Radiated Susceptibility
Measurements on Analogue Temperature Sensors
JACOB CEDERLUND
KTH ROYAL INSTITUTE OF TECHNOLOGY
SCHOOL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCE
Measurements on Analogue Temperature Sensors
JACOB CEDERLUND
Master in Embedded Systems Date: June 23, 2020
Supervisor: Mark Smith
Examiner: Carl-Mikael Zetterling
School of Electrical Engineering and Computer Science Host company: Prevas AB
Swedish title: Mätningar på analoga temeperatursensorers
känslighet för elektromagnetiska fält
Abstract
The need for electromagnetic compatibility is growing steadily as the usage of electronics in our daily lives is increasing more than ever. A common issue encountered in electromagnetic compatibility testing is analogue sensors that fail when exposed to electromagnetic fields. Testing how well electronics do when exposed to electromagnetic fields is called susceptibility testing, and standards for how to do these tests have been developed to ensure that the results of the tests can be reproduced.
In this thesis work, analogue temperature sensors have been shielded using a few common techniques. The susceptibility of the sensors has been analysed by looking at their output voltage when the sensors were exposed to electro- magnetic fields of different field strengths. The output of the sensors was read by an Arduino that was shielded and tested to make sure it would not be af- fected by the electromagnetic fields used in the sensor tests.
The result of the first set of sensor tests shows that shielding the cables running to the analogue temperature sensors and filtering away disturbances using ferrites gives a considerable decrease in susceptibility against electro- magnetic fields, while twisted cables and RC-filters did not. The results also showed that the introduction of a ground plane increased the susceptibility of the sensors, which most likely was due to it not providing the current with a path of less impedance and only served to increase the length of the uninten- tional antenna, which made it couple to the electromagnetic field more easily.
However, during a second round of testing, the results of all the tests were
hard to reproduce exactly, which calls into question how trustable the results
of standardised susceptibility tests are. Therefore, when designing for the elec-
tromagnetic susceptibility of a product, a rather wide margin should be used
in order to make sure that the product can reliably pass susceptibility tests.
Sammanfattning
Användningen av elektronik ökar i samhället och därför även nödvändighe- ten för testning av elektromagnetisk kompatibilitet. Ett vanligt problem inom elektromagnetisk kompatibilitet är att analoga sensorer lätt blir utstörda av elektromagnetiska fält. Hur man ska testa en elektronisk produkts känslighet mot elektromagnetiska fält styrs av standarder som ser till att resultaten av testerna går att återskapa.
I detta examensarbete har analoga temperatursensorer skärmats med ett par vanliga metoder. Sensorernas känslighet har analyserats genom att undersöka hur deras utspänning påverkas när sensorn blir utsatt för elektromagnetiska fält med olika fältstyrkor. Sensorernas utspänning lästes av en Arduino som skärmades och testades för att se till all att den inte påverkades av de elektro- magnetiska fälten som användes under testandet av sensorerna.
Resultaten från de första sensortesterna visar att använda skärmade kablar
till de analoga temeperatursensorerna och att filtrera bort störningar med ferri-
ter sänkte sensorernas känslighet mot elektromagnetiska fält betydligt medan
tvinnade kablar och RC filter inte gjorde det. Testerna visade också att jord-
plan i detta fall ökade sensorernas känslighet då de inte erbjöd en bättre väg
för strömmen att gå utan endast skapade en längre oavsiktlig antenn, vilket
gjorde att den lättare kunde koppla till det elektromagnetiska fältet. Däremot
visade det sig i en andra testomgång, att resultaten inte gick att återskapa ex-
akt. Detta ifrågasätter hur tillförlitliga dessa standardiserade tester är och visar
att man bör ha en ganska bred marginal när man designar för att minska en
produkts känslighet mot elektromagnetiska fält, så att den på ett tillförlitligt
sätt kommer kunna klara av känslighetstester.
Acknowledgments
I would like to thank my supervisor at Prevas, Thomas Bergkvist for his con- tinuous support during this project, and for teaching me all I needed to know (and more) about EMC. I would also like to thank Maria Månsson at Prevas for making this thesis possible in the first place, and for helping me navigate through the confusing world of EMC jurisdiction.
I would like to thank Carl-Mikael Zetterling and Mark Smith at KTH for their help in setting up this thesis work, and for finding the time to guide me during this project, even during a global pandemic.
Lastly, I would of course want to thank my family as well as my better
half, Angelica, without whom this thesis work would not have been possible
nor nearly as fun to do.
1 Introduction 1
1.1 Research Question . . . . 2
1.2 Scope . . . . 2
1.3 Purpose . . . . 3
1.4 Goals . . . . 3
1.5 Outline . . . . 4
2 Background 5 2.1 EMI and EMC . . . . 5
2.1.1 Source/Victim Concept . . . . 5
2.1.2 Radiative Coupling . . . . 6
2.1.3 Electromagnetic Fields . . . . 7
2.1.4 Standards for Emission and Immunity . . . . 7
2.1.5 The EMC Market . . . . 9
2.2 Protecting Against Electromagnetic Radiation . . . . 9
2.2.1 Shielded Cables . . . . 10
2.2.2 Ground Planes . . . . 11
2.2.3 Shielded Boxes . . . . 12
2.2.4 EMC Filters . . . . 12
2.3 Sensors . . . . 15
2.3.1 Signal Conditioning in Sensors . . . . 15
2.3.2 Resistive sensors . . . . 15
2.3.3 Capacitive sensors . . . . 16
2.3.4 Temperature Sensors . . . . 16
2.4 EMC Issues in Analogue Sensors . . . . 17
2.4.1 Analogue and Digital Sensors . . . . 17
2.4.2 Linear Components . . . . 18
2.4.3 Non-linear Components . . . . 18
vii
3 Methods 19
3.1 Lab Setup . . . . 19
3.1.1 Sensor Interface . . . . 19
3.1.2 Shielded Box . . . . 21
3.1.3 EMC Chamber . . . . 22
3.2 Test of the Analogue Temperature Sensors . . . . 22
3.3 Test of the Arduino and Shielded Box . . . . 26
3.4 Mounting of Sensors . . . . 27
3.5 Field Strengths Used During Test of Radiated Susceptibility . 27 3.6 Tests With Different Ground Plane Configurations . . . . 29
3.6.1 No Ground Plane . . . . 29
3.6.2 Perforated Ground Plane . . . . 30
3.6.3 Ground Plane . . . . 31
3.7 Tests With Filters . . . . 31
3.7.1 Ferrites . . . . 31
3.7.2 RC Filter and Bypass Capacitor . . . . 32
3.8 Tests With Different Cables . . . . 33
3.9 Shielded Cable . . . . 33
3.9.1 Twisted Cables . . . . 34
3.10 Repeatability Tests . . . . 34
4 Results and Discussion 35 4.1 Test Results of the Arduino . . . . 35
4.1.1 Susceptibility Tests . . . . 35
4.1.2 Emission Tests . . . . 37
4.2 Tests With Different Ground Configurations . . . . 41
4.3 Tests With Filters . . . . 43
4.4 Test With Shielded Cables . . . . 44
4.5 Discussion of the Repeatability Test . . . . 46
4.6 Test of Passive RTDs . . . . 47
4.7 Discussion of Field Strength Intervals . . . . 48
4.8 General Discussion . . . . 49
5 Conclusion 51 5.1 Future Work . . . . 52
Bibliography 55
A Example Graphs of Test Results 59
Introduction
The need for electromagnetic compatibility is growing steadily as the usage of electronics in our daily lives is increasing more than ever. In order to en- sure that the electronics function correctly in the electronic environment in which they exist, certain precautions must be taken. The reason for this is that all electronics generate unintentional electromagnetic fields, sometimes referred to as electromagnetic radiation. These electromagnetic fields, as well as the intentionally generated electromagnetic fields, such as those from Wi-Fi, 4G, and radio signals, couple with the conductors found in electronic devices, which can cause disturbances. If these potential disturbances are not consid- ered when designing electronics, they can inhibit the proper function of those devices.
