USF Libraries
USF Digital Collections

Design and Implementation of a 200mm 3C-SiC CVD Reactor

MISSING IMAGE

Material Information

Title:
Design and Implementation of a 200mm 3C-SiC CVD Reactor
Physical Description:
Book
Language:
English
Creator:
Frewin, Christopher L
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Cubic silicon carbide
Chemical vapor deposition reactor
Control system
Crystal growth
System safety
Dissertations, Academic -- Electrical Engineering -- Masters -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Silicon carbide, SiC, is a semiconductor material which has many diverse uses in many of today's leading technologies. The wide band-gap aspect of the material has been utilized to create power and high frequency electronics, its physical hardness enables its use for MEMS devices, and the biological compatibility make perfect for utilization in medical applications. SiC is not a chemical compound normally found in nature and must be artificially generated. One of the methods used for the creation of single crystal, high quality SiC material is provided through the use of a chemical vapor deposition reactor. The University of South Florida currently has a horizontal hot-wallLPCVD reactor used by Dr. S. E.^ Saddow and his group to grow epitaxial SiC material for research grants by ONR and ARL.These agencies have commissioned the construction of a second LPCVD reactor for the primary purpose of growing 3C-SiC, a specific SiC crystal polytype, and this work describes the fabrication of the new reactor, MF2. This reactor was designed using the first reactor, MF1, as a template, but the design was modified to better facilitate single crystalline growth. The environment of the reactor is a very important consideration for crystal growth, and slight variations can cause critical defect incorporation into the crystal lattice. Many conditioning runs were required to facilitate the epitaxial growth of the different polytypes of SiC, and constant switching of the primary hot-zone required for the growth of hexagonal 4H-SiC and 6H-SiC to the hot zone required for 3C-SiC consumed precious resources and time.^ The new reactor uses a single primary control to monitor the three most important environmental concerns; hot-zone temperature, gaseous flow, and chamber pressure. The new reactor has been designed to use 100 mm Si substrates instead of the 50mm Si substrate size currently in use by MF1. The construction, testing, and 3C-SiC epitaxial growth on Si substrate capability of a 200 mm 3C-SiC hot-wall LPCVD reactor are demonstrated through this work.
Thesis:
Thesis (M.S.E.E.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Christopher L. Frewin.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 189 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001920864
oclc - 190671940
usfldc doi - E14-SFE0001855
usfldc handle - e14.1855
System ID:
SFS0026173:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001920864
003 fts
005 20080110150507.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080110s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001855
035
(OCoLC)190671940
040
FHM
c FHM
049
FHMM
090
TK145 (ONLINE)
1 100
Frewin, Christopher L.
0 245
Design and Implementation of a 200mm 3C-SiC CVD Reactor
h [electronic resource] /
by Christopher L. Frewin.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: Silicon carbide, SiC, is a semiconductor material which has many diverse uses in many of today's leading technologies. The wide band-gap aspect of the material has been utilized to create power and high frequency electronics, its physical hardness enables its use for MEMS devices, and the biological compatibility make perfect for utilization in medical applications. SiC is not a chemical compound normally found in nature and must be artificially generated. One of the methods used for the creation of single crystal, high quality SiC material is provided through the use of a chemical vapor deposition reactor. The University of South Florida currently has a horizontal hot-wallLPCVD reactor used by Dr. S. E.^ Saddow and his group to grow epitaxial SiC material for research grants by ONR and ARL.These agencies have commissioned the construction of a second LPCVD reactor for the primary purpose of growing 3C-SiC, a specific SiC crystal polytype, and this work describes the fabrication of the new reactor, MF2. This reactor was designed using the first reactor, MF1, as a template, but the design was modified to better facilitate single crystalline growth. The environment of the reactor is a very important consideration for crystal growth, and slight variations can cause critical defect incorporation into the crystal lattice. Many conditioning runs were required to facilitate the epitaxial growth of the different polytypes of SiC, and constant switching of the primary hot-zone required for the growth of hexagonal 4H-SiC and 6H-SiC to the hot zone required for 3C-SiC consumed precious resources and time.^ The new reactor uses a single primary control to monitor the three most important environmental concerns; hot-zone temperature, gaseous flow, and chamber pressure. The new reactor has been designed to use 100 mm Si substrates instead of the 50mm Si substrate size currently in use by MF1. The construction, testing, and 3C-SiC epitaxial growth on Si substrate capability of a 200 mm 3C-SiC hot-wall LPCVD reactor are demonstrated through this work.
502
Thesis (M.S.E.E.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 189 pages.
590
Adviser: Stephen E. Saddow, Ph. D.
653
Cubic silicon carbide.
Chemical vapor deposition reactor.
Control system.
Crystal growth.
System safety.
690
Dissertations, Academic
z USF
x Electrical Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1855



PAGE 1

Design and Implementation of a 200mm 3C-SiC CVD Reactor by Christopher L. Frewin A thesis submission in partial fulfillment of the requirements for the degree of Master of Science in Electrical Engineering Department of Electrical Engineering College of Engineering University of South Florida Major Professor: Stephen E. Saddow, Ph.D. Andrew M. Hoff, Ph.D. Paris H. Wiley, Ph.D. Date of Approval: October 13, 2006 Keywords: cubic silicon carbide, chemical vapor deposition reactor, control system, crystal growth, system safety Copyright 2006, Christopher L. Frewin

PAGE 2

Dedication I am dedicating this work to my parent s and family, who I w ould like to thank, especially my Father who is an electrical en gineer himself, for helping me when I needed and supporting my decision to come and finish my degree finally after so many years. I would like to thank Steve Bradshaw for constantly pestering me all the years I was not in school, and for being my brother. I give special thanks to Ch ristopher Coules for allowing me to stay at his home and not pa y rent when I was poor, feeding me when I was hungry, and providing entertainment when I needed it most.

PAGE 3

Acknowledgements I would like to acknowledge Ian Haselbarth for all the hard work and help he has given me over the years, especially when it co ncerned the reactor. Without his help, this work could have never been accomplished. I would like to acknowledge my mentor, Dr. Stephen E. Saddow, for he saw in me something that few others have seen and gave me a chance to prove myself. I would like to give acknowledgement to th e crystal growers in the group, Dr. Y. Shishkin, Suzie Harvey, and Me ralys Reyes. I appreciate the assistance and guidance in the growth of 3C-SiC you have imparted to me. I would like to acknowledge Jeremy Walker for his knowledge in electronics, and the assistance he gave me in the creation of the relay box I had to c onstruct for this work. Finally I would like to thank Dr. A. M. Hoff and Dr. P. Wiley for taking the time to serve on my committee and provide feedback on my thesis. This work was supported by grants from the Office of Naval Research under Grant No. W911NF-05-2-0028 (Dr. C. E. C. Wood) and the Army Research Laboratory under Grant No. DAAD19-R-0017 (B. Geil), which is gratefully acknowledged.

PAGE 4

i Table of Contents List of Tables................................................................................................................. ....iii List of Figures................................................................................................................ ....iv Abstract....................................................................................................................... .....viii Chapter 1: Introduction.......................................................................................................1 1.1 Silicon carbide CVD.........................................................................................1 1.2 CVD reactor designs.........................................................................................6 1.3 CVD reactor controls........................................................................................9 1.4 Control system theory.....................................................................................11 1.5 Summary.........................................................................................................14 Chapter 2: Chemical Vapor Deposition Reactor Design..................................................16 2.1 Reactor construction.......................................................................................16 2.2 Gas manifold system.......................................................................................25 2.3 RF heating system...........................................................................................30 2.4 Vacuum system...............................................................................................34 2.5 Primary control system...................................................................................36 2.6 Summary.........................................................................................................42 Chapter 3: System Assembly and Preliminary Testing....................................................44 3.1 System assembly overview.............................................................................44 3.2 Interlock safety tests.......................................................................................46 3.3 Vacuum testing...............................................................................................49 3.4 Gas manifold tests...........................................................................................51 3.5 RF and control system tests............................................................................52 3.6 Summary.........................................................................................................53 Chapter 4: Reactor Growth Experiments..........................................................................55 4.1 Silicon melt test...............................................................................................55 4.2 3C-SiC growth experiments............................................................................56 4.3 Summary.........................................................................................................59 Chapter 5: Conclusions and Future Work.........................................................................60 5.1 Conclusions.....................................................................................................60 5.2 Future work.....................................................................................................62 References..................................................................................................................... ....66

PAGE 5

iiAppendices..................................................................................................................... ...69 Appendix A: Reactor housing and hot-wall chamber drawings..........................70 Appendix B: Breakout box drawings and reactor control wiring........................90 Appendix C: Relay box drawings and electronic IC board design....................115 Appendix D: Pump starter motor wiring schematic..........................................119 Appendix E: Labview VI programs for the primary control system.................121

PAGE 6

iii List of Tables Table 1.1: Electrical pr operties of the most common silicon carbide polytypes.............2

PAGE 7

iv List of Figures Figure 1.1: The stacking sequence of the three commonly produced SiC polytypes..........3 Figure 1.2: The Acheson smelting reactor displaying the reactor construction and final products of the reaction............................................................................4 Figure 1.3: The Lely SiC reactor design..............................................................................5 Figure 1.4: The general CVD chamber geometries.............................................................8 Figure 1.5: The general layout of a typical CVD system....................................................9 Figure 1.6: (a) The open loop response cu rve for a self regulating system and (b) the open loop response curve for a non-regulating system with the Zieglar-Nichols's optimum method calculations............................................14 Figure 2.1: A block diagram of the 100mm LPCVD reactor system................................17 Figure 2.2: A sketch of the 75mm hot-wall reactor showing the ss head plate, inlet liner, RF coil, susceptor, insula tion, outlet liner, and exhaust port.................19 Figure 2.3: 200mm hot-wall reactor showi ng the head plate, inlet liner, RF coil, insulation, susceptor, ou tlet liner, and drum (b ack plate assembly)...............19 Figure 2.4: Three-dimensional render ing of the 200mm LP CVD reactor (i.e., MF2)...............................................................................................................21 Figure 2.5: A cross sectional view of th e current reactor hot zone design including the RF coil, quartz tube, foam, gr aphite adaptors, susceptor, SiC polyplate, and sample......................................................................................23 Figure 2.6: 3-Dimensional gas velocity (a) and temperature profile (b) of the 200mm hot-wall SiC CVD reactor.................................................................24 Figure 2.7: A sketch of the gas handling ma nifold displaying the gasses used in the reactor, the primary valves, and the gas transportation routes........................26 Figure 2.8: A sketch of a purge panel s howing a purge panel with valve positions noted................................................................................................................27

PAGE 8

vFigure 2.9: Reactor vacuum design sc hematic displaying placement of piping, filters, valves, gauges and pump.....................................................................35 Figure 2.10: A block diagram showing the si gnal routing of the control system to the various aspects of the reactor....................................................................39 Figure 2.11: A picture of the operational panel for the Labvie w reactor control program set in default mode...........................................................................43 Figure 3.1: A graph three leak-back rate tests performed on the MF2 LPCVD reactor.............................................................................................................51 Figure 3.2: A graph of the data gath ered during the PID tuning for the RF generator heating system.................................................................................54 Figure 4.1: The baseline growth schedule for 3C-SiC epitaxial processing......................57 Figure 4.2: Photographs of the (a) 50 mm and (b) 100 mm 3C-SiC wafers grown with the MF2 LPCVD reactor.........................................................................59 Figure A.1: Mechanical draw ings displaying the specifica tions of the (a) back wall and (b) base plate for the reactor housing.......................................................71 Figure A.2: Mechanical dr awings displaying the specifi cations of th e (a) reactor ceiling and (b) the Al frame assembly............................................................72 Figure A.3: Mechanical dr awings displaying the specifi cations of th e (a) rear loading door (b) the left door of the reactor housing......................................73 Figure A.4: Mechanical draw ings displaying the specifica tions of the (a) right door and the (b) manifold base................................................................................74 Figure A.5: Mechanical draw ings displaying the specifications of the (a) manifold top cover and (b) the assembly sketch for the chamber housing..................75 Figure A.6: Mechanical dr awings displaying the specifi cations of the (a) front mounting plate and the (b) front mounting plate support for the reaction chamber.............................................................................................76 Figure A.7: Mechanical dr awings displaying the speci fications of the (a) gas diffusion plate and the (b) back ba rrel door for the reaction chamber............77 Figure A.8: Mechanical draw ings displaying the specifica tions of the (a) back door assembly and the (b) drum support mount......................................................78

PAGE 9

viFigure A.9: Mechanical dr awings displaying the speci fications of the reaction chamber for (a) the main drum b ack view and (b) side view........................79 Figure A.10: Mechanical drawings disp laying the specificatio n of the reaction chamber for (a) the main drum bolt threading and (b) the drum-gasket tube stoppers...................................................................................................80 Figure A.11: Mechanical drawings displayi ng the specifications of the (a) gasket flange for the quartz tube and th e (b) front mounting plate for the reaction chamber.............................................................................................81 Figure A.12: Mechanical drawings displa ying the specifications of the (a) bolt threading detail for the front moun ting plate and (b) the front mounting plate support for the reaction chamber............................................................82 Figure A.13: Mechanical drawings displa ying the specifications of the (a) front mounting plate VCR gland welds and th e (b) reactor tube assembly............83 Figure A.14: Mechanical drawings displayi ng the specifications of the (a) 100 mm hot-zone carbon foam insulation and (b) the 100 mm graphite susceptor bottom.............................................................................................84 Figure A.15: Mechanical drawi ngs displaying the specifications of the (a) 100 mm susceptor top and the (b) 100 mm quartz inlet liner.......................................85 Figure A.16: Mechanical drawings displayi ng the specifications of the (a) 100 mm female graphite adaptor for the quartz inlet liner and the (b) 100 mm male adaptor for the carbon foam insulation..................................................86 Figure A.17: Three photographs displaying the assembled (a) reactor housing, (b) the manifold, and (c) the reaction ch amber loaded with the hot-zone and quartz inlet liner.......................................................................................87 Figure A.18: Photographs displaying the (a ) rear, or loading, drum assembly, and (b) a view of the reacto r from inside the cleanro om that displays the pyrometer assembly, loading door and an operational reactor......................88 Figure A.19: Photographs displaying the (a) rear of the reaction zone, and (b) front of the reaction zone which display the reactor in fully operational mode................................................................................................................89 Figure B.1: Mechanical drawing displayi ng the specifications of the front panel for the breakout box..............................................................................................91 Figure B.2: Mechanical drawing displayi ng the specifications of the rear panel for the breakout box..............................................................................................92

PAGE 10

vii Figure B.3: Mechanical drawings disp laying the specifications for the (a) analog I/O drawer front, (b) the digital I/ O drawer front, and (c) the drawer assembly for the digital I/O card.....................................................................93 Figure B.4: Mechanical drawing displa ying the specifications of the analog I/O drawer assembly for the breakout box............................................................94 Figure B.5: Mechanical drawings disp laying the specifications for the (a) rear drawer front, (b) the D-pin connect or port holes, and (c) the rear placement for the D-pin connector port holes.................................................95 Figure B.6: Mechanical drawings displa ying the specifications for the (a) rear door with fan holes, and (b) the bottom side plates for the breakout box...............96 Figure B.7: Mechanical drawings displa ying the specifications for the (a) top side plates for the breakout box, and (b ) the assembly drawing of the breakout box without the drawer s and door plates mounted..........................97 Figure B.8: Pictures of the assemble d breakout box with (a) a view from the cleanroom displaying the mounted drawers, gauges, pressure controller, and PC cables, and (b) a rear view of showing the cooling fans, MFC drawers, and transport cables........................................................98 Figure C.1: Mechanical drawings disp laying the specifications for the front and back plates of the relay box..........................................................................115 Figure C.2: The PC style electronic board 1 for the mounting of the relays needed to activate various controls on the reactor....................................................116 Figure C.3: The PC style electronic board 2 for the mounting of the relays and various electronics needed to activat e various controls on the reactor.........117 Figure C.4: A picture s howing electronic board 2 with mounted relays, breakout wiring, and electronics on the right hand side..............................................118 Figure D.5: A hand drawn sketch of the starting system for the vacuum pump motor.............................................................................................................120

PAGE 11

viii Design and Implementation of a 200 mm 3C-SiC Reactor Christopher L. Frewin ABSTRACT Silicon carbide, SiC, is a semiconductor material which has many diverse uses in many of todays leading techno logies. The wide band-gap aspect of the material has been utilized to create power and high freq uency electronics, its physical hardness enables its use for MEMS devices, and the biological compatibility make perfect for utilization in medical applications. SiC is not a chemical compound normally found in nature and must be artificially generated. One of the methods used for the creation of single crystal, high quality SiC material is provided through the use of a chemical vapor deposition reactor. The University of South Florida currently has a horizontal hot-wall LPCVD reactor used by Dr. S. E. Saddow and his group to grow epitaxial SiC material for research grants by ONR and ARL. These agencies have commissioned the c onstruction of a second LPCVD reactor for the primary purpose of growing 3C-SiC, a specific SiC crystal polytype, and this work describes the fabrication of the new r eactor, MF2. This reactor was designed using the first reactor, MF1, as a template, but the design was modified to better facilitate single crystalline growth. The environment of the r eactor is a very important consideration for crystal growth, and slight variations can cau se critical defect incorporation into the crystal lattice. Many conditioning runs were required to facilitate the epitaxial growth of

PAGE 12

ix the different polytypes of SiC, and constant switching of the primary hot-zone required for the growth of hexagonal 4H-SiC and 6H-S iC to the hot zone required for 3C-SiC consumed precious resources and time. The ne w reactor uses a single primary control to monitor the three most important environmenta l concerns; hot-zone temperature, gaseous flow, and chamber pressure. The new r eactor has been designed to use 100 mm Si substrates instead of the 50mm Si subs trate size currently in use by MF1. The construction, testing, and 3C-SiC epitaxial gr owth on Si substrate capability of a 200 mm 3C-SiC hot-wall LPCVD reactor are demonstrated through this work.

