Development of an Affordable EMC Immunity Assessment Setup Using

Direct Power Injection for Biosignals Instrumentation: Application to

ECG Monitoring

Tiago Nunes

1,3 a

, Hugo Pl

acido da Silva

1,2 b

and Hugo Gamboa

1,3 c

PLUX Wireless Biosignals, Lisbon, Portugal

Instituto de Telecomunicac¸

oes, Lisbon, Portugal

NOVA School of Science and Technology, Almada, Portugal

Keywords:

Electromagnetic, Interference, Immunity, EMC.

Abstract:

The increasing number of connected electronic devices in our daily lives contributes to a more dense electro-

magnetic environment, increasing the challenge of resilience to electromagnetic interference. This is particu-

larly concerning when the context is healthcare and the devices currently used to assess one’s health condition.

It is crucial that the development of new devices for biosignals acquisition takes into consideration the elec-

tromagnetic compatibility of the device from an early stage of the design. In this paper, a methodology to

assess the immunity of a device based on direct power injection is proposed. We describe the setup used and

the PCBs designed for the speciﬁc case of an ECG acquisition device. The validation of the setup is made

with two scenarios previously evaluated in anechoic environment. We show that with the proposed setup we

observe the same effects as in anechoic environment.

1 INTRODUCTION

The increasing number of electronic devices available

to the masses, and especially those capable of wire-

less communication, is the source behind many elec-

tromagnetic disturbances (Alaeldine et al., 2008).

This trend tends to increase, as the number of con-

nected devices is growing everyday. In fact, recent

reports have shown that in the last ﬁve years the num-

ber of connected Internet of Things (IoT) devices has

doubled. This number is expected to keep increasing

yearly in the coming years (Sinha, 2023; Sujay Vail-

shery, 2023).

Many of these devices are wearables, i.e., devices

that are designed to be worn embedded in clothing or

used as accessories, such as smartwatches or wrist-

bands, capable of acquiring different biosignals and

used increasingly in medical applications. Be it in

hospitals or at home, the environment in which a med-

ical device is placed is no longer a controlled one,

and it is becoming increasingly harsh from an Elec-

tromagnetic Interference (EMI) point of view.

https://orcid.org/0000-0001-5195-6668

https://orcid.org/0000-0001-6764-8432

https://orcid.org/0000-0002-4022-7424

Their use is only expected to increase as we transi-

tion to an era of digital medicine. Therefore, it is cru-

cial that good Electromagnetic Compatibility (EMC)

design practices are put in place during the develop-

ment of biomedical devices (Smuck et al., 2021; Lu

et al., 2020).

Ensuring resilience to electromagnetic interfer-

ence when designing a new device is key to make sure

the system will be compliant with the standards, and

immune to the noise in its intended use environment.

In fact, one of the most common causes for Printed

Circuit Board (PCB) redesign are EMC related issues.

While it is true that the smaller form factor of

today’s Integrated Circuits (IC) makes them intrinsi-

cally less prone to be disturbed by radiated and in-

duced disturbances, their placement in a PCB can in-

crease susceptibility as the traces leading to the IC can

pick up noise and carry it to the pins of the compo-

nent (Lavarda et al., 2017; Lavarda and Deutschmann,

2015; Jian-fei et al., 2011). Hence, good practices

when designing a new PCB are essential to ensure a

good performance from the device.

Typically, embedded applications, independently

of the domain of application rely heavily on a micro-

controller. This is arguably the most important ele-

Nunes, T., Plácido da Silva, H. and Gamboa, H.

Development of an Affordable EMC Immunity Assessment Setup Using Direct Power Injection for Biosignals Instrumentation: Application to ECG Monitoring.

DOI: 10.5220/0012588900003657

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 17th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2024) - Volume 1, pages 82-87

ISBN: 978-989-758-688-0; ISSN: 2184-4305

ment of the entire system since it makes the device

achieve its main purpose. However, the microcon-

troller in itself is not able to fulﬁll the device pur-

pose without a series of other modules such as power,

communication and, in the speciﬁc case of a medical

application, a biomedical sensor speciﬁcally designed

for the physiological signal of interest in that particu-

lar equipment.

