High Voltage Breakdown Studies in Electron Beam Evaporation System and Methodologies to Solve the Arc Discharge Problem

By Shakti Kumar Mishra (Enrollment No. ENGG/012007 01041)

Bhabha Atomic Research Centre, Mumbai

A dissertation submitted to the Board of Studies in Engineering Sciences In partial fulfillment of requirements For the Degree of MASTER OF TECHNOLOGY Of

HOMI BHABHA NATIONAL INSTITUTE

August 2011

Homi Bhabha National Institute

Recommendations of the Thesis Examining Committee

As member of the Thesis Examining Committee, we recommended that the dissertation prepared by Shakti Kumar Mishra, Enrollment No. ENGG/012007 01041 entitled High Voltage Breakdown Studies in Electron Beam Evaporation System and Methodologies to Solve the Arc Discharge Problem be accepted as fulfilling the dissertation requirement for the Degree of Master of Technology. Designation Member - 1 Member - 2 Member - 3 Technical advisor Guide (Convener) Chairman Name Dr.S.V.G.Ravindranath AMPD Mr.Gopal Joshi ED Mrs.Namita Maiti L&PTD Mrs.Jyoti Jha SO/E,TDD Mrs.Namita Maiti SO/G,L&PTD Dr.D.N.Badodkar Chairman M.Tech. Committee Electrical (EE1) Final approval and acceptance of this dissertation is contingent upon the candidates submission of the final copies of the thesis to HBNI. Date: Place: BARC, Mumbai.
ii

Signature

Declaration

I, hereby declare that the investigation presented in the thesis has been carried out by me. The work is original and has not been submitted earlier as a whole or in part for a degree / diploma at this or any other Institution / University.

Shakti Kumar Mishra

iii

Dedicated to My Parents

iv

Acknowledgement

I take this opportunity to express my gratitude towards my guide Mrs Namita Maiti for her valuable guidance and comments. I also take the same opportunity to express my gratitude to my technical advisor Mrs. Jyoti Jha for her valuable guidance and extremely useful suggestions. Special thanks to Mr. R.K.Gupta, Head, HCES and Dr.G.Sugilal, Head, PSDS for providing me such a great learning opportunity. I am very much thankful to Mr. K.N.S.Nair, Head Technology Development Division, for his kind suggestions and encouragement. Finally thanks to my Parents, relatives, all senior colleagues, and friends Mr. Seetanshu Shekhar Swain, Mr Puneet Srivastava, Mr. Pabitra Barad, Mr.Sandeep Patil, Mr. Biswajit Swain, Mr T H Gowrishankar, Mr Maruti, Mr. Shiva Sahoo, Mr.Sandeep Singh, Mr. Shiva Prasad Behera and Mr. Ashutosh Mohanty, who gave me untiring support and encouragement.

Shakti Kumar Mishra

Contents
Chapter List of Figures Nomenclature Abstract 1 1.1 1.2 Introduction Electron beam evaporation system Description of different components of Electron beam evaporation system used in the experiment 1.3 Experiments on thin film coating and trouble finding during operation of the system 2 2.1 2.2 2.3 3 3.1 3.2 3.3 4 4.1 Literature overview Vacuum break down Paschen curve Different causes of Breakdown in Electron beam evaporation system Analysis of break down signal . 4 6 6 10 13 15 15 16 17 18 18
vi

Title

Page No. ix xiii xiv 1 2

Observations during arcing in the vacuum chamber FFT Analysis of Break down Signal for Yttrium evaporant Output result of FFT analysis Design and selection of ferrite core Introduction to ferrite core

Chapter 4.2 4.3 4.4 4.5 4.6 4.7 4.8

Title Comparision between inductor and ferrite beads A Closer Look at Ferrite Impedance Material selection for ferrite bead Core Size and Volume Number of Turns Design Consideration for material selection Selection related to frequency characteristics and temperature consideration

Page No. 18 20 21 22 23 23

23 25 27 27 30

4.9 5 5.1 5.2 5.3

Design calculation Mathematical Modelling and simulation Ferrite line structure Travelling Wave transmission line model of ferrite core Simulation of the nonlinear ferrite model with static L (I) Characteristic

30 38 38 38 39 40 44

6 6.1 6.2 6.3 6.4 6.5

FEM analysis of ferrite core Analysis Using Finite Element Method Formulation of basic equations Input parameters used in the simulation Simulation by considering five pieces of ferrite core Simulation by considering ten pieces of ferrite core
vii

6.6

Comparison between five pieces of ferrite core with ten pieces of ferrite core in the finite element method analysis 48 49 49 50 53

7 7.1 7.2 7.3 7.4

Experiment and data analysis Description of experimental setup Experimen without ferrite core Experiment with ferrite core Comparison of breakdown transient signals obtained with ferrite core to that of without ferrite core

59 60 61 63

Conclusion References Appendix-1

viii

List of Figures
Figure No. 1.1 1.2 2.1 2.2 2.3 2.4 2.5 2.6 3.1 3.2 3.3 3.4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 5.1 5.2 5.3 5.4 Title of Figures Electron beam evaporation system Water cooled cupper crucible and deflecting magnet Particle exchange mechanism Anode Heating Mechanism Cathode Heating Mechanism Clump Mechanism Paschen curve Application of ferrite core in computer Breakdown signal during power supply interruption Breakdown transient signal during fault clearance time Breakdown signal with interruption of power supply Output result of FFT applied to fig.3.3 breakdown signal Application of ferrite core for transient suppression Comparision of Impedance characteristic between ferrite beads and air core coil B-H curve for a typical ferrite Graphical representation of the loss tangent Impedance vs Frequency curve Series Reactance vs Frequency curve Series Resistance vs Frequency curve Total Impedance vs Frequency curve Flux Density vs Temperature curve Equivalent diagram for Ferrite line structure. Nonlinear transmission line model with static L /R characteristic Ferrite loaded coaxial transmission line models of equivalent circuit based model Ferrite loaded coaxial transmission line models of transmission line element based model 28 28 19 20 21 22 23 24 24 25 27 27 Page No. 2 4 7 8 8 9 10 14 15 16 16 17 18

ix

Figure No. 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15 5.16 5.17 5.18 6.1 6.2 6.3 6.4 6.5

Title of Figures Induced solenoidal currents in outer conductor Output Wave form for Gaussian input Output Wave form for exponentially damping input signal The simulation result of Fig. 3.3 breakdown signal with passing through ferrite core equivalent model The magnified simulation result of Fig. 3.3 breakdown signal with passing through ferrite core equivalent model The break down signal plotted from CRO data without ferrite core The simulation result of Fig. 5.10 breakdown signal while passing through ferrite core equivalent model The magnified simulation result of Fig. 5.10 breakdown signal while passing through ferrite core equivalent model The break down signal ploted from CRO data without ferrite core The simulation result of Fig.5.13 breakdown signal while passing through ferrite core equivalent model The magnified simulation result of Fig. 5.13 breakdown signal while passing through ferrite core equivalent model The break down signal ploted from CRO data without ferrite core The simulation result of Fig. 5.16 breakdown signal while passing through ferrite core equivalent model The magnified simulation result of Fig. 5.16 breakdown signal while passing through ferrite core equivalent model Output result for meshing of the problem by considering five pieces of ferrite core Output result for Magnetic flux density vector sum using five pieces of ferrite core Output result for Magnetic flux density vector sum with enlarge view using five pieces of ferrite core Output result for 2D Flux line using five pieces of ferrite core Output result for 2D Flux lines with enlarge view using five pieces of ferrite core
x

Page No. 29 31 31 32 32 33 33 34 34 35 35 36 36 37 40 40 41 41 42

Figure No. 6.6 6.7 6.8 6.9 6.10 6.11 6.12 6.13 6.14 6.15 6.16 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11

Title of Figures Output result for magnetic Field Intensity vector sum using five pieces of ferrite core Output result for Total current density vector sum using five pieces of ferrite core Output result for Joule heat generation using five pieces of ferrite core Output result for messing of the problem by considering ten pieces of ferrite core Output result for Magnetic flux density vector sum using ten pieces of ferrite core Output result for Magnetic flux density vector sum with enlarge view using ten pieces of ferrite core Output result for 2D Flux line using ten pieces of ferrite core Output result for 2D Flux lines with enlarge view using ten pieces of ferrite core Output result for Magnetic Field Intensity vector sum using ten pieces of ferrite core Output result for Total current density vector sum using ten pieces of ferrite core Output result for Joule heat generation using ten pieces of ferrite core Experimental setup The break down signal taken from CRO without ferrite core The output result of FFT analysis of fig. 7.2 The break down signal taken from CRO without ferrite core The output result of FFT analysis of fig. 7.4 The break down signal taken from CRO without ferrite core The output result of FFT analysis of fig. 7.6 The break down signal taken from CRO with ferrite core The output result of FFT analysis of fig. 7.8 The break down signal taken from CRO with ferrite core The output result of FFT analysis of fig. 7.10
xi

Page No.