Electromagnetic compatibility (EMC) is a term describing an electronic system’s ability to function in the presence of electromagnetic radiation and its ability to not emit unacceptable amounts of electromagnetic radiation. To ensure that electronic devices meet the criteria for EMC tests must be con- ducted. These tests consist of looking at the emissions of electronic devices and their susceptibility towards electromagnetic radiation, where the opposite of susceptibility is usually referred to as immunity. To ensure that a piece of electronic equipment meets the local rules and regulations regarding EMC, a compliance test is performed by a certified EMC testing company. With the steady rise of digitisation, smart homes and internet of things devices, the need for EMC testing is growing and therefore, the need for compliance tests. Fail- ing such a compliance test is often very expensive since it requires redesigning the product, which can take many engineering hours.
One commonly occurring issue concerning EMC is the use of analogue sensors that are not properly shielded. Because of the nature of analogue sig-
1
nals and circuits, these types of sensors are more susceptible to electromag- netic fields than their digital counterparts, which means that more care must be taken into account when designing electronic devices that include analogue sensors.
There are multiple ways of shielding analogue sensors that will be dis- cussed in this thesis, such as:
• Using shielded cables
• Using twisted cables
• Ground planes
• RC filters
• Using ferrites
Testing the efficiency of implementing these shielding techniques is best done in an environment used in compliance testing since they are designed to mimic real-world conditions. One of the ways of testing electronic equip- ment’s susceptibility towards electromagnetic fields is by using an EMC cham- ber. They have walls that are designed to absorb electromagnetic radiation, which ensures that the electromagnetic environment inside the chamber is con- trollable. In such a chamber, a sensor can be exposed to controllable levels of radiation, which may cause the output from the sensor to become faulty. In or- der to see when a sensor fails, their output has to be measured, so implement- ing different shielding techniques while looking at the output of the sensors could give a good idea of which shielding techniques are best to implement for certain types of analogue sensors.
1.1 Research Question
Analogue sensors are susceptible to electromagnetic radiation, and there are different techniques of shielding them. The purpose of this thesis is to ex- pose sensors to electromagnetic fields in an EMC chamber and determine the effectiveness of shielding techniques by looking at the output of the sensors.
1.2 Scope
This thesis work will use an EMC testing chamber provided by the company
Prevas and the data from the sensors will be gathered by looking at their output
voltage with an Arduino. The Arduino may have to be shielded in order to not interfere with the testing of the sensors. The data from the experiment will be analysed to find how much the high field strengths the sensors can be exposed to while certain shielding techniques are implemented. The measuring tech- niques that are to be used will follow the EMC standard for electromagnetic field immunity tests called SS-EN 61000-4-3 and SS-EN 61000-6-1 [1], [2].
1.3 Purpose
The area of EMC is often seen as rather hard to predict when it comes to electronics design. This is likely due to the fact that there are often many different complex interactions within every EMC issue, so that it may seem like issues occur at random. This thesis aims to do a case study of a simplified EMC scenario that is similar to a regularly occurring real-world issue and explore what can be done in that specific scenario. This will hopefully provide insight and knowledge of what techniques are most useful to implement when dealing with such a scenario in the future. Towards the end of the experiments, the data will be evaluated by, for example, redoing certain or all of the tests to see if the results can be replicated.
1.4 Goals
The goals of this project are as following:
• Find if and how an Arduino can be used for gathering sensor output data during an EMC test in an EMC chamber
• Implement ground planes on the sensor platform and measure its effects
• Implement filtering before the sensors in- and outputs and measure its effects
• Implement cable shielding and measure its effects
• Implement the above-mentioned techniques on analogue temperature sensors
• Implement the above-mentioned techniques on other kinds of analogue sensors and attempt to find connections between them
• Evaluate the effectiveness of testing EMC techniques on sensors by look-
ing at their outputs
1.5 Outline
The rest of this thesis is structured as follows. Chapter 2 aims to give the reader
sufficient background information on the EMC field and sensor technology to
understand the work that has been done. Chapter 3 will provide a detailed
explanation of the methods used in this thesis work to achieve the results that
will be shown and discussed in chapter 4. Chapter 5 will briefly summarise the
work and the results and draw conclusions based on them as well as looking
at what future work can be done related to this thesis.
Background
This chapter will cover the essential concepts and theories that are related to this thesis.
2.1 EMI and EMC
Electromagnetic interference (EMI) is a term describing a situation where electrical disturbances are interfering with electronic equipment. The ori- gin of the electrical disturbances is often from other electronic equipment but can also come from natural phenomena like lightning and electrostatic dis- charge (ESD). To combat EMI, the discipline of electromagnetic compatibil- ity (EMC) is introduced, where the issues caused by EMI are analysed and prevented or fixed.
