IOP Conference Series:

Materials Science and

Engineering

PAPER • OPEN ACCESS You may also like

- The effect of opioid leu-enkephalin on the
Design of multi-channel wireless array wearable agranulocyte link of the blood system in
animals with hypothyroidism under stress
surface electromyography testing equipment D V Garmaeva, D S Adushinov, A I
Kuznetsov et al.

- Two-junction superconductor-normal metal

To cite this article: Jinquan Li et al 2019 IOP Conf. Ser.: Mater. Sci. Eng. 563 042040 single-electron trap in a combined on-chip
RC environment
Sergey V Lotkhov and Alexander B Zorin

- Seed distribution by the coulter for the

subsoil broadcast seeding
View the article online for updates and enhancements. E P Alekseev, I I Maksimov, A G Terentev
et al.

This content was downloaded from IP address 14.139.162.2 on 05/06/2025 at 05:22

AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

Design of multi-channel wireless array wearable surface

electromyography testing equipment

Jinquan Li1, Tong Zhang1* and Shuai Yan1

1
Automation School, Beijing University of Posts and Telecommunications BUPT,
Beijing, 100876, China
*Corresponding author’s e-mail: [email protected]

Abstract. Since there are few researches on multi-channel surface electromyography (EMG)
devices, this paper presents a new design scheme. The system uses highly integrated chips and
designs a new wearable electrode array. Also, the data is transmitted by wireless WiFi with a
high bandwidth. Scalable modular architecture is also applied to make the system suitable for
different scenarios. Through experimental tests, the device can work for more than 5 hours,
effectively collecting the EMG signals generated when the hand muscles move.

1. Introduction
The surface EMG signal refers to the detection of the EMG signal obtained on the surface of the skin.
Surface electromyography has great application value in human-computer interaction, medical
diagnosis, human biomechanics and other fields. For example, Jian Wu et al. use surface
electromyography and accelerometer fusion to identify sign language[1]. P. Sbriccoli. used surface
electromyography to detect muscle damage and recovery[2]. Catherine Disselhorst-Klug uses surface
electromyography techniques to measure the muscle strength[3].
The signal detected by the surface electromyography is the superposition of the electrical signals
when the multiple muscles contract. The more the number of channels collected, the better information
collection and better noise reduction. Array multi-channel surface electromyography can be used for
spatial filtering and feature extraction, and improve system resolution[4]. The array electrode can
effectively record the spatial information of muscle activity in addition to the time domain information
of the measured muscle activity, making it possible to detect new muscle characteristics, especially in
the muscle fiber conduction velocity and single motion unit (Single motor unit).It has unique
advantages in feature evaluation[5]. In summary, the multi-channel array acquisition method is very
important for surface myoelectric testing equipment.
Wearable devices can enhance portability and make human-machine synergy closer[6]. For the
application of surface electromyography, it is necessary to design a wireless wearable device. Many
surface electromyography devices currently do not for multi-channel array acquisition and has
wearable feature at the same time. For example, Delsys' Trigno series wireless surface
electromyography system, although wirelessly transmitted, is small enough to meet certain wearability,
but it has few acquisition channels. Italy's OT Bioeletronica EMG-USB2+ collector has 256 channels
of acquisition, but the device and electrodes are connected by wires, during measurement the tester's
activity is limited.
In the traditional surface electromyography system design, it is generally done in the hardware to
complete the amplification, filtering, remove the power frequency interference, and achieve the

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

acquisition of signals, which makes the system more components, large volume, and complicated
debugging process. Such as the design of the 32-channel surface EMG acquisition system designed by
Pengfei Zuo[7].
This paper designs a multi-channel wireless transmission wearable surface myoelectric detection
device. The system uses flexible wearable array electrodes and the data is transmitted wirelessly. The
acquisition module uses a highly integrated chip that reduces volume and enhances stability and
integrity. Scalable modular design for applications in different scenarios. It provides a stable and
reliable surface electromyography detection system for applications such as human-computer
interaction, motion detection, and medical diagnosis.