Traditionally, these systems have been using

PCBs to assemble the components and interconnected

these modules via traces.It is important to mention

that all of the aforementioned elements, from the in-

dividual components to the PCB that brings them to-

gether, are susceptible to electromagnetic interference

which can disturb the system.

Several studies have tried to demonstrate this fact

by analyzing the various potential coupling victims

using the Direct Power Injection (DPI) method.

Established by the standard IEC62132-4 (IEC,

2006), DPI is one of the most reproducible methods

to evaluate a systems susceptibility to electromagnetic

interference. It allows to characterize the immunity

of a system in the presence of RF disturbances by in-

jecting them capacitively in the circuit (Chang et al.,

2013). It is widely adopted as it allows for rapid and

easy assessment of a PCB. In fact, as soon as the ﬁrst

prototypes are ready, DPI can be immediately per-

formed in a simple and intuitive way. This is possi-

ble as it doesn’t require advanced knowledge on EMC

(Pues and Pissoort, 2012; Miropolsky and Frei, 2011).

In a study carried out by (Dai et al., 2021), the

conducted immunity of a microcontroller was inves-

tigated exposing the IC to a continuous-wave electro-

magnetic interference using DPI. They observed the

conditions under which the IC failed and veriﬁed with

an electron microscope the damage inﬂicted.

In (Jian-fei et al., 2011), the authors were in-

terested in evaluating the susceptibility of a Low

Dropout Voltage regulator (LDO) using direct power

injection. They demonstrated via simulation and also

experimentally how a Radio Frequency (RF) distur-

bance injected using DPI generates an offset in the

output of the LDO.

In other works, the effects of electromagnetic in-

terference in ampliﬁers were investigated using DPI

to inject a disturbance through the ground plane and

output pins of ampliﬁers in various topologies and

conﬁgurations, such as the consequences of distur-

bances on precision voltage references (Lavarda et al.,

2017; Deutschmann and Winkler, 2023; Richelli

et al., 2020; Richelli et al., 2016; Richelli et al., 2017).

All of the modules mentioned above and previ-

ously investigated are key elements for a biomedical

system to operate. It is of paramount importance that

all these matters are taken into consideration during

the design of a new device. While anechoic chambers

are not easily accessible to everyone, and in partic-

ular to Small and Medium Enterprises (SME)s, DPI

testing can be easily and affordable to conduct.

In this paper we present a system developed to

evaluate the behavior of a device designed to acquire

biosignals when in the presence of a disturbing sig-

nal. Our solution provides a simple approach to as-

sess EMI immunity for SMEs and researchers who

do not have the means or access to more complex and

expensive solutions.

For this we use the DPI method as the noise in-

jection method, and a simulated Electrocardiogram

(ECG) as desired signal; in section 2 we present the

equipment used for this setup; in section 3 we present

two examples of application using this assessment

method; and in section 4, we outline the main con-

clusions and future work prospects.

2 MATERIALS AND METHODS

The purpose of this setup is to input a simulated ECG

signal for the Device Under Test (DUT) to acquire

while a disturbance is injected in the system capac-

itively using DPI. To fulﬁll these requirements, two

PCBs were designed: one with the purpose of simu-

lating a differential ECG signal for the DUT to sample

and another one to generate the interference to be in-

jected in the system. The entire setup is placed inside

a metallic enclosure to provide shielding against ex-

ternal electromagnetic noise. The 3D models of the

two PCBs are presented in ﬁgures 1 & 2.

2.1 ECG Signal Simulator

To generate the ECG signal, an Analog Discovery 2

arbitrary waveform generator was used. This device

featuring two Digital to Analog Converter (DAC) is

able to generate common signals, such as sinusoidal

waves, but also arbitrary signals with a 14-bit reso-

lution and voltages up to 5 V. We make use of this

feature to create the disturbing signal and the ECG

signal: we use MATLAB 2023b to create the differ-

ent signals that we want to test on the setup, and using

a toolbox by Digilent we are able to send the signals

directly to the signal generator where they are con-

verted into analog single-ended signals.