2.1.1 Source/Victim Concept
Since the interactions between the many equipments and other elements in an EMC situation is so complex, a straightforward way of looking at the problem of who does what is to simplify the problem into a source, a victim and a cou- pling path between the two. The source is, for example, electronic equipment, natural phenomena like lightning, or ESD. The victim is the electronic equip- ment that is disturbed. The coupling path between the source and the victim is what enables the EMI and can consist of, for example, wires, electromagnetic radiation, or crosstalk. To be able to reach a satisfying level of EMC between the source and the victim, a good understanding of source through coupling path to victim is needed. Performing actions on any or all of these three ac- tors are the key to controlling EMI. Acting on the source is most often quite
5
hard to accomplish since it would require control over natural events or autho- rised electronic equipment that intentionally operate at allocated frequencies.
However, if the source of the EMI is, for example, the power converter on the device that is being designed, the disturbances can be reduced by changing or redesigning that component. Acting on the victim is done by making the vic- tim less vulnerable to EMI. This is most often hard to implement since it could mean the essential characteristics of the device like the detection level, band- width, or time constant must be changed. Acting on the coupling path between the source and the victim is however an area where there are many possibilities for design in regards to EMC, for example by implementing shielding, filter- ing, grounding, physical separation, changing orientation and loop reduction, as will be explained in section 2.2 [3].
Source
Radiative Coupling
Conductive Coupling
Crosstalk Victim
Figure 2.1: Source, victim, and the coupling path between them.
2.1.2 Radiative Coupling
Radiative coupling is the coupling path of interest when looking at radiated susceptibility in electronics. In order to understand how radiative coupling works, it is essential to understand the relationship between conductors in an electrical system and antennas. When a conductor is exposed to an electro- magnetic field, a current will be formed with an associated voltage, and there- fore, one can view all conductors as unintentional antennae. If a conductor is open-ended, it is only sensitive to E-fields. If a conductor is formed as a loop, it will be sensitive to magnetic fields (H-fields) and/or electric fields (E-fields) depending on its orientation. In real-life applications, very few conductors act purely as loops or as open wires, so both E-fields and H-fields must be considered [3].
There are three factors that have to be taken into account when avoiding
radiative coupling which are; the dimensions of the receiving loop or wire, the
frequency or rise-time of the threat, and the amplitude of the electromagnetic
field. Given an electromagnetic field with a wavelength λ, closed loop con-
ductors with a length of λ/2 m and open-ended conductors with a dimension
of λ/4 m will have the best ability to pick up that electromagnetic field. If the length of the conductor is less than λ/20 m however, the conductor can generally be seen as not coupling at all to the electromagnetic field. Since the frequency of an electromagnetic field is expressed as f (MHz) = 300/λ (m), a field with higher frequency will couple more easily with a conductor of shorter length. Therefore, high frequency fields are generally more dangerous for electronic devices [3], [4].
2.1.3 Electromagnetic Fields
Electromagnetic fields are everywhere, organised in different frequency bands, from longwave radio to gamma rays. Even visible light, which we humans see, is classified as electromagnetic waves. However, the frequencies that are in- teresting from an EMC point of view are those in the radio frequency (RF) band, which is generally said to be around 30 kHz - 300 GHz. To create an electromagnetic field, a radiating antenna is needed. There are two types of antennae, the dipole antenna, and the loop antenna. In the far field, the dipole antenna creates an electrical field, which consequently generates a magnetic field of equal magnitude. The loop antenna, on the other hand, generates a magnetic field which in turn generates an electrical field of equal magnitude.
In the near field, however, the situation looks a bit different, where the dipole antenna mostly generates an electrical field, and the loop antenna mostly gen- erates a magnetic field. The boundary between the near and the far fields for a loop antenna lies around 2D
2/λ where D is the size of the antenna [4], [5].
2.1.4 Standards for Emission and Immunity
In order to sell a product on the European Internal Market, the product must be
CE marked. To CE mark a product, the manufacturer (the product owner) must
show that the product follows the harmonised standards in the area (directives)
that are applicable to that product. One such area is EMC, where standardisa-
tion organisations like IEC and CISPR develop standards that are referred to as
harmonised standards if they are published by The European Commission. It
is the manufacturer’s responsibility to prove that the harmonised standards are
being followed; however, products are usually tested by a third-party certified
EMC test lab. These tests are often quite expensive, and failing a compliance
test means they must be redone, and sometimes the product needs a major re-
design to comply. Because of this, pre-compliance testing is recommendable,
where the product is tested before the compliance testing, often during the de-
velopment of the product itself. This means EMC issues can be found earlier and for less cost than finding them during the compliance test, which happens at the very end of the development phase of a product. Since Sweden is a part of the EU, all products sold to the Swedish market also fall under this jurisdic- tion and must, therefore, be CE marked. The EMC standards in Sweden are distributed by SEK Svensk Elstandard, and are identical to standards created by IEC and CISPR except for a preface in Swedish [6], [7].
What is being tested in the EMC standards can be categorised into three areas, emissions, susceptibility against electromagnetic fields, and suscepti- bility against transients. The emission levels are tested by measuring the am- plitude of high frequency noise coming from a device. The emissions are measured via antennae in an EMC chamber and also measured on the power and telecommunication cables. The measured fields are sent into spectrum analysers. These tests are done to ensure that the device that is being tested is not a potential source of interference to other systems. A device’s suscepti- bility to electromagnetic fields is measured by exposing it to electromagnetic disturbances via an antenna in an EMC chamber and by direct injection on the cables. Testing a device’s susceptibility towards transients is also done by injecting disturbances on the cables, but in pulses that mimic things like ESD and lightning discharge [3].
One of the main series of standards for the EMC of electronic devices is called the SS-EN 61000 series. It contains information on how immunity tests in an EMC chamber are to be carried out (SS-EN 61000-4-3), what levels are acceptable for the emissions of commercial electronics (SS-EN 61000-6-1), and how high levels of immunity commercial electronics should have (SS- EN 61000-6-3). There are many standards produced with information about specific product categories, but since the sensors that are to be tested in this thesis do not align with any of these categories, the generic standard will be used [1], [2], [8].
Immunity is in the EMC world defined as the opposite of susceptibility and is, therefore, what is most interesting for this project. The generic standard for immunity of commercial electronics says that equipment like the ones that are being tested in this thesis shall operate as intended during and after the test and that the level of performance is specified by the manufacturer. The equipment shall be tested in its most susceptible operating mode when used as intended.