42 43 43 44 44 45 45 46 46 47 47 49 50 50 51 51 52 52 53 53 54 54

Figure No. 7.12 7.13 7.14 7.15 7.16 7.17 7.18 7.19

Title of Figures The break down signal taken from CRO with ferrite core The output result of FFT analysis of fig. 7.12 The break down signal taken from CRO with ferrite core The output result of FFT analysis of fig. 7.14 The breakdown signal taken from CRO with ferrite core The output result of FFT analysis of fig. 7.16 The break down signal taken from CRO with ferrite core The output result of FFT analysis of fig. 7.18

Page No. 55 55 56 56 57 57 58 58

xii

Nomenclature
Symbol A B Name and Unit Magnetic vector potential Magnetic flux density[Wb/ m2] Maximum operating flux density Saturation flux density of the core C L J o dM/dt Lo r Capacitance Magnetic flux, (Weber) Resistivity, (ohm-cm) Inductance, (H) Current density, in A/ m2 Permeability, (H/m) Permeability of free space, (H/m) Magnetisation Switching Rate Ohm Conductivity, (S/m) Slope inductance per unit length [H/m] Relative permittivity Skin depth, (m) Permittivity, (F/m) Input source impedance, (ohm) h Frequency, (Hz) Height [m] Output resistance (ohm) Jt XL ZL Total current Density Inductive reactance, (ohm) Output load impedance, (ohm) Core gap (mm) Transit time M
xiii

Axial magnetisation componenet

Discrete axial length of ferrite core

Abstract
Project Title- High Voltage breakdown studies in electron beam evaporation system and methodologies to solve the arc discharge problem. Project Summary Electron guns typically include an electron emitting source (emitter) for emitting a stream of electrons, high voltage source to accelerate the electrons onto a crucible for holding evaporant material, and a magnet for creating a magnetic field that directs the stream of electrons from the emitter onto the surface of the evaporant material. The incident electron stream evaporates the material to fill the chamber with a gaseous cloud for deposititon onto substrates positioned in the chamber. Vaccum breakdown is a very common problem during the operations of any electron gun. This not only spoils the job but also causes lot of damage to the equipment and its related accessories. The breakdown can be prevented in many ways. So it is important to study the actual cause of breakdown. A systematic study in this direction can zero down to the source of problem. Methodologies can be adapted to solve the arc discharge problem. Arcing occurs especially as the pressure inside the sealed vacuum chamber increases. More specifically, as the material in the crucible evaporates, a portion of it decomposes to positive ions plus some gas. These positive ions are emitted with a certain kinetic energy into the magnetic field, which guides the ions back toward the emitter along a similar path taken by the electron stream. In this project a systematic studies for cause of the breakdown problem, simulation studies and a new approach by inserting ferrite core to solve the breakdown problem has been carried out.

xiv

Chapter-1
Introduction

Introduction

Thin film coatings play an important role ranging from under sea to space applications, including communications, sensors, satellites, optics and aerospace industries [1]. Thin film studies have provided impetus to many new areas of research in solid state physics, chemistry and surface physics, which are based on phenomena uniquely characteristic of the geometry, thickness and structure of thin films. Coating processes can be broadly classified into three groups: vapor phase, which includes physical vapor deposition (PVD) and chemical vapor deposition (CVD); liquid phase, which includes painting, dipping, and electroplating; and solid phase, which includes plasma spray processes and laser cladding [2]. Each process can again be subclassified based on the source of energy used for the deposition of coatings. Each of these processes has its advantages and disadvantages. Resistance heating is the simplest form of vapour source for evaporation. In this method, the material to be evaporated is raised to the required temperature by electrical resistance heating. Here the source is placed in intimate physical contact with the heating element so as to ensure heat economy. Because of such contact there is a possibility of chemical reaction between the source and the material to be evaporated. This would cause film contamination or source destruction. An alternative means of heating is by eddy current or induction heating. In this method the material to be evaporated is raised to evaporation temperature by a radio frequency field. This method is very useful in evaporating large quantities of material but a fine control of the evaporation rate is difficult. Also the radiated radio frequency can cause electrical interference with adjacent electronic equipment and result in severe damage [3]. Deposition by sputtering technique has been quite popular. In the sputtering process a gaseous discharge is created between the materials to be deposited (cathode) and a suitable electrode (anode) placed in the vacuum chamber. By sputtering technique, deposition of refractory materials can be carried out but it offers limitation in terms of contamination of deposited film due to relatively low vacuum of Torr.

RF sputtering is used to sputter insulating or semiconducting targets. The main disadvantages of the RF sputtering method are related to the poor vacuum (> Torr) in

the deposition chamber, to low deposition rate as compared to electron beam evaporation and
1

to the presence of an intense plasma discharge in the neighbourhood of the substrates. The plasma discharge may cause interference to the deposition process. Laser beam and electron beam both provides high power density. In comparison with laser beam, electron beam offers the following advantages a) b) The efficiency of beam generation and energy conversion at the working site is higher The guidance of electron beam on to the job is simpler. In electron beam evaporation system the electron beam generated by thermionic emission is accelerated and bent by 270 degree in an EB field, which subsequently hits the material to be evaporated placed on a crucible. The beam power is translated into heat energy on the material to be evaporated, there by generating stream of vapour. The substrate to be coated is placed above the vapour stream. To obtain desired thickness and property of the coating, the evaporation rate and surrounding medium has to be suitably controlled. The electron beam evaporation system is shown in Fig. 1.1. by one order of magnitude. 1.1 Electron Beam Evaporation System

Fig. 1.1 Electron beam evaporation system.

2

1.2 Description of different components of Electron beam evaporation system used in the experiment There are four main components in the electron beam evaporation system namely, electron-beam (EB) gun assembly, water cooled copper crucible, which contains the material to be evaporated, the substrate (part to be coated), and the vacuum chamber unit. The electron gun is the most important component in the electron beam evaporation system. In electron gun, the beam is generated by thermionic emission from the filament and accelerated by the anode. Electron gun and crucible is mounted inside the work chamber. The electron gun is immersed in an EB field. This field bends the beam by 270 degree and make to fall on the crucible. Thus the vapours generated have very little probability of reaching the gun. This eliminates undesirable electric arcing in the cathode anode region and the consequent process interruption. The cathode is made of tungsten wire formed in helical shape. In order to obtain desired beam shape a focussing electrode is used which is held at filament potential. The shape of focusing electron and the distance between anode and the focussing electrode decide the initial formation of the beam. The filament is floated at -10kV with respect to the anode. The electric field between cathode and anode accelerates the electrons generated at the hot filament and in presence of the magnetic field guides it towards the crucible.The magnetic field is obtained from Nd-Fe-B permanent magnet with suitable magnetic shunts. Electromagnets have been used to position the beam at the desired location and to sweep the beam across the evaporant surface by changing magnetic field. The beam diameter can also be changed using these electromagnets. These electromagnets are placed in longitudinal and lateral direction around the square opening of the crucible. Through this opening the emergent beam from the gun which is placed below the crucible takes 270 degree bend as shown in Fig.1.2. By sweeping the beam over the evaporant surface, not only the evaporation rate can be maximized but also the uniformity in coating thickness can be controlled. The crucible has to be water cooled for high melting point materials and the crucible material should have a good thermal conductivity to ensure that the molten evaporant does not heat the crucible surface. Effective water cooling to crucible can avoid contamination problems which can prevent undesired material to be coated in the substrate. The cupper crucible has low vapour pressure as well as low susceptibility to reactions with the evaporant and the gas in the work chamber at high evaporation temperatures.

Fig. 1.2 Water cooled cupper crucible and deflecting magnet. The substrate is placed in the substrate holder above the water cooled cupper crucible such that the vapour of the evaporant deposits on the substrate. The evaporation rate is to be controlled so that desired thickness of the deposited film is achieved. The vacuum chamber is a small cylindrical chamber placed horizontally with 300mm diameter and 600mm length having ports to accommodate several feed through. The vacuum system has diffusion pump of pumping capacity 1000 lit/sec backed by rotary pump. Vacuum of the order of 3 Torr could be obtained within 40mins from cold start.

The electron beam evaporation system used in the experiment has maximum power capacity of 10kW. Almost 50 percent of the power is expected to be lost in backscattering and radiation. Different evaporant material requires different power like aluminium requires low power and niobium requires relatively higher power according to their melting temperature. 1.3 Experiments on thin film coating and trouble finding during operation of the system The evaporation system is shown in fig. 1.1. The gun was tested at various power levels. Initially a stationary beam i.e. the beam without oscillation was made to fall on an aluminium block mounted inside the water cooled copper crucible. As the aluminium charge used in the experiment was of commercial grade, it was necessary to degas the charge slowly
4

under beam oscillation. Once the charge was completely degassed as indicated by the vacuum level (no increase in pressure on increasing the power), the beam oscillation was stopped and power was increased. The power was maintained for 15 minutes and a vigorous melting was observed. Also a keyhole formation was observed. The charge was subsequently changed to Yttrium oxide and then to Tantalum. During these evaporation experiments, it was observed that the electrical breakdown occurs between anode-cathode gaps. This interrupted the power supply. Such high voltage breakdown adversely affects the evaporation rate. It was found that the number of breakdowns were more during initial heating of the charge due to more degas generated. As the time duration for heating the charge increased, then the subsequent number of breakdown decreases. Also it was observed that when a less degassing material is evaporated like Yttrium oxides then the number of breakdowns decreases compared to high degassing material like Aluminum in particular time duration of operation. To ensure that the deposition rate does not get affected, specialized electrical components required or innovative modification required in the power supply to reduce the potential at the cathode for a short length of time if flashover occurs. Also sufficient protections have been incorporated in the power supply, so that the high voltage surge by pass without affecting the control circuit. Objective The objective of this project was to find out a solution to solve the arc discharge problem in electron beam evaporation system and reduce the number of interruption in the power supply.