Naturally, the ECG signal has a very low ampli-

tude (+/- 1 mV) and for that reason, ECG sensors have

a very high ampliﬁcation gain (+/- 60 dB)(Singh et al.,

2012). Therefore, we need to create a very low am-

plitude signal for both the ECG signal and the distur-

Development of an Affordable EMC Immunity Assessment Setup Using Direct Power Injection for Biosignals Instrumentation: Application

to ECG Monitoring

Figure 1: PCB designed to convert a simulated single ended

ECG signal into a differential one.

bance. When generating low amplitude values with

this generator, only the Least Signiﬁcant Bits of the

DAC are used, rendering a very low resolution sig-

nal. To overcome this, we use all the DAC’s 14 bits to

generate a high resolution signal and we decrease the

amplitude using a voltage divider with the designed

PCB’s.

For the ECG signal, besides the need to reduce the

amplitude, we also need to convert the single ended

signal into a differential one for the sensor to acquire

it. The IC used for this purpose is th LTC6363. The

two outputs of this module will be centered over a

reference voltage provided by the ECG sensor. The

PCB presents snap connectors for the DUT to plug in

by the means of electrode lead wires.

2.2 Direct Power Injection Board

For this speciﬁc application, a PCB was carefully de-

signed with the purpose of delivering the disturbance

to the DUT. This PCB, presented in Figure 2, is used

to connect the signal generator to the DUT, in partic-

ular the ECG sensor. This kind of sensor presents a

high gain on its ampliﬁcation stage as biosignals have

a very low amplitude. For that reason, when inject-

ing a noise signal in this port, if the amplitude is too

elevated, the ampliﬁer will saturate and no particular

conclusions can be deduced from the tests.

This particular setup is intended to be used with

different types of disturbances while the effects are

evaluated in the time domain by analyzing the inﬂu-

ence they have on the ECG signals. To be able to see

the effect of a small disturbance on the ECG signal,

one needs to reduce the amplitude of the disturbance

being generated by the signal generator. We use the

same method as for the ECG signal generator, i.e., we

make use of the full scale of the DAC to produce a

high resolution signal and then decrease its amplitude

with an voltage divider.

2.3 Considerations for PCB Design

During the design phase of any PCB in general, and

PCBs for electromagnetic compliance assessment in

Figure 2: PCB for direct power injection.

particular, it is important to make sure that they oper-

ate with no major disturbances from the surroundings

nor from the PCB itself. In the development of these

PCBs, however simple they might be, several tech-

niques were used in order to minimize interference.

4-layer PCB. In the earliest stage of the design of

the PCB’s, the decision was made to use a 4-layer

PCB instead of a more common 2 layer one. While

the complexity is slightly increased when the num-

ber of layers increase, the beneﬁts extracted from it

compensate the effort put into the design of the prod-

uct. Opting for a 4-layer PCB not only provides more

ﬂexibility for routing traces but it also, and more im-

portantly, allows for signal integrity optimization with

the use of power and ground planes which effectively

reduce crosstalk and electromagnetic interference. In

the example of our application, layers were organized

as follows:

• Layer 1 is where all the components are placed.

Small traces connecting pins close to each other

are routed directly on this layer. Pins spaced far

from each other are routed through Layer 3.

• Layer 2 and Layer 4 are ground planes. Layer 2

creates a stable ground reference for all the ele-

ments of the PCB even in the higher frequencies,

where a simple ground trace would not be sufﬁ-

cient (Armstrong, 1999). The two ground planes,

top and bottom are connected though a dense net-

work of vias. In combination with via fencing,

this conﬁguration turns the PCB into a Faraday

cage, in which signals that can penetrate the struc-

ture are limited by the distance between vias.

• Layer 3 is used to route the traces connecting pins

placed away from each other. Additionally this

plane serves as a power plane providing power di-

rectly to the power pin of any component through

a via. This plane is placed in the middle of

two ground planes ensuring shielding of both the

power connections and traces routed in it.

Via Spacing and Fencing. The downsize in the

form factor of components has lead to increasingly

BIODEVICES 2024 - 17th International Conference on Biomedical Electronics and Devices

Figure 3: Cross section of the PCB showing the different

layers that compose it. All units are in millimeters.

Figure 4: Fencing used around the PCB’s to prevent radially

propagated electromagnetic emissions.

complex and denser layouts in PCBs, which in turn

leads to signal integrity problems such as crosstalk.