The electromagnetic field should be swept from 80 to 1000 MHz and from
1,4 GHz to 6 GHz, have an amplitude of 3 V/m, and an amplitude modulation
of 80% at 1kHz. The device under test shall be placed 2 meters from the
antenna that is generating the electromagnetic field. In industrial electronics,
the equipment must be handle higher field strengths, so the amplitude of the electromagnetic field should instead be 10 V/m. If the equipment under test has functioned correctly at all tested frequencies, it has passed the radiated susceptibility test [2].
SS-EN 61000-4-3 goes through how immunity compliance tests are to be performed in an EMC chamber. It states that both anechoic and modified semi- anechoic chambers can be used for compliance testing. Anechoic chambers are shielded enclosures whose walls, floor, and ceiling are completely covered in material that absorbs electromagnetic fields. Modified semi-anechoic cham- bers are shielded enclosures only the walls and ceiling are covered in material that absorbs electromagnetic fields. The floor in a modified semi-anechoic is a reflective ground plane that has absorbent cones installed on it. This standard also states that the field strengths used in susceptibility testing are allowed to be 6dB higher to ensure that the field strengths do not go below the nominal field strength that is being tested [1].
The generic standard for emissions commercial electronics will be used to test if the Arduino is suitable for immunity testing and states that the equipment under test shall be in the operating mode producing the largest emissions in the frequency band being investigated. The acceptable levels are given in different frequency bands as either peak, quasi-peak, or average values. These tests are made with the test object placed three meters from the antenna [8].
2.1.5 The EMC Market
With the steady rise of digitisation, smart homes, and internet of things de- vices, the demand for testing for EMC is growing steadily. The market for testing and measuring EMC is expected to grow by 10.7% each year between 2019 and 2022 [9] and the market for shielding and filtering is deemed to grow 5% per year between the years 2019 and 2024 [10].
2.2 Protecting Against Electromagnetic Ra- diation
To reduce the effects of radiative coupling by sources that are not controllable,
two main methods can be implemented; shielding and filtering. Examples of
shielding are using shielded cables, shielded boxes, ground planes, RC filters,
and ferrites.
(a) Cutaway of a coax cable [11]. (b) Cross-section of three types of shielded twisted pair cable. Each of the coloured cir- cles in the image is a twisted pair [12].
Figure 2.2
2.2.1 Shielded Cables
Cables running to and from electronic equipment that are longer than the equipment itself is often the largest contributor to EMC issues. Shielding a cable means that the cable is enclosed some kind of conductive material, like a metal, that acts as a barrier against E-fields and H-fields. In most applica- tions, this metal enclosure is not a solid piece of rigid metal but is instead a more flexible material like braided metal or sheets of metal foil. All cables that are wrapped in a conductive foil or braid is a shielded cable, but other than that, there are two general families of shielded cables, coaxial cables and shielded twisted pairs. The coaxial cable seen in figure 2.2a, consists of an in- ner core of conductive metal that is insulated with a dielectric material, often plastic. Outside of that is a sheet or braid of conductive metal which acts as the shield, and the outer layer consists of a plastic jacket that protects from the threats of the outside world. What makes the coaxial cable distinct from a nor- mal shielded cable is that it is mainly intended to be used as a high frequency transmission line where the shield is used as a path for the return current. The shielded pair cable consists of pairs of insulated copper wires that are twisted around each other. This protects against crosstalk and also cancels out some of the differential mode noise that the cable picks up from ambient electric fields.
There are generally three types of shielded twisted pair cables as seen in figure
2.2b, F/UTP where the whole core is wrapped in metal foil or braid, U/FTP
where each pair is individually wrapped in a metal foil or braid and F/FTP
where both the pair and the whole core is wrapped in metal foil or braid. On
the outside of the twisted pair cables is a protective plastic jacket. If the type of
Figure 2.3: Figure showing how return currents move at different frequencies
shielding is a braid, sometimes the nomenclature S/UTP is used, and if there is both a braid and a foil, SF/UTP can be used. There is also U/UTP which is an unshielded twisted cable [3].
To reduce the impact of interference caused by electromagnetic fields, the cable shields are typically connected to ground. It is a common misconcep- tion that grounding cables only on one end is better from a design perspective since it eliminates ground loops. This can be effective in low-frequency cir- cuits, for example, to eliminate 50 Hz humming noise. However, Waldron and Armstrong [13] have shown that with the use of proper grounding techniques, grounding cables on both ends is superior when it comes to tolerance against EMI. Shielded cables are also effective in reducing emissions. Hatsukade and Nagata [14] tested shielded cables and compared them to the non-shielded cables that were used in trains in Japan, and showed that emissions were de- creased by up to 10 dB.
2.2.2 Ground Planes
Grounding generally refers to the act of connecting a conductor to an area with
a common voltage in an electrical system, which is referred to as 0 V. Ground
planes are areas in a circuit that all hold a voltage of 0 V and is used as a return
path for current. Current follows the path of least impedance, which for DC
and low frequency signals means the path of least resistance, which is gener-
ally when the return path is as short as possible. For high frequency signals, however, the return path will follow the path of least inductance, which means it will follow beneath the original signal trace in order to decrease the loop area. An illustration of this can be seen in figure 2.3. As explained in section 2.1.2, reducing the loop area of an antenna will make it less prone to couple with electromagnetic fields, which means better EMC performance. However, in reality, ground planes are not always uniform because of via placement and other design constraints, which means the current cannot always take the path of least impedance, which creates loops, leading to susceptibility to electro- magnetic radiation [15].
2.2.3 Shielded Boxes
A shielded box is an enclosure made of a conductive material. It acts as a barrier against electromagnetic fields, and their effectiveness is defined as the ratio between the incoming field and the part of the field that goes through the barrier. The properties of a shield are governed by its conductivity, magnetic permeability, and thickness, since they are coupled with the barriers surface impedance and penetration depth. These factors are what control a shield’s ability to reflect and absorb electromagnetic fields, where a shield’s absorption increases with conductivity, thickness, frequency, and permeability, and its reflection increases with surface conductivity and wave impedance [3], [16].
2.2.4 EMC Filters
EMC filters are generally built from passive components and are used to reduce
unwanted noise on signals. Since electromagnetic noise is mostly distributed
towards frequencies that are higher than the intended signal, a low pass filter
is used to filter out all the frequencies above what is expected to be found in
the intended signal. EMC filters are mostly used in order to reduce conducted
emissions but can also reduce radiated emissions and susceptibility, since ra-
diated emissions often couple with the conductors in electronics. EMC filters
are often implemented on switching power supplies since they tend to generate
high amounts of conducted emissions but can be used everywhere where high
frequency noise is expected to occur [3].