PAGE 13

1 Chapter 1: Introduction 1.1 Silicon carbide CVD Silicon carbide has long been proven to be a very useful material for various human endeavors. As the third hardest s ubstance known to man with a Youngs modulus of 424 GPa [1], it has been used as an abra sive, cutting material and a strengthening material in protective ballistic armor [2]. Silicon carbide has a high temperature resistance and has no liquid phase but rather su blimates at 1800C [3]. Silicon carbide is also chemically inert and has li ttle or no reactivity to most chemicals at room temperature [3]. This last property enables SiC to be used in harsh environments. Most recently interest has sh ifted from using silicon carbide as only a mechanical and chemical material to using it as an electronic device material since SiC is a semiconductor. Silicon carbide has many elect rical properties that are useful for high temperature, power, and high-frequency device applications. Silicon carbide is a group IV-IV compound semiconductor with a large, indirect bandgap, which ranges from 2.39 eV (for the 3C-SiC polytype) to 3.33 eV (for the 2H-SiC polytype) [4]. The breakdown electric field, Emax, which is the largest electric field a material can be submitted to before catastrophic breakdown occurs, is 2.49 MV/cm for SiC doped at 1016 cm-3 [5]. When compared to silicon, which has an Emax of 0.401 MV/cm at a doping of 1016 cm-3, silicon carbide has demonstrated that it can more effectively block higher voltages for less

PAGE 14

2 material volume [6]. Silicon carbide also has a high saturated dr ift velocity of 2x107 cm/sec, which enables the material to obtain high channel currents for microwave devices [7, 8]. Silicon carbide also displays excelle nt thermal capabilities. Because the bandgap is larger than silicon, silicon carbide has a much lower intrinsic carrier concentration and can therefore operate at higher temperatures [3]. It also has a thermal conductivity ranging from 3.0 W/cm-K to 5 W/cm-K depe nding on crystal orientation, polytype, and carrier concentration [8]. These values, which are comparable to coppers thermal conductivity of 4.0 W/cm-K, enable silicon ca rbide to dissipate heat that causes a decrease in free-carrier mobility [3]. Because of these properties, silicon carbide makes an ideal candidate material for power and fast switching devices. Table 1.1 displays some of the various electrical properties of sili con carbide and Figure 1.1 shows three of the most popular polyty pe stacking sequences. Table 1.1: Electrical properties of the most common silicon carbide polytypes [9, 10]. 3C-SiC 4H-SiC 6H-SiC Bandgap (eV) 2.36 3.23 3.0 Breakdownfield (MV cm-1) 1 3-5 3-5 Electron Mobility (cm2 V-1s-1) at 300K <800 <900 <400 Hole Mobility (cm2 V-1s-1) at 300K <320 <120 <90 Dielectric Constant (Static) dimensionless unit 9.71 9.66 and 10.03 to c-axis Dielectric Constant (High Frequency) dimensionless unit 6.52 6.52 and 6.70 to c-axis Thermal conductivity (W/cm-K) 5 4.9 4.9

PAGE 15

3 Figure 1.1: The stacking sequence of the three commonly produced SiC polytypes [11]. Silicon carbide is a material that was not discovered until the late 1800s, although its existence was speculated by Jns Jacob Be rzelius [12]. The terr estrial formation of silicon carbide is extremely rare and the material has only been found in diamond inclusions and other volcanic rocks as early as the 1950s [13]. In 1891 near Pittsburgh, Pennsylvania, Acheson developed a process to create a substitute for diamond for cutting and abrasive material using an electric sm elting furnace [14, 15]. The furnaces general design is illustrated in Figure 1.2. By mixing coke and aluminum silicate, he created a silicon carbide mass with larger, hexagonal crysta ls that formed within voids in this mass [14, 15]. The substance was mistakenly dubbed carborundum because it was believed to be derived from corundum, the hi storical name for alumnia, Al2O3 [15]. This belief held that SiC was formed from the combination of Al and C in the furnace, however, it was later found to be created from the bond between Si and C, but the name still remained attached to it [15].

PAGE 16

4 Figure 1.2: The Acheson smelting reactor displaying the reactor cons truction and final products of the reaction. The SiC crystals formed in the voids [14]. Investigation into the electri cal uses of silicon carbide began as early as 1892 with Nicola Tesla and a lamp which contained ca rborundum, but he never fully investigated the materials true electrical properties [15]. In 1907, the worlds first LED was created from silicon carbide, but crystal purity and the extraction process from the Acheson style reactors complicated the further study of th e material [16]. A new reactor design was developed by J. A. Lely in 1955 that mi micked the growth conditions found in the Acheson process voids where the large, he xagonal crystals were formed [17]. His reactor, displayed in Figure 1.3, enabled him to modify th e crystal purity and some of the inherent properties through environmental control [17]. Silicon carbide became the subject of great interest as a semiconductor material, and even became more popular than silicon and germanium. Although the envi ronment could be cont rolled to a certain extent, the material needed long time period s to produce and had many crystalline defects

PAGE 17

5 [17]. The interest to use the material in commercial applications dwindled through the 1960s and 1970s, but material scientists a nd physicists in the United States and the former Soviet Union continued to investigate the material [3]. Figure 1.3: The Lely SiC reactor design [16]. Silicon carbide has seen a great resurgence since the late 1980s when Cree Inc. discovered that high quality epitaxy could be achieved at low temperatures using off-axis substrates [18]. This break through has reinvigorated the scientific community and has lead to many breakthroughs for the material silicon carbide. Silicon carbide power diodes and LEDs are regularly manufactured in todays marketplace, and high frequency MESFETs and power MOSFETs are also being developed [3]. These advances lie in part to being able to control the reactor e nvironment. As Lely demonstrated with his reactor, controlling temperature, pressure, and material influx play an important role in

PAGE 18

6 generating a silicon carbide crystal, and as t echnologies become better able to manipulate these conditions, silicon carbide will become a larger part of the world in the areas of power electronics, sensors, and high frequency devices. Epitaxy is used to form the semiconducto r layers necessary to produce electronic devices. One method of creating an epitaxial layer is using chemical vapor deposition, commonly known as CVD. No rmally, silicon carbide CVD uses silane and a hydrocarbon as the precursor gasses, and hydroge n gas as a carrier [3]. These gasses are passed over a heated graphite su sceptor, usually coated with silicon carbide or tantalum carbide [3]. In the CVD process there are four major steps to growing an epitaxial crystal layer [19]. The first step involves the transpor t of the gas precursors, via a carrier, to the growth site. The carrier gas provides the velocity for the gasses to achieve laminar flow [20]. Laminar flow creates a stagnant barrier layer near th e surface of the susceptor, and the heat within the chamber decomposes th e precursors into simpler molecules. The second step involves the diffusion of the reactants through the boundary layer to the surface of the substrate. The reactants then chemically react with the surface of the substrate and find sites to nucleate solid ma tter and create reaction byproducts. Finally, the waste byproducts and remaining reactants desorb from the surface and through the boundary layer to be transported away by the ca rrier gas. If the conditions for reaction are favorable, a new layer of crystal will be formed on the substrate. 1.2 CVD reactor designs Lelys bulk crystal growth reactor and the numerous reactors that have followed it are used for the bulk growth of large semi conductor crystals. Most of the electronic

PAGE 19

7 devices used today need multiple layers with specific electrical properties in order to achieve their designed functions Thin, single-crystal films are grown on top of the base substrates to realize these device layers, and CVD is one of the processes used to accomplish this job. The process of successful CVD epitaxy is strictly governed by various conditions, and many reactors have been constructed to meet these needs and solve various problems that have hampered single crystalline epitaxial growth process. Pressure is an important aspect of CVD reactors and in the past CVD wa s preformed at atmospheric pressure [21]. The APCVD, or atmospheric pressure CVD re actor, has been replaced with LPCVD, low pressure CVD reactors, or ULPCVD, ultra-lo w pressure CVD reactors, because the lower pressures restrict unwanted gas phase reactions and improve overall film uniformity [21]. Temperature is another important factor in CVD reactions as it governs film uniformity and precursor cracking. Cold wall CVD chambers used a design that cooled the walls of the chamber, but the precursor gasses react non-uniformly and have a low through-put [22]. Hot-wall CVD reactors heat the subs trate area and the surrounding chamber, and this gives the reactants a wider area to decompose over [22]. Th e effect allows a reduction in gas phase reactions which can creat e particle formation on the substrate, and creates a more suitable condi tion for highly uniform films in both thickness and doping [23]. Another important fact or in CVD reactions are the chemical precursors and the ease in which they decompose. Plasma e nhanced CVD reactors, PECVD, uses plasma reactions to enhance the decomposition rates of the precursors [21]. MOCVD, metal organic CVD, and DLICVD, direct liquid in jection CVD, reactors use compounds which contain metals [21].

PAGE 20

8 For the vapor-phase CVD reactor, physical construction occurs mainly in two basic geometries, either horizon tal or vertical [24]. The ma in difference between the two reactors is the precursor delivery system The horizontal C VD places the reaction chamber parallel to the ground, and the substrates are laid flat inside [24]. The gas enters one side of the chamber and exits the other, traveling over the face of each substrate. A vertical CVD reactor transports the gas to th e chamber either from the top or the bottom [24]. The gasses usually flow down onto the substrate, where they pass radially over the substrate surface, and then pass out the bottom of the chamber. Illust rations of these two chamber geometries can be found in Figure 1.4. Figure 1.4: The general CVD chamber geometries. The (a) Horizontal hot-wall design, and (b) vertical design, with (1) precursor inlet, (2 ) substrates, (3) heater or furnace, and (4) exhaust. The general CVD chamber geometries [24]. Regardless of the design selected, al l CVD reactors share the same general operation parameters. The first item on the CVD system is the gas delivery system, which can include gas cabinets, purge systems, bubblers, vaporizers, and injectors [25]. The second system is a metering system that controls the amounts of each gas injected into the chamber [25]. The third part of the reactor is the actual reaction chamber. The chamber requires a method of loading and unloading samples, a sample placement susceptor, a chamber cooling system, and a su sceptor heating system [25]. The heating

PAGE 21

9 system can be implemented through RF induction, thermal lamps, or plasma. The final element in the reactor is the exhaust system which pr ovides a method of safely eliminating waste reactants from the system to the outside world [25]. If harmful, toxic, or dangerous waste gases can be produced by the reactor, a scrubber or a burn bottle is required, and a vacuum system is inserted in to the exhaust system for LPCVD to achieve low pressure operation. Figure 1.5 displays the basic layout of a CVD system as discussed previously. Figure 1.5: The general layout of a typical CVD system [25]. 1.3 CVD reactor controls To obtain a uniform epitaxial layer, in terms of both doping and thickness, a CVD reactor must have fine degree of control ove r gas flow, sample temperature, and process pressure. Gas is controlled through two methods; a metering system to change the volumetric amount that flows into the chambe r and susceptor geometry which generates the laminar flow condition required for actual growth [10]. The volume of gas can be

PAGE 22

10 controlled through an adjustable valve, either manually or automatically. The automatic valve is known as a mass flow controller, or MFC, which uses sensors to determine the amount of flow. The sensors compare their me asured value to a calibrated value for the particular gas that is being used, and the valv e is adjusted to maintain the flow set-point designated by the operator. The second aspect of gas control i nvolves reactor chamber geometry and susceptor design. Fluid flow calculations an d simulations must be created during the design phase of a reactor to ensu re laminar flow is present an d turbulence is minimized or eliminated [10]. As reactant concentration decreases over the susceptor due to reactant consumption, a condition known as depletion results in a tapering of the epi layer thickness over the sample [3]. This conditi on can be relieved by phys ically tapering the susceptor to increase gas velocities in the di rection of the depletion, or through designing the susceptor to mechanically rotate the subs trate [3]. The reaction chamber can also be used to influence gas flow through tape ring to increase gas velocities [10]. Temperature is controlled by imparting the susceptor with energy from a power source. RF induction requires a tank circuit made from an inductor and a capacitor and an AC generator [26]. The circuit is matc hed to the load at the frequency of the generator, and the susceptor couples with the magnetic fiel d produced by the inductor and heats through IR2 heating [26]. Temperature cont rol is achieved by increasing or decreasing the power emitted from the generator. In lamp heating, the susceptor heats through infrared emission from the filament, and increasing or d ecreasing the current through the lamp controls the infrared output [27]. Either temperature generation method

PAGE 23

11 requires the implementation of a feedback l oop to compare the actual temperature of the device to a desired op erator set-point. Exhaust and pressure control are contro lled though the use of vacuum pumps. The removal of the waste products is importa nt to avoid precipitation and to maintain laminar flow [25]. The exhaust system has a blower pump or fan system to create a small pressure differential from atmospheric pressure and the result allows the waste gasses to flow from the higher chamber pressure poten tial to the lower ve nt potential for gas elimination [25]. Vacuum pumps usually have one pumpi ng speed, or through-put, determined by the pump type and the power supplied to the pump. The pressure is controlled in a similar fashion as the MFC. A throttle valve allows the pump line to be constricted to restrict flow from the chambe r to the pump, and can be controlled manually or automatically by using ga uges to read the pressure. 1.4 Control system theory One method used in industry to implemen t automatic control of a system is the use of a Proportional-Integral-Differentiator (PID) controller. The PID system can be easily implemented for each of the CVD contro l processes discussed previously. PID is implemented through a simple feedback clos ed-loop system [28]. The PID controller works by adjusting a control signal based on the error generated between a measured value, PV, and a set-point, SP [28]. The ac tual performance of the PID depends on three separate variable conditions that modify th e control variable, CV. The proportional gain, Kp, is an amplifier responsible for process stab ility [28]. A low value causes the error to drift away from the set-point, and a large va lue will induce oscillations in the control