Chapter -2
The difficulty in electron beam evaporation systems

Literature overview

Arcing that occurs during operation is high voltage break down of cathode anode gap. This affects the evaporation rate and also the electronic circuits used in the power supply. The Pressure inside the sealed vacuum chamber increases due to evaporation of the evaporant material. More specifically as the material in the crucible evaporates, a portion of it decomposes to positive ions plus some gases. So the Electrical break down strength of vacuum decreases due to increase of pressure inside the chamber and also for positive ions generated during evaporation. Some of these ions can enter and damage the emitter. Before finding out of any solution for solving the arcing problem, it is required to know the electrical breakdown in vacuum and different causes of breakdown in electron beam evaporation system. 2.1 Vacuum break down [5] In a high vacuum, even if the electrodes are separeted a few centimetres, an electron crosses the gap without encountering any collisions. Therefore, the current growth prior to breakdown cannot be due to the formation of electron avalanches. How ever, if a gas is liberated in the vacuum gap then breakdown can occur. Different mechanisims for breakdown in vacuum have been proposed. These can be broadly divided into three categories a) b) c) A) Particle exchange mechanism. Field emission mechanism. Clump theory. Particle exchange mechanism: In this mechanism it is assumed that a charged particle would be emitted from one electrode under the action of the high electric field, and when it impinges on the other electrode, it liberates oppositely charged particles. These particles are accelerated by the applied voltage back to the first electrode where they release more of the original type of particles. When this process becomes cumulative, a chain reaction occurs which leads to the breakdown of the gap.

Fig. 2.1 Particle exchange mechanism. The particle exchange mechanism involves electrons, positive ions, photons and the absorbed gases at the electrode surfaces. Qualitatively, an electron present in the vacuum gap is accelerated towards the anode, and on impact releases A positive ions and C photons. These positive ions are accelerated towards the cathode, and on impact each positive ion liberates B electrons and each photon liberates D electrons. The breakdown will occur if the coefficient of production of secondary electrons exceeds unity. Mathematically, the condition for breakdown can be written as (AB+CD) > 1 Later, Trump and Van de graaff measured these coefficients and showed that they were too small for this process to take place. Accordingly, this theory was modified to allow for the presence of negative ions and the criterion for breakdown then becomes (AB+EF) > 1 Where A and B are the same as before and E and F represent the coefficients for negative and positive ion liberation by positive and negative ions. It was experimentally found that the values of the product EF were close enough to unity for copper, aluminium and stainless steel electrodes to make this mechanism applicable at voltages above 250 kV. B) (i) Field Emission Theory Anode Heating Mechanism This theory postulates that electrons produced at small micro-projections on the cathode due to field emission bombard the anode causing a local rise in temperaure and realease gases and vapours into the vacuum gap. These electrons ionise the atoms of the gas and produce positive ions. These positive ions arrive at the cathode, increase the primary electron emission due to space charge formation and produce secondary electrons by bombarding the surface. The process continues until a sufficient number of electrons are produced to give rise to breakdown, as in the case of a low pressure Townsend type gas discharge.
7

Fig. 2.2 Anode Heating Mechanism. (ii) Cathode Heating Mechanism This mechanism postulates that near the breakdown voltages of the gap, sharp points on the cathode surface are responsible for the existence of the pre breakdown current, which is generated according to the field emission process described below.

Fig. 2.3 Cathode Heating Mechanism. Breakdown in vacuum caused by the heating of a microprojection on the cathode This current causes resistive heating at the tip of a point and when a critical current density is reached, the tip melts and explodes, thus initiating vacuum discharge. This mechanism is called field emission. Thus the initiation of breakdown depends on the conditions and the properties of the cathode surface. Experimental evidence shows that breakdown takes place by this process when the effective cathode electric field is of the order of to V/cm.
8

C) (I)

Clump Mechanism Basically this theory has been devoloped on the following assumptions. A loosely bound particle (clump) exists on one of the electrode surfaces.

(II) On the application of high voltage, this particle gets charged, subsequently gets detached from the mother electrode and is accelerated across the gap. (III) The breakdown occurs due to a discharge in the vapour or gas release by the impact of particle at the target electrode. Cranberg was the first to propose this theory. He initially assumed that the breakdown will occur when energy per unity area (W) delivered to the target electode by the Clump exceeds a constant value C, characteristics of given pair of electrodes. The quantity W is the product of gap voltage (V) and charge density on the Clump. Latter is propertional to the electric field E at the electrode of origin. The criterion for breakdown, therefore, is VE= C.

Fig. 2.4 Clump Mechanism. Clump mechanism of vacuum breakdown In case of parallel plate electrodes the field E = V/d, where d is the distance between the electrodes. So the generalised criterion for breakdown becomes V= (
9

Where C is another constant involving C and the electrode surface conditions. Cranberg presented a summary of the experimental results which satisfied this breakdown criterion with reasonable accuracy. He stated that the origin of the clump was the cathode and obtained a value for the constant C as 60x1010 V2/cm (for iron

particles).However the equation was later modified as V= Cd, where varies between 0.2 and 1.2 depending on the gap length and the electrode material, with a maximum at 0.6. The dependence of V on the electrode material comes from the observations of markings on the electrode surfaces. Craters were observed on the anode and melted regions on the cathode or vice versa after a single breakdown. The breakdown voltage in vacuum is depending upon gas pressure and gap between the electrodes. Paschen curve explains in detail the vacuum breakdown strength. 2.2 Paschen curve [4]

Fig. 2.5 Paschen curve. The Paschen curve is a plot of breakdown voltage for gas as a function of the product of gas pressure and electrode separation. When pressure is given in mBar (thousand times of an atmosphere) and electrode separation in mm the product is mBar-mm. For pressure-distance values above than 25mBar-mm, the breakdown voltage increases with increasing pressure and or separation. Below 25mBar-mm the curve begins to flatten out and as the pressuredistance product drops lower around 7mBar-mm the breakdown voltage begins to rise very rapidly. It happens due to gas become conductive when they are converted into plasma. Plasma also referred to as the fourth state of matter, is a fluid composed not of molecules but of
10

positive ions (cations) and electrons. Plasmas are formed when energy is put into a collection of molecules faster than energy is lost by radiation when the cations and electrons recombine. Recombination of the cations and electrons causes energy to be released as light, this is how fluorescent lights and neon signs work. In a simple atmospheric pressure spark gap, as the potential difference (voltage) between the electrodes is increased (or as the distance between the electrodes is decreased), electrons begin to be emitted by the cathode (negative electrode) and travel to the anode (positive electrode). As they travel, some of the electrons will collide with gas molecules, knocking electrons loose its energy and forming cations and more free electrons. Near the electrodes, where the concentration of traveling electrons is highest, a faint glow caused by the recombination of ions and electrons will become visible. This glow is called a corona discharge and removes energy from the ionized gas at a high enough rate to prevent the formation of plasma. At higher voltages (or smaller gaps) the energy put into the molecules by electron collision exceeds the ability of the corona discharge to dissipate the energy and plasma is formed. The electric field between the electrodes will then separate the cations and electrons. The electrons will flow towards the anode, while the cations will flow towards the cathode. When the cations impact the cathode, they recombine with electrons from the surface of the cathode, completing the electric circuit. In an atmospheric pressure spark gap, the flow of cations and electrons tends to be confined to a fairly narrow channel which is called a spark. At the point on the cathode where the spark connects, a large amount of heat is generated by the impacting cations, damaging the electrode surface (more current = more damage). Electrons impacting the anode surface do not cause much damage, since they are thousands of times lighter than the cations and thus have much less kinetic energy. The rate of collision between gas molecules changes as the pressure changes. The rate at which molecules collide depends on the concentration of molecules (how many there are in a given volume) and the average velocity of the molecules. The temperature of a gas is actually a measure of the average kinetic energy of the molecules which make up the gas. The velocity of the molecules then depends on temperature and the mass of the individual molecules. The pressure of a gas is the amount of force exerted by the gas per unit area. This is a combination of the kinetic energy of the individual molecules and the total number of molecules in a unit of volume. If a sample of gas is at constant temperature and volume, it is clear that a reduction in pressure means that there are fewer molecules and fewer collisions. Another way to think of this is that the distance a molecule can travel (keeping the
11