2. Design scheme
The work of collecting myoelectric signals is divided into two steps: hardware signal acquisition and
upper computer processing. The system flow is shown in Figure 1. The measurement site here is the
surface EMG signal of the human arm part. After the array electrode collects the original surface
electromyogram signal, the preprocessing circuit of the data acquisition board filters out some of the
low frequency noise and performs overvoltage protection. The data acquisition board then performs

Figure 1 System workflow diagram

acquisition amplification and analog-to-digital conversion on the surface EMG signals. The MCU
control board controls the multiple data acquisition boards to work together and obtain the values
measured by the data acquisition board via SPI bus, and the data is summarized and transfer the data
to the upper computer wirelessly. The upper computer uses a digital filter for high-pass filtering, uses a
power frequency notch filter to suppress 50Hz power frequency and its harmonic interference, finally
obtains real-time surface EMG signals and displays and saves them in files.
The hardware part has a flexible electrode array, a data acquisition board and an MCU control
board. The data acquisition board integrates the front-end second-order filter circuit and the ADS1298
system circuit. The MCU control board integrates the STM32 microcontroller system circuit, WiFi
wireless communication system and so on.
The software part is the MCU program and the host computer program. The MCU program is
responsible for controlling the data acquisition board and transmitting the collected data to the upper
computer wirelessly. The upper computer performs digital filtering to display waveforms, storage, and
the like.

2.1. Flexible electrode array design

Figure 2 shows a set of flexible electrode arrays including a flexible substrate, a silver chloride
electrode, and a wiring port. A set of flexible electrodes has 20 electrodes that can be connected to the

2
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

data acquisition board through the adapter plate. When measuring, the flexible electrode is bound to
the tester's arm by a strap.
The substrate used a 150μm polyimide(PI) flexible material. Polyimide can be stretched to make
the substrate follow the skin for small amplitude movement, reducing motion artifacts. The flexible
substrate is more comfortable to wear than a rigid, non-stretchable substrate.
The electrode adopts the traditional JK-1 type silver chloride electrode, which has a signal-to-noise
ratio of about 45 dB at the main frequency of the surface electromyography signal, and the noise is
small[8]. The silver chloride electrode has a diameter of 9 mm and a thickness of 2 mm. Each of the
flexible electrode arrays has 20 channels, and the silver chloride electrode is fixed to form an electrode
array on the pads of the substrate.

Figure 2 flexible electrode array

The size of the polyimide substrate is 30mm× 150mm, and the 20 copper pads on the substrate are
divided into upper and lower rows, each row is spaced.
15mm apart, and the pads are arranged at a lateral interval of 15mm. The inner diameter of the pad
is 3 mm and the outer diameter is 6 mm. There are contact points on the upper and lower sides. The
electrodes are attached to the pads by snaps. In order to ensure the conduction between the pad and the
electrode, the contact point is coated with a conductive silver paste. The middle layer of the substrate
is provided with wires, and the pads are led out through the wires to the cable interface on one side.
After measurement, the resistance of the silver chloride pad to the cable interface is 15-20Ω, which
can meet the surface electromyography detection requirements.

2.2. Data acquisition board

The data acquisition board consists of two parts: the front-end preprocessing circuit and the ADS1298
system circuit. The pre-processing circuit performs preliminary processing on the surface myoelectric
raw signal. The ADS1298 system circuit is responsible for controlling the initialization of the chip and
communicating with the MCU control board.
The pre-processing circuit consists of a second-order passive RC filter circuit and a Zener diode
circuit. The pre-processing circuit functions to filter out part of the high-frequency noise and
over-voltage protection. The second-order RC filter removes high-frequency noise above 30 kHz. The
Zener diode turns on when the input voltage is too large, protecting the data acquisition board.
The data acquisition board needs to collect high-quality surface EMG signals with hardware
filtering noise, and should have a scalable data interface. This paper selects the ADS1298 chip as the
acquisition chip. The ADS1298 chip is an analog front end of a high-performance test biopotential
released by TI. It has an 8-channel analog acquisition channel. The SPI data interface has a built-in
right-leg circuit that suppresses the power frequency, and the power frequency noise is suppressed on
the hardware. ADS1298 can be used to collect weak physiological signals such as ECG signals and

3
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

surface EMG signals. Such as Dunqiang Lu, Wei Wang and other surface electromyography
acquisition systems designed using ADS1294 chip[9].
The ADS1298 system circuit uses the officially provided system circuit. Use 0Ω resistor isolation
between analog ground and digital ground. The data acquisition board has a 20-pin interface, of which
16 pins are 8 positive and negative input interfaces, and 4 pins are right leg circuit interfaces. The
adapter board is connected to the flexible electrode, and the MCU control board and the data
acquisition board are connected by a pin header. The data acquisition board has a sampling frequency
of 500 Hz and a preamplifier amplification factor of 6.