PAGE 24

12 system [28]. The integral gain, Ki, is used to drive the erro r drift from the proportional gain to zero, but it can also create oscillations if set too high [28]. The final value is the derivative gain, Kd, which is responsible for the rate of change in the error to prevent the controlled value from overshoot or undershoot of the desired set-point [28]. A low derivative value will cause the system to react slowly, a high value will cause the system to oscillate, and the system noise is amplif ied through this variable [28]. The simplest implementation of a PID algorithm is given by the following equation: t t e K t t e K t e K CVd i p 1 (1.1) where CV is the control variable, Kp is the proportional gain, Ki is the integral gain, Kd is the derivative gain, and e(t) is th e error with respect to time [28]. Many systems use modern electronics, sp ecifically digital processors, to implement the PID algorithm. Integration in digital processing us es various processing techniques to manipulate various step functi ons read over a specific amount of time. If the digital processor uses re ctangular integration, the base equation given in (1.1) becomes: k k k k k k k k k k k d k k i k k k p kPV SP e D I P CV CV e e e T K D Te K I e e K P 1 2 1 12 (1.2a) (1.2b) (1.2c) (1.2d) (1.2e) where Kp, Ki, and Kd are the gain values of the pr oportional, integral, and derivative control parameters, e is the error of the read ing between the setpoint, SP, and the process variable, PV, T is the sampling interval, and k is the current interval [28]. The previous

PAGE 25

13 design is referred to as a two-degree of free dom design because all of the controllers actions use the error, e, which is the composite of the set-point and pr ocess variable [28]. The design can multiply transient errors in the system by a quickly changing set-point through applications in the derivative and proportional actions [ 28]. An alternative method to the design uses the one-degree of freedom design where the proportional and derivative actions only work on the process vari able instead and is given in the following equation: 2 1 12 k k k d k k i k k k p kPV PV PV T K D Te K I PV PV K P (1.3a) (1.3b) (1.3c) where the variables are the same as explained in (1.2) [28]. Various alternative methods and complexities exist for dig ital integration methods, and syst em designers must take the methods used by hardware into consideration in PID design. Many methods exist for tuning a PID syst em including Zieglar-Nichols's, CohenCoon's, Chien, Hrones, and Reswicks, Takaha shis, and F. G. Shinskeys tuning rules [29]. Each of these systems performs an open-loop manual system test to gather data on the systems response to a change in the pro cess variable. The system is stabilized and the controllers output is recorded and then the controller is increased by 5 to 20 percent [29]. If the system is self regulating, as in pressure and temperature systems, a steady state will be reached with this new stimulus but if the system is non-regulating, the output will keep increasing at a steady rate [29] The data can then be graphed and one of the tuning methods can be used to optimize the PID system. Zieglar-Nichols's method is a simple and easy method commonly used to ga in 25% damping in a system [29]. Using

PAGE 26

14 the open loop response data, the change in pr ocess variable over time compared to the change in controller output is calculated to gain a process rate per change in controller output, RR [29]. The process dead time, or th e time required for the system to accelerate to the response, is calculated for the system and the PID values are generated as shown in Figure 1.6. (a) (b) Figure 1.6: (a) The open loop response curve for a self regulating system and (b) the open loop response curve for a non-regulating system with the Ziegla r-Nichols's optimum method calculations [29]. 1.5 Summary Silicon carbide is a promising wide bandga p semiconductor material that displays electrical properties well suited to power de vices and high frequency applications, and it

PAGE 27

15 is physically suited to be used in harsh envi ronments and displays superior strength for MEMS devices. While bulk crysta l growth provides the base material used in substrates, the complexity of electrical devices requires other methods of growth, namely chemical vapor deposition. To minimize crystalline def ects which can hamper a devices operation, the CVD reactor must be designed to provide a well controlled environment. Control of gas flow, pressure, and temper ature must be considered when building a CVD reactor, either physically through the use of susceptors or electronically th rough the use of PID algorithms. In Chapter 2, the design of a 200 mm 3C-SiC reactor wi ll be discussed in detail. Specifics concerning the control algor ithm implemented will also be discussed in that chapter, allong with details of the CV D reactor designed, built and tested in this work. The tuning and testing results of the sy stem are covered in Chapter 3. Chapter 4 will discuss the epitaxial growth of 3C-SiC grown with the reactor, and Chapter 5 will outline suggested future work to improve the reactor further.

PAGE 28

16 Chapter 2: Chemical Vapor Deposition Reactor Design 2.1 Reactor construction The LPCVD hot-wall reactor is a system co mposed of three main systems, the gas delivery system, the heating system, and th e vacuum system. As discussed in the previous chapter, these three systems can act independently through their own control systems and parameters, however, to control the reactor environment effectively for the growth of single crystal SiC epitaxial films, th e three systems must be integrated together and act in concert. The purpose of this work was to build a LPCVD hot-wall reactor which would be under the control of one contro l system, and to enable this requirement, a primary control system was added to oversee the three major sub-systems of the SiC reactor. The block diagram showing the impl ementation of this system is displayed in Figure 2.1. Each section of this chapter will co ver each separate subset of the reactor as mentioned above, and a detailed description of the design process behind each of the individual systems. This first section c overs the design process for the reactor housing and the hot-wall construction. The next sect ions describe the ga s manifold, the RF heating system, the vacuum system, and the primary control system. The entire system is based on the original 75mm APCVD cold-wall reactor designed by Thomas E. Schattner during his masters research at Mississippi State University [10]. The system was later conve rted into an LPCVD reactor by Matthew T. Smith for his masters research at the Univ ersity of South Florida [30], and hot-wall

PAGE 29

17 development, as well as an increase to 200mm size, was developed by Rachel L. MyersWard as part of her Ph.D. dissertation in elec trical engineering at th e University of South Florida [31]. These three works cover many of the current CVD reactor system design parameters, and the new reactor design takes the knowledge that was built up during these projects into consideration. Figure 2.1: A block diagram of the 100mm LPCVD react or system. The major gas lines are split into three sections dictated by blue, red, and purple lines, and the black lines show the path of the exhaust gasses. Analog and digital signals are in orange and green respectively. The reactor housing was modeled after the first reactor built at USF, called MF1 by the USF SiC Group. The new reactor is re ferred to as MF2 as it is the second generation in the MF series. The MF2 ch amber housing was made with 1.25cm thick anodized aluminum plates to both house the reactor and protect the operator from the

PAGE 30

18 chamber heat and gases in the unlikely event of a leak [10]. The plates were mounted onto an aluminum frame which was grooved to allow hex bolts and wing nuts to fasten the plates firmly to the frame. The reaction tube was mounted length wise in the center of the housing, and cooling water fed to the cold -wall reaction tube an d the stainless steel inlet head plate from holes drilled in the bottom of the box [10]. The reaction tube was accessed through a door along one of the long sides of the reactor, and fastened with a single, turn-screw latch [10]. The gas system was transported to the head plate from a manifold through holes in the ceiling of the chamber. Two QF40 tubes were welded to holes in the roof of the box to allow access to the vent through the gas manifold, where the gas vent system was attached [10]. Holes in the sides of the short walls provided a pathway for outside air to refresh the ch amber, with the vent above the chamber providing the drawing action [ 10]. This sturdy system has proven very reliable, and is still in use to house the MF1 reactor. The original reactor housi ng design was suitable for APCVD cold-wall, however beginning in the summer of 2004, the reactor was redesigned for a hot -wall configuration [31]. With the cold-wall de sign, a large temperature gradient developed between the susceptor and the quartz tube walls, so in order to increase temperature uniformity, the susceptor was converted into a TaC coated base and a SiC coated ceiling. The susceptor was enclosed in porous gra phite insulation. A cross-secti on of the original hot-wall reaction tube is shown in Figure 2.2 [31]. The water jacket to the quartz walls was left intact to provide cooling to the gaskets and quartz walls to prevent O-ring failure [31]. After testing the new system, Ian Haselbarth converted the 75mm hot-wall system to a 200mm hot-wall system through direct sca ling of the smaller sy stem [33]. The new

PAGE 31

19 Figure 2.2: A sketch of the 75 mm hot-wall reactor showing the ss he ad plate, inlet liner, RF coil, susceptor, insulation, outlet liner, and exhaust port. (1) Denotes the quartz adaptor, (2) the water cooled jacket, and (3) the qu artz end cap. Sketch provided by I. Haselbarth, University of South Florida [31] system eliminated the cooling jacket from the quartz tube, and a dded a stainless steel base plate for system support and exhaust [31] The system that was developed is shown schematically in cross-section in Figure 2.3. Although the head and exhaust plates were still cooled by water, fans were added to the wall of the reactor Figure 2.3: 200mm hot-wall reactor showing the head plate, inlet liner RF coil, insulation, susceptor, outlet liner, and drum (back plate assemb ly). The ss flanges for attaching th e quartz to the plates are at (1), the back plate is (2), the access door is (3), and a viewport is at (4). Sketch by I. Haselbarth, USF [31]. opposite the door and the wall nearest the exha ust plate. The fans were to create a constant air flow over the parts of the reac tor that became hottest during operation. Fans

PAGE 32

20 were never added to the interior door to cons erve the integrity of the cleanroom space the reactor was housed in, so as a final measure to create more ai rflow, the housing wall near the reactor head plate was removed. The new reactor housing was designed with two factors in mind, reactor footprint space and reactor functionality. I. Haselbarth created the designs on Autocad Autodesk 2005 [32] with guidance from S. Saddow, Y. Shishkin, and the author. Because cleanroom space was a limited commodity, there was only 2 m of space along the cleanroom wall for the reactor, which would not allow side access as the current reactor provides. To accommodate this factor, the re actor was to be placed so the back of the reactor intersected with the cleanroom wall. This desi gn had an added benefit of enabling the reactor to be accessed directly through the door on the barrel assembly, unlike the original design where the operator had to awkwardly load the reactor from the side. The reactor would still be constructe d out of anodized aluminum, but to cut down on unnecessary weight and monetary cost, the th ickness of the plates would be reduced to 6.4 mm. The frame was acquired from the same source as the first reac tor, and the plates were also attached in the sa me manner. To assist in th e support of the reaction tube assembly and the manifold, two crossbeams we re added to the frame to support ceiling and the base of the chamber. A few modifications were made in the desi gn to address proble ms encountered in the original reactor. First, the old reactor had only one access point, and often during the upgrade process and maintenance, the side wa lls had to be removed, which was a long and tedious process. The new reactor include s three main access points, the back end and both sides, to allow for easy maintenance and operator access. The door inside the

PAGE 33

21 cleanroom slides upward on a track and is held in the open position with two pins. Both sides of the reactor are hinged at the base w ith sturdy hinges and held closed with 2 screw latches each. To provide airflow, but still keep the reactor enclosed, eight fans were placed into the system. Each fan blows into the reactor to prevent gasses from escaping during an emergency. Three fans are mounted in each door, one in the front panel, and one in the base plate near the gas inlet. The air is still transported to the vent through the gas manifold box, but the manifold QF40 tube supports were increased to four for better airflow and increased support of the manifold box. Finally a ledge was added onto the back end of the reactor to house the pyrometer and optics and provide a working area for the operator. The back door has a slit to al low the hot zone to be observable through the viewport in the barrel assembly. Both the pyrometer and optics are placed onto tracks so Figure 2.4: Three-dimensional re ndering of the 200mm LP CVD reactor (i.e., MF2). The back access door is denoted by (1), the side doors with fan holes at (2), the manifold box with support tubing at (3), the pyrometer ledge at (4), and the frame at (5). Drawing by I. Haselbarth, USF. they can be moved out of the way during load ing and unloading of the reactor hot zone. A three-dimensional Autocad Autodesk 20 05 drawing of the reactor housing is displayed in Figure 2.4.

PAGE 34

22 The reaction tube has very few modifica tions from the original 200 mm reactor, (i.e., MF1); however, the hot zone has under gone many revisions [31]. The flanges for the head plate and barrel assembly have been tapered to more effectively create a vacuum seal with the quartz tube. The barrel asse mbly was modified first by placing latches on the loading door and second by placing a s upport bar under the barrel to alleviate pressure from the vacuum lines. The final addition was to place a 100 mm diffuser on the inlet plate where the gas enters the chamber. The original 50 mm diffuser was developed during R. Myers-Ward dissertation work, and it was constructed to eliminate gas jetting effects and evenly distribute gas through the inlet liner [31]. However, the original design had a straight wall with a gasket to create the seal between the quartz and the metal [31]. When the inlet was slid over th e diffuser, it would sometimes not be aligned perfectly, causing chipping and cracking of the quartz. The new diffuser was tapered before the gasket to allow the inlet liner to slide into place and create the seal. The development of the hot zone of the reactor, which consists of the foam and the susceptor, began with the original 75 mm reactor hot-wall upgrad e [31]. When the susceptor was scaled up to 200 mm, there we re a few problems that were addressed and corrected through experimentation. The original graphite foam insulation was cylindrical and fit within the outlet liner [31]. While this configuration worked for the 75 mm reactor, the 200 mm reactors foam caused a fr equency mismatch in the RF generator due to the increased load [31]. The solution was to remove part of the foam to reduce the mass and reduce the total load to enable the RF generator to matc h the frequencies and provide adequate power to the system [31]. The top and bottom and part of the sides were sliced off with a knife, creating a flat hexagonal stru cture, however currently the

PAGE 35

23 foam has only a small portion removed from the top and bottom due to changes in the RF generator system which will be cove red later in section 2.3 [31]. The second modification was the sloping of the ceiling of the susceptor. The original susceptor design was derived from the 75mm reactor and had a flat ceiling and base, but the 200 mm susceptor scaled up from this design showed heavy reactant depletion toward the back of the hot zone [31]. A slope was added to the ceiling as depicted in Figure 2.5 [31]. The slope allowed a tapering in the boundary layer over the susceptor by creating a larger gas volume in th e front of the zone, a nd this effect allows for slower gas velocity at the front of th e susceptor as compared to the back [31]. Figure 2.5: A cross sectional view of the current reactor hot zone design including the RF coil, quartz tube, foam, graphite adaptors, susceptor, SiC polyplate, and sample [31]. Simulations on the reactor growth condi tions were presented at the ECSCRM conference in Newcastle, England September 2006 [33]. The simulations combined fluid dynamics, temperature profiling, and chemi cal reactant kinetics to derive threedimensional images of the reaction proce sses [33]. The threedimensional reactor simulations, based off the current 200mm design as depicted in Figure 2.5, show the best placement of the sample in the hot zone for the best depositio n rate. Temperature and gas velocity simulations for the hot zone are provided in Figure 2.6.

PAGE 36

24 (a) (b) Figure 2.6: 3-Dimensional gas velocity (a) and temperature profile (b) of the 200mm hot-wall SiC CVD reactor. The black lines show the current placement of the substrates [33]. Simulations courtesy of Y. Makarov of Semitech. The modifications described pr eviously in the present system (MF1) were used to create the 100mm susceptor and foam. Wh ile the size of the reaction tube did not change, the 50 mm design was scaled up two times to support the 100 mm wafer size. Because the tube is 200 mm in diameter, th is increase in size did not require a full restructuring of the entire system. Only the foam, susceptor, sample polyplates, graphite adaptors, inlet liner, and diffuser were scaled to accommodate the new substrate size. Detailed drawings of the reactor housing a nd hot-wall system are included in Appendix A.