C R
Figure 2.4: Circuit diagram of an RC filter.
The most common type of low-pass filter is the RC filter. As can be seen in figure 2.4, an RC filter consists of a resistor that is in series with a load and a capacitor in parallel to that load that is connected on one end to ground. What this filter relies upon is the fact that capacitors have reactance. Reactance is similar to resistance, but it varies with frequency. The reactance of a capacitor decreases with frequency, meaning it acts more and more like a closed circuit at higher frequencies. This means high frequency signals will choose the path of least resistance and short through the capacitor to ground, while low fre- quency signals will pass through the resistor. The frequency at which the filter reaches an attenuation of -3 dB is called the cut-off frequency, and it occurs at f =
2πRC1Hz. After the cut-off frequency, the gain keeps dropping with -20 dB per decade. A plot of a typical low-pass filter can be seen in figure 2.5.
Something that needs to be taken into account when designing low pass RC filters is reducing the parasitic inductances. Having too much parasitic induc- tance causes the reactance to increase at very high frequencies, negating the function of the capacitors. Parasitic inductances are inherent in any capacitor and in any conductive material, like wires and traces. This means that when designing an RC filter, it is better to use surface mounted components instead of through hole components and also to connect the components via copper planes instead of wires or traces.
Bypass capacitors are often implemented between the ground and Vdd rails in electronics as short-term energy storage, but they also work as filters against high frequency noise. This is because the reactance of the capacitors means that they, at higher frequency, act close to a short circuit. This means that high frequency noise will short through the capacitor while low frequency and DC signals will be virtually unaffected.
Ferrite beads are modified iron cores that are often used as a first-aid against
EMC issues since they can be easily mounted on cables that are suspected to
be picking up EMI. They filter away high frequency noise by acting as a re-
sistor in its intended frequency range, which turns the noise in that frequency
Figure 2.5: A typical low-pass filter [17].
band into heat energy. They are usually mounted on cables or on PCBs and can be implemented together with decoupling capacitors and damping resis- tors to avoid unwanted resonances. When mounted on cables, the cable can be looped around the ferrite bead, which increases the high frequency impedance, but also introduces capacitance, which causes the impedance to drop quicker at higher frequencies. A simplified equivalent circuit model of a ferrite can be seen in figure 2.6 [6], [18].
R
DCL
R
ACC
Figure 2.6: Equivalent circuit diagram of a ferrite bead.
2.3 Sensors
Sensors operate as an electrical interface to the physical world by translating information from the physical domain into electrical signals. These electrical signals can then be interpreted and used by electronic devices like computers and telephones [19], [20].
2.3.1 Signal Conditioning in Sensors
Signal conditioning in sensors has the function of making the signal coming from a sensing element readable by a processing instrument. This usually consists of boosting the often quite weak output signal with amplifiers and changing the output format, for example, by converting analogue signals to digital. One of the main amplifiers used in sensors are operational amplifiers (op amps), which are active components that are mainly made of transistors, diodes, capacitors, and resistors. The op amp can be used in very many differ- ent applications and is usually implemented with a feedback mechanism where the output is connected to its inverting input, either directly or via resistors or capacitors. Depending on how this feedback mechanism is implemented, the gain of the amplifier can be adjusted [20], [21].
2.3.2 Resistive sensors
Resistive sensors interface with the physical world by turning physical phe- nomenon into a variation of resistance. In general, the physical phenomenon directly changes the resistance of the sensor, by either changing the sensors material properties or its geometry. Resistance can be expressed by
R = ρl
A [Ω] (2.1)
where ρ is the resistivity, l is the length, and A is the cross-sectional area
through which current flows. As can be seen from equation 2.1, changing
the geometry with the parameters A and l or changing the material properties
with parameter ρ, results in a change in resistance of the resistive sensor. This
change in resistance can be measured with an ohm-meter or by measuring the
resulting change in voltage or current that is caused by the change in resistance
[22].
2.3.3 Capacitive sensors
Capacitive sensors convert physical parameters into electricity into a change in capacitance via a transduction element. A capacitor is a passive electrical component that stores electrical energy. In its basic construction, a capacitor is formed of two metal surfaces. The capacitance created by the two metal surfaces can be described with the following equation
C = ε S
d [F ] (2.2)
where ε is the dielectric constant, S is the area of the metal surfaces, and d is the distance between the two metal surfaces. As can be seen from 2.2, in order to change the capacitance C, either the distance between the surfaces needs to be changed, by moving either of the surfaces or the dielectric constant can be changed via a change in the permittivity of the material [22].
2.3.4 Temperature Sensors
Temperature sensors translate the heat energy of molecules into another pa- rameter, which can be electrically measured, such as voltage, resistance, or current. There are many different types of temperature sensors, and a few of the generic ones will be described below [23].
• Semiconductor sensors can be produced at a low cost and use the fact that pn-junctions have a high temperature dependence. When the for- ward biased junction of a diode or a bi-polar transistor is connected to a current generator, the voltage across it decreases almost linearly with temperature. They do however have a limited temperature range [21], [23], [24].
• Thermally Sensitive Resistors (thermistors) have good resolution and sensitivity but are non-linear and have a low temperature range. They are often made from metal oxides or silicon and work by changing their resistance in relation to the surrounding temperature. Thermistors are divided into two groups, those with a negative temperature coefficient (NTC) and those with a positive temperature coefficient (PTC). Ceramic thermistors have a highly non-linear relationship between resistance and temperature, so the temperature needs to be estimated with models, like the simple model, the Fraden model, and the Steinhart and Hart model.
[21], [23], [25].
• Resistive Temperature Devices (RTDs) are accurate and stable but are often larger and more expensive. Their resistance also changes in re- sponse to changes in temperatures but consists of a sensing element that usually contains a film or wire of conductive material with conductors etched or cut into it. The most common type of RTD is the so-called PT100 sensor [21], [23].
• Thermocouples have a high temperature range and are durable. They work by having two different conductive metals, one that is held at a known temperature and one that is exposed to the temperature that is being measured. This creates a difference in voltage between the two metals, which is proportional to the temperature. Since this voltage dif- ference is quite small, the signal needs to be amplified to be measured properly [23], [26].