To ensure a good connection between the top and bot-

tom ground planes, as well as to prevent radially prop-

agated electromagnetic emissions through the PCB’s

edges, a dense grid of vias is used. Additionally, it is

also common practice to use via fencing as a measure

of preventing radially propagated energy from the

sides of a PCB. In fact, this constitutes one of the main

sources of radiated emissions in a PCB. These propa-

gate thanks to pseudo-waveguides formed by ground

and/or power planes leading to emissions from the

PCB’s edge (Lindseth, 2016; Suntives et al., ).

Power Supply Decoupling. The power supply

must be as stable as possible. As demonstrated by

previous works mentioned in Section 1, voltage refer-

ences and power/ground connections are susceptible

to inducing disturbances on the device components.

To be sure that the power lines are as stable as pos-

sible, decoupling capacitors are used to stabilize the

supply and avoid high frequency noise. Their place-

ment on the PCB is equally important. These must be

placed as close as possible to the pins of the power

supply of the components.

Coaxial Connections. The disturbance and simu-

lated ECG signals created by the signal generator

Figure 5: Enclosure used to protect the setup.

must arrive at the PCB undisturbed. The best way to

do so is by using coaxial cables. Typically the signal

generator’s output is, by default, a coaxial connection

BNC. Therefore, by applying a SMA connector to the

PCB side, we can use a coaxial cable to connect the

two devices. This is achieved by using a BNC to SMA

cable ensuring that the signal produced by the signal

generator arrives uncorrupted at the PCB.

3 RESULTS

The validation of the setup was performed with time

domain measurements using a BITalino (r)evolution

(PLUX Wireless Biosignals, Lisbon, Portugal) as

DUT. This device was designed to acquire multiple

biomedical signals (in particular, the ECG) and trans-

mit them via Bluetooth to a computer nearby.

A simulated ECG signal is generated with the

setup previously presented and then sampled by the

DUT. The latter, transmits the signal to a nearby com-

puter where the signal can be visualized and stored for

further post processing. The DPI board injects then a

signal in one of the pins of the board while the acqui-

sition is ongoing. The effects of the disturbance are

evaluated on the signal acquired.

The system itself is quite sensitive to the inher-

ent noise from the mains supply at 50 Hz. Such a

signal can be of an order of magnitude big enough

to completely mask all the other disturbances and

even the ECG signal itself. In an attempt to con-

tain such effects and focus on the disturbance be-

ing injected, we placed the setup inside a metallic

box. Although the container is not completely sealed,

the biggest apertures are small enough to prevent the

50 Hz noise from arriving at the DUT while allowing

for Bluetooth communication to be established with

the nearby computer.

This enclosure was ﬁtted with BNC feedthroughs

so that the signals from the waveform generator could

be delivered to the PCBs using fully shielded cables

Development of an Affordable EMC Immunity Assessment Setup Using Direct Power Injection for Biosignals Instrumentation: Application

to ECG Monitoring

Figure 6: Acquired ECG in the presence of a disturbance

modulated at 1.1 kHz.

thus preventing any interference from coupling to the

signals. On the inside, an SMA cable carries the sig-

nal from the generator to each PCB as illustrated by

Figure 5. Finally, the electrode lead wires connect to

the PCB via the snap connectors and the disturbance

is injected using an SMA probe.

3.1 Use Cases

Case 1. In standardized test procedures such as

IEC61000-4-3, the disturbing signal modulates the

carrier signal using a 1 kHz sine wave modulated at

80% depth (CENELEC, 2015). In previous works

(Bastian et al., 2023), it has been demonstrated, by

changing the modulating frequency with small incre-

ments, that when the modulating frequency matches

the sampling rate some failure modes are missed.

Here, we attempt to reproduce the same phenom-

ena by injecting a disturbance modulated in amplitude

at 80% with a carrier of 1 kHz. This frequency is then

incremented by 0.1 kHz and the effects are observed;

in Figure 6 we can clearly see the moment in which

the signal gets disturbed. At this point, the frequency

of the carrier shifts from a multiple of the sampling

rate to value slightly different.

Case 2. Another example used to validate the setup

consisted in injecting a modulated signal with a low

frequency AM modulation, in particular a 1 kHz sine

wave modulated at 20 Hz with a 80% depth. The ef-

fects of such a signal on the same device for biomed-

ical acquisition have previously been demonstrated

in an anechoic chamber with radiated interference

(Nunes et al., 2023).