PAGE 37

25 2.2 Gas manifold system The nature of SiC epitaxial growth in a CVD reactor requires the use of hazardous gasses, therefore a system needs to be de signed to both handle the gasses and provide safety for the operator and the facility the reac tor is housed in. Each gas needs to have a system for transport, valves to control the start and stop of flow, and for the hazardous gasses, a system to purge the lines safely. Al so as mentioned in the previous chapter, the system needs the ability to control the flow of gases accurately for single crystalline epitaxial growth. The gas system design was se parated into three part s. The first begins at the gas bottle and describes the necessary eq uipment to transport the gas to the reactor area; the second is the transfer panel to f acilitate the transport of the gasses to two separate reactors, and the final system is the gas manifold itself. A schematic of the manifold system is shown in Figure 2.7. The SiC LPCVD reactor is designed for f our primary gasses and four secondary gasses. The primary gasses are hydrogen, H2, argon, Ar, propane, C3H8, and silane, 10% SiH4 in H2, while the secondary gasses are nitrogen, 3% N2 in H2, hydrogen chloride, 10% HCl in Ar, and methyl chloride, CH3Cl, and an auxiliary setup for future expansion. Most of these gasses are hazardous. H2 and the N2 balanced in H2 are explosive in an oxygen environment with the provision of a spark or heat source, SiH4 is pyroforic even in small quantities, and HCl and CH3Cl are corrosives when exposed to water. The system used by industry to transport pressu rized gasses is welded 316 stainless steel piping connected by VCR fittings. The VCR system chosen for the reactor was constructed by Swagelok and consists of two glands, each with a knife edge, a male and female nut, and a stainless steel gasket [34]. As the glands are pr essed together through

PAGE 38

26 the action of the two nuts, they cut into the ga sket and create a tight, metal to metal seal [34]. The entire transport of the gas system was designed with VCR fittings to prevent leaks that could cause the co mplications mentioned above. Figure 2.7: A sketch of the gas handling manifold displaying the gasse s used in the reactor, the primary valves, and the gas transportation routes. Sketch provided by I. Haselbarth, University of South Florida. The gas delivery begins at the gas bottle. For safety, each bottle is firmly attached to a wall, or inside a vented cabinet for the hazardous gasses. The bottles can have a pressure output of up to 2500 psi, so a regulator is necessary to step the pressure down to a manageable level and maintain the pressure to levels ar ound 20 to 40 psi. After the regulator, all gasses with th e exception of Ar are designed to have a normally closed, pneumatic valve that would only open during normal, safe operations. For the three

PAGE 39

27 systems, SiH4, HCl, and CH3Cl, a purge system is necessary to prevent corrosion from the chlorides, or blockage created from SiOx particulates from the reaction of silane and oxygen. An example of a purge panel is shown in Figure 2.8 which uses a Venturi tube to create a vacuum to expel the gasses in the transport lines. The valves can be manipulated to open the low or high pressure sides of the regulator to the vent line, and Ar can be admitted into the high pressure side of the system to purge the line. Various check valves allow the gas to only flow in predetermined directions. After the purge Figure 2.8: A sketch of a purge panel showing a purge panel with valve positions noted. Sketch from Maetheson Tri-Gas website [35]. panel, a filter is added to help prevent line blockage. Silane has an oxygen filter, while the chlorides have a moisture filter.

PAGE 40

28 The gasses are transported to an interm ediate processing ar ea between the two reactors. This area contains two vented va lve boxes and a hydrogen purification system. Because the purity of the carrier gas, H2, is crucial for preventi ng crystalline and point defects, a further processing of the gas is needed to remove impurities. The purifier contains a palladium cell which is heated to 400C. At this temperature, palladium allows the small H2 molecules through, but it blocks larger impurities [ 10]. Because H2 is the main carrier gas in the reactor, and a typical A sized cylinder only lasts a few days of normal processing, each reactor wa s designed with its own separate H2 source to accommodate the increased load. The other gasses are routed to the vented valve boxes where they encounter another regulator to re duce the gas pressure to 15 psi. After the regulator in the valve box, there is a part iculate filter and a manual needle valve to control gas flow into the manifold. The original valve box contained C3H8, SiH4, N2, and an auxiliary connection [10]. The research le ad by R. Myers-Ward and Meralys Reyes for their dissertations added the gasses HCl and CH3Cl, but these gasses were directly routed to the first reactor due to space considerations [31]. Th e addition of the new reactor provided the space necessary to add another panel box and in clude these two gasses with the same connections as constructed in the original valve box. The new design moves the needle valves after a T junction placed beyond the filter to allow flow control between both reactors. The needle valves allow isolation for the reactors from the gas transportation lines, and they allow gas flow only to only the reactor that is in operation (or both). The gases are transported from this transfer area directly into the manifold above the reactor through a singl e stainless steel ga s line without any fittings or seams.

PAGE 41

29 The final system, the gas manifold, serves three purposes. First, this system provides a purging system for the reaction cham ber. Second it provides a system to flow gasses to either the vent or to the reacti on chamber. Finally, the system provides a method of controlling the amount of gas allowe d into the chamber. The manifold system is based off the work done by both T. Schattner and M. Smith in their thesis work on the first reactor [10, 30]. The two gasses H2 and Ar are split between the purge systems and processing system; Ar is further divided to supply the topside cooling purge and also to keep the pyrometer viewport clear of conde nsate. Each line contains a pneumatic valve that is activated when purge conditi ons are required. The H2 line has a normally closed pneumatic valve and is used primarily to back fill the chamber in preparation for crystal growth or for hydrogen termination of the cr ystal surface during cool down. The main Ar line has a normally open pneumatic valve and is used to purge the chamber of gasses during emergencies and provides additional c ooling after processing. The normally open valve allows Ar to flow into the chamber even during a power failure. The final purge line, the topside Ar cooling line, has a normally closed pneumatic valve followed by a regulator to control flow. This valve is activated to allow Ar to pass through the insulation to provide cooling during processin g. The flow rate of the purge gasses is controlled by a 10slm, H2 calibrated MFC. The rate of flow through the MFC is controlled by the primary control system covered in section 2.5. The second part of the manifold is the ma in process control f unction. The gasses flow from the transfer panel to the MFC. The specific MFCs used in the manifold are Celerity Analog/Digital Unit 1661 w ith flow rates of 50 slm for H2 and Ar, 500 sccm

PAGE 42

30 for SiH4, and 50 sccm for the remaining gasses [ 36]. The MFCs requ ire a +15/-15 VDC differential power supply for operation of the internal electronics and motors for the needle valve. The systems have an integrated PID control system that takes a 0-5 VDC control signal, provided by the primary cont rol system, and a mass flow sensor that detects the amount of gas passing through the un it as compared to a calibrated value of N2. The internal PID system adjusts the valve to allow the gas to reach the requested setpoint. The flow sensor re turns a 0-5 VDC value to the control system so gas flow through the MFC can be monitored. After the MFC is a dual, normally closed pneumatic valve. One valve connects to the vent li ne, and the other conne cts to the reactor assembly. The vent valve is used by the pr imary control system to allow venting at any time during the process, but the main process lines can only be activated after the system is in growth mode. The actual actions of these valves will be covered in the primary control system section. After the last MFC, a normally clos ed pneumatic valve controls the flow of the main process gasses into th e chamber, and as a maintenance provision, a manual hand valve is added to the main line. This valve is used to stop all gas flow into the reactor, and it can also be used to isolat e the reaction chamber from the gas system for leak and vacuum tests. 2.3 RF heating system The next important design consideration for a CVD re actor is the method of heating the reaction zone. The heating syst em chosen for the reactor uses magnetic induction from an RF field to couple with th e graphite susceptor. This is achieved through the generation of a RF field from an AC powe r generator with a specific

PAGE 43

31 frequency which is connected to a tank ci rcuit consisting of a capacitor and a wound spiral inductor. The tank circuit can be pl aced in parallel or series depending on the electronic design of the RF generator. As w ith all RF circuits, the LC tank circuit must be matched to the generator resonant frequenc y to achieve the best Q factor for the best power transfer capability. The original reactor used an Amerith erm ISM/L-80 80 KW RF generator unit with a variable frequency range of 50-150 KHz [37]. The input power for the unit was 3 410/480 VAC power which requires a fuse d knife switch disconnect for safe operation, and the generator and tank circuit both required a cooling water system to prevent overheating of the unit an d melting of the circuit through IR2 heating. This generator was discovered to have too high a frequency range to match to the 200 mm reactor induction load during the testing phase that was conducted during R. Myers dissertation research [31]. A new AC generator was chosen to provide the RF induction field for the system from Mesta, Inc. [38]. This generator provides 50 KW at a frequency range of 9-11 KHz. The power requirements we re the same as the first reactor, so a power disconnect was also provided for the gene rator. This system is air cooled, so only the tank circuit requires cooling water. The lower frequency more easily matched the larger load of the 200 mm hot zone, so this generator was also chosen to supply the power for the new reactor MF2. The Mesta generator uses a LC tank circu it in the parallel c onfiguration. Using Kirchoffs voltage and current laws, the resona nt frequency of the system is given by the following equation:

PAGE 44

32 LC f2 1 (2.1) where f is the resonant freque ncy, L is the inductance, and C is the capacitance of the circuit [39]. The complex im pedance of the circuit can be calculated through the parallel sum of the complex inductance and capacitan ce, and the circuit be comes a first order band pass filter with the equation: 12 L C w Lwj Z (2.2) where Z is the complex impedance, L is the inductance, and C is the capacitance of the circuit [39]. The impedance of the system is greatest at the resonant frequency, therefore the values of L and C must be carefully chosen to allow for the great est transfer of power to the system. The Mesta capacitor allows for selective tuning through a parallel series of capacitors accessed through posts which are conn ected to a copper bus shorting bar. The first six posts have a value of 6 F each, the seventh post is half this value at 3 F, and the eighth post has a value half again of 1.5 F. By connecting the posts to the bar, a variable capacitance of 1.5-40.5 F can be achieved for load impedance matching. The inductor is wound by caref ully bending 3.1mm copper tubing around the 200 mm reaction tube. The length of the coil is dete rmined by the length of the hot zone, and the diameter of the coil is slightly larger th an the reaction tube OD. The inductance of a single layer air coil is gi ven through the equation: l r N r L 10 9 39 02 2 (2.3)

PAGE 45

33 where L is the coil inductance in H, r is the radius of the coil in cm, N is the number of turns, and l is the length of the coil in cm [ 39]. Before the turns can be calculated, the lengths of wire from the RF generator to th e tank circuit must be considered since they add a small amount of inductance to the system. At low frequencies, a straight solid wire has an inductance given by the equation: 75 0 2 ln 2 r z z L (2.4) where L is the inductance in nH, z is the length of the wire in cm, and r is the radius of the wire in cm [39]. The tota l inductance is the sum of th e wire inductances calculated from equations 2.4 and the coil inductance from equation 2.3. The turns on the coil can then be designed using equation 2.1. At th e center frequency the RF generator generates a 10 KHz RF field. Because no coil is perf ectly ideal, the Mesta has onboard electronics that can actively adjust the output frequency to match the tank circuit and create the best power output efficiency. The control system fo r the Mesta is part of the primary control and will be covered in section 2.5. The final part of the heating system is the optical pyrometer. The pyrometer measures 900nm infrared radiation emitted by the susceptor, along with built-in instrumentation, to estimate the temperature of the target. As described in the first section of this chapter, the pyr ometer is mounted on a track in the back of the reactor. It is aimed through a viewport in the reactor housing door, thr ough a viewport in the reactor door on the drum and finally at a hole in the back-side top section of the susceptor. The pyrometer provides the feedback to the prim ary control system and is used in the temperature control PID sub-system. The pyrometer chosen for the reactor was the Exactus optical pyrometer from Engelhard, Inc. because it provided a range of

PAGE 46

34 temperature reading from 400-3000 C with a 4mm focal spot at target dist ance of 85 cm [40]. 2.4 Vacuum system The LPCVD reactor needs an efficient pum ping system to maintain low pressure operation during crystal growth processing when a large volume of gas is flowing. M. Smiths main thrust of his Thesis was th e conversion of the 75 mm APCVD reactor to a 75 mm LPCVD reactor [ 30]. His design remains in use after the 200mm expansion and still provides adequate vacuum capabilities for 100Torr operati on. Because of this, the author used a similar system with few modi fications when designing the reactor as to keep flow, conductance, and pump throughput similar to the orig inal design [30]. Because of space considerations, and to make the conductance path as short as possible, the entire pumping system had to f it under the reactor housing. The final design of the vacuum system is given in the sketch in Figure 2.9. The new system uses the Edwards DP-40 dry rotary pump which prov ides 20CFM pumping cap acity at 100 Torr operation [41]. The piping on the main line uses QF40 quick connect 316 ss piping with 40mm radius to provide the same conductance as designed for the fi rst reactor, and the bypass line uses 316 ss 13 mm stainless VCR line [30]. The bypass line has smaller conductance to provide for a soft-start of th e pumping so movement of the samples and hot zone foam is eliminated in the main chamber during pumpdown. The poppet valve seals completely the main line from the chamber, and is opened only after the system is pumped down to below 200 Torr. At that point the Lesker Type 153 downstream throttle valve is allowed to control the conductance of the ma in line to provide pressure

PAGE 47

35 Figure 2.9: Reactor vacuum design schematic displa ying placement of piping, filters, valves, gauges and pump. control [42]. The throttle valve control is provided th rough a closed loop system consisting of the Lesker Type 655 pressu re controller and a diaphragm manometer (model KJL902020). The pressure contro ller has onboard PID control, but is connected to the reac tor primary control system thr ough a 25 pin digita l signal control cable. Many other considerations were used in the design of the vacuum system for the new reactor. With the use of the chloride gasses comes the chance for corrosion, so

PAGE 48

36 proper measures must be taken to contain this problem. Whenever po ssible, straight solid QF40 316 ss lines were used for the construction because of their thick walls and flexible lines were used sparingly because their thin walls have developed leaks in the past on MF1. An extra chemical filter was installed upstream to help filter the chlorides out of the line and prevent corrosion. Finally, to enable atmospheric pressure processing, the vent line was installed after the first two filters. This line has a normally open valve to allow venting of the chamber in case of a power outage. The final design consideration for the vacuum system involved measuring the pressure at crucial points along the path of c onductance. As previously mentioned, there is a diaphragm manometer measuring pressure at the throttle valve to provide closed loop control. The diaphragm manometer reads pr essures from 1-1500 Torr. A T.C. gauge is placed at the pump inlet for testing the pumpi ng status, and it can measure from 1 mTorr to atmospheric pressure (i.e., 760 Torr). At th e head plate of the reactor, two gauges are placed to read pressure in th e reaction tube. These gauges are important to check for the leak-back rate within the tube before proce ssing and to indicate th e pressure difference between the reaction tube and vacuum lines. A diaphragm manometer is coupled with a Pirani gauge to read pressure s from 1 mTorr to 1500 Torr. 2.5 Primary control system The original reactor developed by T. Schattner was designed with a semiautomatic control system composed of a PL C logic controller for control of safety systems and valves and a computer with anal og I/O cards for cont rol of gas flow and temperature regulation [10]. Th e decision behind this sepa ration of control systems was

PAGE 49

37 due to technological capabiliti es at the time the system was created (late 1990s). The computer operating system at the time was Microsoft Windows 98, a system that was highly unstable with a registry that often b ecame corrupted over time, so to ensure safe operation the reliable PLC system was chos en for all systems concerning safety and system regulation [10]. The PC was incl uded to provide user access and flexibility during processing because the PLC is highly in flexible towards complex controls [10]. This dual brain control system, although it provided stability, made system process automation nearly impossible, so each experi ment had to be performed manually. With advances in both PC computer technology and operating systems, the entire system could be designed under the control of one brain, thus enabling the prospect of total automation of the system. The new reactor system was designed for total control under one PC system. The PC has three PCI cards from National Inst ruments to enable control of the gas, temperature, and pressure systems [43]. The NI6221, which has 16 analog inputs, and the NI6703, which has 16 analog outputs, were used to follow the original reactors control design for the gas MFCs and the RF generator [10]. These cards in the new design provide the means to measure and cont rol the pressure system. The NI6509 96 digital I/O 5VDC TTL/CMOS compatible PCI card replaces the PLC to provide a means of activating pneumatics and safety monitori ng. The computer uses Microsoft Windows XP Professional, which is a much more stable system than Windows98 [44], and Labview 7.1 [43] provides the flexibility necessary to control the entire system process.