2.3. Wireless transmission module and MCU control board

Bluetooth can be used for communication in devices with a small amount of data transmission, such as
a wristband real-time gesture recognition device made by Shuo Jiang using Bluetooth to transmit the
acquired four-channel surface EMG signals and single-channel IMU signals to the host computer[10].
Bluetooth transmission bandwidth cannot meet the needs of more channels to transmit data. The
Bluetooth transmission rate is up to 115200 baud and can transfer 14KB of data per second. Under the
sampling frequency of 500Hz/s, the 8-channel data of each data acquisition board of the system needs
to occupy 13.5KB/s transmission bandwidth, and the Bluetooth data transmission speed only supports
one data acquisition board (8 channels) data transmission. Using the WiFi module to wirelessly
transmit data baud rate up to 921600, the transmission rate can reach 112.5KB / s, can support 8 data
acquisition board (64 channels) data transmission at 500Hz / s sampling frequency.
The MCU control board needs to configure the ADS1298 chip register through the SPI protocol,
read the measurement data, control multiple data acquisition chips to work together, and transmit the
data to the host computer through the WiFi. The MCU processor should have an SPI interface, a serial
interface, and a faster main frequency. The MCU processor selects the STM32F103 series, which can
achieve a processing speed of 72MHz and an internal SPI interface. The data on the data acquisition
board can be received simultaneously via the SPI.

2.4. Implementation of multi-channel acquisition

The single data acquisition board only has 8 surface EMG acquisition channels. For more channel
application scenarios, this paper designs a multi-data acquisition board to work together. The
ADS1298 chip itself supports multi-chip cooperative working mode and can be connected through
Daisy-chain or Cascade. The Daisy-chain method uses less wiring, but the right leg circuit cannot be
used to remove the power frequency, which reduces the quality of the surface EMG signal. Cascade's
method has good versatility and simple principle, but there are many wiring ports and complicated
circuits. It is necessary to use the MCU control board to control the timing of multiple data acquisition
boards, which requires high synchronization and program processing. Based on the quality of the
acquired signal, the Cascade connection is selected to achieve multi-channel acquisition.

4
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

Figure 3 multi-chip cascade circuit connection diagram

When using Cascade connection mode, you need to select each ADS1298 chip from the MCU
control board, and connect the three SPI data transmission lines of each chip and the common START
and CLK signal lines together. The circuit connection diagram is shown in Figure 3-1.
When Cascade is connected, it needs to use an external clock frequency for synchronization. The
data acquisition board must use the internal clock to use the external clock when configuring the
register. The MCU control board program controls the switching data acquisition board to use the
clock to realize the multi-data acquisition board coordination jobs.

Figure 4 Multi-channel acquisition system

The SPI data line of each data acquisition board, the common START and CLK signal lines are
connected by pin headers. An acquisition system consists of the topmost layer of the MCU control
board and the following layers of data acquisition boards. Figure 4 shows the physical connection
diagram of the 16-channel acquisition system of two data acquisition boards. Changing the number of
connected data acquisition boards, modify the MCU control board and the host computer program to
meet the needs of different numbers of acquisition channels.

2.5 Software programming

The MCU control board completes the data acquisition board, collects data through the SPI interface,
sends data to the upper computer through WiFi, and controls multiple data acquisition boards to work
together. After each system startup, the MCU control board needs to re-initialize the settings, and then

5
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

the data is read through the SPI interface. The ADS1298 chip generates an interrupt by setting the
DRDT pin high level, and informs the MCU control board that the EMG data acquisition of the 8
channels is completed, and the MCU control board continuously reads 27 bytes of data information to
complete the acquisition.
The upper computer program is responsible for sending instructions to set the working mode of the
data acquisition board, accepting the data packet uploaded by the MCU module, and then performing
filtering processing to display in real time. Increase the cross-platform and portability of the host
computer, and write it on the QT platform in C++.The upper computer program sends instructions to
control the MCU control board. The command protocol is similar to the data packet sent by the MCU
control board. Send different command characters to configure the data acquisition board mode. In
order to make the signal processing speed and fast real-time performance, the system uses an
infinite-length unit impulse response digital filter. The model of the high-pass filter and the notch filter
is constructed using MATLAB, and the model parameters are calculated, which is used in the upper
computer program. A part of the high-frequency noise in the surface EMG signal has been filtered by
hardware, and the low-frequency noise in the host computer is filtered by the high-pass filter. The
high-pass filter has a cutoff frequency of 5Hz and an order of 5, and the acquisition frequency of the
system is between 5Hz and 500Hz. In this interval, the 50Hz power frequency interference has been
suppressed by the right leg circuit, but the upper computer uses the digital notch filter to filter again.
The notch order is 7, and the quality factor is 5.