temperature constant, then the average velocity of the molecules is not changed) without colliding with another molecule has increased. This distance is called the "mean free path". In air at one atmosphere of pressure and a temperature of 0 C (standard temperature and pressure or STP) the mean free path is 1 X 10-7meters (or 1 X 10-4 millimeters, 0.1 micrometers, 100 nanometers). In the atmospheric pressure spark gap for a moment, can be realized that one electron, traveling a centimeter or so from the cathode to the anode, will collide with about 100,000 gas molecules. Whether or not a particular molecule will be ionized depends on how much kinetic energy the electron has when it hits the molecule. If the energy of the incoming electron is greater than the binding force of the electrons in the molecule (the "ionization potential"), one or more electrons can be knocked out. Unlike the gas molecules, whose kinetic energy depends on the gas temperature, the kinetic energy of the electrons is determined by the acceleration caused by the electric field between the electrodes and the distance over which this accelerating force acts. The distance that the electron "falls" in the electric field (and hence its kinetic energy) is limited by the mean free path of the gas. Every time the electron collides with a gas molecule, it loses most of its kinetic energy. If the distance the electron accelerates through is very small, it will never have enough kinetic energy to ionize the gas molecules. This is why higher pressure requires a greater voltage before a plasma can form. Longer gaps reduce the electric field strength and consequently reduce the kinetic energy of the electrons which is why longer gaps require higher voltages to initiate plasmas. So on the "left side of the Paschen curve" signifies that as the pressure drops; the mean free path of the gas increases and the kinetic energy of the electrons will also increase, meaning that a collision with a gas molecule will be more likely to result in ionization. But the x-axis of the Paschen curve is not pressure, but the product of pressure and distance. What has happened is that the distance between the electrodes is now smaller than the mean free path of the gas. The electrons have plenty of kinetic energy, but they are no longer colliding with any gas molecules so no ionization occurs. Summary The most significant experimental factors which influence the breakdown mechanism are: gap length, geometry and material of the electrodes, surface uniformity and treatment of the surface, presence of extraneous particles and residual gas pressure in the vacuum gap.

12

2.3 Different causes of Breakdown in Electron beam evaporation system [5] 1) Cathode initiation of breakdown When high energy electron beam falls on the evaporant material then evaporation occurs due to excessive heating. This metal vapour has a high probability of being ionized in the very high density. Production of positive ions in front of the evaporated material reduces the electron space-charge effect and therefore increases the field emitted current. Furthermore the cathode tip is bombarded by these positive ions which can liberate more electrons so that these two regenerative processes lead to an extremely rapid development of a full vacuum arc in times of the order of 10 nanoseconds. 2) Anode versus cathode control of breakdown The field emitted electrons, accelerated by the field across the gap, cause anode heating on impact with the anode. Breakdown is likely to occur when the temperature at either electrode becomes high enough locally to cause significant evaporation of electrode material into the path of the electron beam. Whether this critical condition first occurs at the cathode or at the anode depends on the gap geometry, on the field enhancement factor, on the relative electrode materials and on whether the gap voltage is dc or pulsed. In electron beam evaporation system dc voltage is used. So in that case it is virtually certain that excessive heating will occur at the anode well before the electron current density reaches a level sufficient for excessive protrusion self heating at the cathode. 3) The micro particle breakdown initiation hypothesis In 1952 Cranberg put forward the hypothesis that breakdown is initiated by tiny bits of material or clumps, impinging on the electrodes under the effect of the electric field. The breakdown, in this clump theory, is the consequence of the electrical gas discharge taking place in the vapour generated when a high speed micro particle collides with one of the electrode. Assuming that the micro particle gets detached from one electrode, it carries along an electrical charge q, which is proportional to the field, and it has energy W=q.V after crossing the gap when it impinges on the other electrode. A criterion for breakdown is obtained by simply stating that breakdown will take place when this energy is larger than a critical value. The micro discharges are observed at voltages lower than the breakdown voltage. They are essentially self extinguishing bursts of current that is they do not involve a complete collapse of the voltage across the gap. On the basis of the clump hypothesis, a micro discharge and a full discharge can be caused by the same process that is by a more or less
13

complete vaporization of the micro particle when it impinges on the opposite electrode. In the case of a micro discharge one can assume that the amount of material vaporized does not generate a bubble of vapour large enough to induce a diverging gas discharge process. In the clump theory the micro discharge and the full discharge are considered as being unrelated to the steady current that exists simultaneously between electrodes since no proposed clump extraction or generation process under the action of this steady current can account for the very large variation of the micro discharge threshold current as a function of the gap in the centimetre range. The clump can rather be considered as detached by the action of the electrostatic pressure from the electrode to which it is loosely bound.

Fig. 2.6 Application of ferrite core in computer. For solve the breakdown problem nothing cannot be control related to evaporation system like vacuum pressure, gap between anode and cathode etc. So only if the transient current can be suppressed externally by providing in power supply system, then it can possible to reduce the high voltage breakdown. In many applications ferrite core is used to suppress high frequency current likely data transmission and power supply cables used in computer and electronic devices as shown in fig. 2.8.The "bumps" are used in fig. 2.8 called ferrite beads or sometimes ferrite chokes. Their goal in life is to reduce EMI (electromagnetic interference) and RFI (Radio frequency interference). The ferrite core can be selected for suppression of particular frequency band [6].

14

Chapter-3

Analysis of Breakdown Signal

For the selection of ferrite core, frequency response of the transient signal is required. For solving the breakdown problem in electron beam evaporation system by inserting ferrite core in the high voltage cable, transient signal response was required during breakdown. Selection of ferrite core depends on the frequency spectrum. The experiment was done in the existing Electron Beam Evaporation system by putting Yttrium as evaporant material. A peak up coil was placed in the high voltage cable for taking the breakdown signal. The breakdown signal was captured and analysed on 2 GS/s, 200 MHz, Tektronix make storage oscilloscope. 3.1 Observations during arcing inside the vacuum chamber During evaporation degassing occurs which leads to the increase of pressure and ions inside the chamber, consequently breakdown strength in vacuum decreases and arcing occur. A transient current was flowing during high voltage breakdown as shown in fig. 3.1.

Fig. 3.1 Breakdown signal during power supply intruption. It was observed that some arcing occured and disappeared without interruption of the power supply. The transient current flowing during without interruption of power supply is shown in fig. 3.2 and fig. 3.3. Reason of non interruption of power supply is that the time
15

duration and peak value of arcing current is less than the power supply trip current and also due to continuous operation of vacuum pump the pressure decreases within the fault clearance time.

Fig. 3.2 Breakdown transient signal during fault clearance time. Analysis of breakdown signal The actual Breakdown signal with interruption of power supply was plotted in a graph plotting software using CRO generated data.

Fig. 3.3 Breakdown signal with interruption of power supply. 3.2 FFT Analysis of Breakdown Signal for Yttrium evaporant So the transient signal taken through CRO during breakdown was analysed by Fast Fourier Transform (FFT) for finding maximum magnitude of frequency content. Different

16

number of breakdown signal was taken for finding frequency spectrum and it was found that the variation of maximum magnitude of frequency content is in 4 to 16 MHz range.

Fig. 3.4 Output result of FFT applies to fig. 3.3 breakdown signal. 3.3 Output result of FFT analysis From the output, it was observed that the maximum magnitude of frequency occurs at 13MHz and other peak frequency occurs in the band of 1MHz to16 MHz range. The maximum magnitude of frequency component was nearly same for different transient signal taken during high voltage breakdown in electron beam evaporation system. Here it was required that the selected ferrite core must suppress the frequency from 1MHz to16 MHz range.

17

Chapter-4
4.1 Introduction to ferrite core [7]

Design and selection of Ferrite core

A ferrite bead or choke is a non-resistant electronic member composed of ferromagnetic compounds that contain iron and trace amounts of nickel, zinc, or manganese oxides. The impedance properties of the ferrite choke allow it to act as high resistance to higher frequency signals and low resistance to lower frequency signals. This way, the high frequency noise is damp out and the absorbed energy is converted to a very small amount of heat as shown in fig. 4.1. The type of ferrite material used in the bead will determine the range of frequency suppression and the physical dimensions and shape of the ferrite bead determine the amount of suppression possible.

Fig.4.1 Application of ferrite core for transient suppression. The beads themselves can be nothing more than a dowel-type device that has a hole in the center or something that looks like a metal blob, which is really a multi-layer bead utilized in surface mount applications. Electromagnetic Interference (EMI) suppression beads, which are comparable to inductors, are designed to raise the transmission of signals with frequency until a resonance point. While a regular inductor's wave propagation would begin to drop as frequency continued to increase, the ferrite bead inductance flattens out and acts as a perfect muffler for a wide frequency band. 4.2 Comparison between inductor and ferrite beads [6] Ferrite beads are used (in a way similar to inductors) as a passive low-pass filter. The geometry and electromagnetic properties of coiled wire over the ferrite bead result in a high
18

resistive impedance (resistance) for high-frequency signals, attenuating high frequency EMI/RFI electronic noise. The energy is either reflected back up the cable, or absorbed resistively within the ferrite core and dissipated as heat. Only in extreme cases, the heat will be noticeable.

