has changed or is different
from another channel under test. In general, each EMG channel is a separate unit in
the recording chain. As a result, errors or defects in an EMG instrumentation system
tend to affect only one channel rather than every channel in the system. A multi-
channel equipment test is therefore a useful tool to detect problems in individual
EMG channels that might otherwise be missed.

18 • Using the Whisper Whisper User Manual

 In a multi-channel equipment test the Whisper is connected to each EMG channel
under investigation and the response of each channel to the standardized EMG test
signal is recorded simultaneously. You can make a quick assessment of the EMG
channel quality by looking at these recordings. If the recordings are in phase and the
amplitude of the individual signals is similar then it's likely that the performance of
the EMG channels will be acceptable.
In order to obtain a more A full multi-channel equipment test consists of a number of recordings. For a bipolar
precise measurement, we can EMG instrument it is necessary to make three recordings: a Mode 0 recording to test
calculate a number of the timing properties and the dynamic properties of each channel, a Mode 2
standard parameter values recording to test the Common Mode Rejection Ratio of each channel and a Mode 3
for each EMG channel. By recording for noise and offset testing. Figure 12shows an example of the results of a
comparing these values to full equipment test of a bipolar raw EMG instrument using the EMG-Equipment-
each other, it is easy to draw Tester software, supplied with the Whisper.
evaluate your EMG channel
performance.

Figure 12 - Test results of the full equipment test of a 4 channel raw-EMG instrument.

As you can see, four separate EMG channels have been tested. Eleven unique
parameter values have been calculated for each channel and the impulse response has
been recorded. This report may be printed out and saved as a record of the test and
the EMG system functionality.

The system performance test

The goal of a system performance test comparison to detect any degradation,
malfunction, or change in the EMG system and recording instrumentation in order
detect these problems at an early stage. Therefore, a system performance test will
compare the results of a recent EMG system test to the results of an earlier test. In
this way, invalid EMG measurements can be avoided and appropriate maintenance
actions can be undertaken.
The starting point of a systems performance test are the results from an earlier
recorded test of the EMG instrumentation. The saved results can be compared to the
results of a new equipment test. If you perform a full EMG system performance test
on a regular basis (for example once a week), you should find that the week to week

Whisper User Manual Using the Whisper • 19

 changes of the test parameter values are very small (< 1.0 %), indicating the EMG
instrumentation and data collection system are working properly.
It is recommended that the Any large deviation in the test parameter values will suggest that the performance of
Whisper tests be run on a your EMG instrumentation has changed in some way. It is almost impossible to
regular basis in order to make any general statements here about the values of specific deviations in
maintain a history of the parameter values since they depend a lot on the type of instrumentation and the
performance of your EMG environment in which you execute your measurements, etc. When any potential
instrumentation. problem is observed, the test data produced by the Whisper is invaluable in assisting
the biomedical engineer in determining the cause of the problem.

The system specifications test

The system specifications test is similar to the system performance test, but instead
of comparing the test result of an equipment test with results from earlier recordings
made in your own laboratory, you compare your test results with results provided by
the EMG instrumentation manufacturer.
If you obtain a set of Whisper tests that have been recorded by the EMG system
manufacturer then you can use these to confirm that the EMG system conforms to
the specifications. Such a recording can be used as a 'standard' for verification of the
specifications of your EMG instrumentation. This is especially useful when an EMG
system is first installed in a laboratory since this allows you to verify that the EMG
instrumentation has been installed properly and is functional.

The system comparability test

The system comparability test allows users in separate laboratories with different
EMG systems and data collection instrumentation to verify the performance of their
systems and determine if the instrumentation in both laboratories is functionally
similar enough to allow exchange of EMG records. This test is based on the concept
that the separate laboratories apply the same standardized test signal to their
equipment, calculate the same test parameters from the recordings and compare the
values of these test parameters with each other. Differences between test parameter
values can only be attributed to differences in the EMG instrumentation in both
laboratories.
If you conclude that the output of the instrumentation of the separate laboratories is
different, you may conclude that the equipment is functionally different and that data
recorded and processed in these two laboratories may not be directly comparable.
If you conclude that the output of the separate EMG instruments is the same, then
you can consider this as a reasonable indication that the equipment is functionally
similar and that the data recorded and processed in these two laboratories may be
directly comparable.

20 • Using the Whisper Whisper User Manual

A Whisper Tutorial

Introduction
It is very simple to record EMG test data from the Whisper but care must be taken to
ensure that all the parameters of the recording are carefully controlled in order to
produce meaningful results. This chapter will illustrate this using a pair of different
EMG systems (Motion Lab Systems MA-100 and Motion Lab Systems MA-300
EMG systems) together with a Dataq analog data acquisition system on an IBM
compatible PC (100MHz 486 with Windows 95 and 64Mb of memory).
The Whisper can also be used Analog data recording is a complex subject and it is very helpful to consider all the
as a quick check of an EMG parameters that may affect your test data before you start to use the Whisper. This is
system by simply reviewing especially helpful as most of the parameters that affect the test recordings that you
the output of a Mode 0 EMG make with the Whisper will also affect your day to day analog recordings in the
test signal via any convenient laboratory. It is important to remember that the tests that you will be performing will
display device such as an allow you to investigate the performance on not only your EMG system but also
oscilloscope or analog signal your analog data acquisition system as well as the parameters that you use to sample
monitor. your analog data during your normal day to day operations.

Setting the Gain

Many EMG systems have a number of different gain settings and it is important to
select the optimal gain setting to record the test signals that are produced by the
Whisper. The largest EMG test signal is produced when the Whisper is in Mode 0 or
1, depending on the type of test that you are performing. For a typical surface EMG
system this will be a Mode 1 signal which has a maximum differential EMG signal
amplitude of 500uV. There are three factors that must be considered:

Electrode gain
Both the EMG systems used in this tutorial use pre-amplified surface electrodes that
have an internal gain (amplification) factor (MA-100 pre-amplified electrodes have a
gain of x320 while the MA-300 pre-amplified electrodes have a gain of x20). It is
important to know the gain of any pre-amplified electrode.
When testing any EMG system that uses pre-amplifiers you must connect the
Whisper EMG test set directly to the electrode disks that are normally placed on the
subjects skin. If you are testing an EMG system that uses passive electrodes then
you must connect the Whisper directly to the EMG system inputs that would
normally connect to Gel or other passive electrodes on the subject.

Whisper User Manual A Whisper Tutorial • 21

 System Gain
Both the MA-100 and MA-300 systems support a range of different gain settings for
each EMG channel.
The MA-100 EMG system uses an individual variable potentiometer gain control for
each EMG channel on the back-pack to allow the user to adjust the EMG level that
the system detects on the subject. This gain control is adjusted by the user to record
an optimum maximal EMG signal on each channel, as determined by the individual
LED displays. As a result, the channel gain control must be set by generating a
Mode 0 EMG test signal from the Whisper and adjusting the gain control so that the
maximal signal is recorded without overloading the system. Once the gain control is
set, the precise channel gain can be calculated by making a test recording from the
internal calibration source built into the MA-100 back-pack.
The MA-300 EMG system uses preset switches to determine the individual channel
gains. Channel gain settings are marked in terms of the maximum peak to peak input
that can be applied for a given switch setting. The MA-300 switch setting #3
indicates that this setting can handle EMG input signals of up to ±875μV peak to
peak. This is the optimum input range on an MA-300 system for the Whisper tests
which generate a differential EMG signal in the range of +742μV to -670μV.

ADC input range

The analog recording system must be able to accurately record the signals generated
by the equipment under test. The maximum output signal generated by an MA-300
EMG system is ± 5 Volts which means that the recording ADC sub-system must be
set for a range that will handle signals of this amplitude. Most common ADC sub-
systems will handle signals in the range of ± 5 Volts or ± 10 Volts. Depending on
the ADC system, you may be able to directly select a range of ± 5 Volts, or have a
fixed range of ± 10 Volts and the option to apply a gain of x2 to provide an
equivalent range.
The Dataq analog data acquisition system supports the sampling of analog data with
a range of ± 5 Volts – this is appropriate for the EMG signal levels that are produced
by both the MA-100 and MA-300 systems.

Frequency Response
It is important to consider the frequency response of both the Whisper test signal and
the EMG system under test as both of these items will determine the minimum
sampling rate that will be required from your data recording system. Many EMG
systems include a set of user-adjustable filters that allow the user to tailor the
frequency response of the system to match the analog recording system or filter the
EMG signal to remove noise and other artifact before recording and measurement.

High Pass Filters

If the EMG system contains variable High Pass filters (these remove low frequency
artifact components from the raw EMG) then they should be set to the lowest
frequency setting available. The MA-100 contains a variable High Pass filter that is
controlled via a connector on the rear of the unit (CONTROL) or via an optional
switch box. The filter should be set to the minimum value of 20Hz. The standard
MA-300 system does not contain a High Pass filter so this setting can be disregarded
unless you are using the optional High Pass filter or an external filter. In either case
the High Pass filter should be set to the lowest frequency available or turned off.

22 • A Whisper Tutorial Whisper User Manual

 Low Pass Filters
If the EMG system contains variable Low Pass filters (these remove high frequency
noise and other components from the raw EMG signal) then they should be set to an
appropriate value. Figure 13 shows the spectral content of the test signal generated
by the Whisper and demonstrates that the signal does not contain any components
above 300Hz. If you only consider the spectral response of the EMG signal then you
could choose a Low Pass filter point that is too low and removes part of the Whisper
test signal.

Figure 13 - Whisper EMG spectrum (Hz)

However, the test signal generated by the Whisper also contains a number of short
timing pulses that are also affected by this filter setting. As a result it make sense to
choose a higher filter setting to avoid unnecessarily attenuating the five leading
pulses that precede the EMG test signal. A suitable setting for the internal MA-300
Low Pass filter would be #5 (750Hz) or #4 (1,000 Hz) while the MA-100 can be set
to either 600 Hz or 1,300 Hz via an optional filter control box.
Some EMG systems do not have adjustable Low Pass filters in which case you can
consider the entire EMG system to have a Low Pass frequency response that is
equivalent to the maximum frequency that the system can reproduce. You may need
to consult the manufacturers literature to determine the exact bandwidth of the EMG
system under test.

Choosing a Sample rate

It is important to select a suitable sample rate to record the test signals generated by
the Whisper. In order to do this you must know the distribution of the signal
spectrum that you intend to sample. Figure 13 shows the spectral content of the test
signal generated by the Whisper and demonstrates that the signal does not contain
any components above 300Hz. Given this information, analog sampling theory
dictates that the EMG signal must be sampled at a rate of at least 600Hz to prevent
the generation of aliasing artifact in the recorded signal. Since you intend to analyze
the recorded EMG signal you should select an analog sampling rate that is at least
double the minimum required. If at all possible you should set your analog recording
system to sample the EMG data at a rate of at least 1800 to 2000 samples per second
to provide a signal bandwidth of 900 to 1000Hz in practice.

Whisper User Manual A Whisper Tutorial • 23

 Sampling Characteristics
There are a wide range of different sampling techniques and methods and it is
important to understand the different options that your ADC card offers in order to
generate clean and accurate data recordings. Many of these options are unique to
particular manufacturers or data collection cards so it is quite possible that you may
have additional options to those discussed below. In addition, many inexpensive
ADC cards offer only a simple interface without any of these options.