3. experimental design and results

3.1. Experimental design

In order to verify that the system can effectively collect surface EMG signals, a set of experiments was
designed.
Four data acquisition boards (32 channels total) and MCU control board were used, and the flexible
electrode plus position machine was used to build the experimental system. The hardware acquisition
system operates at approximately 90mA and can operate for more than 5 hours with a 500mA battery.
The volume is 90mm × 80mm × 50mm, and the weight is 300g. The PC software runs in the
environment where QT is configured.
In the test acquisition, four sets of flexible electrode arrays were fixed to the left forearm of the
subject by a band. The first set of electrodes (1 channel to 8 channels) are attached to the shallow and
deep flexors. The second set of electrodes (9 to 16 channels) were attached to the radial flexor of the
wrist. The third set of electrodes (17 to 24 channels) are attached to the extension finger muscles. The
fourth set of electrodes (25 to 32 channels) is attached to the elbow muscle. The subjects were in a
standing state when tested, and the left arm was a pronation. In 18 seconds, the fist is made, the palm
is opened, the palm is bent forward, the palm is bent backwards, the palm is bent to the inside of the
body, and the palm is bent to the outside of the body for six movements.

3.2. Experimental results and analysis

The raw data of one channel surface EMG signal (1 channel, 9 channels, 17 channels, 25 channels)
from each of the four data acquisition boards is integrated into a graph, as shown in figure 5.

6
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

Figure 5 Multi-channel surface EMG signal

It can be seen from the electromyogram that the myoelectric signal of the shallow, deep flexor and
extensor muscles is obvious when the fist is applied and the palm movement is extended. The muscles
that stretch the finger muscles and allow the fingers to straighten out. The shallow and deep flexors are
responsible for pulling the tendons and making the hands grasp the fists. This is consistent with the
moment when the sEMG1-8 channel signal has a peak, and the waveform is obvious and the noise
signal is less. Experiments show that the system is effective in collecting surface electromyography.

3.3. Conclusion
This paper designed a wearable multi-channel array surface electromyography acquisition device. The
system effectively filters out high frequency, low frequency and power frequency noise, and can
collect real-time stable surface EMG signals. The flexible electrode array makes it easy to wear and is
suitable for collecting surface EMG signals from complex muscle groups. It is convenient for the
measurer to use in the form of wireless transmission. The system can be applied in the fields of
human-computer interaction gesture recognition, medical diagnosis, rehabilitation medicine, motion
detection and the like.

References
[1] Wu J, Sun L, Jafari R. (2016) A wearable system for recognizing American sign language in
real-time using IMU and surface EMG sensors. J. IEEE journal of biomedical and health
informatics, 20(5): 1281-1290.
[2] Sbriccoli P, Felici F, Rosponi A, et al. (2001) Exercise induced muscle damage and recovery
assessed by means of linear and non-linear sEMG analysis and ultrasonography. J. J
Electromyogr Kinesiol, 11(2):73-83.
[3] Disselhorstklug C, Schmitzrode T, Rau G. (2009) Surface electromyography and muscle force:
limits in sEMG-force relationship and new approaches for applications. J. Clinical
Biomechanics, 24(3):225-235.
[4] Zhao Zhangbiao, Chen Xiang, Lei Peiyuan, et al. (2009) Array surface electromyography signal
acquisition instrument[J]. Journal of Electronic Measurement and Instrument, 23(12): 88-93.
[5] Zhang Hengyi, Chen Xiang. (2014)Design of active flexible surface electromyography electrode
array[J]. Journal of Electronic Measurement and Instrument, 28(8): 901-908.
[6] Nanxiang Yu, Dongyi Chen, Shiji Xiahou. (2012) New developments in wearable computing
technology and its applications. J. Digital Communication.
[7] Pengfei Zuo, Zhiyuan Lu, Yongqiang Zhang. (2013) Design of 32-Channel Wireless Surfa
ce Electromyography and Acceleration Signals Acquisition System[J]. Chinese Journal
of Sensors and Actuators.
[8] Zhangyan Zhao. (2010) Research on Some Key Technologies in Surface EMG Signal Det
ection and Processing[D]. University of Science and Technology of China.

7
AEMCME 2019 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 563 (2019) 042040 doi:10.1088/1757-899X/563/4/042040

[9] Dunqiang Lu, Wei Wang, Yongpeng Chen. (2015) Design of ADS1294 SEMG acquisition
system[J]. Journal of Tianjin Normal University
[10] Jiang S, Lv B, Guo W, et al. (2017) Feasibility of Wrist-worn, Real-time Hand and Surface
Gesture Recognition via sEMG and IMU Sensing.J. IEEE Transactions on Industrial
Informatics, PP(99):1-1.
[11] Zhongfei Dong, Xiang Chen, Hao Deng. (2012) Design of differential concentric ring sEMG
electrode array based on flexible printed circuit.J. University of Science and Technology of
China