Fig.4.2 Comparision of Impedance characteristic between ferrite bead and air core coil. A pure inductor does not dissipate energy, it merely absorbs energy from the circuit and returns it at a later time. A ferrite bead, by design, filters out the high frequency noise in the circuit by dissipating it as heat. The ferrite bead is effectively an inductor with a very small Q factor as shown in fig. 4.2. When electrical noise within the target frequency range travels in the signal cable a back emf is induced in the ferrite bead because of its high inductance. The material used to construct the ferrite bead however, becomes highly resistive at the design
19

frequency range (the magnetic field within the bead is unable to establish properly at that specific frequency range) and the induced current inside the bead is dissipated as heat instead of inducing an opposing current back in the signal cable. It is for this reason that the specific circuit characteristics as well as the frequency band of the noise need to be taken into account when the ferrite bead is installed as a noise filter. 4.3 A Closer Look at Ferrite Impedance [6] Complex impedance of ferrites can be analyzed further if the situation considered is limited to small applied magnetization forces (i.e., small forward current, few turns of conductor around /through the core). In such cases, the application of incremental increases in magnetizing force H to a ferrite coe will result in a corresponding increase in magnetic flux density B in the core. This operation typically displayed graphically via a devices B-H curve, as shown in Figure 4.3

Fig. 4.3 B-H curve for a typical ferrite. With the previously mentioned restrictions, the impedance of a given ferrite bead or core can be expressed as: Z = R (f) + j L (f) The frequency dependent loss term arises from the loss of energy incurred as a result of oscillation of microscopic magnetic regions (called domains) within the ferrite. The loss and the ferrite impedance can be expressed in terms of a complex permeability as: Z = K {j Z Z Where: (f) = the real component of the frequency dependent series complex relative permeability.
20

[( (f) j (f))]} (f)

=K

(f) + j K

= R (f) + j L (f)

(f) = the imaginary component of the frequency dependent series complex relative permeability. K = a constant corresponding to the number of windings and the core dimensions = permeability of free space = radian frequency = 2 f The loss tangent (tan ) of a ferrite material can be defined as the ratio of the imaginary part to the real part of the materials relative permeability. ta n = (f)/ (f)

Figure- 4.4 gives a graphical representation of the loss tangent. As is true with the permeability, the loss tangent is frequency dependent. The loss tangent is an intrinsic property of a given ferrite material formulation. Choosing a particular ferrite material corresponds to choosing a particular loss tangent and associated impedance versus frequency characteristic.

Fig.4.4 Graphical representation of the loss tangent. Ferrite beads are one of the simplest and least expensive types of interference filters to install on pre-existing electronic cabling. For a simple ferrite ring, the wire is simply wrapped around the core through the centre typically 5 or 7 times. Clamp-on cores are also available, which can be attached without wrapping the wire at all. Although the wire is not coiled around the core for this type of ferrite beads, the introduction of the ferrite core around the wire increases the self-inductance of the wire thus still has the effect of absorbing energy from the noise traveling in the wire. 4.4 Material selection for ferrite bead [8] The electrical fast transient spectrum usually contains the frequency band from 10 kHz
21

to 50MHz. To provide proper attenuation, the impedance of the inductor must be sufficiently high over this frequency band. Total impedance of the ferrites is composed of two parts, the series inductive reactance (Xs) and series resistance (Rs). At low frequencies, the reactance is the primary contributor to impedance, but as the frequency increases, the real part of the permeability drops and losses within the core rise, as per Impedance vs Frequency curve shown in figure 4.5. These two factors combine to help acceptable impedance (Zs) over the entire frequency spectrum.

Frequency in MHz Fig.4.5 Impedance vs. Frequency curve (in MHz). Ferrites materials are divided into two groups: Nickel Zinc and Manganese Zinc. Nickel Zinc materials are characterized by low initial permeabilities (<1000), but they maintain their permeabilities at very high frequencies (>100MHz). Manganese Zinc materials, on the other hand, can attain permeabilities in excess of 15,000 but may start to roll-off at frequencies as low as 20 kHz. Because of their low initial permeabilities, nickel zinc materials will not produce high impedance at low frequencies. They are basically used when the majority of un-wanted transient signal is greater than 10 to 20MHz. Manganese Zinc materials, however offer very high permeabilties at low frequencies and are very well suited to EMI suppression in the 10kHz to 50MHz range. 4.5 Core Size and Volume Once the ferrite material and approximate part dimensions are selected for a given application, in circuit impedance and noise suppression performance can be optimized by: 1) Increasing the length of the portion of the conductor surrounded by the ferrite. 2) Increasing the cross sectional area of the ferrite (especially for power applications).

22

3) Selecting a ferrite with an inner diameter most closely matching the outer diameter of the wire or wire bundle to be filtered In general, the best ferrite for a particular application is the longest, thickest device that can be accommodated and whose inner aperture is closely matched to the outer dimensions of the cable to be treated. 4.6 Number of Turns [6] The series impedance of a high frequency ferrite device can be increased by running two or more turns of the treated conductor through the ferrites core. Magnetic theory predicts that the impedance of the device will increase with the square of the number of turns. However, due to the lossy and nonlinear nature of EMI suppression ferrites, a ferrite bead with two turns will yield somewhat less than four times the impedance of an identical part wound with only one turn of the conductor. Since inter winding capacitance will increase along with the number of added turns, Steward recommends that a maximum of two conductor turns be wound on a single part. Increasing the number of turns beyond two will tend to degrade performance at higher frequencies where inter winding capacitance dominates the characteristics of the device. 4.7 Design Consideration for material selection The highest frequency component in the breakdown transient signal is 15MHz from the fast Fourier transform (FFT) analysis. So the selected ferrite core required to provide high impedance at 15MHz frequency. Manganese Zinc material is suitable beacause it offers very high permeabilties in the 10 kHz to 50MHz range. 4.8 Selection related to frequency characteristics and temperature consideration [9] Manganese Zinc ferrites exhibit high permeability at low frequencies (<500 kHz), but roll-off as the frequency increases. The higher the permeability, lower the frequency, where this rolls-off occurs. Fortunately, these materials also become very lossy at high frequencies, and these resistive losses keep the total impedance of the inductor high, beyond 100MHz. Figures 4.6 and 4.7 show how the series inductive reactance (Xs) and series resistance (Rs) change over frequency for the three high permeability materials (J, W and H). Figure 4.8 displays the total impedance versus frequency plot for each material.

23

Frequency in MHz Fig.4.6 Series Reactance vs. Frequency curve.

Frequency in MHz Fig.4.7 Series Resistance vs. Frequency curve.

Frequency in MHz Fig.4.8 Total Impedance vs.Frequency curve. The graphs indicate that H material has a distinct advantage over W and J at low frequencies. However, between 100 kHz and 200 kHz its permeability has dropped low enough so that the total impedance has fallen below the W material impedance. The W material then has the highest impedance until 2MHz where the J material takes over. In addition, all magnetic materials have a Curie temperature, the point where magnetic activity stops. High permeability ferrites usually have Curie temperatures between 120C and 175C. It is important to know where this Curie point is and to maintain the core operating
24

temperature below this limit. Ferrites are not damaged if the Curie temperature is exceeded, but they will become non-magnetic if the Curie temperature is reached during operation. The Flux Density vs Temperature curve explains how the temperature affects the permeability of the ferrite core.

Temperature (Centigrade) Fig.4.9 Flux Density vs Temperature curve. Design Consideration for Frequency Characteristics and Curie temperature The highest frequency component of the breakdown signal is in the range of 1 MHz to 20MHz. So H material ferrite core is suitable according to the impedance vs. frequency curve as shown in fig. 4.5. The heat produced in the ferrite bead depends upon the duration of transient current flowing which is short duration (momentary). So temperature is not a problem for this design. 4.9 Design Calculation [10] The selected material is Manganese Zinc ferrite bead which has high permeability in the An impedance of 1000 is needed in the 1MHz to 20MHz frequency range according range of 1 MHz to 50MHz range. to the total Impedance vs Frequency chart shown in fig.4.8. So the selected impedance of value 1000 can suppress the frequency range from 1MHz to 20 MHz effectively. Lmin =1000 /2 (10,000,000 Hz) =15.9 H So the designed ferrite core will require 15.9 H for eliminating efficiently the transient signal. According to the formula [10]
25

N=1000 (L/AL) Where N= no. of turns inside the ferrite core. L = inductance offered by the ferrite core. AL = Inductance factor in MHz/1000 turns. The high voltage cable connected to electron beam evaporation system has 24mm outer diameter. As per application requirement the selected ferrite core was 46.7mm OD, 24.1mm ID, 18.0mm HT, Kool M type powder cores, easily available in the market, supplied by MAGNETIC Company. The selected powder cores part number was 55438 A2 (77439) which have AL value of 135. By putting and N=10 in the above formula, the L can be calculated as 13.5

H, which is nearly equal to the minimum inductance required. For providing 10 turns, the cable must be inserted 10 times inside the powder core which was not possible due to the 24mm diameter of cable. So if 10 numbers of designed powder core will be inserted in the cable, it is equivalent to 10 turns. Because only the ferrite core equivalent resistance is affect to EMI suppression not the inductance. Also the resistance is added when a high frequency transient current flow. So ferrite core resistance is not dependent on square of number of turn like inductance of a coil [6]. The ferrite core with SL. No.55438A2, Kool M type, part Number 77439 manufactured by MAGNETICS Company (Appendix-1) of 10 numbers can satisfy the design requirement.