We try to reproduce these effects by injecting a

similar signal using DPI. In our setup, while the DUT

was acquiring the ECG signal, the DPI board injected

the disturbing signal in the input pin of the ECG am-

Figure 7: Acquired ECG in the presence of a disturbance

modulated at 20 Hz.

pliﬁer. On the computer placed nearby, the signal be-

ing acquired was registered and is presented in Fig-

ure 7. It is possible to see the moment in which the

disturbance is activated; the effects of a disturbance

on the ground plane are immediately visible on signal

being acquired.

4 CONCLUSIONS

Assessment of electromagnetic compatibility is a cru-

cial step to guarantee a product behaves as expected,

and to be able to certify the device before putting

it in the market. Typical testing is performed in an

anechoic environment, which is not always accessi-

ble small and medium enterprises and/or students and

researchers. A more affordable method to assess elec-

tromagnetic immunity of devices is DPI.

We proposed a solution based on the DPI method-

ology, in which a device for biomedical acquisition

can be tested against conducted immunity. Two PCBs

were developed providing a simulated biosignal for

the device to acquire and a disturbing signal to inter-

fere with the ﬁrst.

This approach provides students, researchers and

SMEs with an affordable solution to verify and vali-

date devices for biosignals acquisition against electro-

magnetic interference from an early stage of the de-

sign phase as it eliminates the need for expensive and

specialized solutions such as anechoic environments.

Our system was validated reproducing tests previ-

ously conducted leading to the same conclusions.

In future works, it is foreseeable the addition of

frequency analysis as a complement to better under-

stand how the different disturbances affect the DUT

while performing an acquisition.

BIODEVICES 2024 - 17th International Conference on Biomedical Electronics and Devices

ACKNOWLEDGEMENTS

The research leading to these

results has received funding

from the European Union’s EU

Framework Programme for Re-

search and Innovation Horizon

2020 under Grant Agreement No.

955.816. Project website: https://eternity-project.eu

REFERENCES

Alaeldine, A., Perdriau, R., Ramdani, M., Levant, J.-L., and

Drissi, M. (2008). A Direct Power Injection Model

for Immunity Prediction in Integrated Circuits. IEEE

Trans. on Electromagnetic Compatibility, 50(1):52–

62.

Armstrong, M. K. (1999). PCB design techniques for

lowest-cost EMC compliance. Part 1. Electron-

ics & Communication Engineering Journal,

11(4):185–194.

Bastian, G. G., Pinto Nunes, T., Qu

ılez, M., Fern

andez-

Chimeno, M., and Silva, F. (2023). Analysis of the

Effect of Deviated Modulating Signal Characteristics

on the Susceptibility of a Small Medical Device. In

Proc. of Intl. Symp. on Electromagnetic Compatibility

– EMC Europe 2023, pages 1–6.

CENELEC (2015). Medical Electrical Equipment—Part 1–

2: General Requirements for Basic Safety and Essen-

tial Performance—Collateral Standard: Electromag-

netic Disturbances—Requirements and Tests.

Chang, Y.-C., Hsu, S. S. H., Chang, Y.-T., Chen, C.-K.,

Cheng, H.-C., and Chang, D.-C. (2013). The direct

RF power injection method up to 18 GHz for inves-

tigating IC’s susceptibility. In Proc. of the 9th Intl.

Workshop on Electromagnetic Compatibility of Inte-

grated Circuits (EMC Compo 2013), pages 167–170.

Dai, S., Lu, X., Zhang, Y., Liu, L., and Fang, W. (2021).

Electromagnetic Conductive Immunity of a Micro-

controller by Direct Power Injection. In Proc. of 6th

Intl. Conf. on Integrated Circuits and Microsystems

(ICICM), pages 280–284.

Deutschmann, B. and Winkler, G. (2023). Characterizing

the Electromagnetic Immunity of Operational Ampli-

ﬁers based on EMIRR and DPI. In Proc. of Intl.

Symp. on Electromagnetic Compatibility – EMC Eu-

rope 2023, pages 1–4.

IEC (2006). IEC 62132-4:2006 Integrated circuits - Mea-

surement of electromagnetic immunity 150 kHz to 1

GHz - Part 4: Direct RF power injection method.