Oversampling
Many Analog Data Collection systems allow you to sample the analog data at very
high rates – often much higher than you would normally record the sampled data.
This is a very useful feature that allows the ADC card to provide some measure of
anti-alias filtering of the input data as the system can process a larger number of
analog samples to generate a single sample for recording. Oversampling usually
offers a number of options for selecting a data sample:
• Average – a number of data points are averaged together to generate a
sample data point. This method is very effective in removing high
frequency noise from the sampled data that would otherwise appear as
aliasing artifact.
• Last Point – this simply returns the most recent data point as a sample.
• Maximum – this returns the data point with the greatest amplitude as
the sample point.
• Minimum – this returns the smallest data point as the sample point.
When recording test data for use with the Whisper you will find that the Average
and Maximum sampling methods produce the most accurate data. The Maximum
method will usually provide a better record of the five timing pulses when your data
recording sample rate is low. In general the Average method is the most appropriate
for day to day EMG recording.

Differential Inputs
Most Analog Data Collection systems offer the option to recording the data using a
differential input – this is generally only appropriate if the EMG system under test
produces a differential output signal. Most common ADC systems will use a single
ended input arrangement.

Preparing for an EMG Test

If you use an EMG system in regular day to day data acquisition then all you should
have to do in order to make a recording is to connect the Whisper EMG Out cable to
the two differential EMG inputs and connect the separate ground lead to a suitable
ground connection on the EMG unit. In the case of the MA-100 and MA-300 EMG
systems used in this tutorial this means connecting the EMG Out cable to the pre-
amplifier inputs (stainless steel disks on the pre-amplifier body) and connecting the
separate ground lead to the green DIN connectors on the subject back-back. This is
illustrated in Figure 8 on page 6.

24 • A Whisper Tutorial Whisper User Manual

 Checking the Signal Levels
If this is the first time that you've made a recording using the Whisper then you may
want to connect up the system and try a few trial recordings to check that you have
the Whisper EMG Out cable polarity correct. The correct signal is shown in Figure
14 where the five leading pulses in a Mode 0 or Mode 1 test signal are positive going
signals although you may see some overshoot of the signal depending on your EMG
system characteristics. In general the positive leading edge of the signal will have a
greater amplitude than the –ve going overshoot.

Figure 14 - the five leading pulses of a Mode 0 or Mode 1 signal should be +ve.

Because the signal levels For first time recordings it is also a good idea to make a trial Mode 0 or Mode 1
generated by the Whisper are recording and check that the recorded signal levels are correct – the Whisper EMG
very low it is essential that all test signal should be clearly seen and should utilize a significant portion of the ADC
the equipment used in the test input range as shown in Figure 14 if the EMG system gain is set correctly and the
is grounded and shielded. analog data acquisition system is setup correctly.
There are two possible problem areas with the EMG test signal at this point – the
gain may be too high or it may be too low. In each case you should check that the
analog data acquisition system is set up for the expected EMG system output range
before adjusting any part of the EMG system. Both the MA-100 and MA-300 EMG
systems used in this tutorial have user adjustable gain controls and these can be
adjusted if necessary to produce a good test signal.

Figure 15 - Too much gain causes signal clipping

Whisper User Manual A Whisper Tutorial • 25

 If the EMG system gain is set too high then the EMG test signal generated by the
Whisper will be clipped as shown in Figure 15 – note the extremely flat appearance
of the EMG signal. This produces distortion of the test signal and will cause the
EMG testing software to produce erroneous results that do not accurately reflect the
system performance. Signal clipping can be caused by an incorrect choice of data
acquisition input levels as well as by the selection of too large an EMG channel gain.
If the EMG system gain is insufficient then the EMG signal will appear to be very
low as shown in Figure 16. While this is not so much of a problem as having too
much gain it is still not an optimal situation for either day to day operation of your
EMG system or testing with the Whisper EMG system. If your EMG system has
variable gains for the EMG channels then you should increase the gain to try and
obtain a signal closer to that illustrated in Figure 14. This problem can also be
caused by an incorrect choice of data acquisition input levels where the ADC card
range is much greater than the output level of the EMG system.

Figure 16 - Too little gain causes the signal to be lost in the background noise.

If you are working with an EMG system that does not have adjustable gain controls
then it is likely that you will either see a signal that is the correct level or you will see
a signal that appears to be too low. While it is possible to use the Whisper and
obtain accurate results with a low EMG test signal this is not an optimal
configuration for either testing or regular use.

Gain Calibration
One of the principal test performed by the Whisper is to report the EMG recording
channel gain. This test requires that the recording levels are correctly set for the
entire EMG signal recording chain. In other words, we expect that an input signal of
one millivolt will be reported as a signal level of one millivolt. This would be a
simple task except that most EMG systems produce an output level that is set at a
convenient level for computers to record. Both the MA-100 and MA-300 EMG
systems used in this tutorial produce output signals in the range of ±5 Volts so this
signal must be scaled to accurately represent the original input level.
This tutorial will discuss the methods used to calculate this scaling factor for each of
the two EMG systems (MA-100 and MA-300) in two different recording situations:
• Stand-alone tests using a Dataq ADC card recordings where the EMG
information is recorded directly to a PC file.

26 • A Whisper Tutorial Whisper User Manual

 • System tests using a Motion Capture system that generates a C3D file –
this is a common file format standard that is in use in many
biomechanics and gait analysis laboratories worldwide.

Stand-alone Tests
This section will discuss the gain calibration of sample recordings made with the
MA-100 and MA-300 EMG systems using a stand-alone Dataq ADC card. For the
purposes of this tutorial, these two EMG systems differ in their treatment of the
signal in that the MA-100 system has a continuously variable signal gain while the
MA-300 system offers a number of discrete switched gain levels. While this
example uses a Dataq ADC card the principals described here can be applied to
almost any analog data collection system.

Variable Gain Systems

The MA-100 has a continuously variable gain control for each EMG channel. This
allows the user considerable flexibility in selecting an optimal EMG recording level
but means that the amplitude of the recorded signal is not accurately known. Any
variable gain system like this can be calibrated by applying a known input to the
system and measuring the output level using the Dataq software.
The MA-100 includes a built-in calibration source that generates an 87Hz sine wave
calibration signal at a level that is equivalent to 200μV RMS at the EMG pre-
amplifier inputs. This calibration source can be used to calculate the system gain
after the variable gain control has been adjusted to produce a suitable EMG test
signal as discussed earlier. Once the variable gain controls have been set to the
desired position they must not be moved or adjusted unless you are prepared to
repeat the gain calibration procedure.
The calibration procedure is simple. A calibration recording is made once the
individual gains have been set. The MA-100 uses its internal calibration source
when the "CAL" button is pressed applies a known calibration source to each EMG
input. Since we know the input signal (200μV RMS sine wave) and can directly
measure the recorded level of the resulting output signal we can easily calibrate the
system to record data scaled in μV or directly calculate the system gain.

Figure 17 - A calibration level recording using a 200uV RMS input.

Whisper User Manual A Whisper Tutorial • 27

 The calibration recording shown in Figure 17 using the Dataq Browser software that
is supplied free of charge with the Dataq ADC cards. It shows the recorded signal
from an MA-100 calibration signal - at this point the ADC system is calibrated in
Volts – the ± 5 Volt range is shown on the left side of the display. The cursor has
been moved to measure the maximum signal present in the display (Ctl-K using the
Dataq software) and this value (1.5759 Volts) is displayed in the center of the range
display on the left. We can use the built-in calibration function of the Dataq software
(Function key F11 or assessable via the Edit menu) to scale the signal. We know
that the input signal is 200μV RMS and that the output signal is 1.5759 Volts Peak.

Figure 18 - Dataq peak level calibration.

It is vital that you do not Applying the RMS conversion factor (1.4144) give us a peak input signal of
calibrate a measured "peak" 282.88μV so we can use the Dataq Browser software to directly re-scale the
value with an RMS value. recorded signal and display the new data units. This is done by entering a "High
Calibration" value that is equivalent to the peak value of the input signal and
changing the "Engr. Units" display to "uV".

Figure 19 - The Dataq Browser calibrated in uV.

Since the MA-100 EMG System has individual gain controls on each channel, this
procedure must be performed on each channel that is to be evaluated using the
Whisper. Once the "High Calibration" has been performed, the Dataq Browser
software will directly indicate the range of recorded signal for the channel in μV at
the skin surface or pre-amplified electrode input – in the case of the illustration in
Figure 19, this range is ± 897.5μV.

28 • A Whisper Tutorial Whisper User Manual

It is important to note that the This value can be entered into the Dataq Data Acquisition software as a pair of
actual "calibration" of the "Fixed Calibration" values (Function F12 or accessible via the Edit menu) to allow
EMG system for data the Dataq Acquisition system to create output data files that contain test data that is
collection is performed when scaled directly in microvolts (μV) as required by the Whisper EMG test software.
the range values are entering Dataq files that are recorded in microvolts can be quickly converted by RData2 to
in the Dataq Analog Data create the DST files required for analysis and reporting.
Acquisition software.

Fixed gain systems

Fixed gain EMG systems may offer no user gain adjustment or may offer a limited
number of preset gains, The gains may be chosen as a single value that applied to all
the EMG channels or may be individually preset for each EMG channel.
MA-300 EMG Systems offers a range of individual gain settings for each EMG
channel – these may either be preset to record a fixed level or preset to individual
levels that are considered optimal for the EMG source under investigation. This
allows the user a great deal of flexibility in selecting an optimal EMG recording
level while maintaining a known gain for each EMG signal that is recorded.
The calibration procedure for fixed gain systems is very quick when using the Dataq
Data Acquisition software. All that needs to be known is the input range of the signal
for a given channel gain setting – in the case of an MA-300 EMG system, this value
can be read directly from the manual.
If the EMG system under test provides a channel gain factor (rather than an input
range) then the input range can be calculated by dividing the output range of the
device by the gain. For a system with a gain of 8,000 and an output range of ± 5V
this figure would be 5 Volts / 8,000 = 0.000625 Volt or 625μV peak. This value can
be used to set a fixed calibration for the EMG channel by entering the values as the
positive and negative limits for the data in the Dataq Data Acquisition software.

Figure 20 - The fixed Calibration can scale the data directly in uV.

The Fixed Calibration is accessed via the F12 function key or via the Edit menu
option. Your ADC system is ready to record calibrated data in microvolts (μV) once
this value has been set for each EMG channel that is to be tested. Remember that a
system with individual gain controls for each channel will require the correct fixed
calibration values for each channel.
An MA-300 system with a gain switch setting of #3 ( a gain of x5714 ) and
connected to a ±5 volt ADC will require a Fixed Calibration value of 0.000875 Volts
which can be entered as shown in Figure 20.