26

Chapter-5
5.1 Ferrite line structure [11]-[13]

Mathematical modelling and simulation

The ferrite cores used in the high voltage cable can be represented as ferrite loaded coaxial line. In the experiment ferrite cores were used side by side around the cable, which can be represented in fig.5.1. The ferrite core used in the experiment has 46.7mm OD, 24.1mm ID and 18.0mm HT.Total ten numbers of ferrite core were used in the high voltage cable. So the length of the equivalent ferrite line structure was 180mm.

Fig.5.1 Equivalent diagram for ferrite line structure. The equivalent circuit of the ferrite loaded coaxial line can be represented as distributed lines in fig.5.2. The discrete L and C values represents per unit length inductance and capacitance values multiplied by the section length. The ferrite core nonlinerity is represented by a suitable non linear L (I) function. A direct representation of the magnetization switching rate (dM/dt) in the ferrite is ideally required.

Fig.5.2 Nonlinear transmission line model with static L/I characteristic.

27

The value of

for a coaxial line is ............ (5.1)

Where Lo is the saturated slope inductance per unit length, I is current and dM/dt is the rate of change of the circumferential component of M Where and are the coaxial conductor inner and outer diameters.

Fig.5.3 Ferrite loaded coaxial transmission line models of equivalent circuit based model.

Fig.5.4 Ferrite loaded coaxial transmission line models of tranmission line element based. For numerical solution, the line is spatially discretised into elements of axial length z as shown in fig.5.3. The line current acts on the ferrite magnetization through the component in the L-L magnetisation equation. In turn dM/dt determines the instantaneous magnitudes of the voltage sources , . for each section, given by ............ (5.2) Induced axial flux linkage term The coaxial outer conductor effectively forms an extended short circuited turn linked by the static axial magnetisation flux in the ferrite.The radial dimensions are small compared to the wave lengths of interest and application of Lenzs law indicates that any change in axial
28

flux will induce circumferential currents in the outer coaxial conductor which act to maintain the net axial flux linkage. The effect of the induced component of axial H field is to reduce the angle of swing of M away from its initial axial alignment as shown in fig.5.5.

i)

Solenoid currents are induced to maintain constant .

ii)

is angle calculated using

and

is angle calculated when induced axial H

field term given by -kM is included. Fig.5.5 Induced solenoidal currents in outer conductor. The shock front permeability is therefore reduced and the propagation velocity is increased. Estimation of magnitude of axial induced field Consider a circular cross section of the line. Let the ferrite and the coaxial cable cross sectional areas are denoted and respectively. A change in the axial magnetisation

componenet of M causes a change in axial flux linkage given by =M ............ (5.3)

If the coupling factor is taken as 100% due to the short time scales, the eddy current induced flux component c is equal and opposite c = - ............ (5.4)
29

Making the simplifying assumption that the induced flux density is uniform over the coaxial cross section gives the corresponding induced field strength He as He = ............ (5.5)

An estimate of the persistence time scale of circumferential eddy currents due to variations in axial magnetisation can be gauged from consideration of the relevant L/R time constant. Circumferential eddy currents in the coaxial outer conductor effectively flow in a single turn. 5.2 Travelling Wave transmission line model of ferrite core A direct time stepping solution of the equivalent circuit based ferrite core model was found to be often highly osillatory, numerically stable. This is due to the equivalent circuit LC ladder structure interacting with the gyromagnetic resonance effects. It was important to avoid the possibility of confusion between any real resonant behaviour produced by actual gyromagnetic phenomena in the ferrite line and resonance introduced in the modelling by equivalent circuit. It was realised that the latter phenomenon could be eliminated by replacing each L-C pair by a linear transmission line section with transit time = characteristic impedance Zo shown in fig.5.4. The transit time and

in transmission line element based model as is then equivalent to the time step used for the numerical

solution. Solution of this model proceeds by time steping solution of the L-L magnetisation equation for each of the n elements. The voltage steps propagating along the transmission line is the combination of (a) Changes in the source voltage at the line input at each time step and (b) changes in each of the n voltage sources. Solution of the travelling wave network was found to be computationally more time efficient than solution of the equivalent L-C network. 5.3 Simulation of the nonlinear ferrite model with static L (I) characteristic [14]-[19] The breakdown transient signal was initially simulated by mathematical model of selected feerite core according to the design. The ferrite core was considered as a distributed transmission line model of equally spaced inductance and capacitance for a discrete axial length. The net inductance and capacitance of ferrite core was cosidered as the same for equivalent transmission line model. In the simulation the practically obtained breakdown transient signal was passed through equivalent model of ferrite core and the output gave the

30

suppressed transient signal whose rise time, peak amplitude, fall time and frequency changed with respect to input. Gaussian and exponentially damping input signal was simulated in the mathematical model of equivalent ferrite core for finding out the variation between input and output signal like rise time, fall time, peak amplitude. In the mathematical model the whole ferrite bead inductance was divided into ten numbers of series inductance and total capacitance was divided into ten numbers of parallel capacitance. Individual impedance was taken as 100 Ohm according to the value of selected ferrite core to be used in the experiment [9]. The designed impedance is 1000 Ohm for effective suppression of 1 to 10 MHz as shown in fig.4.8. The mathematical model was divided into ten numbers of series impedance and each was taken as 100 Ohm. Different breakdown signals taken from high voltage breakdown in electron beam evaporation system through pick up coil were simulated in the model and the output results were given below. (i) Simulation of Gaussian signal as input to mathematical model A Gausian input signal was given to equivalent ferrite core transmission line model to scrutinize the variation in the output. Gaussian input signal Output signal

Fig.5.6 Output Waveform for Gaussian signal as input to mathematical model. Obsevation in the output It was found that the output Gaussian signal has low peak amplitude and more rise time compared to input Gaussian signal. (ii) Simulation of exponentially damping signal as input to mathematical model An exponentially damping input signal was given to equivalent ferrite core transmission line model to scrutinize the variation in the output. Exponentially damping input signal
31

Output signal

Fig.5.7 Output Waveform for exponentially damping input signal. Obsevation in the output It was found that the output exponentially damping signal has low peak amplitude and more rise time compared to input exponentially damping signal. (iii) Simulation of Sample breakdown signals as input to mathematical model. A) Sample breakdown signal No-1. A sample breakdown signal taken from experimental setup without ferrite core was given as input to the equivalent ferrite core transmission line model.

Fig.5.8 The simulation result of Fig. 3.3 breakdown signal after passing through ferrite core equivalent model. Actual breakdown signal Output signal

32

Fig.5.9 The magnified simulation result of Fig. 3.3 breakdown signal after passing through ferrite core equivalent model. The output of ferrite core equivalent model shows that the peak value, settling time of actual breakdown signal decreases and rise time, fall time increases as it passes through ferrite core equivalent model. B) Sample breakdown signal No-2. A Sample breakdown signal taken from experimental setup without ferrite core was given to the equivalent ferrite core transmission line model.

Fig.5.10 The breakdown signal ploted from CRO data without ferrite core.

33

Fig.5.11 The simulation result of Fig. 5.10 breakdown signal after passing through ferrite core equivalent model.

Actual breakdown signal

Output signal

Fig.5.12 The magnified simulation result of Fig. 5.10 breakdown signal after passing through ferrite core equivalent model. The output of ferrite core equivalent model shows that the peak value, settling time of actual breakdown signal decreases and rise time, fall time increses as it passes through ferrite core equivalent model. C) Sample breakdown signal No-3.

34

A Sample breakdown signal taken from experimental setup without ferrite core was given to the equivalent ferrite core transmission line model.

Fig.5.13 The breakdown signal ploted from CRO data without ferrite core.

Fig.5.14 The simulation result of Fig. 5.13 breakdown signal after passing through ferrite core equivalent model. Actual breakdown signal from experimental setup Output signal

35

Fig.5.15 The magnified simulation result of Fig. 5.13 breakdown signal after passing through ferrite core equivalent model. The output of ferrite core equivalent model shows that the peak value, settling time of actual breakdown signal decreases and rise time, fall time increses as it passes through ferrite core equivalent model.

D)

Sample breakdown signal No-4. A Sample breakdown signal taken from experimental setup without ferrite core was

given to the equivalent ferrite core transmission line model.

36

Fig.5.16 The breakdown signal ploted from CRO data without ferrite core.

Fig.5.17 The simulation result of Fig. 5.16 breakdown signal after passing through ferrite core equivalent model.

Actual breakdown signal

Output signal

Fig.5.18 The magnified simulation result of Fig. 5.16 breakdown signal after passing through ferrite core equivalent model.

37

The output of ferrite core equivalent model shows that the peak value, settling time of actual breakdown signal decreases and rise time, fall time increses as it passes through ferrite core equivalent model.