PAGE 50

38 A problem with the original reactor was th e single wires needed for the activation panel. Each valve or safety process in the original system required a physical switch connected to the PLC for activation and displa y of status, which inevitably created a large and confusing wiring scheme [10] Because of this, the old system was hard to maintain and required a lot of wire tracing to locate problems. The new system was designed to bundle wires into groups to create as few pr oblems as possible. The amount of inputs and outputs going to different de stinations required a system of organization, and two boxes were created to enable the organization of these signa ls. The first box, which was located in the wall of the cleanroom a bove the MF2 computer, houses the physical vacuum gauge monitors, pressure contro l system, a linear +15/-15/+5 VCD power supply, and organizational breakout boards. This box also includes a physical emergency switch and system power switch. The brea kout boards enable th e 32 analog I/O and 96 digital I/O routing to the appr opriate areas on the reactor. In this box the analog I/O and power are grouped for routing to each individua l MFC, the RF generator, vacuum gauges, and the pressure control system. For ease in transport and to match the connections on the attached systems, 9-pin and 15-pin cable assemblies were used for these groupings. The digital wires were grouped into groups of 25-pin wires and 15 pin wiring systems. A 25-pin wire system was used to connect to th e pressure control system inside the box to the digital control to allow computer cont rol. The wiring scheme and design of the breakout box is covered in Appendix B. The digital signal wires and RF control wires were routed to a secondary system which contains the electronics necessary for proper system operation. The secondary box contains two manufactured PC boards, one cont aining DC solid state relays and the other

PAGE 51

39 containing DC solid state relays, AC to DC re lays, and electronics. The board design and relay box schematics are displayed in Appe ndix C. Even though the PC digital card provided 5 VDC from each of the digital supp lies, the power behind each was severely limited, so the relays allowed the power from the 5 VDC, 12 A linear power supply to be routed upon activation of the digital signal. The relays also take the load off of the switching power supply in the computer. The wiring scheme is displayed in the block diagram of the primary control system in Figure 2.10. Figure 2.10: A block diagram showing the signal routing of the control system to the various aspects of the reactor. Analog output signals are in blue, analog input signals are in red, digital output signals in purple, and digital input signals are in green. Located on the second board ar e the electronics necessary to control the vacuum system and RF generator. The vacuum pump motor re quires a starter motor and safety system to operate safely. This system, placed within reach of the vacuum pump to meet OSHA requirements, has a starter solenoid which can only be activated if all the safety conditions are met. The safety conditions ar e provided by various sensors wired in series along the solenoids electrical pa th. There is an overload se nsor, a low current sensor, a

PAGE 52

40 thermal snap switch for detecting pump ove rheating, a water flow sensor, and the emergency interlock relay. The system of re lays and sensors are located near or on the pump with the exception of three relays. The relay box contains th ese three components, two of which allow the computer to start a nd stop the pump with digital signals, and one that relays a signal to the com puter indicating the pump is in operation. The schematic of the pump starter system is located in Appendix D. The final electronics on the board involve the RF generator. The RF generator onboard control system use both relays and anal og signals for remote control. Two 4-20 mA output signals can display the voltage, current, or power levels of the reactor and the operator can choose which two signals are disp layed through a computer monitor on the front of the reactor. One 4-20 mA input sign al controls the output of the generator. Because the computer analog output card is designed for voltage, a dual current mirror system was created on the relay PC board to transfer a 0-5 VDC signal into a 4-20 mA signal. Reactor safety status is controlled by two relays, which are attached to two of the digital inputs of the PC digital I/O card, and a digital output is transferred to a coil DC relay for the inverter enable control. Because the inverter enable signal required a true open switch, a solid state relay could not be used because it still provided a conductive path between the junctions, thus the use of the coil relay. The digital output signals tr ansferred through the relays, safety relay digital input signals, and fan 120 VAC are transferred thro ugh 25-pin and 9-pin cables that attach to the pneumatic plate located inside the manifo ld box. The wires are attached to breakout boards located on this plate where they are ro uted to either the pneumatics, the safety systems located on the reactor systems, or the fans on the reactor housing. The

PAGE 53

41 pneumatics are activated through 5 VDC solenoi ds which require 0.94 W to activate and they control the various valves for gas and pressure control desc ribed earlier in the previous sections. The safety system, c onsisting of various normally closed relays, includes two gas detection units, one mode l 2001-01 indoor combustible gas sensor, 1000 ppm, from Sierra Monitor, Co. [45], and one Spectrum On-Line HCl monitor, 5 ppm, from Enmet, Co [46]. The normally clos ed relays are connected in series so either sensor activation will disrupt the circuit a nd cause a fault. The water sensor FPR126, a 1.5 to 12 Gal/min adjustable paddle style ac tive water sensor, from Omega Inc. [47], is used to ensure that cooling water is runni ng through the system before RF activation. The final safety system routed from the cont rol panel is the door sy stem, consisting of two rocker switches, one on each side door, a nd one magnetic switch for the main sliding door. Once again the system is connected in series so one breach of a door during operation will cause a break in the circuit and activate the control system emergency mode. The computer system has the Labview7.2 platform loaded to provide the means to control the entire reactor process. Labview provides a real-time software environment and is considered a powerful control program. The original reactor system designed by T. Schattner used Labview6.0 to control the MFC s and RF generator [10]. During the authors senior design project for the BSEE degree, a new control algorithm in Labview was developed to add the chlorinated gasses used for R. MyersWard dissertation work [31]. The Labview VIs (virtual instruments) developed in this project were expanded in the fall of 2005 to meet the requirements of the new reactor, namely by adding the digital controls and in cluding all the processe s controlled by the

PAGE 54

42 PLC in the first reactor. The program was al so expanded to have total control of the RF generator and pumping system, which the first reactor control has lim ited or no input, so as to enable a future total auto mation of the growth process. Figure 2.11 shows the final user interface as displayed from the computer system. Dual monitors were required to display all of the reactor functi ons for the user to safely vi ew them without difficulty. Appendix E contains the entire reactor program developed for the reactor control system. 2.6 Summary The design of the new CVD reactor system, fondly named MF2, was the accumulation of knowledge and research done fo r the original MF1 SiC reactor built at USF. This new reactor combines the best qu alities from various sub-systems in the first reactor and was designed to ove rcome problems that the MF1 reactor had encountered in its many years of operation. The new system provides a single computer control to vary temperature, pressure, and gas flow and inco rporates new hot zone designs to improve single crystal SiC epitaxial growth, and e xpand the size of samples from 50 mm to 100 mm. With the system under th e control of a primary contro l unit, eventu al automation can be achieved which is currently underway by I. Haslebarth of our group who is designing a fully-automatic control system as part of his capstone design course (fall 2006, implementation winter 2 006-2007). Now that the system has been constructed, each sub-system must be thoroughly tested before SiC growth and materials processing can commence, which will be discussed in the next chapter.

PAGE 55

43 (a) (b) Figure 2.11: A picture of the operational panel for the Labview reactor contro l program set in default mode. The system requires a dual monitor video system to display the entire front panel. The left monitor displays the panel in (a), and the right monito r displays the panel displayed in (b). Each process of the reactor is separated in to blocks of differ ent color to help the oper ator during processing.

PAGE 56

44 Chapter 3: System Assembly and Preliminary Testing 3.1 System assembly overview As described in the previous chapter, a SiC reactor is the collection of numerous components, each having to be either desi gned and custom built, or located through numerous vendors of specialty scientific and industrial pr oducts. The summer of 2005 was spent locating, planning, and ordering th e various parts necessary for the reactor construction. By the fall of 2005, the Pentium Celeron computer had arrived along with the majority of the control components. The co mputer system was assembled in the student laboratory to provide access to the internet, which was not available at the time in the cleanroom, for purposes of accessing the help and files provided by National Instruments [43]. The National Instruments Labview cont rol program was created at this time and tested through detection of the voltage leve ls of the input and output channels. Fall brought about the machining of the anodized Al panels and stainless steel custom parts required for the systems housing and hot wall assembly. As the year closed out, the frame, housing and stainless pa rts were installed in the cl eanroom chase way by S. E. Saddow, I Haselbarth, and the author. A few modifications had to be made on th e reactor housing due to slight problems with the design and machine work. The side do ors had to be mounted with sliding inserts between the door plate and the reactor frame to allow the hinges to swing freely. The

PAGE 57

45 manifold housing was constructed to be att ached with screws, but to provide easier maintenance access, piano hinges were added to allow the doors to swing upward. With the construction of the housing and reactor frame completed, the vacuum system and gas manifold system could be adde d to the existing frame easily. The author completed the addition of the vacuum system while R. Myers-Ward and I. Haselbarth worked on the construction of the gas manifold and gas lines. Y. Shishkin and S. Harvey were called upon to lend assistance to the va cuum and gas manifold systems because of the complexity and amount of work to be co mpleted. Soon after th e completion of these tasks, the new lines were leak checked by R. Everly, a cleanroom specialist employed at the University of South Florida NNRC facility He uses a mass spectrometer specially tuned to detect He gas flow rates inside a vacuumed line. The VCR fittings have bleed holes on all fittings which He gas is passed over. If a leak exists, the He will be sucked into the vacuumed lines and can be detected at leak rates down to ~10-9 Torr. All welds, gaskets, and QF type fittings on the reactor were checked in this manner, and all leaks detected were repaired as they were detected. December, 2005, through March of 2006 marked the assembly of the control system wiring breakout and relay boxes. Bo th of these boxes were designed to be welded, but due to the increased load on the CoE machine shop, the boxes were assembled by the author and S. E. Saddow. Al requires special welding environments, and after a failed attempt at welding the box with an arc-welder, the author decided to use nuts and bolts to rivet the side s of the box together. When the wiring and soldering of the two boxes was finished, they were mounted to the cleanroom wall and the cables were attached to the various pressure, gas flow, a nd RF devices necessary to run the reactor.

PAGE 58

46 The computer system was attached to the breakout box, and each wire was tested with a voltmeter to ensure proper connection. The computer La bview program was initiated and briefly tested to ensure continuous c ontrol system operation (long run times of days) without major problems. By the end of March 2006, the reactor was ready to begin functional testing, starting with the safety system. 3.2 Interlock safety tests As briefly described in the previous chap ter, the primary control system consists of various sensors and devices used for the detection of da ngerous conditions within the reactor environment. The Labview program is written to work in three separate running modes: safe, process, and emergency. The interlocks are designed to move the program from safe, or process, mode into emergency mode. The interlocks are door opening, water flow, vacuum pressure, gas leak RF trip, electrical power interruption, and a user-initiated emergency button nicknamed the OS button. The safe mode is the mode which the system is placed in when pr ocessing is not happening. During safe mode interaction, the gasses and RF generator ca nnot be activated, and only the vacuum pump can be activated for preparation of the reacti on chamber. In this mode, only the power, gas, and OS emergency interlocks are active. Processing mode enables all activities of the reactor, therefore all interloc ks have the ability to shut down the process. When the system is placed in emergency mode, all valves are closed, with the exception of the two normally-open vent and the Ar purge valves and all gasse s and the RF generator are switched off. All gases and RF power are reset to a zero setpoint level to avoid reactivation of the system to a full processing level without reinitializing the entire

PAGE 59

47 system. As an additive measure, the cooling fans are activated to assist in gas control, but they of course will not work in a power fault situation. The Ar gas will purge the reaction tube and force the gasse s out of the system through the vent even in the case of a power outage. The emergency situation must be deactivated, and all the interlocks reset before the system returns into safe operation mode. However, if the unsafe condition still exists (i.e., a particular interlock has not b een cleared), the system will be placed in emergency again when the interlocks activate again. The major initiator of the emergency condi tion is the OS button, a virtual button located on the computer interface, and a physical button located on the breakout box. Activation of this button overri des all reactor activities, no matter what mode it is in. Both of these buttons were tested in safe and processing mode to ensure this system worked correctly before any testing of the syst em continued. The next major part of the emergency system is provided by a Tripp Lite Internet Office [48], 700 VA, 425 W UPS battery backup system which will provide more than fourteen minutes of power to the PC system if a power fault occurs. This system powers not only the PC, but the power to the solenoids and MFCs through the linear power supply so the computer can still control them in the event of the power loss. This system was tested to verify at least five minutes of power was available wh en not connected to the main power. The next tests involved the doors. All th ree doors when opened either break the spring switch connection for the side doors, or the magnetic connection for the loading door, and the interlock immediately activates. The water system was tested in a similar manner by activating and deactivating the water supply. The pressure system was tested by turning on the pump, although this system also needs the water turned on as well to

PAGE 60

48 activate the starter circuit (see Appendix D). The RF system faults are detected through a system of two relays from the main system boa rd inside the RF generator. A series of logical status commands indica te if the system is functioning correctly; however the RF system cannot be safely forced into these c onditions to test the sy stem. Instead, testing involved using the software to test the conditions for activation, and the author verified that the relays worked with a voltmeter. The final two system tests were not as straight forward to achieve because activation of these systems could mean expos ing the system and tester to dangerous conditions. The power system relay is conne cted to power that supplies one of the cleanroom wall outlets, so instead of turning o ff the power to the outlet, this system was tested by disconnecting the relay from the pow er. The PC UPS system was unplugged to verify at least five minutes of running time as well. The gas detection systems were not truly testable because releasing the gasses into the system was not viable. Instead, these systems have a test mode, so the reaction to ac tivation of the sensor to gas could at least be tested. Now that all the systems were tested and working, the system program had to be modified slightly to avoid immediate activation and deactivation of the interlocks. To prevent the system from shutting down due to a slight loss of water supply, a five second timer was added so the system could have tim e to correct itself. Power was limited to thirty seconds to avoid small power brownouts. The RF syst em, due to switching in the relay system, would cause a fault when the system was powered on, so a five second delay was added to this interlock. The re maining interlocks were left non-delayed and will throw the system into emergency if activated. With the interlocks all tested and

PAGE 61

49 operating correctly, the reactor could now begi n to be tested in earnest, beginning with the vacuum system. 3.3 Vacuum testing The vacuum system tests not only evalua te if the system can be pumped down to operating pressures, but if the system can maintain a low vacuum level for an extended period of time. This test no t only evaluates the integrity of the vacuum lines, but it also tests the various gaskets and seals used to m ount the quartz tube to the steel faceplates in the reactor housing. The controls and pum p interlocks were tested during the interlock tests. Because these tests were successf ul, the pumping sequen ce was initiated by activating the bypass line. This line is developed with a smaller conductance to prevent a large outflow of gasses which can cause the hot-wall insulation and the samples inside the susceptor to move toward the drum outlet. When the pressure is reduced to around the 200 Torr range, the poppet valve was opened, and the main vacuum line could now be used to pump down the reactor at an increase d pumping speed. With both of these valves tested and working correctly, the pressure on the gauges was monitored until it read below 2 Torr in the chamber. The Pirani gauge was activated so as to read pressures in the mTorr range. The main pumping test involves measurement of the leak-back rate, or the rate at which pressure leaks back into the system at low pressure. Oxygen is a problem for system operation because of the hazardous gass es involved in processing, so it must be eliminated from the system and not allowed to reenter for the duration of the growth process. For the reactor system, a leak-back ra te of 10 mTorr/min at a system pressure of

PAGE 62

50 500 mTorr is required for safe operation [49]. To test this, all valves are closed in the vacuum system, and the vacuum is monitored to evaluate the leak ra te per minute. The initial tests at the end of April 2006 had the le ak-back rate at greater than 30 mTorr/min. After inspection of the chamber, it was discover ed that the machining of the reactor drum and faceplate were suspect, so they had to be machined to make th em smoother. After careful reconstruction of the reactor chamber, tests were once again performed at the end of July 2006. The tests were preformed on the chamber alone, and the chamber with the insulating foam and susceptor. The results of the test are in Figure 3.1. The difference in the leak-back between the empty chamber and the chamber with the hot-zone is due to the out-gassing of gasses trapped in the porous fo am. This effect is seen with new foam, but lessens over time as the foam is heated as seen in the graphs displayed in Figure 3.1. The pressure control system also require d testing and adjust ment to correctly control the low pressures needed for crystal gr owth. First, computer digital and analog control of the pressure controller was eval uated, and then the system was adjusted through the flow of Ar gas. The system needed to adjust the internal PID system to the working load of the system, using an on-boa rd set of parameters similar to the PID optimization parameters discussed in Chapter 1, so the maximum current process flow of 30 slm Ar was activated to allow the controller to adjust the system pressure When the controller was finished detecting the system pressure range, the pressure could be maintained at any value under 200 Torr with an accuracy of 2 Torr.