Whisper User Manual A Whisper Tutorial • 29

 C3D Recordings
This section will discuss the gain calibration of sample recordings made with the
MA-100 and MA-300 EMG systems using a commercial motion capture system with
a proprietary ADC card. For the purposes of this tutorial, the two EMG systems
differ in their treatment of the EMG signal as the MA-100 system has a continuously
variable signal gain while the MA-300 system offers discrete switched gain levels.
This section assumes that the reader is familiar with the stand-alone calibration
protocols described in the preceding pages and as a result, descriptions of the
individual features of the example EMG systems will not be repeated.
Full details of the public C3D C3D files are a common biomechanics standard that allows analog and 3D data to be
file format description recorded in data files, together with various parameters that describe the data. A
together with common standard analog channel consists of a number of analog samples together with two
parameter information is gain factors that are applied to the analog signal to transform the binary samples into
available from the C3D web real-world values. These gain factors are called GEN_SCALE and SCALE. There
site at http://www.c3d.org is a single GEN_SCALE parameter that is applied to every analog channel – the
value of the GEN_SCALE parameter is usually set when the data capture system is
installed. Since this scale factor is applied to all analog data it is usually not changed
very often – we will need to know its value but will not change it when we calibrate
our individual analog channels.
The calibration for each analog data channels in a C3D file is controlled by an
SCALE factor parameter – this parameter has an individual value for each channel
that is directly related to the gain or input range of the channel in combination with
the GEN_SCALE value. While the GEN_SCALE and individual SCALE
parameters can take any value, it is normal to choose these parameters such that
results in the analog channel are scaled in some known value – normally this will
default to Volts. The C3D analog parameter UNITS should normally describe the
measurement system for each channel.
Common values for GEN_SCALE are:
• 0.0048828 – the value of a single bit of data from a 12-bit ADC that is
measuring a ± 10V input range. An individual channel SCALE value
would then be 1.00 to obtain the analog data scaled in Volts.
• 0.0024414 – the value of a single bit of data from a 12-bit ADC that is
measuring a ± 5V input range. An individual channel SCALE value
would then be 1.00 to obtain the analog data scaled in Volts.
• 1.000000 – individual channel SCALE values must be set to 0.0048828
or 0.0024414 to obtain analog data scaled in Volts when sampled by a
12-bit ADC that is measuring a ± 10V or ± 5V input range respectively.
Many commercial motion capture systems use a standard naming procedure for data
recording trials that may not allow you to specify the standard names used by the
EMG Testing software. If you are using one of these systems then it may be
necessary to copy and rename the C3D files after recording to facilitate the use of the
Whisper EMG test software.

Variable Gain Systems

As described earlier, the calibration procedure is for an MA-100 system is simple. A
calibration recording is made using the internal calibration source to apply a known
calibration source to each EMG input. Since we know the input signal level (200μV
RMS sine wave) all we need to do is directly measure the recorded level of the
resulting output signal to calculate a scaling factor to record data scaled in μV. The
exact method of generating a C3D file will vary depending on your data collection
software however it is important that the analog parameters GEN_SCALE and

30 • A Whisper Tutorial Whisper User Manual

 SCALE are correctly set to produce a calibration file output scaled in Volts as well
will need to be able to read the system output levels directly.
The contents of the C3D files The exact method for setting these values will depend on the motion capture system
are easily viewed using the that is used to collect the data – questions about controlling and viewing these
MLSviewer software from parameters via your system software must be addressed to your data collection
Motion Lab Systems web site system manufacturer. However, independent software is available from Motion Lab
http://www.motion-labs.com. Systems, Inc. that displays the values of the parameters and data within any C3D file
(MLSviewer) and allows the user to edit any parameter or other component of a C3D
file (C3dEditor).
The C3D file displayed in Figure 21 contains a calibration signal recorded by an
MA-100 system – the C3D file has a GEN_SCALE parameter value of 1.00 and an
analog SCALE value of .00488 to produce an analog display that can be read directly
in Volts. The scale illustrates that the peak calibration signal is 2.23 Volts.

Figure 21 - Viewing an EMG calibration signal in a C3D file.

Since we know that the calibration signal is 282.88μV peak (200μV RMS) we can
calculate the correct value for the ANALOG:SCALE parameter for this channel
using the following formula:
ADCrange
SCALE =
resolution * ADCgain * EMGgain
GENSCALE
Where:
EMGgain = (Vout / Vin) = 2.23 / 0.00028288 = 7883
ADCrange = 20 Volts
ADCgain = x1
resolution = 4096 (12-bit resolution)

Whisper User Manual A Whisper Tutorial • 31

Warning: At the time of Using an ADCrange of 20 (the ADC channel range is ± 10Volts), an EMG system
writing, Oxford Metrics Vicon gain of 7883 (Vout / Vin) and a GEN_SCALE values of 1.00 yields a value of
Workstation applications 0.00000062 for the ANALOG:SCALE parameter to calibrate this channel in
have a bug that causes all microvolts at skin surface. You should change the ANALOG:SCALE value in the
SCALE factors of less than C3D file to this value using the facilities provide in your C3D file creation software.
0.000001 to be stored This calculation must be performed for each channel in the C3D file that is to be
incorrectly. Please contact calibrated. Note that this calculation should only be done for the EMG channels –
Motion Lab system or your the ANALOG:SCALE parameter for force platforms and other analog channels
supplier if you encounter should not be altered without careful consideration of the effects of any change.
difficulties creating the
correct SCALE values.
Normally the analog SCALE parameters are created with the correct values when the
C3D file is created using the software application supplied by your vendor. The
SCALE parameter, and other parameters may also be directly modified using
programs such as C3dEditor, available from Motion Lab Systems, Inc.

Fixed Gain Systems

An Excel spreadsheet is MA-300 EMG Systems offer individual gain settings for each EMG channel that
available from Motion Lab may be preset to record specific gain levels for each EMG channel. This type of
Systems that documents the system, with a known gain factor for each channel or switch setting can be used to
precise calculations needed to generate calibrated C3D file data very quickly. All that needs to be known is the
calculate C3D scale factors. input range of the signal for a given channel gain setting – in the case of an MA-300
EMG system, this value can be read directly from the manual or calculated using the
formula recommended by your C3D data collection system manufacturer.
As in calculations for variable gain systems, the calibration scaling factor used for a
fixed gain EMG system will be applied to the individual ANALOG:SCALE
parameters for the channels under investigation while leaving the universal
ANALOG:GEN_SCALE parameter unchanged as this affects all the analog channels
in the C3D file.
For C3D file recordings made with an MA-300 using the #3 gain setting using an
ADC with a ±5 volt range, and a GEN_SCALE parameter value of 0.00488, the
individual channel ANALOG:SCALE factors should be set to the value of
0.00009083 to scale the recorded EMG test signal from the Whisper in microvolts.
The #3 gain setting is selected because this is the optimum input range on an MA-
300 system for the Whisper tests which generate an EMG signal in the range of
+742μV to -670μV. A higher gain setting would cause the Whisper test signal to
exceed the input range of the EMG system under test (see Figure 15 on page 25 for
an example of this type of problem).
The individual SCALE value for the EMG channels can be determined from
information provided with the EMG system user manual or calculated directly as
follows:
ADCrange
SCALE =
resolution * ADCgain * EMGgain
GENSCALE
Where:
EMGgain = 5508
ADCrange = 20 Volts
ADCgain = x2 (±5 volt range)
resolution = 4096 (12-bit resolution)

32 • A Whisper Tutorial Whisper User Manual

 Thus:
20.0 20.0
SCALE = = = 0.0000004
4096 * 2.0 * 5508.0 4096 * 2.0 * 5508.0
1.0
The EMGgain figure is the total amount of amplification applied to the signal from
skin surface to output of the EMG system, as applied to the input of the ADC. If any
additional signal processing equipment is placed in-line between the EMG system
output and the ADC data sampling system then the gain (if any) of this item must be
included in the EMGgain figure. Note that the gain figure used here is a nominal
gain – due to component tolerances this figure may vary from the typical quoted gain
value for a specific setting. Consult your EMG system manual or equipment
manufacturer for details of acceptable tolerances for this figure.
Note that the value of the system gain of the EMG system being tested is easily
estimated by dividing the maximum output voltage of the system by the maximum
input voltage for a given gain setting – in this case 10 Volts divided by 0.00175
Volts or 5,714 – however, in this case various component tolerances are known to
reduce the gain for this setting to a figure of 5508.
The ADCrange is considered to be 20 Volts in this instance although the actual range
of the ADC card is quoted at ±5 volts. This is due to an ADCgain factor of x2 that is
applied within the ADC card. This has the effect of amplifying the incoming signal
to effectively reduce the ADC input voltage range to 10 volts (+5 Volts to –5 Volts)
which accounts for the manufacturers quoted range of ±5 volts.
The ADC card has 12 bits of sampling resolution resulting in a total resolution of
4096. GEN_SCALE is a (usually fixed) universal scaling factor in the C3D file and
usually has a value of 0.00488 or 1.00 although other values are acceptable.
It is important that the SCALE value is calculated and entered correctly as this
determines the scaling of the EMG signals that are recorded in the C3D file although
it does not otherwise affect the quality of the signal. Incorrect scaling of the test
signals from the Whisper is a common problem when users first start to examine the
EMG signals. It is recommended that anyone using the C3D file format should
carefully examine the data scaling and verify that it is correct before attempting to
analyze the data with the Whisper EMG Analysis Software. A C3D and DST file
viewer (MLSviewer, illustrated at the end of this chapter) is available from the
Motion Lab Systems, Inc web site at http://www.motion-labs.com.

Recording the Test Signals

Once you have made some initial basic tests to verify the correct connections, gain
levels and the recording parameters, you are ready to start making some test runs or
recordings. The Whisper Mode 0 provides the best test signal to the initial set up of
your recording or test environment. This mode produces five leading pulses at
intervals of one second. Each pulse is a single positive going square wave with an
amplitude of 500μV and a duration of one millisecond. These timing pulses are used
by the Whisper EMG Test Software to determine the sample rate of the system under
investigation.
Note that the high frequency component of the rising and falling edges of the square
wave timing pulse will be altered by the frequency response of the system
components under test. As the timing pulse contains a DC component you will not
be able to measure the pulse amplitude accurately with any EMG system that
incorporates a High Pass filter.

Whisper User Manual A Whisper Tutorial • 33

 Figure 22 - Whisper timing pulse (20Hz - 2,000Hz bandwidth)

A brief switching transient The digitized Whisper EMG signal starts exactly one second after the fifth timing
may be seen halfway between pulse and includes a common mode component. Select Whisper Mode 1 to generate
the last timing pulse and the a test signal that does not contain any common mode signals – this is useful if you
start of the EMG signal – this are testing an EMG system that has a mono-polar input or has a very low common
is normal and can be ignored. mode rejection ratio.

Figure 23 - Whisper timing pulse (20Hz - 500Hz bandwidth)

The following items are good indicators of a quality EMG test recording using a test
signal generated by the Whisper in either Mode 0 or Mode 1.
• Five positive going timing pulses should appear at one second intervals
prior to the start of an EMG signal. The amplitude of the timing pulses
is not critical as this will be affected by the frequency response of the
system under test. If the timing pulses appear to be negative then you
will need to reverse the EMG signal connections to the inputs of the
device that is being tested.
• The EMG signal should start one second after the last timing pulse and
last for fifteen seconds followed by a quiet baseline signal. If you have

34 • A Whisper Tutorial Whisper User Manual

 recorded five timing pulses then this will require a recording that is at
least twenty two seconds long.
• The EMG signal should be clean and should not clip at the maximum
and minimum recording signal range. There should be no significant
difference in EMG amplitude between a Mode 0 and Mode 1 EMG test
signal. If you find that Mode 0 EMG test signals are very large while
Mode 1 signals appear to be much lower in amplitude then the
equipment under test may have a common mode problem.
• The peak EMG signal amplitude should be approximately
Once you have established that the signal generated by the Whisper appears to be
appropriate for testing the system you will need to make three separate data
recording for the EMG testing software. These will normally be:
• Mode 0 – five timing pulses followed by fifteen seconds of EMG with
a common mode component.
• Mode 2 – fifteen seconds of common mode signal without any EMG
signal.
• Mode 3 – ten seconds of baseline without and common mode or EMG
signal.
You must use the correct file If you are testing an EMG system that does not have a differential input and cannot
name format for your EMG reject a common mode signal, or has problem with large common mode inputs then
data files if you wish to use you should substitute a Mode 1 test for the Mode 0 test. If you want the EMG
the EMG-Equipment-Tester testing software to automatically recognize the file types then you should choose file
software to analyze the data. names that end in En for Mode 0 tests (EMG), Cn for Mode 2 tests (CMRR) and Nn
for Mode 3 tests (noise/offset) where n is a number from 0 to 9 that can indicate the
test number for multiple tests. It is common to use the date of the test as part of the
filename so a typical set of filenames for a test would be:
000618E1.C3D – EMG test data recorded on 18th June, 2000.
000618C1.C3D – CMRR test data recorded on 18th June, 2000.
000618N1.C3D – Noise and offset test data recorded on 18th June, 2000.