Chapter-6

Finite Element based analysis of ferrite core

6.1 Analysis Using Finite Element Method [20] Application of commertially available code : For actual simulation purpose commercially available code was used. In this code first analysis in magnetic nodal is obtained. To simulate a problem following steps are involved, Material of different parts of the problem. Equations used for different boundary condition. Assignment of boundary condition. Loading with actual excitation. Selection of steady, harmonic or transient nature of the excitation or combination of all. In the commercially available code there are inbuilt elements which solve the fixed equation assigned to them. Condition of model can be selected in elements. Element used for loading high voltage cable, ferrite core and sourrounding will solve the magnetic vector
38

potential. Material properties to different zones was assisgned. Excitation to high voltage cable was current density in harmonic condition. After solving the problem results were read in the post processor. From their all data as magnetic flux density vector sum, magnetic field intensity vector sum , current density vector sum , 2D flux lines and amount of Joule heat was obtained. 6.2 Formulation of basic equations Magnetic vector potential was used for basic formulation. Magnetic flux density vector sum B can be expressed as curl of magnetic vector potential A. B= A (1) The equation was derived in cylindrical co-ordinate system. H =Jt i.e (1/) B = Jt A. (2) Now B =

Since magnetic vector potential A and total current Jt has component in the direction of . Hence B can be evaluated as follows. B= A = r/r 0 rA z/r 0 (3) (4) /r / /z = -r (A/z) + z (1/r) (r A)/ r Hence component of B in the direction of r can be expressed as Br = - (A/z) And component of B in the direction of z can be expressed as BZ = (1/r) (r A)/ r Now solving equation (2) with the values obtained in equation (3) and (4). r/r z/r = Jt

(1/) B = /r / /z (1/)Br 0 (1/) Bz

6.3 Input parameters used in the simulation The size of the ferrite core and electrical properties used in the simulation was same as the ferrite core used in the experiment. The excitation transient current density was given as 4.42e6 A/ in the simulation.

The applied frequency was 200 kHz. The inner and outer diameter of the ferrite core was 24mm and 48mm respectively. The height of the ferrite core was taken as 18mm. The cable
39

length was taken as 800mm. The simulation was done by considering the problem as axisymmetric plane. The permeability was taken as 10000 H/m. The resistivity of the powder core was taken as 1e-6 Ohm-m. The cable conductor material was taken as copper. The gap between each ferrite core was 0.5mm and cable to ferrite core was 0.3mm taken for simulation, which was nearly equal to actual dimension of the ferrite core structure used in the experiment. To study the effect of number of ferrite cores to high frequency transient signal, two models with five numbers and ten numbers of ferrite core was modelled. The effect of ferrite core to the high frequency transient current was verified by observing the flux profile, flux density and joule heat generation by simulation.

6.4 Simulation by considering five piece of ferrite core in the high voltage cable 2D axi-symmetric of the system was modelled and the model was sub-divided into triangular elements. Fine meshing of the ferrite core was done for better and accurate results.

40

Fig.6.1 Output result for meshing of the system by considering five pieces of ferrite core. Magnetic Flux Density Profile

Fig.6.2 Output result for Magnetic flux density vector sum using five pieces of ferrite core. From the FEM analysis the flux density in the ferrite core was obtained. Flux density was high at the inner surface of the ferrite core as predicted and very less flux density at the outer surface of the core.

41

Fig.6.3 Output result for Magnetic flux density vector sum with enlarges view using five pieces of ferrite core. Magnetic Flux Line- From the finite element based method analysis the flux line in the ferrite core can be seen as shown in fig.6.4 and fig.6.5.

Fig.6.4 Output result for 2D Flux line using five pieces of ferrite core.

42

Fig.6.5 Output result for 2D Flux lines with enlarges view using five pieces of ferrite core. The flux line profile shows that a higher concentration of flux occure between each ferrite cores contact surface compare to other ferrite core area and also the sorrounding space. Magnetic Field intensity

Fig.6.6 Output result for Magnetic Field Intensity vector sum using five pieces of ferrite core.
43

Magnetic field intensity of the model shows that the magnetic field intensity decreases from centre of the high voltage cable to outer surface of ferrite core. Current density vector sum- From the finite element method analysis the current density vector sum in the ferrite core can be obtained.

Fig.6.7 Output result for total current density vector sum using five pieces of ferrite core. A high concentration of induced eddy current in the ferrite cores was obtained on the inner surface of the ferrite core. Joule heat generation- From the finite element method analysis the Joule heat generation in the ferrite core can be obtained.

Fig.6.8 Output result for Joule heat generation using five pieces of ferrite core.
44

Maximum heat loss due to the high frequency transient current was at the corner of the ferrite core. 6.5 Simulation by considering ten piece of ferrite core in the high voltage cable which is the design value 2D axi-symmetric of the system was modelled and the model was sub-divided into triangular elements. Fine meshing of the ferrite core was done for better and accurate results.

Fig.6.9 Output result for messing of the problem by considering ten pieces of ferrite core. Magnetic Flux Density

Fig.6.10 Output result for Magnetic flux density vector sum using ten pieces of ferrite core.

45

Fig.6.11 Output result for Magnetic flux density vector sum with enlarges view using ten pieces of ferrite core. From the FEM analysis the flux density in the ferrite core was obtained. Flux density was high at the inner surface of the ferrite core as predicted and very less flux density at the outer surface of the core. Magnetic Flux Line- From the finite element based analysis the flux line in the ferrite core can be seen as shown in fig.6.12 and fig.6.13.

Fig.6.12 Output result for 2D Flux line using ten pieces of ferrite core.
46

Fig.6.13 Output result for 2D Flux lines with enlarges view using ten pieces of ferrite core. The flux profile shows that a higher concentration of flux occure between each ferrite cores contact surface compare to other ferrite core area and also the sorrounding space. Magnetic Field intensity- From the finite element method analysis the magnetic field intensity in the ferrite core can be obtained.

Fig.6.14 Output result for Magnetic Field Intensity vector sum using ten pieces of ferrite core.

47

Magnetic field intensity of the model shows that the magnetic field intensity decreases from centre of the high voltage cable to outer surface of ferrite core. Current density vector sum- From the finite element method analysis the current density vector sum in the ferrite core can be obtained.

Fig.6.15 Output result for Total current density vector sum using ten pieces of ferrite core. A high concentration of induced eddy current in the ferrite cores was obtained on the inner surface of the ferrite core. Joule heat generation- From the finite element method analysis the Joule heat generation in the ferrite core can be obtained.

Fig.6.16 Output result for Joule heat generation using ten pieces of ferrite core.
48

From the finite element based analysis the Joule heat generated in the ferrite core was obtained. The permeability of the ferrite core depends upon the temperature rise. Joule heat generation value in the simulation is very less which cannot affect the permeability value of the ferrite core. The joule heat is generated only in breakdown period which is not continious (momentary) with time. The less temperature rise is too low on the ferrite core. 6.6 Comparison between five pieces of ferrite core with ten pieces of ferrite core in the finite element method analysis Ten pieces of ferrite core have more joule heat generation compared to the five cores. Hence ten pieces of ferrite core have more suppression effect to high frequency transient signal compared to that of five pieces of ferrite core. The current density vector sum area in the ferrite core obtained in the simulation was more in ten numbers of ferrite core compared to that of five numbers of ferrite core. The magnetic flux density vector sum obtained in the simulation was more uniform in ten numbers of ferrite core compared to that of five numbers of ferrite core. This represents the equivalent impedance of ten numbers of ferrite core has matched the applied high frequency component of the transient signal. So maximum power from transient signal can transfer efficiently to the ten numbers of ferrite core compared to a case with any other numbers of ferrite core. In ten numbers of ferrite core maximum power from transient signal was transferred to the ferrite core and the energy was dissipated in the form of heat. The impedance of ten numbers of ferrite core matches to the high voltage cable impedance when input transient signal has a frequency range 1 to 15 MHz. Impedance matching can not be obtained by inserting higher or lower number of ferrite cores in the high voltage cable.

49

Chapter-7
7.1 Description of experimental setup

Experiment and Data analysis

The experiment was done in the existing Electron beam evaporation system. A pickup coil was placed in the high voltage cable for taking the high voltage breakdown transient signal. The breakdown signal was observed on 2 GS/s, 200 MHz, Tektronix make storage oscilloscope. The breakdown signal was taken without ferrite core initially with aluminium as evaporant material and different samples of breakdown signal were taken. 10 pieces of ferrite core were inserted in the high voltage cable according to the design value for maximum suppression of transient signal. The breakdown signal was taken with insertion of ferrite core and was observed that the beam current was increased to 1.7 times of beam current obtained without ferrite core. Fig. 7.1 shows the experimental setup, showing the location of sensor coil and ferrite beads on the incoming cable to the gun of the evaporation system.

Fig.7.1 Experimental setup

50

7.2 Experiment without ferrite core The fast Fourier transform (FFT) of different breakdown transient signals collected during experiment without ferrite core is given below. A. Sample breakdown signal taken from experimental setup without ferrite core

Fig.7.2 The breakdown signal taken on CRO without ferrite core. The volt per division is 10V in the above transient signal. So the maximum peak voltage is 30V.

Fig.7.3 The output result for FFT analysis of fig.7.2. Analysis: The Fast Fourier Transform gives 14 MHz as highest frequency component in the breakdown signal taken on CRO without ferrite core experiment.
51

B. Sample breakdown signal taken from experimental setup without ferrite core

Fig.7.4 The breakdown signal taken on CRO without ferrite core. The volt per division is 20V in the above signal. So the maximum peak voltage is 60V.

Fig.7.5 The output result for FFT analysis of fig.7.4. Analysis: The fast Fourier transform gives 15 MHz as the highest frequency component in the breakdown signal taken on CRO without ferrite core experiment.