Jian-fei, W., Sicard, E., Ndoye, A. C., Lafon, F., Jian-cheng,

L., and Rong-jun, S. (2011). Investigation on DPI ef-

fects in a low dropout voltage regulator. In 2011 8th

Workshop on Electromagnetic Compatibility of Inte-

grated Circuits, pages 153–158.

Lavarda, A. and Deutschmann, B. (2015). Direct power in-

jection (DPI) simulation framework and postprocess-

ing. In Proc. of IEEE Intl. Symp. on Electromagnetic

Compatibility - EMC Europe, pages 1248–1253.

Lavarda, A., Deutschmann, B., and Haerle, D. (2017). En-

hancement of the DPI method for IC immunity char-

acterization. In Proc. of 11th Intl. Workshop on the

Electromagnetic Compatibility of Integrated Circuits

(EMCCompo) 2017, pages 178–183.

Lindseth, W. (2016). Effectiveness of PCB perimeter Via

fencing: Radially propagating EMC emissions reduc-

tion technique. In Proc. of IEEE Intl. Symp. on Elec-

tromagnetic Compatibility - EMC Europe 2016, pages

627–632.

Lu, L., Zhang, J., Xie, Y., Gao, F., Xu, S., Wu, X., and

Ye, Z. (2020). Wearable Health Devices in Health

Care: Narrative Systematic Review. JMIR mHealth

and uHealth, 8(11):e18907.

Miropolsky, S. and Frei, S. (2011). Comparability of RF im-

munity test methods for IC design purposes. In 2011

8th Workshop on Electromagnetic Compatibility of In-

tegrated Circuits, pages 59–64.

Nunes, T. P., Qu

ılez, M., Fern

andez-Chimeno, M., Silva, F.,

and da Silva, H. P. (2023). Stage-by-stage evaluation

of a biomedical system regarding its electromagnetic

susceptibility. In Proc. of Intl. Symp. on Electromag-

netic Compatibility – EMC Europe 2023, pages 1–6.

Pues, H. and Pissoort, D. (2012). Design of IEC 62132-4

compliant DPI test Boards that work up to 2 GHz. In

Proc. of IEEE Intl. Symp. on Electromagnetic Com-

patibility - EMC EUROPE 2012, pages 1–4.

Richelli, A., Colalongo, L., and Kovacs-Vajna, Z. (2020).

EMI Susceptibility of the Output Pin in CMOS Am-

pliﬁers. Electronics, 9(2):304.

Richelli, A., Colalongo, L., Toninelli, L., Rusu, I., and Red-

out

e, J.-M. (2017). Measurements of EMI suscepti-

bility of precision voltage references. In Proc. of 11th

Intl. Workshop on the Electromagnetic Compatibility

of Integrated Circuits (EMCCompo), pages 162–167.

Richelli, A., Delaini, G., Grassi, M., and Redout

e, J.-M.

(2016). Susceptibility of Operational Ampliﬁers to

Conducted EMI Injected Through the Ground Plane

into Their Output Terminal. IEEE Transactions on

Reliability, 65(3):1369–1379.

Singh, Y. N., Singh, S. K., and Ray, A. K. (2012). Bio-

electrical Signals as Emerging Biometrics: Issues and

Challenges. International Scholarly Research No-

tices, 2012:e712032.

Sinha, S. (2023). State of IoT 2023: Number of con-

nected IoT devices growing 16% to 16.7 billion glob-

ally. Technical report.

Smuck, M., Odonkor, C. A., Wilt, J. K., Schmidt, N., and

Swiernik, M. A. (2021). The emerging clinical role of

wearables: Factors for successful implementation in

healthcare. npj Digital Medicine, 4(1):1–8.

Sujay Vailshery, L. (2023). IoT connected devices world-

wide 2019-2030. Technical report.

Suntives, A., Khajooeizadeh, A., and Abhari, R. Using via

fences for crosstalk reduction in PCB circuits. In Proc.

of IEEE Intl. Symp. on Electromagnetic Compatibility,

EMC Europe 2006., volume 1, pages 34–37.

Development of an Affordable EMC Immunity Assessment Setup Using Direct Power Injection for Biosignals Instrumentation: Application

to ECG Monitoring