Figure 24 - EMG showing the first 5 timing pulses.

The raw data files can be reviewed using any suitable data file viewer – the contents
of C3D files may be checked using software supplied by your data capture system

Whisper User Manual A Whisper Tutorial • 35

 manufacturer or by using the MLSviewer software available from Motion Lab
Systems web site at http://www.motion-labs.com. Data recorded using Dataq
software can be viewed using the Codas WINDAQ browser software. A quick check
of the start of any Mode 0 or Mode 1 test should produce an EMG data display
similar to that shown in Figure 24 – it may be necessary to adjust the horizontal
time-base to view the entire trial.

Converting the data

DST files use a standard file The EMG Equipment Test Software supplied with the Whisper requires that the
format developed by data files submitted to it are in DST format. This is a standard ASCII text file format
CAMARC in the 1980's. that can be used to describe a broad range of data with the exact format of the
Detailed information on the contents of a DST file being specified by a lexicon. In this case the Whisper lexicon
file format and the various describes a header that must exist at the start of the file to provide some basic
lexicons is available from the information about the test situation while the test data is stored as a series of integers
Motion Lab Systems web site. with a count of "1" representing one microvolt. An example of this format can be
http://www.motion-labs.com found on page 63 in Appendix A1 at the end of this manual. If your data recording
and collection system produces data in this format then you can immediately start
using the EMG-Equipment-Tester software supplied with the Whisper.
However, while the DST format commonly used for storing processed data, it is not
widely used to store raw data so we must usually convert the recorded EMG data
from the Whisper before we can use the EMG-Equipment-Tester software to analyze
the information and produce a report. Many common file formats, including C3D
and Dataq CODAS formats, can be converted by RData2, an ASCII software
conversion tool that is included with your purchase of the Whisper from Motion Lab
Systems. RData2 can read many common binary biomechanics data formats and
generate a wide range of ASCII output data, including DST file using the Whisper
lexicon.

Configuring RData2
At the time of writing, RData2 The RData2 application from Motion Lab Systems Inc., is supplied with its own user
supports the generation of guide and registration information and must be installed separately from the Whisper
Whisper DST files from C3D EMG-Equipment-Tester software. The registration information provided with your
files, Motion Analysis copy of RData2 must be entered as described in the RData2 User Manual before you
Corporation .ANA files, and can use the application to generate DST formatted data files suitable for analysis.
Dataq ADC files only. Once the RData2 application has been registered you will need to specify the DST
lexicon to be used (RData2 supports several different lexicons) and customize it with
information about your testing environment and location.

Selecting the Whisper Lexicon

Start the RData2 application and select "Defaults…" from the "Export" menu and
choose the "DST Export" option tab. In this tab you will find the Lexicon option
which will initially default to "MLS". Click on the "Lexicon" option until it displays
"Whisper" – Rdata2 will now default to producing DST files using the Whisper
lexicon whenever a data file is exported.
DST files generated using the "Whisper" lexicon will have a specific file header and
contain EMG data scaled in microvolts (μV). An illustration of this file lexicon can
be found in this manual in Appendix A1 on page 63.

36 • A Whisper Tutorial Whisper User Manual

 In addition to selecting the Whisper lexicon, make sure that the Export option "DST
File" is checked so that the application will produce DST format files and not plain
ASCII text files.

Using RData2 with Motion Analysis data files

In addition to selecting the Whisper lexicon, Motion Analysis users who are
recording data in the ANA file format will need to select an appropriate scaling
factor for the data. This is set in the "MAC Selection" tab in the options menu.

Whisper User Manual A Whisper Tutorial • 37

 Entering your Test Information
The final configuration option that is need is the information for the Whisper DST
file header. The only required item in this menu (ADC Resolution) is set by default
to "12" – a common value. The remaining information is optional and can be entered
as required via the RData2 "Export" menu selection "Whisper Information…"

Figure 25 - The RData2 "Whisper Information" menu.

The menu also contains a setting for "ADC Resolution" which is set to the value of
12 by default when RData2 is installed. If your data collection hardware is not a 12-
bit resolution system then you must change this value to reflect your specific ADC
capabilities.
All of the data in the Whisper The menu allows you to enter some information about the location of the testing site,
Information menu can be a serial number and a description of the testing configuration. This information is
changed by double clicking saved in all Whisper formatted DST files and can be used for quality control and
on the item and entering the configuration purposes. Full details of these header options can be found on page 57
appropriate information. in the chapter on File Descriptions.

Using RData2
It is a simple process to generate the files required by the EMG-Equipment-Tester
software once RData2 has been registered and configured to generate DST formatted
files using the Whisper lexicon. The process is very simple and consists of opening
the raw EMG data file and then saving the file – this process automatically exports
the data to a DST format file with the same file name in the same directory as the
original data file. The original data file is unchanged.
A progress bar is displayed in You can use the "Save As" option to change the file name and the location of the
the bottom status line of exported data and change the type of data file that will be written out. RData2 can
RData2 during all data be closed once you have converted all the raw Whisper data files to DST file. The
export operations. resulting DST files can be viewed using the MLSviewer software, available from
Motion Lab Systems, and can be opened by the EMG-Equipment-Tester software.
Note, that for C3D files and Dataq WDQ files, it is essential that all required scaling
factors have been applied to the raw EMG data file before RData2 is used to export
the Whisper DST file. If you find that you have to change an EMG channel scaling
factor then this must be done in the original data file and the data exported again to
overwrite the older DST file.

38 • A Whisper Tutorial Whisper User Manual

 Using RData2 with ANA files
ANA files, collected with a EMG data files generated using a Motion Analysis Corporation system must be in
Motion Analysis Corporation the "ANA" compressed ASCII format in order to be read by RData2. This file
system require that a specific format does not have any provision for scaling the individual data channels and so
gain factor is entered into the scaling factor must be determined and entered in the RData2 "MAC" format
RData2 before generating a options tab. In general, data collected on a Motion Analysis Corporation system in
Whisper DST file. the .ANA format uses a fixed gain range – normally this is ±5 Volts although gains
corresponding to ±10V, ±5V, ±2.5V and ±1.25V can be used.

Figure 26 – The MAC selection tab includes a specific scale factor for Whisper output.

Viewing the Whisper DST file

The EMG-Equipment-Tester software can be run once a DST file has been created.
This software application provides the ability to view (via the Quickview option) the
contents of any suitable DST file. However, if the DST file contains data or
parameters that are out of the expected range then the EMG-Equipment-Tester
software may have problems opening and viewing the data. This type of problem
typically occurs if the DST file has been created with the wrong format or scaled
incorrectly so that it contains data that is outside the range that the software is
expecting to analyze.

Whisper User Manual A Whisper Tutorial • 39

 Figure 27 - MLSviewer can display the DST file structure, parameters and file header.

MLSviewer is a general purpose DST file viewer that can open almost any DST
formatted file and display the contents of the header and all parameters and data
within the file. This application can often be useful if the EMG-Equipment-Tester
software fails to open a new DST file for any reason.

Figure 28 - MLSviewer can display the raw data in an easy to understand view.

This allows the user to check any DST file to verify that the format is correct and
that the DST file header contains accurate information about the data. A typical
Mode 0 or Mode file should contain EMG data in the range of +800 to –800 μV and
be preceded by five positive going timing pulses. The MLSviewer has a flexible
viewing mode that can be used to display the entire data recording as well as
allowing the user to zoom in to view individual data samples in an easy to use
graphical display.
The MLSviewer data display is auto-scaled and allows the user to rapidly determine
that the Whisper data has been scaled correctly - Figure 28 illustrates the transition

40 • A Whisper Tutorial Whisper User Manual

 from the initial Whisper timing pulses to the start of the simulated EMG signal. The
data in this illustration (a DST file) is shown scaled in the range of +756 to –718
(vertical scale) over a sample range of 5416 to 7042.
Most of the initial problems with generating DST files from raw Dataq or C3D files
can be traced directly to errors in the file format or the scaling of the data. The
MLSviewer is a very useful tool that can provide a great deal of useful information
about any DST file, as well as its built-in ability to display the contents of industry
standard C3D files.

Whisper User Manual A Whisper Tutorial • 41

42 • A Whisper Tutorial Whisper User Manual
The EMG Test software

Introduction
This manual and tutorial The EMG-Equipment-Tester software has been developed by Roessingh Research
uses, as an example, data and Development as a tool to analyze test records of EMG-Equipment using the
from a Whisper EMG signal standardized EMG signals generated by the Whisper. The software is able to view
generator recorded and and analyze the data from up to four EMG channels simultaneously and displays the
analyzed using an EMG EMG, common mode, noise, and DC offset parameters. The software also allows the
amplifier in the R.R.D. user to compare the current test results with any prior test results using the same
laboratory in Enschede, the equipment or with test results obtained in a colleague's laboratory. As such it is not
Netherlands. only possible to test the performance of your own equipment but it is also possible to
compare the performance of your system with other EMG recording systems or
environments.
The following chapters explain how to use the EMG-Equipment-Tester software and
assumes that you have read the earlier chapters explaining the Whisper EMG signal
generator. If you have not done so, please read the earlier chapters that explain the
Whisper functionality and operating modes as this information will not be repeated
here.
The EMG-Equipment-Tester software reads EMG data that has been stored in files
using the DST file format defined by CAMARC. If your data collection system does
not produce EMG data files that can be stored in the DST format then you may need
to use the accompanying RData2 software package. This creates DST files in the
correct format from C3D files and WDQ files created by Dataq data acquisition
software programs (CODAS and WinDaq/Pro).

Software installation
The best way to avoid The EMG-Equipment-Tester software has been developed under LabVIEW and runs
problems during installation under the Microsoft Windows 95, Windows 98 and Windows NT operating systems.
is to read the manual before Before installing and using the EMG-Equipment-Tester software, you therefore need
you start the installation. to understand the basic skills for working with your version of the Microsoft
Windows operating system.
The typical requirements for an IBM compatible PC to run this software are:
• At least a 100MHz Pentium processor with a minimum of 32 Mb of
main RAM memory and 10 Mb of free disk space - the software may
be installed on a network drive.

Whisper User Manual The EMG Test software • 43

 • An SVGA monitor and graphic adapter that supports at least 256 colors
and a color printer supported by the operating system.
You can use the Windows Add/Remove Programs option in the Control Panel to
install the application from a floppy disk. This will install the program on your
system together with some example files that you can use to demonstrate the
program.
If you accept the default options during installation, then you will have created a
short-cut in the Gait Lab menu group called Whisper together with a desktop short-
cut and icon.