52

C. Sample breakdown signal taken from experimental setup without ferrite core

Fig.7.6 The breakdown signal taken on CRO without ferrite core. The volt per division is 20V in the above signal. So the maximum peak voltage is 80V.

Fig.7.7 The output result for FFT analysis of fig.7.6. Analysis: The fast Fourier transform gives 14 MHz as the highest frequency component in the breakdown signal taken on CRO without ferrite core experiment.

53

7.3 Experiment with ferrite core The breakdown transient signal was taken with ferrite core by inserting 10 pieces of ferrite core in the high voltage cable. Different samples of breakdown signal were taken and corresponding FFT analysis was carried out for finding highest frequency component. A. Sample breakdown signal taken from experimental setup with ferrite core

Fig.7.8 The breakdown signal taken on CRO with ferrite core. The volt per division is 50V in the above signal. So the maximum peak voltage is 50V.

Fig.7.9 The output result for FFT analysis of fig.7.8.

54

Analysis: The fast Fourier transform gives 18 kHz as highest frequency component in the breakdown signal taken on CRO with ferrite core experiment. B. core Sample breakdown signal taken from experimental setup with ferrite

Fig.7.10 The breakdown signal taken on CRO with ferrite core. The volt per division is 50V in the above signal. So the maximum peak voltage is 50V.

Fig.7.11 The output result for FFT analysis of fig.7.10. Analysis: The fast Fourier transform gives 20 kHz as the highest frequency component in the breakdown signal taken on CRO with ferrite core experiment.
55

C. Sample breakdown signal taken from experimental setup with ferrite core The breakdown signal was taken through a pickup coil, which was placed on the surface of the high voltage cable. The output of the pickup coil was connected to cathode ray oscilloscope (CRO) through a 1:10 probe.

Fig.7.12 The breakdown signal taken on CRO with ferrite core. The volt per division is 50V in the above signal. Thus the maximum peak voltage is 40V.

Fig.7.13 The output result for FFT analysis of fig.7.12.

56

Analysis: The fast Fourier transform gives 19 kHz as the highest frequency component in the breakdown signal taken on CRO with ferrite core experiment. D. Sample breakdown signal taken from experimental setup with ferrite core

Fig.7.14 The breakdown signal taken on CRO with ferrite core. The volt per division is 50V in the above signal. So the maximum peak voltage is 60V.

Fig.7.15 The output result for FFT analysis of fig.7.14. Analysis: The fast Fourier transform gives 18 kHz as highest frequency component in the break down signal taken on CRO with ferrite core experiment.
57

E. Sample break down signal taken from experimental setup with ferrite core

Fig.7.16 The breakdown signal taken on CRO with ferrite core. The volt per division is 50V in the above signal. So the maximum peak voltage is 60V.

Fig.7.17 The output result of FFT analysis for fig.7.16. Analysis: The fast Fourier transform gives 20 kHz as highest frequency component in the breakdown signal taken on CRO with ferrite core experiment.
58

F. Sample breakdown signal taken from experimental setup with ferrite core

Fig.7.18 The breakdown signal taken on CRO with ferrite core. The volt per division is 50V in the above signal. So the maximum peak voltage is 60V.

Fig.7.19 The output result of FFT analysis for fig.7.18. Analysis: The fast Fourier transform gives 19 kHz as the highest frequency component in the breakdown signal taken on CRO with ferrite core experiment.
59

7.4 Comparison of breakdown transient signals obtained with ferrite core to that of without ferrite core The following points were observed from analysis of the transient signal taken from experiment with ferrite core to that of without ferrite core. The peak amplitude of transient current with ferrite core was less compared to that without ferrite core. The frequency components in the transient signal were in the range of MHz without ferrite core, where as it was the range of kHz with ferrite core. The settling time of the transient signal with ferrite core was less as compared with without ferrite core. The rise time and fall time of the transient signal with ferrite core were more compared to that of without ferrite core. The number of high voltage breakdown decreased with ferrite core insertion in the high voltage cable for a given time duration compared with that of without ferrite core.

60

Chapter-8
concluded for high voltage breakdown in electron beam evaporation system.

Conclusion

From practical observations and mathematical modelling, the following points were When high voltage electrical breakdown occurs, a transient current flows for a short duration which contains high frequency components. Different transient current signals were taken through current sensor for different types of evaporant material on storage oscilloscope. A fast Fourier transform (FFT) analysis gives the frequency components of the transient signal, which help for design and selection of ferrite core to solve the breakdown problem in electron beam evaporation system. Practically obtained high voltage breakdown transient signal was given as input to equivalent mathematical ferrite core model. According to the simulation result, the transient signal with ferrite core has more rise time, fall time and peak amplitude, high frequency components reduces compared to transient signal without ferrite core. Modelling of the ferrite core in finite element method shows flux density distribution and joule heat generation within the ferrite core. Hence properly designed and seleced ferrite cores are the best solution for solving the high voltage breakdown problem in electron beam evaporation system. Here ten numbers of ferrite core with SL. No.55438A2, Kool M type, part Number 77439 manufactured by MAGNETICS Company, were used in the high voltage cable for maximum suppression of the transient signal. The practical observation shows that breakdown transient current peak amplitude with ferrite core reduces to half the value as that of without ferrite core. Also the rise time, fall time increases and settling time, high frequency components reduces by inserting ten numbers of ferrite core in the high voltage cable. Mathematical modelling of the high voltage breakdown with ferrite core also confirms the practical observation. It was also observed that electrical breakdown with ferrite core occurs at 1.7 time higher value of beam current as that of without ferrite core. The number of high voltage breakdown decreases with ferrite core for a given evaporant material in a given time duration as compared to that of without ferrite core.

61

References
[1] D. Wolfe, Synthesis and characterization of TiC, TiBCN, TiB2 /TiC and TiC/CrC multilayer coatings by reactive and ion beam assisted, electron beam-physical vapor deposition (EB-PVD), Thesis 2001 D Wolfe, The Pennsylvania State University, 1996. [2] J. Sobota, G. Sorensen, H. Jensen, Z. Bochnicek, and V. Holy, C-N/ MeN Nanocomposite Coatings, Deposition and Testing of Performance, Surf. Coat. Technol., Vol 142-144, 2001, p 590-595. [3] M. Brogren, G.L. Harding, R. Karmhag, C.G. Ribbing, G.A. Niklasson, and L. Stenmark, Titanium-Aluminum-Nitride Coatings for Satellite Temperature Control, Thin Solid Films, Vol. 370 (No. 1-2,) 2000, p 268-277. [4] [5] [6] [7] [8] [9] A. A. Avdienko and M.D.Malev, Flashover in vacuum, Vacuum, Vol. 27, No.12. C Germain and F Rohrbach, High voltage breakdown in vacuum, Vacuum, Vol. 18, No. 7. The Use of Ferrites in EMI Suppression, Steward, U.S.A. How to choose ferrite components for EMI suppression, Ferrite Products Corp. Technical Information Catalog. MAGNETICS Ferrite Core Catalog, FC-601, 1994. MAGNETICS Technical Bulletin, FC-S2, 1994. to Core Material Selection, Magnetics, Division of Spang & Co., Butler, Pennsylvania. [11] J.E.Dolan and H.R.Bolton, Shock front development in ferrite-loaded coaxial lines with axial bias, IEE Proc.-Sci. Meas. Technol., Vol. 147, No. 5, September 2000. [12] Seizo Furuya, Haruo Matsumoto, Hitoshi Fukuda, Takao Ohboshi, Saburo Takano and Juichi Irisawa,Simulation of Nonlinear Co-axial Line Using Ferrite beads, Jpn.J.Phys., volume 41. [13] I.G.Kataev, Electromagnetic shock waves, Lliffe Books Ltd., London 1966. [14] Maurice Weiner, Pulse sharpening effects in ferrites, IEEE transaction on magnetics, Vol. Mag-17, NO. 4, July1981. [15] N.C JOSHI, etc all, Application of Ni substituted Li-Zn-Mn Ferrite for the suppression of transients, Electronics Regional test laboratory report.
62

[10] Robert West, Common Mode Inductors for EMI Filters Require Careful Attention

[16] J.E.Dolan, H.R.Bolton and A.J .Shapland, Development of 60ps rise-time ferriteloaded coaxial line, Electron. Lett. 1997, 33. (24), pp. 2049-2050. [17] L.A Ostrovskii, Formation and development of electromagnetic shock waves in transmission lines containing unsaturated ferrite, Sov.Phys. - Tech. Pllys., 1964, 8, (9), pp. 805-B 13. [18] A.J Baden fuller, Ferrite at microwave frequencies, IEEE electromagnetic Waves Series 23 (Peter Peregrinus. London, 1987), p.23. [19] J Smit and H.P.J Wijn Ferrites, Physical properties of ferromagnetic oxides in relation to their technical applications, Philips Technical Library, Eindhoven, 1959. [20] J.E.Dolan, H.R.Bolton and A.J .Shapland, Simulation of ferrite loaded coaxial lines, Electron. Lett.1993, 29, (9), pp.762-763.

63

Appendix-1

Ferrite core Catalog provided by MAGNETICS Company used for selection of suitable ferrite core.

64