PAGE 63

51 Figure 3.1: A graph three leak-back rate tests performed on the MF2 LPCVD reactor. Leak-back tests performed at a pressure of 150 mTorr to 500 mTorr. 3.4 Gas manifold tests The gas manifold was the next item to be tested to prepare for reactor operation, and there were two modes within the computer program to test the proper activation of the gasses. The system was designed to only allow purging gasse s and the venting of gasses during safe mode operation, and the fl ow of all gasses to the chamber would only be allowed during processing mode. The initial testing began in the safe mode. The purge gasses we re tested first to make sure that the chamber could be purged e ffectively in the case of an emergency. Ar is the main safety purge gas, and H2 is used as a purge for the backfilling of the chamber after low-pressure pumping. These two gasse s were successful in the test, so each MFC could now be tested through the vent valve lo cated in the manifold. To begin the gas

PAGE 64

52 flow through a line, the safety valve needs to be activated to open the valve at the gas bottle. Then, each of the MFCs were sent a control signal of 0-5 VDC corresponding to a flow setpoint for the particular gas tested and the vent valve was opened to allow flow through the MFC to the vent. The MFCs were able to control the flow of each gas with only a 10 mVDC swing in return voltage. The final testing involved pl acing the system into pro cess mode. Within this mode, as mentioned before, all interlocks were active to ensure system safety. To begin the flow of gasses into the chamber, a pro cess enable switch must be activated. This switch enables both the process valves, the ma in valve for the chamber, and enables the activation of the RF generator. The process valves were tested, a nd gas flow was tested into the chamber, to ensure that the program and valves were working correctly. With these tests done, the system was ready for the final tests and the first heating of the hot zone/graphite foam. 3.5 RF and control system tests The final testing of the system would invo lve the first heating of the system to fully test the control system in process mode. This test would ensure the control system could effectively control all aspects of the react or system and provide the final test before the system was used to grow an epitaxial film. For this test, the reactor was pumped down, backfilled with Ar, and placed into process mode. To avoid complications during the test, the heating of the system was only conducted under Ar flow The RF generator was activated and the power to the generato r was slowly increased, and the power output of the RF generator monitored with each in crease. Roughly, the current mirror provides a

PAGE 65

53 1 KW power output increase for every 0.1 VDC in crease in the control input. The optical mirrors, viewing telescope, and pyrometer sy stem were adjusted to provide optimum temperature readings from the pyrometer. The optimum signal is achieved through a careful adjustment of the Gi mble style mount, aiming the pyrometer at the hole machined in the ceiling plate of the susceptor, until a maximum signal is located. The viewing telescope and mirrors are adjusted until the fu ll growth area inside the susceptor can be viewed by the operator. The output of the pyrometer was graphed and the value recorded through the Labview software, and evaluation for the PID values was derived from this data and the Zieglar-Nichols's optimizatio n method. The graph of the data and parameters for the method are given in Figure 3.2, and the values for the PID calculated are P = 1.32, I = 11.9, and D = 2.98. The values of the PID produce an acceptable control of the temperature with only a 1C swing. 3.6 Summary Testing the LPCVD reactor system, MF2, involves a systematic evaluation of all components of the system. The safety of the system, being a foremost consideration for the operator and the environment the system is housed in, was performed first. After the successful testing of this system, the reactor needed to be tested for leak-back rate and pressure control through a series of tests w ith the vacuum system. Gas viability within the purge and process system were tested ne xt, followed by a full system test using the RF heating system. With each of these syst ems tested, and the control system proving its ability to manage these systems, the reactor was ready to begin the growth of 3C-SiC epitaxial films on Si substrates. The next chapter will explain the epitaxial growth process

PAGE 66

54 for 3C-SiC, and how this process, which was developed in MF1, was used to develop a new growth process to grow a 100 mm wafer in MF2. Figure 3.2: A graph of the data gathered during the PID tuning for the RF generator heating system. The graph was used to evaluate values for the PID control using the Zieglar-Nichols's optimization method.

PAGE 67

55 Chapter 4: Reactor Growth Experiments 4.1 Silicon melt test The first step in growing 3C-SiC on a Si substrate involves calibrating the system for the error in temperature read through the pyrometer system. 3C-SiC has the advantage of using large-area, low-cost Si as a substrate for growth making it less expensive than 4H and 6H-SiC which requ ire higher growth temperatures and more expensive substrates. Although the pyromet er system is an accurate method of measuring the temperature of the substrate because it has been constructed to match the CVD system optical path length between the pyrometer and the susceptor, small variations in the reactor environment can cause a discrepancy in the actual temperature of the system. The method to measure this error and calibrate the growth process accordingly is to perform the so-called Si melt test. Si melts at 1410C under standard pressure and temperature (STP) conditions, so this fact is exploited to calibrate the temper ature monitoring system. A small piece of Si is placed on a 4H-SiC sample and placed in the reactor hot z one in the same location as a substrate is placed for growth. The system is heated under H2 gas flow, again mimicking the growth flow rate of the carrier, until ~1300 C, when the sample is monitored visually by the operator to note when melting occurs. The temperature is then slowly increased until the Si sample melts, and then imme diately the system is cooled under H2/Ar flow. The swift cooling allows surface tension in the melted Si to form a small peak, which is

PAGE 68

56 then used to find the temperature difference with the actual melting point of Si and the pyrometer temperature reading during the next heat-up cycle. Several tests are performed at both atmospheric and low pressure to es tablish the error in temperature readings. These tests are preformed routinely when a hot-zone is first placed within the reaction chamber. Usually, there are multip le pieces of the Si/4H-SiC samples placed onto the testing susceptor to test the temperat ure difference across the entire growth zone for both 50 mm and 100 mm diameter wafers. The pieces are placed in a cross or an xpattern to determine the temperature differences from front to back and side to side in the hot zone. The initial tests of the growth zone in the reacto r showed a 40C difference in temperature from front to back, and a 1C diffe rence from side to side within the growth zone. Uniformity in the temperature across the growth zone is important to proper epitaxial growth and to prevent the Si s ubstrate (when growing 3C-SiC on Si) from melting to the SiC polyplate that is placed into the hot zone to facilitate sample insertion/extraction. By repositioning the RF induction coil to be more compressed in the back end of the hot zone than the front, th is difference in temperature measured was reduced to 7C from front to back, and 7C si de to side. With the temperature accuracy of the hot zone measured the reactor was now ready to begin 3C-SiC growth. 4.2 3C-SiC growth experiments The growth of 3C-SiC in a LPCVD r eactor follows guidelines previously established through the efforts of research done at the University of South Florida on the original reactor, MF1. A baseline process w ith two distinct processes was developed to successfully grow 3C-SiC on Si substrat es [50, 51]. The process shown in Figure 4.1

PAGE 69

57 shows the time, temperature, and gasses used during the first and second growth steps, named carbonization and 3C-SiC growth, respec tively. The first temperature ramp and plateau make up the carbonization phase wh ich introduces carbon in to the system at atmospheric pressure to initia te the growth of a few monol ayers of 3C-SiC. The second ramp and plateau begin introducing silane slow ly to facilitate the continuation of the crystal growth, at the system is switched to low pressure at the end of the ramp. After growth, the gasses are turned off and H2 gas is purged initially to perform a slight etching of the grown material, and then cool down proceeds normally in Ar flow. Figure 4.1: The baseline growth schedule for 3C-SiC epitaxial processing. Time is represented on the xaxis, temperature on the y-axis, and precursor and transport flow are shown with separate graphs with the y-axis translating to flow rates. Orig inal picture provided by M. Reyes [51]. Many experiments needed to be conducte d to establish the baseline growth process. After the init ial melt tests, small samples of (100) Si were diced into 10 mm x 8 die and used to first establish the pr oper flows of precurs or gasses for a good

PAGE 70

58 carbonization. Carbonization is considered good when there are few to no etch pits after carbonization as viewed under an optical micr oscope. Once the carbonization process was deemed to have been established, the silane ramp was introduced to determine the proper saline flow necessary for 3C-SiC nucl eation. Too much silane produces a white, milky like film, but not enough produces more etch pits as seen in the carbonization ramp. A proper 3C-SiC thin film will have color bands (i.e., Newton s fringes) and their spacing depends on the thickness of the pr oduced layer. When the process was established, a 50 mm wafer was loaded a nd a fifteen minute run produced a good quality film of ~5 m in thickness. The reactor was re-configured for 100 mm growth by extracting the 50 mm hotzone and inlet liner. The new insulation fo am, susceptor, and 100 mm inlet liner were loaded into the reactor, and a new RF inducti on coil constructed to c over the entire length of the newly installed (and l onger) hot-zone. The adjustment s to the reactor required a repeat of the vacuum test to insure prope r seals and the melt test to establish the temperature difference across the entire 100 mm growth zone. Beginning with the process flow established by the 50 mm 3C-S iC growth, the carbonization and silane ramps were adjusted to meet the new size of the sample. Figure 4.2 displays a picture of the 50 mm and 100 mm wafer grown in the co mpleted 200 mm 3C-SiC LPCVD reactor. The 100 mm wafer shows the 3C-SiC layer did not completely cover the entire surface. Unfortunately, the gradient difference experi enced within the hot-zone was over 61 C, and this large gradient required a much lower temperature for growth to ensure the Si did not melt. Simply scaling the design us ed for the 50 mm hot-zone did not produce a

PAGE 71

59 viable 100 mm process, and much more si mulation and redesign will have to be conducted in the future for successful epitax ial growth on the larg er substrate size. (a) (b) Figure 4.2: Photographs of the (a) 50 mm and (b) 100 mm 3C-SiC wafers grown with the MF2 LPCVD reactor. Newtons fringes ar e visible as the green, yellow, and violet bands of color on the wafer surface. 4.3 Summary The growth process of 3C-SiC on a Si substrate was ba sed from the basic process developed at the University of South Florida through the use of the original r eactor, MF1. This process has been developed over th e years through a great number of hours of research by S. E. Saddow, T. Schattner, R. My ers, S. Rao, Y. Shishki n, M. Reyes, and S. Harvey. This process was modified empiri cally to produce the growth of a 3C-SiC crystal film, first on a small 10 mm x 8 mm samples, followed by a 50 mm wafer, and finally a 100 mm wafer. Although the 100 mm gr owth was only partially successful, the reactor has proven it can successf ully grow on the other two smaller sizes. The initial reason to build and cons truct this reactor was to separate the 4H/6H-SiC growth from the 3C-SiC growth. The reactor control system provides a flexibility not possible with the previous reactor, and this flexibility aided in the quick development of a 3C-SiC baseline process.

PAGE 72

60 Chapter 5: Conclusions and Future Work 5.1 Conclusions The semiconductor material, SiC, has seen an increase in interest by researchers and industry alike because of its inherent chemical resistance, physical hardness, and electrical properties. A lthough the polytypes 4H and 6H-S iC lend themselves to highpower and high-frequency electronics because of a larger bandgap, the 3C-SiC polytype has begun to receive interest in the MEMS a nd biological fields of study. The 3C-SiC polytype, unlike its counterparts of 4H and 6H-SiC, can be grown at relatively low temperatures and on large-area Si substrates. This makes the grow th of 3C-SiC less expensive simply because the expense of a SiC substrate is not needed for growth. Some of the current research grants the SiC group at University of South Florida have involved 3C-SiC growth on Si molds for MEMS appli cations and biological applications through research pioneered by Camilla Colletti for her doctorial dissertation [52]. A grant through ONR in 2005 enabled th e construction of a second LPCVD reactor for the SiC group at the University of South Florida. At this time, the author, who had worked on an upgrade of the pres ent LPCVD system, MF1, was given the opportunity to design and build th is new reactor. The author proposed a system with one primary central control that would facilitate the growth variables be tter than the current system which split the system into a two c ontrol system. The design of the housing and control of the new LPCVD reactor system began in the summer of 2005 using the

PAGE 73

61 original reactor as a template. The new de sign made general func tionality improvements in the reactor housing an d used a PC computer as the basi s of the primary control system. This system was built through the design guideli nes developed in the winter of 2005, and construction was completed in the spring of 2006. Testing and initial growth processing was conducted in the summer of 2006, and the system made its first 50 mm diameter whole-wafer growth on a 3C-SiC film on September 11, 2006. The new reactor design demonstrates that it is possible to c ontrol the numerous aspects of the reactor, such as process pressure, gas flow, and temperature, with one central system while maintaining the support system necessary for the safety of the operator and the environment. The mistakes made in the numerous upgrades of the first reactor from 75mm to 200mm were eliminat ed in the new system. Technological advances in the PC microprocessor, operati ng system, and support software enabled the change in system design that was considered unfeasible during the crea tion of the original reactor. The new design also resulted in a reduction of reactor components, a problem that made the original reactor difficult to repair and maintain. The panel of lighted switches, along with the wiring to connect these activation sw itches to the PLC, has been transformed into a virtual device displayed on a computer screen and controlled with the use of a mouse. Incorporati on of the pressure control an d indicator devices into the reactor control system gives a dimension of freed om not available in the original reactor. Finally, the RF generator and MFC control, little changed from the original upgrade done by the author for his senior design project [53], were improved by adding the valve and safety controls in tandem to make a better overall control system.

PAGE 74

62 The new reactor design also facilitated the increase in size of substrate size from 50 mm to 100 mm diamet er wafers. Real estate on a wafe r is an important consideration in device manufacture. More space on a wa fer reduces processing costs by enabling the same amount of processing to gain more devi ces. SiC needs to increase the wafer size to help reduce costs from processing, especial ly since SiC takes much more to produce single crystal materials than the standard Si wafer. The reactor was designed not only to produce more material for the various re search programs the SiC group conducts each year, but to produce this material more effi ciently for these experiments. The new LPCVD reactor has proven that it can grow 3C -SiC effectively and wi ll be a signiffigant asset to the SiC group at USF for years to come. 5.2 Future work The new reactor, having a single primar y control system, has a major advantage over the first reactor constructed at USF. Automation of the en tire 3C-SiC growth process is currently being implemented by I. Haselbarth of the USF SiC group for his senior design project to earn his BSEE degree [54]. The proj ect involves a dding another mode to the Labview program th at was written by the author that allows a recipe to be executed by the PC computer system. This m ode follows the baseline process developed on the reactor during the summer of 2006 a nd consists of a pop-up menu and execution button. Because all systems are under the ultim ate control of the system, this revision of the program is possible and should be completed by the end of 2006. Although most of the reactor modifications were successful, one system does not work as required. The ledge and railing system that enable the pyrometer and optics to

PAGE 75

63 be maneuvered out of the way when the hot-zone is loaded does not function as expected, as predicted by S. Saddow, but unfortunately the advice was ignored by the author. The pyrometer must be adjusted every time it is moved back into position to read the hot-zone temperature because sliding the pyrometer results in vibrations that misalign the pyrometer optics. These misalignments either require another melt te st to be performed to ensure the temperature is known to a suffi cient degree of accuracy, or the system that causes the misalignment needs to be redesigned. The redesign would require th at the pyrometer is secure ly fastened in a stable position as to facilitate temperature acquisiti on repeatability. The door system needs to be redesigned to stabilize the pyrometer. The solution the author would like to see implemented is to remove the current loadi ng door of the reactor and replace it with a sturdy door connected directly to the frame by a large hinge. Opening the larger door vertically, as is done with the current design, would be difficult due to the greater weight, and the possibility of someone slamming the door down by accident would be increased. So the door will be made to swing out, as Dr. S. Saddow originally suggested, and the pyrometer could then be fixed to a nonmovable mount attached to the door, and therefore, the sturdier frame. The only phys ical motion that the pyrometer would sense will be from the Gimble mount which is used for precision aiming of the instrument. This would free up the ledge for use as a loading and unloading dock, and prevent the accidental damage of the pyrometer in this h eavily trafficked zone. This proposition will need careful planning and execution to acqui re the parts strong enough to withstand the constant handling. The pyrometer is an expe nsive and delicate piece of equipment, so all precautions to protect it from damage must be taken.