Getting Started
The EMG-Equipment Tester is the analysis tool which comes with the Whisper and
was developed specifically to analyze the data recordings produced by the Whisper.
The analysis results from equipment tests can be stored in files on disk – up to four
EMG channels can be analyzed simultaneously.
Each stored analysis is informally known as a Spot Check and the results of the
analysis are stored in a .res file. A sample results file called YOURSIM.RES is
included with the software and is documented in Appendix A1.
The EMG-Equipment Tester is able to compare two different .res file analysis
records with each other. The Spot Check feature can be used to provide a history of
EMG equipment tests of your EMG instrumentation as well as documenting the
performance of your EMG instrumentation against the EMG equipment in another
laboratory.

The structure of the EMG-Equipment-Tester software is reflected in the diagram

shown below:

Main Screen level 0

Spot Record Analysis Spot Check level 1

Spot Record Report Report level 2

Quickview Analyzer Generator Spot Check Generator

Table 3 - Function Diagram of the EMG-Equipment-Tester software.

From the main menu you can choose to do a Spot Record Analysis or to do a Spot
Check.
Within a Spot Record Analysis you are able to:
• View the data files recorded using the Whisper (Quickview).
• Make a Spot Record of a data file which has been recorded using the Whisper.
You can extract a number of EMG parameters from the data and store these
parameters into a file for further investigation (Spot Record Analyzer).

44 • The EMG Test software Whisper User Manual

 • Print the results of a Spot Record to the default Windows printer (Report
Generator).
Spot Record Analysis will analyze data from up to four different EMG channels at
one time for multi-channel equipment tests.
Within a Spot Check you are able to:
• Compare two Spot Records with each other in order to check the performance of
your EMG instrumentation (Spot Check).
• Print the results of this Spot Check to the printer (Report Generator).
The Spot Check allows you to analyze data that provides detailed information about
your system performance and specifications. Data from independent tests made over
a period of time can be compared to verify continued system performance.

Figure 29 – The main screen of the EMG-Equipment-Tester software.

If the EMG-Equipment-Tester software has been installed correctly, it can be run by

double clicking on the EMG-Tester icon on the desktop or by running it from the
Start menu applications group that you specified when the program was installed.
This is normally the Gait Lab group.

Figure 30 - The software "splash" screen showing the licensing information.

Whisper User Manual The EMG Test software • 45

 When the application first starts it will display a "splash" screen that displays some
information about the software such as the registration number and site location.
This screen will disappear after twenty seconds, or as soon as the Continue button is
pressed. The EMG-Equipment-Tester software will then display the main screen as
shown in Figure 29. You are now ready to analyze test recordings that have been
collected using the Whisper.

Using the EMG-Equipment-Tester software

This chapter is a tutorial in the use of the EMG-Equipment-Tester software that uses
a number of sample data files that are provided as part of the software installation.
This tutorial starts from the main screen of the EMG-Equipment Tester shown in
Figure 29. This screen contains three windows, the Selection window, the File
Information window and the Control window.
You can exit the EMG- While using the EMG-Equipment-Tester software, you can expect buttons and
Equipment Tester by clicking selection switches to pop-up in the Selection window. These buttons and switches
on the Quit button in the main allow you to select the task and data-analysis operations that you want the software
menu. The software will then to perform. Depending on the analysis or operation requested, the EMG-Equipment
return you to the Windows Tester will prompt the user for the data files needed to complete the operation.
operating system. These requests will appear in the File Information window. The analysis operation
can not start until the appropriate information is supplied. Buttons in the Control
window allow you to start the analysis or move to a different application option.
The next two sections discuss the basics to perform the Spot Record Analysis and
Spot Check options which are entered by clicking on the corresponding button in the
Control window.

The Spot Record Analysis screen

Click on the Spot Record Analysis button in the Control window of the Main Screen
to enter the Spot Record Analysis – you should see the screen shown in Figure 31.
The Exit to Main button in the Control Window allows you to return to the Main
screen to switch to Spot Check option or exiting the application if you wish.

Figure 31 - Spot Record Analysis screen of the EMG-Equipment Tester

46 • The EMG Test software Whisper User Manual

 The default Spot Record Analysis option is Quickview and this is the option that is
initially displayed on the selection button in the Spot Record Option. Clicking on
this button will display a menu that allows you to select one of three possible Spot
Record Analysis options: Quickview, Spot Record Analyzer and Report Generator.

Viewing a file in Quickview

If you select Quickview in the Selection window, the File Information window will
contain a single file request dialog. Enter the name and the path of the file that you
want to view by typing it directly into the space provided or use the control on the
left of the dialog box to open a window and browse for the file.
When using the browse file For the purposes of this tutorial select the file 951001e1.dst which has been installed
selection method the "open" in the \data sub-directory. To select the file you must click on the file and then click
window will only display data the open button. The EMG-Equipment-Tester software will then select this file for
files with the appropriate file display and will cause the "Start Analysis" button to appear at the bottom of the
type – in this case only .dst control window. If any problems are detected in the selection procedure then the
filenames will be shown. Message <Not A Path> will appear in the File Information dialog box. If this
happens the Start Analysis button will not appear. In that case you have to repeat the
file selection procedure until the file has been selected.

Figure 32 - The file open dialog.

Once the file has been selected the EMG-Equipment-Tester software is ready to view
the data contained within file. Click Start Analysis to start the Quickview option.
The selected EMG data file will then be loaded and the Quickview screen displayed.
After a brief pause (depending on your CPU and computer speed), four graphs will
be drawn on the screen.
The demonstration file that you have selected contains raw EMG data recorded from
a four-channel equipment test. Therefore, you will see four graphs in the Quickview
screen as soon as the data loads. The display will show the first second of each data
recording on the screen. The Y-values of all graphs are in the same range. The
minimum and maximum values of this range are shown.
The initial view of the sample file will display four flat lines as the start of each
channel of the sample recording containing the five timing pulses that occur in each
Mode 0 or Mode 1 Whisper signal.
Quickview allows you to Quickview allows you to scroll quickly through your recordings. It can be used to get
easily check that all the a quick overall view of the data quality and allows the user to check that the
channels have recorded the recorded EMG signals are appropriate to be used in the Spot Record Analyzer. It has
standard test signal in the an easy to use interface containing only three buttons. By clicking on the <<< or the
same way. >>> button the graphs will scroll to the left or the right.

Whisper User Manual The EMG Test software • 47

 If you scroll through the sample recording by continuously clicking on the >>>
button, you will easily see that the standard Whisper test signal has been recorded
correctly. The start of the recording will display the timing pulses, one second apart,
quickly followed by the digitized EMG signal. Note that the signal has been recorded
at very similar levels in each channel.

Figure 33 - Quickview screen of the EMG-Equipment Tester.

You can leave Quickview by clicking on the EXIT button in the center of the top
three buttons. Then you will return to the Spot Record Analysis screen.

Preparing a Spot Record Analysis

You can perform a multi-channel equipment test by selecting the Spot Record
Analyzer option in the pop-up-menu button in the Selection window of the Spot
Record Analysis screen. Selecting the Spot Record Analyzer causes five buttons to
appear in the Selection window.
The five buttons allow you to select which parameters, which you want to measure
and display. For the tutorial please select all parameters by clicking on each button.
As you select the buttons you will see the appropriate filename dialog boxes appear
in the File Information window. Note that when you have selected all five
parameters there will be three separate file name dialog boxes displayed. These are
EMG file, Common Mode file and Noise/Offset file. You can either type the filename
and path into each dialog box directly or use the browse feature to point and click on
the appropriate filename.
The EMG file is the file that contains the Mode 0 / Mode 1 EMG test recording from
the Whisper. A Mode 0 or Mode 1 recording provides information on EMG
frequency and amplitude data properties, as well as timing and dynamic information.
The EMG-Equipment-Tester As a result selecting any one or more of the first three buttons will display a single
software will only open files filename dialog request for the EMG file. By convention the EMG filename will
with the correct filename have the format ??????E?.DST and only filenames with this format will appear in
format, e.g. nnnnnnEn.DST the file open window.
To display the sample data supplied with the EMG-Equipment-Tester software you
should select the file name 951001E1.DST that is provided with the installation.

48 • The EMG Test software Whisper User Manual

 Figure 34 - Spot Record Analysis window when preparing a Spot Record Analysis.

The Common Mode file is the file that contains the Mode 2 Whisper recording. This
file is generated when the Whisper is used to apply a common mode test signal to the
EMG instrumentation. This test enables the Common Mode Rejection Ratio
(CMRR) to be measured.
The dialog box requesting the Common Mode File will appear as soon the Common
Mode Rejection Ratio button is selected in the Selection window. You should enter
a filename in the Common Mode file dialog box – by convention, only those files
which have the format ??????C?.DST will be displayed in the file open browse
window. A file can be selected by double clicking on its name – for the purposes of
this tutorial select the file 951001C1.DST that can be found in the data directory that
is created when the application is installed.
The Noise/Offset file is the file that contains a Mode 3 recording. This recording is
used to measure the noise and DC offset levels the EMG instrumentation. The
dialog box for this filename will appear as soon as you select the Noise and Offset
Parameters option in the Selection window. Only files that have names in the format
??????N?.DST will be displayed. A file can be selected by double clicking on its
name.
In order to demonstrate a full Spot Record Analysis, 3 demonstration files have been
provided on the data diskette that comes with your Whisper. These are
951001e1.DST, 951001c1.DST and 951001n1.DST. These 3 files contain EMG data,
Common Mode data and Noise/Offset data measured with bipolar raw EMG
instrumentation.
Select all five test parameter options and select the three demonstration files. As
soon as all three filenames are loaded, the software will display the Start Analysis
button in the Control window. Click on this button in order to start the Spot Record
Analyzer.

Whisper User Manual The EMG Test software • 49

 The Spot Record Analyzer screen
Use the "Spot Record As soon as the Spot Record Analyzer is started, the screen which is shown in Figure
Analyzer" to view the data 35 should pop up and will be initialized. The screen is divided into five windows: the
analysis results and the EMG parameter window, the Common Mode parameter window, the Noise/Offset
"Report Generator" to send parameter window, the File Information window and a message/controller window.
copies of the analysis results The first three windows will show the results of the data analysis. Each column
to the current default printer. relates to one separate EMG channel and each window is related to one of the files
which has been selected in the Spot Record Analysis screen. The names of these
files are shown in the File Information window.

Figure 35 - The Spot Record Analyzer screen at the end of an analysis.

During an analysis these files are loaded and the selected parameters will be
calculated and will appear on the screen. The message box in the message/controller
window shows which task is under execution. Since a complete test analyzes three
files with lengths varying from ten to twenty-two seconds, each containing the data
of up to four EMG instrumentation channels, it can take up to several minutes before
a Spot Record Analysis will be completed if you do not have a fast computer.
During a complete Spot Record Analysis, performed with all three data files, the
following parameters will be calculated:

Amplitude parameters
• The Root Mean Square (RMS) value of the recorded EMG signal (in
microvolts)
• The Mean value of the recorded EMG signal (in microvolts)
Both parameters will be calculated over the fifteen seconds of data in the EMG file
that contain the EMG signal. The EMG data is located by first reading the file and
looking for the five leading pulses. The position of the five pulses within the file is
measured and the average number of samples between successive pulses is
calculated. The Whisper generates these pulses at one-second intervals so the
average number of samples measured enables us to determine the EMG sample rate.

50 • The EMG Test software Whisper User Manual

 Since we know that the EMG signal starts exactly one second after the last leading
pulse and that the EMG signal is 15 seconds long, it is easy to extract the EMG
signal data from the EMG file and calculate the RMS and Mean values.