2.4 EMC Issues in Analogue Sensors
The effect of EMI on digital and analogue circuits has been investigated in recent years. Both in theory and experiments, it has been shown that the most sensitive circuits are analogue. In general, non-linear components such as transistors and diodes cause issues in electronics due to rectification. More specifically, the op amps seem to be the components that are most susceptible to EMI [27]–[29].
2.4.1 Analogue and Digital Sensors
The difference between an analogue and a digital circuit can be understood by looking at the types of output signals they produce. A digital circuit creates a digital output signal of varying voltage that is interpreted as being one of two logic states, a one or a zero, whereas an analogue signal is interpreted as containing any value, generally between a lower and an upper limit. Digital circuits have an inherently higher tolerance against EMI since they have the benefit of using thresholds between the two logic levels, meaning, for example, that an input voltage of 0.9 V and an input of 1.1 V are both interpreted as a one.
The EMI issues that arise for digital circuits are when the EMI levels are so high that either the logic state changes or delays start occurring, which disrupts the timing of the digital circuit. These errors can lead to false information processing, which is hard for systems to recover from and may require a reset.
However, in reality, the levels of EMI required for this to happen is much too
high to impact circuits in normal circumstances, given that some basic EMC techniques have been implemented. When it comes to sensors, most digital circuits have analogue parts to them, so when it is claimed that a sensor is digital, it can still have analogue parts, which makes them more sensitive than an all-digital circuit otherwise would be [4].
2.4.2 Linear Components
Passive, linear components like resistances, capacitors, and inductors are gen- erally seen as not being susceptible to EMI, and in ideal cases, this is true.
However, in real-world applications, these components cannot always be treated as ideal, and in the case of passive components, there will always be parasitic capacitances and inductances to consider. At higher frequencies, these par- asitic values may change the impedance of passive components which may disrupt the intended operation of the circuit [30].
2.4.3 Non-linear Components
The response that electronics most commonly have to EMI is that non-linear
devices cause failure. One common example of this is when transistors in op
amps convert high-frequency EMI into a DC voltage via rectification. When
this DC voltage appears on the output, it acts as an offset voltage, which
changes the behaviour of the op amp. This means that the entire signal be-
comes shifted and adds an offset error and is particularly harmful since it can
lead to a change in the operating point of a circuit. To combat this, filtering
out high frequency noise before it reaches non-linear components is of high
importance [4], [29], [31]–[33].
Methods
In this chapter, the methods used to achieve the results produced in this thesis work will be presented. An Arduino Uno, together with an Ethernet interface board, was used to gather the data that the analogue sensors picked up. The sensors were shielded by implementing a few techniques and then put into an EMC chamber where they were exposed to electromagnetic radiation. The data was then analysed to determine how effective the shielding techniques were.
3.1 Lab Setup
The standard for testing radiated susceptibility SS-EN 61000-4-3 states that the tests shall be made in a shielded enclosure (EMC chamber) in order to protect the test setup from local ambient electromagnetic fields and also to not leak any electromagnetic fields into the local area that would disrupt radio communication and potentially break various laws regarding emission levels.
The equipment under test shall be installed and used as intended during the test, and any data gathering methods shall impact the test as little as possible.
The lab setup described below has tried to follow what is stated in the stan- dard as closely as possible to ensure the data gathered from the experiment is repeatable and matches actual sensor applications as closely as possible.
3.1.1 Sensor Interface
To gather the data from the sensors during the experiments, some kind of pro- cessing unit needs to be attached to the sensor’s output in order to read data.
An Arduino Uno has a simple processor that can be programmed to read both
19
Table 3.1: Components used to make the interface that was used to send the sensor from the chamber to a PC.
Processing unit Arduino Uno rev. 3 SMD Ethernet interface board Arduino Ethernet Shield 2 Power supply Nordic Power ATS012T-W120E
analogue and digital inputs. It has a built-in analogue to digital converter, which turns an analogue signals voltage into a digital value between 0 and 1023 relative to a reference value. This reference value is 5 V by default but can be changed to an internal 1.1 V or to an external reference voltage that can be connected. Converting the output of the Arduino to voltage is therefore done with the following equation: V
out= A
out∗ A
ref/1024 V, where A
outis the digital output value of the Arduino and A
refis the reference voltage of the Arduino. The Arduino also has connections to 5 V and ground, which was needed to power the sensors. A cable had to be connected to the Arduino to send the output of the sensors from inside the EMC chamber to a computer outside of the chamber. By itself, an Arduino only has a USB connector, but USB cables have a length restriction of around 5 meters, which was too short to send out a signal from within the chamber. Therefore, an Ethernet cable was used instead. An Ethernet Shield 2 was attached to the Arduino board, which made it possible to connect the Arduino to the local area network via Ethernet. A local server was set up with XAMPP where the data was stored.
The data that was stored in the local server was presented in graph format that was periodically updated every 10 seconds. Seeing the data in the form of a chart made it very clear when and how much the sensor was being disturbed by the electromagnetic field. After this, the test result was analysed, and it was noted down if the sensor failed or not. Then the raw test data was deleted from the database in order to facilitate a new measurement, and if it was needed to be saved for later, Microsoft Excel Get Data From Web feature was used.
Since the connection between the Arduino and the local server was made
via the local area network, the IP addresses of the Arduino and the computer
where the database was located had to be known. Since these IP addresses
change every now and then by default, the Arduino’s code had to be updated
every time they changed. For this reason, the computer hosting the local server
was given a fixed IP address, and the Arduino was given a fixed IP address via
the code it ran. This meant that the Arduino could simply be connected to the
power cable and Ethernet cable inside the EMC chamber and would connect
with the server without having to change the code which it ran.
(a) Shielded box without the lid. (b) Shielded box without the lid.
Figure 3.1
3.1.2 Shielded Box
Table 3.2: Components used to make the box that was used to shield the Ar- duino.
Box Hammond Manufacturing 1550G
USB Socket Neutrik NAUSB-W-B Ethernet Socket Neutrik NE8FDX-P6
Power Socket Marushin Electric MJ-14SR D-sub Socket Amphenol L717DA15P
To make sure the Arduino did not produce errors when it was exposed to the electromagnetic fields in the EMC chamber during the tests and to make sure it did not emit an unacceptable amount of radiation of its own, an aluminium box with outputs for power, USB, Ethernet and signals were created. An image of the finished shielded box can be seen in figure 3.1. The ground plane of the Arduino was connected to the box chassis by scraping the Arduino PCB near the screw holes. Holes were drilled in the chassis to connect the ground plane of the Arduino to the box chassis via spacers. Holes were also drilled to attach the sockets to the box. To make sure the ground planes and chassis of all the components were connected to each other, a multimeter was used to check for continuity.