PAGE 76

64 The unsuccessful growth of 3C-SiC re quires a redesign of the hot-zone, specifically the susceptor, to alleviate the in credible temperature gradient within the 100 mm growth zone. The first step to achieve a correction of this problem will be to characterize the temperature along the entire le ngth and width of the growth zone. With this data, the geometrical design of the zone and the gas flow regimen used during the growth process, a detailed simulation can be performed to characterize the reaction chamber. Modifications can then be made a nd simulated to determine the most efficient design for the susceptor. Semitech shall be used again because they have demonstrated with the first reactor, MF1, the ability to accurately determine the location of the most favorable growth zone, which was verified th rough later experiments. Armed with this knowledge, a new susceptor can be created and tested that will allow the final actualization of 3C-SiC epitaxial growth on a 100 mm Si substrate. The reactor will also be used to provide material for the upcoming biological and MEMS research within the SiC group at the Un iversity of South Florida. SiC is known to be a biocompatible material and has been used in many applicati ons in the bio-medical field, however, much is not understood about the relationship of SiC and biological interactions, so there is much more to learn and explore. MEMS a pplications, especially in relation to biological sensors, may also be investigated, and 3C-SiC will be used as the prospective material. 3C-SiC is inherently harder than Si, and chemically inert, so complications may be created in working with the material, but, because it can be grown using Si as a substrate, the 3C-SiC material can be used to coat already manufactured molds. The SiC group has also had past suc cess in using 3C-SiC coated MEMS devices created in this fashion. The author has curr ently been accepted into the doctorial program

PAGE 77

65 at USF and will be involved in some of the biological and MEMS applications within the SiC group.

PAGE 78

66 References [1] Matus, L. G., et al., Inst. Phys. Conf. Ser., No 137, 1993, p 185-188. [2] http://www.ceradyne.com/info/Silicon_Carbide_info.asp [3] Saddow, S. E., and A. Agarwal, editors, Advances in Silicon Carbide Processing and Applications, Boston, MA: Artech House, Inc, 2004. pp. 2-3, 7-8, 18. [4] Choyke, W. J., and G. Pensl, Physical Properties of SiC, MRS Bulletin March 1997, pp. 25-29. [5] Konstantinov, A. O., et al., Ionizatio n Rates and Critical Fields in 4H SiC, Appl. Phys. Lett., Vol 71, No. 1, 1997, pp. 90-92. [6] Jrrendahl, K., and R. Davis, Materia l Properties and Char acterization of SiC, Semiconductors and Semimetals, SiC Materials and Devices, Vol. 52, Y.S. Park, (ed.), 1998. [7] http://www.cree.com/Products/sic_silicarb.html [8] Nordling, C., and J. sterman, Physics Handbook, 5th ed., Lund, Sweeden: Studentlitteratur, 2002, pp. 28-29. [9] http://www.onr.navy.mil/sci_tech/31/312/ ncsr/materials/sic.asp#Properties [10] T. E. Schattner, M.S.E.E., Th esis, Mississippi State University, 2000. [11] Daulton, T. L., et al., Polytype dist ribution of circumstellar silicon carbide: Microstructural characterization by transmission electron microscopy, Geochimica et Cosmochimica Acta Vol. 67, No. 24, pp. 4743, 2003. [12] Berzelius, J. J., Ann. Phys., Lpz., Vol. 1, 1824, p. 169. [13] Di Pierro, S., et al., Rock-formi ng moissanite (natural -silicon carbide), American Mineralogist Vol. 88, 2003, pp. 1817. [14] Acheson, A. G., Engl. Patent 17911, 1892.

PAGE 79

67 [15] Fissel, A., Artificially layered heteropolytypic structures based on SiC polytypes: molecular beam epitaxy, characterization and properties, Physics Reports, Vol. 379, pp. 149, 2003. [16] Round, H. J., Electrical World Vol. 19, 1907, p. 309. [17] Lely, J. A., Berichte der Deutschen Keramischen Gesellshaft e.V., Vol 32, 1955, p. 229. [18] Kurida, N., et al., Ext. Abstr., 19th Conf. on Solid State Devices and Mater., 1987, p. 227. [19] http://www.uccs.edu/~tchriste/co urses/PHYS549/549lectures/cvd.html [20] Leys, M. R., et al, Mat. Sci. Forum Vols. 264-268, 1998, pp. 103-106. [21] http://en.wikipedia.org/wiki/Chemical_vapor_deposition [22] http://www.fysik.uu.se/cmt/documents/LicHakan/node17.html [23] Paisley, M. J., et. al., Mat. Res. Soc. Symp. Proc. Vol. 572, 199, p. 167. [24] http://chiuserv.ac.nctu.edu. tw/~htchiu/cvd/reactor.htm [25] http://www.timedomaincvd.com/ [26] http://www.ameritherm.com [27] http://www.infraredheaters.com/basic.htm [28] http://www.tcnj.edu/~rgraham/PID-tuning.html [29] http://aabi.tripod.com/ [30] M. T. Smith, M.S.Ch.E., Thesis University of South Florida, 2003. [31] R. L. Myers-Ward, Ph. D.E.E., Dissert ation, University of South Florida, 2006. [32] http://usa.autodesk.com/adsk/ servlet/home?siteID=123112&id=129446 [33] Shishkin, Y., et. al., Analysis of SiC CVD Growth in a Horizontal Hot-wall Reactor by Experiment and 3D Modeling, ECSCRM Newcastle, England, September 2006. [34] http://www.swagelok.com/downloads/webcatalogs/EN/MS-13-150.PDF

PAGE 80

68 [35] https://www.mathesontrigas.com/index.aspx [36] http://www.celerity.net/ [37] http://www.ameritherm.com/ [38] http://www.mesta.com/ [39] http://www.rfcafe.com/referen ces/electrical/inductance.htm [40] http://www.engelhard.com [41] http://www.bocedwards.com [42] http://www.lesker.com/newweb/index.cfm [43] http://www.ni.com/ [44] http://www.microsoft.com/ [45] http://www.sierramonitor.com/ [46] http://www.enmet.com [47] http://www.omega.com/ppt/pptsc.asp?ref=FPR100&Nav=gref06 [48] http://www.tripplite.com/ [49] A. M. Hoff, private conversation. [50] Harvey, S., M.S.E.E., Thesis, University of South Florida, 2006. [51] Reyes, M., et. al., Development of a high-growth rate 3C-SiC on Si CVD process, Mater. Res. Soc. Symp. Proc ., Vol. 911, 2006. [52] Coletti, C., et. al, Surface mor phology and structure of hydrogen etched 3CSiC(001) on Si(001), Mater. Res. Soc. Symp. Proc. Vol. 911, 2006. [53] Frewin, C., B.S.E.E., Project, Un iversity of South Florida, 2005. [54] Haselbarth, I., B.S.E.E., Project, University of South Florida, 2005.

PAGE 81

69 Appendices

PAGE 82

70 Appendix A: Reactor housing and hot-wall chamber drawings The 3C-SiC reactor, MF2, needed a large amount of planning and precise mechanical drawings. The reactor housing and chamber are all custom machined by various machining companies to meet the st ringent requirements a full 200 mm reactor requires. The complete reactor housing and hot-wall drawings (S eptember 2006) that were used in the creation of the reactor custom made com ponents are provided in this Appendix. The original drawings for th e use by these machinists to fashion the components of the reactor were drafted w ith the aid of AutoCAD-autodesk2005 by I. Haselbarth and converted to .jpeg format fo r viewing without the AutoCAD software. The first part of this appendix covers the ex terior housing of the r eactor, the second part covers the internal reactor chamber, and the final part displays the hot-zone components.

PAGE 83

71 Appendix A: (Continued) Reactor housing mechanical drawings: (a) (b) Figure A.1: Mechanical drawings displaying the speci fications of the (a) back wall and (b) base plate for the reactor housing.

PAGE 84

72 Appendix A: (Continued) (a) (b) Figure A.2: Mechanical drawings displaying the sp ecifications of the (a) react or ceiling and (b) the Al frame assembly.

PAGE 85

73 Appendix A: (Continued) (a) (b) Figure A.3: Mechanical drawings displaying the specifi cations of the (a) rear loading door (b) the left door of the reactor housing.

PAGE 86

74 Appendix A: (Continued) (a) (b) Figure A.4: Mechanical drawings displaying the specifications of the (a) right door and the (b) manifold base.

PAGE 87

75 Appendix A: (Continued) (a) (b) Figure A.5: Mechanical drawings displaying the specifications of the (a) manifold top cover and (b) the assembly sketch for the chamber housing.

PAGE 88

76 Appendix A: (Continued) Reaction chamber mechanical drawings: (a) (b) Figure A.6: Mechanical drawings displaying the specifications of the (a) front mounting plate and the (b) front mounting plate support for the reaction chamber.

PAGE 89

77 Appendix A: (Continued) (a) (b) Figure A.7: Mechanical drawings displaying the sp ecifications of the (a) gas diffusion plate and the (b) back barrel door for the reaction chamber.

PAGE 90

78 Appendix A: (Continued) (a) (b) Figure A.8: Mechanical drawings displaying the speci fications of the (a) back door assembly and the (b) drum support mount.

PAGE 91

79 Appendix A: (Continued) (a) (b) Figure A.9: Mechanical drawings displaying the sp ecifications of the reaction chamber for (a) the main drum back view and (b) side view.

PAGE 92

80 Appendix A: (Continued) (a) (b) Figure A.10: Mechanical drawings displaying the specification of the reaction chamber for (a) the main drum bolt threading and (b) the drum-gasket tube stoppers.

PAGE 93

81 Appendix A: (Continued) (a) (b) Figure A.11: Mechanical drawings displaying the specifications of the (a) gasket flange for the quartz tube and the (b) front mounting plate for the reaction chamber.

PAGE 94

82 Appendix A: (Continued) (a) (b) Figure A.12: Mechanical drawings displaying the specifications of the (a) bolt threading detail for the front mounting plate and (b) the front mounting plate support for the reaction chamber.

PAGE 95

83 Appendix A: (Continued) (a) (b) Figure A.13: Mechanical drawings displaying the specifications of the (a) front mounting plate VCR gland welds and the (b) reactor tube assembly.

PAGE 96

84 Appendix A: (Continued) Hot-zone mechanical drawings: (a) (b) Figure A.14: Mechanical drawings displaying the specifications of the (a) 100 mm hot-zone carbon foam insulation and (b) the 100 mm graphite susceptor bottom.

PAGE 97

85 Appendix A: (Continued) (a) (b) Figure A.15: Mechanical drawings displaying the specifications of the (a) 100 mm susceptor top and the (b) 100 mm quartz inlet liner.

PAGE 98

86 Appendix A: (Continued) (a) (b) Figure A.16: Mechanical drawings displaying the specifications of the (a) 100 mm female graphite adaptor for the quartz inlet liner and the (b) 100 mm male adaptor for the carbon foam insulation.

PAGE 99

87 Appendix A: (Continued) Assembled reactor photographs: (a) (b) (c) Figure A.17: Three photographs displaying the assemb led (a) reactor housing, (b ) the manifold, and (c) the reaction chamber loaded with the hot-zone and quartz inlet liner.

PAGE 100

88 Appendix A: (Continued) (a) (b) Figure A.18: Photographs displaying the (a) rear, or loading, drum assembly, and (b) a view of the reactor from inside the cleanroom that disp lays the pyrometer assembly, loading door, and an operational reactor.

PAGE 101

89 Appendix A: (Continued) (a) (b) Figure A.19: Photographs displaying the (a) rear of the reaction zone, and (b) front of the reaction zone which display the reactor in fully operational mode.

PAGE 102

90 Appendix B: Breakout box drawin gs and reactor control wiring The breakout box houses the main control sy stem components of the reactor that are not contained within the PC computer. The box is designed to separate the main DC power supply from the more sensitive contro l components, such as the vacuum gauge indicators. It was designed to separate th e analog and digital lines from the PCI cards within the PC computer and reorganize them into 9, 15, and 25-pin cables for transport across the long distances betw een system components. This section contains the mechanical drawings originally sketched w ith paper and pen by the author, then later transferred to AutoCAD-autodesk2005 drawings by I. Haselbarth. The latter part of the section details the wiring scheme used throughout the c ontrol system, beginning with the connections and assignments of each computer digital/ana log card and extends throughout the control network system.

PAGE 103

91 Appendix B: (Continued) Breakout box mechanical drawings: Figure B.1: Mechanical drawing displaying the specifications of the front panel for the breakout box.

PAGE 104

92 Appendix B: (Continued) Figure B.2: Mechanical drawing displaying the specifications of the rear panel for the breakout box.

PAGE 105

93 Appendix B: (Continued) (a) (b) (c) Figure B.3: Mechanical drawings displaying the sp ecifications for the (a) analog I/O drawer front, (b) the digital I/O drawer front, and (c) the drawer assembly for the digital I/O card.

PAGE 106

94 Appendix B: (Continued) Figure B.4: Mechanical drawing displaying the specifications of the analog I/O drawer assembly for the breakout box.

PAGE 107

95 Appendix B: (Continued) (b) (A) (c) Figure B.5: Mechanical drawings displaying the speci fications for the (a) rear drawer front, (b) the D-pin connector port holes, and (c) the rear placem ent for the D-pin conn ector port holes.

PAGE 108

96 Appendix B: (Continued) (a) (b) Figure B.6: Mechanical drawings displaying the specifi cations for the (a) rear door with fan holes, and (b) the bottom side plates for the breakout box.

PAGE 109

97 Appendix B: (Continued) (a) (b) Figure B.7: Mechanical drawings displaying the specifications for the (a) top side plates for the breakout box, and (b) the assembly drawing of the breakout box without the drawers and door plates mounted.

PAGE 110

98 Appendix B: (Continued) Pictures of assembled box: (a) (b) Figure B.8: Pictures of the assembled breakout box with (a) a view from the cleanroom displaying the mounted drawers, gauges, pressure controller, and PC cables, and (b) a rear view of showing the cooling fans, MFC drawers, and transport cables.

PAGE 111

99 Appendix B: (Continued) Analog and digital card assignments: Wiring allocation for entire reactor system AI/O# Analog I/O P#L# Port number and line number digital I/O Breakout box (front) PCI6221 (analog input 16, analog ou tput 2, and 16 digital I/O) AI0 V H2 MFC return 9-pin D connector #1 pin 2, 0-5 VDC, brd. pin in. 68 (black), 9pin D connector #1 pin 7-8 gnd. 67 (yellow) AI1 V Ar MFC return 9-pin D connector #2 pin 2, 0-5 VDC, brd. pin in. 33 (black), 9pin D connector #2 pin 7-8 gnd. 32 (yellow) AI2 V C3H3 MFC return 9-pin D connector #3 pin 2, 0-5 VDC, brd. pin in. 65 (black), 9pin D connector #1 pin 7-8 gnd. 64 (yellow) AI3 V SiH4 MFC return 9-pin D connector #4 pin 2, 0-5 VDC, brd. pin in. 30 (black), 9-pin D connector #4 pin 7-8 gnd. 29 (yellow) AI4 V N2 MFC return 9-pin D connector #5 pin 2, 0-5 VDC, brd. pin in. 28 (black), 9-pin D connect or #5 pin 7-8 gnd. 27 (yellow) AI5 V HCl MFC return 9-pin D connector #6 pin 2, 0-5 VDC, brd. pin in. 60 (black), 9-pin D connector #6 pin 7-8 gnd. 59 (yellow) AI6 V CH3Cl MFC return 9-pin D connector #7 pin 2, 0-5 VDC, brd. pin in. 25 (black), 9-pin D connector #7 pin 7-8 gnd. 24 (yellow) AI7 V Aux 1 MFC return 9pin D connector #8 pin 2, 0-5 VDC, brd. pin in. 57 (black), 9-pin D connector #8 pin 7-8 gnd. 56 (yellow) AI8 V T.C. gauge pressure return 2-pin DC jack 0-5 VDC, brd. pin in. 34 (red), gnd. 32 (black) AI9 V D.V. gauge upstream pressure return 2-pin DC jack 0-1.5 VDC, brd. pin in. 66 (red), gnd. 67 (black) AI10 V M.K.S. 651C downstream pressure retu rn 2-pin DC jack 0-1.5 VDC, brd. pin in. 31 (red), gnd. 29 (black) AI11 V RF voltage return reading 15-pin contact #1 Pin 9+,10-, 0-10 VDC, brd. pin in. 63 (white-black stripe), gnd. 64 (green-black stripe) AI12 V RF power return reading 15-pin contact #1 pin 11+,12-, 0-10 VDC, brd. pin in. 61(orange-black stripe), gnd. 59 (blue-black stripe) AI13 V purge MFC return 9-pin D connect or #9 pin 2, 0-5 VDC, brd. pin in. 26 (black), 9-pin D connector #9 pin 7-8 Gnd. 27 (yellow) AI14 Pressure output voltage digital I/O 37-pin connection pressure controller pins 36+, 35-, 0-10 VDC, brd. pin in 58 (white-black red stripe), gnd. 56 (green) Ai15 Position output voltage digital I/O 37-pin connection pre ssure controller pins 37+,35-, 0-10 VDC, brd. pin in 23 (orange-g reen stripe), gnd. 24 (green) AO0 N.C.