Frequency parameters
• The Median frequency (Fmed) of the Power Spectrum of the recorded
EMG signal (in Hz)
• The Mode frequency (Fmode) of the (smoothed) Power Spectrum of the
recorded EMG signal (in Hz)
• The -3 dB points of Fmode (F-3dB left, F-3dB right) of the (smoothed) Power
Spectrum of the recorded EMG signal (in Hz)
These parameters are calculated from the first two seconds of the EMG signal in the
EMG file. The two seconds are extracted from the EMG file using the same
algorithm as the amplitude parameters discussed above.
The Power Spectrum is calculated, and by integrating the Power Spectrum, the total
power is calculated. The median frequency can be determined by integrating the
Power Spectrum until the frequency at which the integrated power equals half of the
total power. This median frequency Fmed thus cuts the Power Spectrum in two pieces
of equal power.
Fmode is the frequency in the smoothed Power Spectrum at which the spectrum is
maximal. Smoothing yields multiplication of the Power Spectrum with a Papoulis
window.
The F-3dB points are related to Fmode. Two points can be separated: F-3dB left and F-3dB
right. They can be defined as follows: for any frequency below F-3dB left and above F-3dB
right, the value in the Power Spectrum will never exceed Fmode/2. The frequency bands
outside the F-3dB points therefore indicate frequency ranges in which the power will
never increase half of the maximum Power (Fmode/2). The frequency band between
the -3dB points can therefore be regarded as a measure of the bandwidth of the
spectrum.

Dynamic properties
• The sample rate of the EMG instrument under test (in Hz)
• The dynamic response of the EMG instrument under test (in
microvolts)
The sample rate of the EMG instrument under test is calculated from the leading
pulses in the EMG record (as mentioned above).
The dynamic response that is shown in the Spot Record Analyzer screen after an
analysis is the response of the EMG instrument to the first leading pulse in a Mode 0
or Mode 1 recording. The graphs that are plotted show the shape of the filter
responses start 10 samples before the pulse arrives and show the response of the
system as long as the absolute value of the amplitude is more then 1 % of its
maximum value.
It is important to note that the leading pulses are only 1mS wide – as a result they
may not always be accurately recorded if the frequency response of the EMG
instrumentation does not extend to the higher frequencies or the analog data
collection sample rate is low.

Whisper User Manual The EMG Test software • 51

 Common Mode parameters
• The Common Mode Rejection Ratio (CMRR) of the EMG instrument
under test (in dB)
• The frequency of the recorded common mode signal (in Hz)
Both parameters are calculated from the Common Mode file, a Whisper Mode 2
recording. Two seconds of data is extracted from the middle of the file (ten to
twelve seconds after the start of the recording) and used to calculate the Power
Spectrum of the data.
The frequency at which the Power Spectrum is maximal is determined first. Since
the common mode signal is a pure 40 Hz sine wave, the maximal frequency should
also be 40 Hz. This is displayed as the estimated frequency of the recorded common
mode signal.
Once the maximal frequency has been determined the common mode signal can be
reconstructed from the same two seconds of Common Mode file data by first
performing a Fast Fourier Transform (FFT) on the signal. The FFT signal is then
bandpass filtered (bandwidth = 25% of Fmode around Fmode) and from this signal the
Inverse Real FFT will be calculated. This signal should be a 40 Hz sine wave. The
Whisper generates a known 1.0 volt common mode signal during a Mode 2 test.
Since we know the amplitude of the common mode signal at the input of the EMG
instrument, the Common Mode Rejection Ratio can be calculated by dividing 1.0 by
the estimated amplitude and taking the 20-log of this result.

Noise/Offset parameters
• The RMS noise of the EMG instrument under test (in microvolts)
• The DC offset of the EMG instrument under test (in microvolts)
These parameters are calculated from the Noise/Offset file. Since in a Noise/Offset
test does not produce any differential or common mode signal during the test, the
Noise and Offset parameters can be calculated over the complete file. The RMS
noise is determined by the RMS value of the recording while the Offset is equal to
the Mean value of the recorded signal.

The Results File

As soon as the Spot Record Analyzer has finish an analysis, it stores the results of the
analysis in a results file. This file will be written in ASCII (text) format. The name of
this file will automatically be generated and has the format YYnnnn.RES, where YY
= Year, and nnnn is a number that is incremented each time a results file is created.
The results file will be written to the same directory as the data.
The Spot Record Analysis is complete as soon as the application has completed
writing the results file to disk. Once the Spot Record Analysis is completed the Exit
button will appear in the message/controller window. If you click on this button you
will return to the Spot Record Analysis screen.

Printing a Spot Record (Report Generator)

The Report Generator option from the Spot Record Analysis menu works in exactly
the same way as the Spot Record Analyzer option. The only difference is that the
Report Generator will print the results to a printer at the end of each analysis.
You can select the Report Generator option from the Selection window in the Spot
Record Analysis screen. The analysis will proceed in exactly the same manner as the
Spot Record Analysis, using the same five buttons to select the Whisper test files as

52 • The EMG Test software Whisper User Manual

 described above. Once a Report Generator analysis has been selected, and the Start
Analysis button pressed, an analysis will be performed and a report printed on the
default Windows printer attached to your computer.

Interpretation of the Results

The EMG-Equipment Test software has been developed test the performance of an
EMG recording system using a Whisper generated EMG test signal. In addition it
the basic test and quality assurance functions, the EMG-Equipment Test software
can be used to compare the behavior of different EMG systems and data collection
instrumentation in separate laboratory or data collection environments.
When you are evaluating the performance of an EMG system it is important to know
what results you expect to obtain from the tests in order to determine the operational
status of the equipment. Since the characteristics of the Whisper generated test
signals are well known it is possible to predict the results of many of the tests.
The characteristics of the standard EMG test signals generated by the Whisper are
covered in extensive detail in Using the Whisper of this manual. Using the Whisper
to generate Mode 0 or Mode 1 test signals applied a signal to the equipment under
test that has well known amplitude and frequency parameters of the EMG signal.
These values are shown in Table 4.
The Whisper Mode 0 and If the EMG instrumentation under test records raw EMG, then the system should
Mode 1 test signals are record the standard Whisper Mode 0 and Mode 1 test signals without significantly
appropriate when testing changing the signal properties. If you run a series of Whisper tests and find that the
systems that record raw EMG results reported from a Spot Record Analysis shows values comparable with those of
signals. Table 4 then you can reasonably conclude that your EMG data collection and
recording system is functioning correctly.
However, if your EMG instrumentation generates smoothed, rectified EMG then you
will find that a Spot Record Analysis reports different numbers. In the ideal case the
RMS value of the recorded EMG has to be compensated with a factor 2 / π and
the average RMS value should be equal to 159.68 microvolts. The Common Mode
frequency has to be equal to the frequency that is shown in Table 4. It is also difficult
to predict the values of Fmed, Fmode and F-3dB since they depend too much on the
bandwidth properties of your smoothed, rectified EMG system. In practice the Fmed,
Fmode and F-3dB parameters are not very useful if the recording instrument produces
smoothed, rectified EMG.
test parameter value unit
RMS (EMG) 200.13 microvolt
mean (EMG) 0.09 microvolt
Freq (Com. mode) 40 Hz
Fmed 59.00 Hz
F-3dB (left) 23.00 Hz
Fmode 31.00 Hz
F-3dB (right) 70.50 Hz
Table 4 - Test parameter values of the standard test signal

The sample rate is determined The sample rate at which the EMG test signals will be recorded is set by the data
by the number of samples sampling system that is used to record EMG signals. The sample rate reported by
recorded between the leading the Spot Record Analyzer must match the value that has been set for the data
pulses in Mode 0 and Mode 1 sampling system, regardless of the type of EMG system (raw or smoothed, rectified
test signals. EMG recordings).

Whisper User Manual The EMG Test software • 53

 The filter response, Common Mode Rejection Rate (CMRR) and the Noise and
Offset of the EMG system depend a lot of the quality and type (raw EMG or
smoothed, rectified EMG) of the EMG system under test. This makes it difficult to
suggest an "ideal" set of numbers for the Spot Record Analyzer to report. The shape
of the dynamic filter response depends on the type of EMG system – raw and
smoothed, rectified EMG systems will produce radically different displays due to the
pre-processing inherent in a smoothed, rectified EMG system.
The CMRR should be as high as possible (usual between 80 to 120 dB). The Noise
and Offset of the EMG instrument under test should be as low as possible (a few
microvolts).
Figure 37 shows the results of a multi-channel Equipment Test of an EMG recording
instrument that records bipolar raw EMG (sample rate 1024 Hz). If you compare
these results with Table 4, you should be able to conclude that the EMG instrument
is functional and behaving as expected.

Spot Check
The Spot Check and Report The Spot Check option allows you to run system performance, specification and
Generator options function in comparison tests. You can enter the Spot Check option by clicking on the Spot Check
the same way except the button in the Control window of the Main menu screen. The Spot Check option will
Report Generator will create display a selection button in the Selection Window – this has two selections, which
a printed copy of the analysis are Spot Check and Report Generator. Spot Check will display the results on the
results on the default screen while Report Generator will also send the results to the default Windows
Windows printer. printer.
The File Information window displays two file selection dialog boxes – these expect
you to enter the names of two result files (.res). A new result file is created by the
Spot Record Analyzer (see page 50) whenever a Spot Analysis is performed and
contains a record of the analysis results.

Figure 36 - The Spot Check option.

Executing a Spot Check is very easy - first select either the Spot Check or Report
Generator option from the pop up menu in the Selection window. Then switch to the
File Information window then and load the name of the 1st Results file. A pair of
sample results file are provided with the EMG-Equipment-Tester software - these are

54 • The EMG Test software Whisper User Manual

 the files 95100101.RES and YOURSIM.RES. Load 95100101.RES as the name of the
1st file and YOURSIM.RES as the name of the 2nd result file.
As soon as both filenames have been loaded the Start Analysis button will be
displayed in the Control window. Click on the Start Analysis button to start the Spot
Check (or the Report Generator) and the screen shown in Figure 37 will appear.
Note that if the file selection procedure is not completed properly the error message
<Not A Path> will appear in the File dialog box. Repeat the file selection procedure
until both files have been selected.

Figure 37 - Spot Check screen of the EMG-Equipment tester.

The Spot Record screen displays the results of two prior Spot Record Analysis
operations. Each Spot Record Analysis results file can contain details from one to
four EMG channel analysis tests. The left side of the window displays the contents
of the first results file - each column displayed in this half of the screen contains the
test results of one EMG channel. At the right of the screen you will find the contents
of the second results file. Again each column contains the test results of a separate
EMG channel.
The Spot Record screen provides an easy way to compare two result files. If an
equipment test has been done according to a standardized measurement protocol, the
contents of two different results files relates to the same EMG channels of the
instrument under test. In that way, the performance of the EMG instrument under test
can be tested by simply comparing the corresponding results in two different results
files. Thus the quality of an EMG system can be checked on a regular basis by
comparing data from tests made several days or weeks apart.
For example, by examining two results files that have just loaded you could
conclude that the performance of the EMG instrument under test has not changed
since both Results files contain very similar test values.
As soon as you have compared two result files, the Spot Check has been completed.
You can leave the Spot Check screen by clicking on the Exit button. If you click on
Exit to Main and on Quit then, you will leave the EMG-Equipment-Tester software
and return to the Windows operating system.