USB, Ethernet, power, and signal cables were connected internally to the
sockets and the Arduino in the box. Ferrites were placed on the unshielded
power cable as close to the socket as possible, to filter out the disturbances
present on the cable and to minimise the risk that disturbances could enter the
box via the power cable and radiate from the power cable inside the box. Since a shielded Ethernet cable was used, it was deemed that no ferrite was necessary on that cable. The USB socket was only used to reprogram the Arduino and was not to be connected while the tests were running inside the EMC chamber, and it was therefore deemed that no ferrite was necessary. A 15-pin D-sub was used to receive the output from the sensors as well as provide the sensors with power. At first, no filters were added to the D-sub, but as will be shown in 4.1.1, ferrites were needed here as well.
3.1.3 EMC Chamber
Table 3.3: Components used in the lab setup.
EMC chamber EMC semi-anechoic chamber, Prevas
“Kummelkammaren” at Prevas EMC lab Turntable Innco DS 1200-HA with Innco controller
CO-3000
RF amplifier Vectawave VBA1000-65 Signal generator IFR 2025
Measuring receiver Rohde & Schwarz ESPI7 Preamplifier Schwarzbeck BBV 9743B
Antenna mast Schwarzbeck AM9104, manual antenna mast Antenna (immunity) EMCO 3142B BiConiLog
Antenna (emission) Schwarzbeck VHA9103 30-300 MHz Antenna (emission) Schwarzbeck UHAALP9107 300-1000MHz Antenna (emission) Schwarzbeck STLP9149 1-6 GHz
The EMC chamber seen in figure 3.2 had equipment for emission tests and for radiative susceptibility tests with field strengths up to 20 V/m in the fre- quency range of 80-1000 MHz and for radiative emission tests in the frequency range of 30-6000 MHz. The equipment used for these tests can be seen in table 3.3.
3.2 Test of the Analogue Temperature Sen- sors
The sensor tests were conducted as follows. The antenna was mounted in one
end of the room, and the sensor and shielded Arduino was placed on the edge
Figure 3.2: A picture of the EMC test chamber when rigged for immunity testing with the antenna in vertical orientation.
of the table at a distance of 2 meters away from the antenna, measured with a tape measure. Absorbent cones were placed on the floor between the table and the antenna, and the door was closed. An electromagnetic field with an amplitude of between 0.1 and 20 V/m was selected, which was then generated by the signal generator, amplified by the RF amplifier and sent to the antenna.
An amplitude modulation (AM) was added to the signal with a frequency of 1 kHz and a depth of 80%. The frequency of the electromagnetic field was swept from 80 MHz to 1000 MHz with a frequency step length of 1%, and the time between each step was 4 seconds. A sketch of the test setup inside the chamber can be seen in figure 3.3.
While the sensor was being radiated inside the chamber, its output was read
by the shielded Arduino and sent to the database. This data was then displayed
via a continually updating graph. As soon as the output of the sensor showed
an error of more than 1
◦C, the test was deemed to be a failure, and a new,
lower value of field strength was chosen. If the sensor managed to output the
temperature with an error of less than 1
◦C throughout the whole frequency
Figure 3.3: Sketch showing the test setup
sweep, the test was deemed to be a success, and a new higher value of field strength was chosen. This was done until the highest possible field strength was found where the sensor did not fail.
Following is a list of all the sensors tested in this thesis where A
outis the output from the Arduino ranging from 0 to 1023:
• LM35DZ/NOPB (LM35) is a semiconductor based analogue sensor that has been on the market for a long time. It produces an output voltage that increases linearly with the temperature in Celsius according to the following equation: 0mV + 10mV /
◦C. The temperature is therefore calculated as T
amb=
Aout∗A1024ref∗100 ◦C. Since the output voltage of the sensor at room temperature is around 200 mV, using A
ref= 1.1 V is suitable, which gives a resolution of roughly 0.1
◦C. The LM35 sensor used in these tests was housed in a TO-92 package, which is a through- hole design.
• LM61CIM3/NOPB (LM61) is a semiconductor based analogue sensor.
It produces an output voltage that increases linearly with the temperature in Celsius according the following equation: 600mV + 10mV /
◦C. The temperature is therefore calculated as T
amb=
Aout∗Aref 1024 −0.6
0.01
◦
C. Since
the output voltage of the sensor at room temperature is around 800 mV,
using A
ref= 1.1 V is suitable, which gives a resolution of roughly
0.1
◦C. The LM61 sensor used in these tests was housed in a SOT-23 package, which is a surface mount design.
• The TMP36GT9Z (TMP36) is a semiconductor based analogue sensor.
It produces an output voltage that increases linearly with the temperature in Celsius according the following equation: 500mV + 10mV /
◦C. The temperature is therefore calculated as T
amb=
Aout∗Aref 1024 −0.5
0.01
◦
C. Since the output voltage of the sensor at room temperature is around 700 mV, using A
ref= 1.1 V is suitable, which gives a resolution of roughly 0.1
◦C. The TMP36 sensor used in these tests was housed in a TO-92 package, which is a through-hole design.
• AD22100KTZ (AD22100) is an RTD based analogue sensor. It has on- board signal conditioning consisting of an op amp, current source and resistances, to produce an output voltage that increases linearly with the temperature in Celsius according the following equation: 1375mV + 22.5mV /
◦C. The temperature is calculated as T
amb=
Aout∗Aref 1024 −1.375
0.0225
◦
C. Since the output voltage of the sensor at room temperature is around 1.8 V, using A
ref= 5 V is suitable, which gives a resolution of roughly 0.2
◦C. The AD22100 sensor used in these tests was housed in a TO-92 package, which is a through-hole design.
• MCP9700T-E/TT (MCP9700) is a thermistor based analogue sensor. It has on-board signal conditioning to produce an output voltage that in- creases linearly with the temperature in Celsius according the following equation: 500mV + 10mV /
◦C. The temperature is therefore calculated as T
amb=
Aout∗Aref 1024 −0.5
0.01
◦
C. Since the output voltage of the sensor at room temperature is around 700 mV, using A
ref= 1.1 V is suitable, which gives a resolution of roughly 0.1
◦C. The MCP9700 sensor used in these tests was housed in a SOT-23 package, which is a surface mount design.
• TC1047AVNBTR (TC1047A) is a semiconductor based analogue sen- sor. It produces an output voltage that increases linearly with the temper- ature in Celsius according the following equation: 500mV + 10mV /
◦C.
The temperature is therefore calculated as T
amb=
Aout∗Aref 1024 −0.5
0.01
◦
C.
Since the output voltage of the sensor at room temperature is around
700 mV, using A
ref= 1.1 V is suitable, which gives a resolution of
roughly 0.1
◦C. The TC1047A sensor used in these tests was housed in a SOT-23 package, which is a surface mount design.
• The KTY81/120 (KTY81) is an RTD and is, therefore, a passive re- sistive element, which resistance changes with the temperature. The nominal resistance of KTY-120 at 25
◦C is 1 kΩ, and the change in re- sistance is not quite linear, but at 20
◦C it is roughly 8Ω/
◦C. To read the temperature sensed by the KTY81-120 a voltage divider was cre- ated with a 1000 Ω resistor, which means the temperature is calculated as T
amb=
128000−230∗AoutAout
◦
C with a sensitivity of around 0.5
◦C. The KTY81 sensor used in these tests was housed in a TO-92 package, which is a through-hole design.
• The FK222 PT100 1/3 DIN (FK222) is a PT100 sensor and therefore an RTD, which is a passive resistive element which resistance changes with temperature. The nominal resistance of a PT100 is 100 Ω, and the change in resistance is linear at roughly 0.385 Ω/
◦C. To read the temperature sensed by the FK222 a voltage divider was created with a 1000 Ω resistor, which means the temperature is calculated as T
amb=
102400−200∗Aout
Aout∗0.385
◦
C with a sensitivity of around 1
◦C. The FK222 sensor used in these tests was housed in a custom package similar to TO-92 and was a through-hole design.
3.3 Test of the Arduino and Shielded Box
Before the testing of the sensors began, the integrity of the shielded Arduino had to be tested to ensure that the results that were gathered showed the sensors failing and not the Arduino failing. This was done by connecting the Arduino to a voltage divider using resistances measured to 328 Ω and 181 Ω bringing the 5 V down to
181+328328∗ 5 = 3.2V . Since resistances are linear compo- nents, they are very resilient to electromagnetic radiation, and therefore only the immunity of the Arduino was tested. The Arduino was tested both inside the shielded box and by itself. The tests were conducted in the same way the susceptibility of the sensor, as explained in section 3.1.3.
The emissions coming from the Arduino were also tested and analysed, ac-
cording to SS-EN 61000-6-3. There were three antennas designed for different
frequency ranges. The measurements were taken in both vertical and horizon-
tal orientation and at 1 m antenna height. The emission tests were made both
when the Arduino was inside the shielded box and without the shielded box.
The measurements were made while the Arduino was 3 m from the Antennae, placed on the turntable that was rotating 360
◦.
3.4 Mounting of Sensors
In order to connect the leads running from the Arduino inside the shielded box, mounting platforms had to be created. The TO-92 package is a through-hole design so it can be mounted without permanently soldering them to a printed wire board (PWB). Therefore platforms with round female pin headers were created where the sensors could be mounted, which can be seen in section 3.6.
The sensors in SOT-23 packages are meant for surface mounting and must, therefore, be soldered in place. The sensors needed to be soldered to a PWB with both surface mount and through-hole possibilities to be able to connect the wires from the Arduino. In order to do this, the sensors were first surface mounted on a smaller PWB and then glued to and connected to a through-hole PWB with patch wires. Round female pin headers were then soldered onto the holes, and continuity was tested with a multimeter to make sure the solder provided adequate connection. An image of one of these sensor platforms can be seen in figure 3.4.
Figure 3.4: The sensor platform made for the SOT-23 packaged sensors. The sensor is the small black package seen in the middle of the sensor platform, the rest is to facilitate connecting the sensor to the leads from the Arduino.
3.5 Field Strengths Used During Test of Ra- diated Susceptibility
The field strength used in the susceptibility test of the sensors was varied to
see how well the shielding techniques improved the immunity of the sensors.
Table 3.4: The field strengths used to test the sensors and the corresponding increase from the value before it. The increase was kept under 25% when possible.
Field Strength (V/m) Increase (%)
0.1 -
0.2 100
0.3 50
0.4 33
0.5 25
0.6 20
0.7 17
0.8 15
1.0 25
1.2 20
1.5 25
1.7 14
2.0 18
2.5 25
3.0 20
3.5 17
4.0 15
4.5 13
5.0 12
6.0 20
7.0 17
8.0 15
10.0 25
12.0 20
15.0 25
17.0 13
20.0 18
The equipment at Kummelkammaren made it possible to set the field strength between 0-20 V/m with steps of 0.1 V/m. In order to set the field strength, Excel files that specified variables like Field strength, Frequency step, Stop and Start Frequency and Dwell Time were loaded into a program that controls the Signal generator. Since each test took around 20 minutes to complete, testing every field strength step was not feasible, and therefore a maximum field strength step size of 25% was chosen. This was, however, not always possible in the lower field strengths, as can be seen in table 3.4. When possible, intervals of 0.5, 1, or 5 were used.
3.6 Tests With Different Ground Plane Con- figurations
Three sensor platforms that could house a sensor in a TO-92 package were created with different ground plane configurations. The devices under test were the sensors in a TO-92 package, which are: LM35, AD22100, TMP36, KTY81, and FK222. If the sensor platforms provided a change in the sensors’
EMI behaviour, the amplitude of the incoming field was changed to find at which field strength the sensors would again start failing. The highest field strengths where the sensors still could perform as intended were then noted down.
3.6.1 No Ground Plane
The first sensor platform had no ground plane and had a direct connection from
the sensors’ ground, Vdd, and output pins to the corresponding leads from the
Arduino box. The platform can be seen in figure 3.5.
(a) The front of the PCB used in the test with- out ground plane.
(b) The back of the PCB used in the test with- out ground plane.
Figure 3.5
3.6.2 Perforated Ground Plane
The second sensor platform had a ground plane that was full of holes, some- thing which is generally not recommended in PCB design. Both the ground pin of the sensors and the ground lead of the shielded Arduino were connected to this ground plane. The Vdd and output pins were connected as in the previous sensor platform. The platform can be seen in figure 3.6.
(a) The front of the PCB used in the test of perforated ground plane.
(b) The back of the PCB used in the test of per- forated ground plane.