Whisper User Manual The EMG Test software • 55

56 • The EMG Test software Whisper User Manual
File Descriptions

Protocol standardization
The EMG-Equipment-Tester software has been developed to automate the analysis
of standardized EMG test signals generated by the Whisper EMG signal generator.
It analyzes recordings from the output of EMG system instrumentation using an
automated, standardized procedure that is guaranteed to be repeatable at any time.
EMG instrumentation can be tested at any time and in any location with the certainty
that the results of the tests can be compared in a meaningful way.
In addition to the basic quality assurance features of the Whisper, the instrument also
allows users of EMG instrumentation in separate laboratories to check that their data
collections systems and EMG equipment is comparable.
It is important that the data recorded from the Whisper is maintained according to a
standard protocol if the Whisper will be used for equipment testing. The signal data
protocols have been discussed earlier in Using the Whisper on page 16 of this
manual however the storage of the information needs to be defined to some extent if
the EMG-Equipment-Tester software will be used to analyze the recordings.
This chapter will therefore discuss the protocols that must be followed in order to use
the EMG-Equipment-Tester software. These are:
• The data format the EMG-Equipment-Tester software can read and
analyze.
• The three basic Whisper data modes - Mode 0/1, Mode 2 and Mode 3
signals.
This chapter will end with a description of a full equipment test and some
recommendations for proper use of the EMG-Equipment-Tester software.

The CAMARC DST format

The following description of the Whisper file formats is provided for the benefit of
anyone who is unable to generate native DST file by either directly exporting them
from their data collection software or by using the Motion Lab Systems program
RData2 that is supplied with the Whisper. The Rdata2 program can generate
Whisper compatible DST files from any C3D or Dataq EMG data file in seconds.

Whisper User Manual File Descriptions • 57

Full details of the CAMARC In order to be able to analyze your test data with the EMG-Equipment Tester, it is
DST file formats may be necessary to write your data in the appropriate data format. The format which will be
found in the files distributed used is an ASCII text format and origins from the CAMARC II DST/EXP lexicon.
with the Motion Lab Systems This means that each test data file should have a header in which the kind of test data
MLSviewer application. that will be described, followed by the actual test data. The following shows an example
is available from the web site: of such a header followed by the first 5 data lines of a test data file.
http://www.motion-labs.com
#!DST-1.0 EXP-1.0 01/10/95 Enschede
$EXPeriment
EMG source: Whisper TM
serial number: RRD/CES-9412-01
Equipment under test: 8 channel EMG amplifier
Employee performing test: BF
Date of test: 01-10-95
Test sequence starts at: 13.24
Channels tested: 1 2 3 4
!AdcSampleRate
1024
$EmgPreProcessing
raw
$EmgUnits
microvolts
!AdcRESolution
12
!EMG-4
7 8 3 -1
7 2 7 9
9 -3 1 4
5 4 1 -2
14 1 12 8
. . . .
. . . .
This header is provided in the data directory as the file HEADER.TXT. I you are
generating DST files using a text editor it is recommended that you copy this header
to the start of the file, above your test data, and then change the information in the
header so that it matches the recorded data.
In order to be able to edit the header in the appropriate way you should know that:
• The header consists of a file-header-line (first line; starts with #!), text-
section-header-lines which indicate the beginning of a text section
(these lines begin with $ and are followed by text lines) and data-
section-header-lines which indicate the beginning of a data section
(these lines begin with ! and will be followed by data lines)
• The file-header-line contains information about the CAMARC II
lexicon and about the date and the place of the experiment.
• The text-section-header-lines and the text lines contain comments on
the experiment. The contents of text-section-header-lines may not be
changed and is case sensitive. The contents of the text lines needs to be
edited in order to fit your test data.
• The data-section-header-lines indicate parameters. The value of these
parameters can be found in the data-line following the data-section-
header-line. The contents of data-section-header-lines may not be
changed and is case sensitive. The contents of the data lines needs to be
edited in order to fit your test data

58 • File Descriptions Whisper User Manual

 • The lines in the header are separated by EndOfLine characters (EOL =
CR (Carriage Return) + LF (Line Feed) = ASCII 13 + ASCII 10)
• The order of the file-header-line, the text-section-header-lines and the
data-section-header-lines as shown above is the prescribed order. If
you change this order, your EMG-Equipment-Tester software may not
be able to analyze your Test date automatically.
In order to avoid problems with the EMG-Equipment-Tester software, it is
recommended that you only change the following information in the header of a Test
date file:
• The name of the city and the date of the experiment in the file-header-
line. Be sure that you have written the date in the appropriate format
and that you separate the date and place by only 1 SPACE (ASCII 32).
• The text in text lines following the text-section-header-lines. You may
include any comment over as many lines as you want in order to
describe the experiment as clear as possible. Be sure that you use the
appropriate EndOfLine characters.
• The sample frequency of your recording system in the line following
!AdcSampleRate. Replace 1024 by the sample rate (in Hz) of your
recording system.
• The resolution (in bits) of your recorded data in the line following
!AdcRESolution. Replace 12 by the resolution (in bits) of your
recording system.
• The number of EMG channels of which your test data file contains the
recorded data. Replace 4 in the line !EMG-4 by the number of EMG
channels in your Test date file (maximum number of channels = 4).
The header is followed by the actual test data. The test data should be written as
follows:
• Each column contains the data of 1 separate EMG channel. TABs
(ASCII 9) separate the columns.
• Each row contains 1 sample value of each recorded EMG channel.
Each line ends with an EOL character (ASCII 13 + ASCII 10).
• The data should be written as text integers (the file is an ASCII text
file). Each test integer should indicate the measured value in microvolt.
• At the end of the test data an EndOfFile marker (EOF = ctrl z) should
be included.
If you have written a test data file in the appropriate format, you can save the file to
disk. This file should have a name according to the following format:
YYMMDDtN.DST
in which:
• YY stands for Year (95 for 1995, 00 for 2000)
• MM stands for Month (01 for January, 12 for December)
• DD stands for Date (01 ... 31)
• t indicates the type of the test date file (e for EMG file; c for Common
Mode file; n for Noise/Offset file)
• N stands for the files (serial) number (0 ... 9)

Whisper User Manual File Descriptions • 59

 EMG Test Protocol (Mode 0, Mode 1)
The amplitude parameters, the frequency parameters and the dynamic properties of
your EMG instrument can be tested as follows:
• Connect the Whisper to your EMG system.
• Setup your EMG recording instrumentation such that it can record the
standard test signal without saturating during at least 22 seconds.
• Select the test mode (mode 0, mode 1) and set the signal switches (DIF
and COM) on the Whisper in the appropriate positions.
• Switch the Whisper on .
• Push the Whisper Start Button, start your EMG data recording
instrumentation and release the Start Button within one second.
Write the data that you have recorded in the appropriate DST-data format and name
the file according to the following format: YYMMDDeN.DST.
• Run the EMG-Equipment-Tester software
• View the file in Quickview and be sure that you have recorded the 5
leading pulses and that the first and only the first leading pulse occurs
in the first 2 seconds of the recording
If both criteria are satisfied, your file is suitable to be analyzed by the EMG-
Equipment-Tester software.

Common Mode Test Protocol (Mode 2)

The test protocol for Common Mode can be used to measure the Common Mode
Rejection Ratio of your EMG instrumentation.
• Connect the Whisper to your EMG instrumentation.
• Setup your EMG recording instrumentation such that it can record the
standard test signal without saturating during at least 15 seconds.
• Select only the common mode signal of the simulator by switching the
COM switch on and the DIF switch off.
• Switch the Whisper on.
• Push the Start Button and start recording the EMG data within one
second.
Write the data that you have recorded in the appropriate DST data format and name
the file as follows: YYMMDDcN.DST.
• Run the EMG-Equipment Tester
• View the file in Quickview and be sure that you don't see the 40 Hz
sinusoidal signal during the first 6 seconds of the recording and that
the 40 Hz sinusoidal signal is present during the 10th .. 12th second of
the recording
If both criteria are satisfied, your file is suitable to be analyzed by the EMG-
Equipment-Tester software.
If you have calculated the CMRR, you need to verify that the estimated frequency of
the Common Mode signal is indeed 40 Hz before you can be sure that the CMRR
calculation is a reliable measurement.

60 • File Descriptions Whisper User Manual

 Noise and Offset Test Protocol (Mode 3)
The test protocol for Noise and offset testing can be used to test the Noise and Offset
of the EMG instrument under test.
• Connect the Whisper to your EMG instrumentation.
• Prepare your recording system to record the standard test signal with
out saturating during at least 10 seconds.
• Set both the COM and DIF switches off.
• Switch the Whisper on.
• Start your recording device after pushing the Start Button.
Write the data that you have recorded in the appropriate DST data format and name
the file as follows: YYMMDDnN.DST.
• Run the EMG-Equipment-Tester software and select the noise and
offset analysis.
If you use the protocol in this manner, the total amount of Noise/Offset that you will
find indicates the Noise/Offset of both your EMG recording instrument and the
Whisper. This amount of Noise/Offset is superimposed on all your test data.

A full Equipment Test

When making connections to A full Equipment Test yields that all facilities of the EMG-Equipment-Tester
the Whisper take special care software will be used simultaneously such that a single analysis is able to show you
to avoid introducing any AC the overall performance of up to four EMG-channels. This means that you have to
noise into your EMG system develop a connection board that can be used to connect the output of the Whisper to
under test. four EMG channel inputs simultaneously.
In case of a bipolar EMG instrument, a full Equipment Test consists of three
recordings (mode 0, mode 2 and mode 3). In case of a mono-polar EMG instrument
the full Equipment Test consists of only two recordings (mode 1 and mode 3). After
you have made the recordings and transformed the data in the appropriate data
format, you present them to the EMG-Equipment-Tester software. Before you are
able to analyze them you should verify in Quickview whether the EMG file (mode 0,
mode 1) and the Common Mode file (mode 2) are indeed suitable to be analyzed. If
this is the case, you can start the analysis. We recommend doing this analysis using
the Spot Record Analyzer (Report Generator). If your EMG recording system
records raw EMG you can select all five parameters to view/test. If your EMG
system records smoothed, rectified EMG, it is useless to select Frequency
parameters and the Common Mode Rejection Ratio.
As soon as the analysis has been done you can compare the behavior of your EMG
recording instrument with a previous recording. We recommend doing this with the
Spot Check (Report Generator). After this check the full Equipment test has been
done and the EMG-Equipment-Tester software can be left.
At the R.R.D. laboratory in Enschede, the Netherlands, a full Equipment Test has
been done with a Whisper using a Klab SPA20 pre-amplifier, a KL100 (raw) EMG
amplifier and a TRI 815 AD-board. The results of this test are provided in the \data
directory of the software YOURSIM.RES and are also shown in Appendix A1.

Whisper User Manual File Descriptions • 61

62 • File Descriptions Whisper User Manual
Appendix A1

EMG-Equipment Tester Output

This appendix shows the outcome of a full equipment test of a Whisper EMG signal
generator at the R.R.D. laboratory in Enschede, the Netherlands. The results of this
test can also be found in the file YOURSIM.RES at the data diskette, which is
provided with the Whisper package.

!#!DST-1.0 EXP-1.0 09/28/95 Enschede

$EXPeriment
CAMARC II Specimen Test Result file, 09/28/95
Ampl/Freq/DynTest file: 950928E2.DST
Common Mode Test file: 950928C2.DST
Noise/Offset Test file: 950928N2.DST
!Results-4[11111]
2.000167E+2 2.011645E+2 1.983335E+2 2.014467E+2
6.542840E-1 5.730922E-1 3.880684E-1 4.920955E-1
5.794341E+1 5.744390E+1 5.794341E+1 5.744390E+1
2.247805E+1 2.247805E+1 2.247805E+1 2.247805E+1
3.096976E+1 3.096976E+1 3.096976E+1 3.096976E+1
7.043122E+1 6.993171E+1 7.043122E+1 6.993171E+1
1.024750E+3 1.024750E+3 1.024750E+3 1.024750E+3
3.300000E+1 3.300000E+1 3.300000E+1 3.300000E+1
2.000000E+0 -1.000000E+0 0.000000E+0 -1.000000E+0
2.000000E+0 -1.000000E+0 1.000000E+0 2.000000E+0
2.000000E+0 1.000000E+0 1.000000E+0 1.000000E+0
-2.000000E+0 0.000000E+0 0.000000E+0 1.000000E+0
3.000000E+0 2.000000E+0 0.000000E+0 0.000000E+0
-3.000000E+0 0.000000E+0 0.000000E+0 -1.000000E+0
-2.000000E+0 4.000000E+0 0.000000E+0 0.000000E+0
0.000000E+0 -2.000000E+0 -2.000000E+0 -1.000000E+0
0.000000E+0 2.000000E+0 0.000000E+0 2.000000E+0
1.000000E+0 3.000000E+0 -1.000000E+0 -2.000000E+0
4.900000E+2 4.920000E+2 4.890000E+2 4.940000E+2
-5.500000E+1 -5.700000E+1 -5.900000E+1 -6.100000E+1
-6.200000E+1 -6.200000E+1 -6.100000E+1 -6.000000E+1
-5.500000E+1 -5.600000E+1 -5.400000E+1 -5.900000E+1
-5.200000E+1 -5.200000E+1 -5.200000E+1 -5.300000E+1
-4.900000E+1 -4.800000E+1 -4.800000E+1 -4.700000E+1
-4.400000E+1 -4.600000E+1 -4.500000E+1 -4.300000E+1

Whisper User Manual Appendix A1 • 63

 -4.500000E+1 -4.400000E+1 -4.100000E+1 -4.000000E+1
-3.800000E+1 -3.800000E+1 -3.900000E+1 -4.000000E+1
-3.500000E+1 -3.400000E+1 -3.400000E+1 -3.400000E+1
-3.200000E+1 -3.300000E+1 -3.200000E+1 -2.800000E+1
-2.900000E+1 -2.900000E+1 -3.000000E+1 -2.900000E+1
-2.500000E+1 -2.700000E+1 -2.700000E+1 -2.500000E+1
-2.400000E+1 -2.100000E+1 -2.300000E+1 -2.600000E+1
-1.900000E+1 -2.100000E+1 -2.100000E+1 -2.100000E+1
-1.600000E+1 -1.800000E+1 -1.800000E+1 -1.600000E+1
-1.600000E+1 -1.700000E+1 -2.000000E+1 -2.000000E+1
-1.500000E+1 -1.600000E+1 -1.300000E+1 -1.400000E+1
-1.200000E+1 -1.300000E+1 -1.400000E+1 -1.600000E+1
0.000000E+0 -1.100000E+1 -9.000000E+0 -9.000000E+0
0.000000E+0 -8.000000E+0 -9.000000E+0 -9.000000E+0
0.000000E+0 -1.100000E+1 0.000000E+0 -6.000000E+0
0.000000E+0 0.000000E+0 0.000000E+0 -5.000000E+0
9.736605E+1 9.674984E+1 9.703971E+1 9.673129E+1
4.000000E+1 4.000000E+1 4.000000E+1 4.000000E+1
1.937604E+0 1.892821E+0 1.895600E+0 1.859045E+0
3.582031E-1 2.833984E-1 1.171875E-1 2.051758E-1

64 • Appendix A1 Whisper User Manual

Appendix A2

Technical Issues
Service and Support
The Whisper and the EMG-Equipment-Tester software are guaranteed and supported
for one year from the date of purchase. This means that in case of problems and
questions you can always contact us by mail, phone, fax or e-mail.
It is suggested that US customers who purchased the Whisper through Motion Lab
Systems should first contact Motion Lab Systems to resolve any problems. Please
mention the serial number of your Whisper when you contact us:
US distributor:
Motion Lab Systems, Inc.
4326 Pine Park Drive, Baton Rouge, LA 70809 USA
Tel: +1 225 928-4248
Fax: +1 225 928-0261
e-mail: support@motion-labs.com
Internet: http://www.motion-labs.com
Manufacturer:
Roessingh Research and Development
Roessinghsbleekweg 33, 7522 AH Enschede, The Netherlands
Tel. :+31 53 4875777
Fax :+31 53 4340849
e-mail: rrd@rrd.nl
Internet: http://www.rrd.nl

Electrical Specifications

Power Source 4 AA batteries, 1.25 Volt, 500 or 720 mAh.

Average current dissipation stand-by mode: 50 mA
run mode: 300 mA
Output Impedance 22 Kohm [+/- 2 %]
Common mode output level [ -1.0, +1.0 ] Volt [+/- 5 %]
Differential mode output level [ -2.048, +2.047] millivolt [+/- 5 %]
Table 5 - Whisper technical specifications.

Whisper User Manual Appendix A2 • 65

Troubleshooting the Whisper
The Whisper is a precision EMG signal generator and should give many years of
trouble free service. The following fault finding instructions are provided in the
unlikely event that you experience problems with your Whisper.
The LED does not light when you switch on the Whisper:
• Check that the batteries are not discharged.
• Check that the batteries are placed correctly in the battery compartment
(see the illustration on page 4).
• Check the size of the batteries. If you are using rechargeable batteries
(not recommended) you may find that they are shorter then regular AA
batteries. This may give some problems.
The Whisper does not produce a test signal:
• Check that the Whisper has been switched on and the LED is
illuminated.
• Check that the batteries are charged – the LED will not turn on if the
batteries are discharged.
• Check that the LED on the power switch turns red / orange as soon as
you release the START switch.
• Check that the START switch releases after being pressed – if the
switch sticks down then the LED will stay red /orange and no output
will be produced.
• Check that at least one of the COM and DIF mode switches was set on.
The Whisper only produces a differential output signal in Mode 0 or
Mode 1.
• Check the connections between the Whisper and your EMG
instrumentation.
• Check that the Whisper and your EMG instrumentation have been
grounded correctly. Incorrect grounding could short-circuit the test
signal.
• Check that your EMG instrumentation has been setup correctly and
responds to an EMG level input.
The EMG signal does not look like EMG:
• Check all the connections between the Whisper, your EMG
instrumentation and pay special attention to the ground connections.
• Check that you are using an appropriate gain level for your EMG
system and recording or data sampling equipment.
• Check that you are using an appropriate sampling rate. If in doubt use
a rate close to 1,000 samples per second per channel. If you are
sampling multiple channels then the total sample rate must be increased
appropriately.
The leading pulses in Mode 0 and Mode 1are negative:
• The connection cable between the Whisper and your EMG
instrumentation is not correct. Disconnect the Whisper signal cable
(Figure 6 on page 1) and connect the cable the other way around.

66 • Appendix A2 Whisper User Manual

 The test signal saturates your EMG system:
• Check all the connections between the Whisper, your EMG
instrumentation and ground.
• Verify whether the connection cable between the Whisper and your
recording EMG instrumentation is symmetrical. Even the smallest
(common mode) asymmetry at the input of your EMG instrument can
cause your EMG instrument to saturate.
• Check that your EMG instrumentation has been set correctly.
• If the EMG system appears to work in Mode 1 but not Mode 0 then
you may have a very poor CMRR – perform a Mode 2 test to determine
your systems CMRR response.
The recording contains line frequency signals:
• There is a lot of common mode interference. Check all the connections
between the Whisper, your EMG instrumentation and the laboratory
ground.
• If you are running a Mode 0 test then check that your EMG inputs are
actually differential inputs. If the system appears to work with a Mode
1 test but fails a Mode 0 test then it is likely that your EMG system is
either not a differential input system or does not have a good CMRR.
The recording is very noisy:
• Check all the connection between the Whisper, your EMG
instrumentation and ground.
• Verify that the Whisper and/or your recording instrumentation have not
been placed near to noise sources such as computer screens or
equipment which generates electrical or magnetic fields.
Your recording contains more than five leading pulses:
• If you see a spike 1/2 second after the 5th leading pulse this is due to an
internal switch in the Whisper. This spike will not cause any problems.
• If you see a spike at the beginning of your EMG recording this is due to
pushing the start button. You probably did not push the Start Button
before you started your EMG recording instrument. If you only view
your EMG recording this won't be a problem. If (in a later stage) you
are going to use the EMG-Equipment-Tester software, the spike will
cause problems if its amplitude exceeds the amplitude of the leading
pulses.
Your recording contains less then the five leading pulses:
• The sample rate of your recording device is low compared to the
impulse response of your EMG instrumentation. Increase the sample
rate of your recording device.
• Check to see if you started the Whisper prior to starting recording.
Some recording systems may take a second or so after the commond to
start before they actually record any data.
When you start the EMG-Equipment-Tester software a window pops containing the
message: Could not allocate the total amount of memory requested [20480000
bytes]. Continuing with ... bytes.

Whisper User Manual Appendix A2 • 67

 • This message indicates that you have less than 20 Mb of RAM memory
available – this may not cause too much of a problem unless you have
less than 16 Mb. The EMG-Equipment-Tester software may run into
serious memory problems if there is not enough memory to perform the
software analysis functions.
While processing data an error message appears Application memory full. Saving
and closing other applications (for instance a mail program running in the
background) can free memory. You may need to increase the Total Memory
allocated to LabVIEW from the EMGEQTST.INI file. This message can be
followed by another message that indicates the location of the memory conflict.
• Restart the computer and run the analysis again using fewer data
files.
• Sample the EMG data at a lower sample rate if possible.
• Add more memory to the computer system that you are using.
The EMG-Equipment Tester will fails to open a data file to perform an Analysis.
• Check that the data files are in the appropriate data format. The
EMG-Equipment Tester will only open DST in the correct format.
You can use the MLSviewer (from Motion Lab Systems) to view
the contents of a DST file and confirm that the format is correct.
• Check that you have named the data files in an appropriate format.
The EMG-Equipment Tester will not open files that have names
that do not conform to its naming convention.
• The EMG-Equipment Tester does not open binary data files –
check that you data file is viewable in Quickview or MLSviewer.
The EMG-Equipment Tester reports the wrong sample rate.
• If you do not find the correct sample rate, this can be due to your
Mode 0/1 data file containing less than five leading pulses
• Check the data file in Quickview or MLSviewer to make sure that
the data file does not contain any unexpected spikes that could
interfere with the software measurements.
• Check that your data collection sampling rate is high enough in
comparison with the length of the recorded filter response. If the
time between two samples is more than the filter response time (at
least a few milliseconds), the analysis software may not see every
pulse.
• Check that the EMG signal gain is appropriate – excessive gain
can cause the software to detect an erroneous sample rate.
The Spot Analysis Report is missing some parameter reports for some channels in a
multi-channel analysis.
• Make sure that each Test data file contains the same number
of EMG channels if you analyze an EMG file, a Common
Mode file and a Noise/Offset file simultaneously.
Unexpected files appear in the EMG-Equipment Tester directory.
• The EMG-Equipment Tester creates some *.vi and *.rsc files
under certain circumstances. These files can safely be deleted
after each analysis.

68 • Appendix A2 Whisper User Manual