PAGE 112

100 Appendix B: (Continued) AO1 N.C. P0L0 N.C. P0L1 N.C. P0L2 N.C. P0L3 N.C. P0L4 N.C. P0L5 N.C. P0L6 N.C. P0L7 N.C. P1L0 N.C. P1L1 N.C. P1L2 N.C. P1L3 N.C. P1L4 N.C. P1L5 N.C. P1L6 N.C. P1L7 N.C. PCI6703 (analog output 16 and digital I/O 8) AO0 V H2 MFC set point 9-pin D connector #1 pin 6, 0-5 VDC, brd. pin in. 34 (brown), 9-pin D connector #1 pin 7-8 gnd. 68 (blue) AO1 V Ar MFC set point 9-pin D connect or #2 pin 6, 0-5 VDC, brd. pin in. 66 (brown), 9-pin D connector #2 pin 7-8 gnd. 33 (blue) AO2 V C3H3 MFC set point 9-pin D connector #3 pin 6, 0-5 VDC, brd. pin in. 31 (brown), 9-pin D connector #3 pin 7-8 gnd. 65 (blue) AO3 V SiH4 MFC set point 9-pin D connector #4 pin 6, 0-5 VDC, brd. pin in. 63 (brown), 9-pin D connector #4 pin 7-8 gnd. 30 (blue) AO4 V N2 MFC set point 9-pin D connector #5 pin 6, 0-5 VDC, brd. pin in. 28 (brown), 9-pin D connector #5 pin 7-8 gnd. 62 (blue) AO5 V HCl MFC set point 9-pin D connect or #6 pin 6, 0-5 VDC, brd. pin in. 60 (brown), 9-pin D connector #6 pin 7-8 gnd. 27 (blue) AO6 V CH3Cl MFC set point 9-pin D connector #7 pin 6, 0-5 VDC, brd. pin in. 25 (brown), 9-pin D connector #7 pin 7-8 gnd. 59 (blue) AO7 V Aux 1 MFC set point 9-pin D connect or #8 pin 6, 0-5 VDC, brd. pin in. 57 (brown), 9-pin D connector #8 pin 7-8 gnd. 24 (blue) AO8 V RF generator control voltage 15-pin connector #1 pins 7+,8-, 0-10 VDC, brd. pin in. 22 (red-black stripe), gnd. 55 (black-white stripe) AO9 V M.K.S. 651C downstream pressure an alog set point digital I/O 37-pin pressure controller pins 33+,34-, 010 VDC, brd. pin in. 54 (red-green stripe), gnd. 20 (black-red white stripe) AO10 V Purge MFC set point 9-pin D connect or #9 pin 6, 0-5 VDC brd. pin in. 52 (brown), 9-pin D connector #9 pin 7-8 gnd. 18 (blue) AO11 N.C.

PAGE 113

101 Appendix B: (Continued) AO12 N.C. AO13 N.C. AO14 N.C. AO15 N.C. P0L0 N.C. P0L1 N.C. P0L2 N.C. P0L3 N.C. P0L4 N.C. P0L5 N.C. P0L6 N.C. P0L7 N.C. PCI6509 (digital I/O 96, 0-5 VDC TTL/CMOS compatable) Breakout board 1 (block 1 pins 1-50) P0L0 25-pin contact #1 pin 4 H2 safety activate DO, brd. pin 47 (green) P0L1 25-pin contact #1 pin 5 C3H8 safety activate DO, brd. pin 45 (orange) P0L2 25-pin contact #1 pin 6 SiH4 safety activate DO, brd. pin 43 (blue) P0L3 25-pin contact #1 pin 7 N2 safety activate DO, brd. pin 41 (red-black stripe) P0L4 25-pin contact #1 pin 8 HCl safety activate DO, brd. pin 39 (black-white stripe) P0L5 25-pin contact #1 pin 9 CH3Cl safety activate DO, brd. pin 37 (whiteblack stripe) P0L6 25-pin contact #1 pin 10 aux 1 sa fety activate DO, brd. pin 35 (greenblack stripe) P0L7 25-pin contact #1 pin 11 H2 process gas activate DO, brd. pin 33 (orangeblack) P1L0 25-pin contact #1 pin 12 Ar pro cess gas activate DO, brd. pin 31 (blueblack stripe) P1L1 25-pin contact #1 pin 13 C3H8 process gas activate DO, brd. pin 29 (redblack white stripe) P1L2 25-pin contact #1 pin 14 SiH4 process gas activate DO, brd. pin 27 (blackwhite red stripes) P1L3 25-pin contact #1 pin 15 N2 process gas activate DO, brd. pin 25 (whiteblack red stripes) P1L4 25-pin contact #1 pin 16 HCl pro cess gas activate DO, brd. pin 23 (greenblack white stripes) P1L5 25-pin contact #1 pin 17 CH3Cl process gas activate DO, brd. pin 21 (orange-red stripe) P1L6 25-pin contact #1 Pin 18 Aux 1 pro cess gas activate DO, brd. pin 19 (bluewhite) P1L7 25-pin contact #1 pin 19 H2 vent gas activate DO, brd. pin 17 (red-white

PAGE 114

102 Appendix B: (Continued) stripe) P2L0 25-pin contact #1 pin 20 Ar vent ga s activate DO, brd. pin 15 (black-red stripe) P2L1 25-pin contact #1 Pin 21 C3H8 vent gas activate DO, brd. pin 13 (white-red stripe) P2L2 25-pin contact #1 pin 22 SiH4 vent gas activate DO, brd. pin 11 (greenwhite stripe) P2L3 25-pin contact #1 pin 23 N2 vent gas activate DO, brd. pin 9 (orangegreen stripe) P2L4 25-pin contact #1 pin 24 HCl vent gas activate DO, brd. pin 7 (blue-red stripe) P2L5 25-pin contact #1 pin 25 CH3Cl vent gas activate DO, brd. pin 5 (redgreen stripe) P2L6 25-pin contact #2 pin 1 Aux 1 ve nt gas activate DO, brd. pin 3 (red) P2L7 25-pin contact #2 pin 12 RF On DO, brd. pin 1 (blue-black stripe) P3L0 25-pin contact #2 pin 2 H2 chamber purge flow DO, brd. pin 48 (black) P3L1 25-pin contact #2 pin 3Ar chambe r purge flow DO, Brd. pin 46 (white) P3L2 25-pin contact #2 pin 4 main ma nifold valve activate DO, brd. pin 44 (green) P3L3 25-pin contact #2 pin 5 top Ar pur ge cooling gas DO, brd. pin 42 (orange) P3L4 25-pin contact #2 pin 6 roug hing valve DO, brd. pin 40 (blue) P3L5 25-pin contact #2 pin 7 poppet va lve DO, brd. pin 38 (red-black stripe) P3L6 25-pin contact #2 pin 8 TC gauge valve DO, brd. pin 36 (black-white stripe) P3L7 25-pin contact #2 pin 10 vent va lve close (NO valve) DO, brd. pin 34 (green-black stripe) P4L0 25-pin contact #2 pin 9 poppet bypass valve (DO), brd. pin 32 (whiteblack stripe) P4L1 25-pin contact #2 pin 11 purge flow valve (DO), brd. pin 30 (orangeblack stripe) P4L2 25-pin contact #2 pin 15 fan power on (DO), brd. pin 28 (whiteblack red stripes) P4L3 25-pin contact #2 pin 14 pump N2 on (DO), brd. pin 26 (black-red white stripes) P4L4 25-pin contact #2 pin 16 Aux 1 pneumatic (DO) Brd. Pin 24 (greenblack white stripes) P4L5 25-pin contact #2 pin 17 Aux 2 pne umatic (DO) brd. pin 22 (orange-red stripe) P4L6 25-pin contact #2 pin 18 Aux 3 pne umatic (DO) brd. pin 20 (blue-white stripe) P4L7 25-pin contact #2 pin 19 Aux 4 pne umatic (DO) brd. pin 18 (red-white stripe) P5L0 25-pin contact #2 pin 20 Aux 5 pne umatic (DO) brd. pin 16 (black-red stripe)

PAGE 115

103 Appendix B: (Continued) P5L1 25-pin contact #2 pin 21 Aux 6 pne umatic (DO) brd. pin 14 (white-red stripe) P5L2 25-pin contact #2 pin 22 Aux 7 pneum atic (DO) brd. pin 12 (green-white stripe) P5L3 25-pin contact #2 pin 23 Aux 8 pneum atic (DO) brd. pin 10 (orange-green stripe) P5L4 25-pin contact #2 pin 23 Aux 9 pne umatic (DO) brd. Pin 8 (blue-red stripe) P5L5 25-pin contact #2 pin 25 Aux 10 pneumatic ( DO) brd. Pin 6 (red-green stripe) P5L6 15-pin contact #2 pin 1 Aux 11 pneumatic (DO) brd. pin 4 (red) P5L7 15-pin contact #2 pin 2 Aux 12 pneumatic (DO) brd. Pin 2 (black) +5VDC power brd. pin 49 (N.C.) Signal ground brd. pin 50 Breakout board 2 (block 2 pins 1-50) P6L0 O.S. emergency button (external bu tton switch) DI, brd. pin 97, BO pin 47 (black) P6L1 15-pin contact #1 pin 1 water inte rlock DI, brd. pin 95, BO pin 45 (red) P6L2 15-pin contact #1 pin 2 door interl ock DI, brd. pin 93, BO pin 43 (black) P6L3 15-pin contact #1 pin 3 gas interlock H2/HCl DI, brd. pin 91, BO pin 41 (white) P6L4 15-pin contact #1 pin 4 pressure in terlock (reverse acting) DI, brd. pin 89 BO pin 39(green) P6L5 15-pin contact #1 pin 5 K1 RF stat us relay return, brd. pin 87, BO pin 37 (orange) P6L6 15-pin contact #1 pin 6 K2 RF stat us relay return, brd. pin 85, BO pin 35 (blue) P6L7 15-pin contact #1 pin 14 power sens e (DI), brd. pin 83, BO pin 33 (greenwhite stripe) P7L0 15-pin contact #1 pin 15Aux. digital input 1, brd. pin 81, BO pin 31 (bluewhite stripe) P7L1 15-pin contact #2 pin 3 Aux 13 pne umatic (DO) brd. pin 79, BO pin 29 (white) P7L2 15-pin contact #2 pin 4 Aux 14 pneuma tic (DO) brd. pin 77, BO pin 27 (green) P7L3 15-pin contact #2 pin 5 Aux 15 pne umatic (DO) brd. pin 75, BO pin 25 (orange) P7L4 Pressure DI/O pin 5 learn syst em (DO) brd. pin 73, BO pin 23 (red) P7L5 Pressure DI/O pin 6 analog set poi nt (DO) brd. pin 71, BO pin 21 (black) P7L6 Pressure DI/O pin 7 softstart (DO) brd. pin 69, BO pin 19 (white) P7L7 Pressure DI/O pin 8 close valv e (DO) brd. pin 67, BO pin 17 (orange) P8L0 Pressure DI/O pin 10 analog scali ng control (DO) brd. pin 65, BO pin 15 (blue) P8L1 pressure DI/O analog set point c ontrol pin 11(DO) brd. pin 63, BO pin 13

PAGE 116

104 Appendix B: (Continued) (red-black stripe) P8L2 Pressure DI/O set point E pin 12 (DO) brd. pin 61, BO pin 11(black-red stripe) P8L3 Pressure DI/O set point D pin 13 ( DO) brd. pin 59, BO pin 9 (white-black stripe) P8L4 Pressure DI/O set point C pin 14 ( DO) brd. pin 57, BO pin 7 (green-black stripe) P8L5 Pressure DI/O set point B pin 15 (DO) brd. pin 55, BO pin 5 (orangeblack stripe) P8L6 Pressure DI/O set point A pin 16 ( DO) brd. pin 53, BO pin 3 (blue-black stripe) P8L7 Pressure DI/O valve open status return pin 19 (DI) brd. pin 51, BO pin 1 (red-white stripe) P9L0 Pressure DI/O valve closed status return pin 23 (DI) brd. pin 98, BO pin 48 (black-white stripe) P9L1 Pressure DI/O remote zero pin 25 (DO) brd. pin 96, BO pin 46 (white-red stripe) P9L2 Pressure DI/O stop valve pin 26 (DO) brd. pin 94, BO pin 44 (green-white black stripe) P9L3 Pressure DI/O open valve pin 27 (DO) brd. pin 92, BO pin 42 (orange-red stripe) P9L4 Pressure DI/O PLO #2 status pin 28 (DI) brd. pin 90, BO pin 40 (blue-white stripe) P9L5 Pressure DI/O PLO #1 status pin 29 (DI) brd. pin 88, BO pin 38 (red-black white stripe) P9L6 N.C. P9L7 N.C. P10L0 N.C. P10L1 N.C. P10L2 N.C. P10L3 N.C. P10L4 N.C. P10L5 N.C. P10L6 N.C. P10L7 N.C. P11L0 N.C. P11L1 N.C. P11L2 N.C. P11L3 N.C. P11L4 N.C. P11L5 N.C. P11L6 N.C. P11L7 N.C.

PAGE 117

105 Appendix B: (Continued) Breakout box (back) MFC pinouts (Celerity model 1661 UDU9 configuration pg. 16 of manual) 9-pin connection MFC (male) to 9-pi n breakout boards (male connection) Wire female/female ends (positions 1-8) 1. valve off (red) 2. output (0-5 VDC) (black) 3. +15 VDC power (white) 4. Power common (green) 5. -15 VDC power (purple) 6. Set point (0-5 VDC) (brown) 7. Signal common (blue) 8. Signal common (yellow) 9. VTP -15-0 VDC (orange) Presure control MKS 651C Transducer control end 15-pin male (pressure control box) 1. +15VDC supply 2. +Pressure input (0-10 VDC) 3. Reserved 4. Reserved 5. Power ground 6. 15 VDC supply 7. +15 VDC supply 8. Reserved 9. -15 VDC supply 10. Reserved 11. Digital ground 12. -Pressure input 13. Reserved 14. Reserved 15. Chassis ground Transducer end (15-pin male from relay block connected to 15pin from relay box) 1. N.A. 2. +Pressure input (0-10 VDC) (red) 3. N.A. 4. N.A. 5. N.A.

PAGE 118

106 Appendix B: (Continued) 6. N.A. 7. N.A. 8. N.A. 9. N.A. 10. N.A. 11. N.A. 12. Pressure input (0-10 VDC) (black) 13. N.A. 14. N.A. 15. Chasis ground (green) Poppet Valve Control Control End 9-Pin Male 1. Motor winding A low (red) 2. Motor winding A high (black) 3. Limit switch ground (white) 4. Open limit switch signal (green) 5. Closed limit switch signal (purple) 6. Motor winding B high (brown) 7. Motor winding B low (blue) 8. +15 VDC 25 mA (for opto-switches) (yellow) 9. N.A. (orange) Valve end 15-pin female (purchased wire) 1. N.C. 2. N.C. 3. Limit switch connection 4. Open limit switch 5. Close limit switch 6. N.C. 7. N.C. 8. Winding A high 9. Winding A low 10. Winding A common 11. N.C. 12. N.C. 13. Winding B high 14. Winding B low 15. Winding B common

PAGE 119

107 Appendix B: (Continued) Digital I/O remote control (connections w ith digital and analog cards listed above) 1. Process limit relay PLO#1 NC contact (N.C.) 2. Process limit relay PLO#1 NO contact (N.C.) 3. Process limit relay PLO#2 NC contact (N.C.) 4. Digital ground (green) 5. Learn system (hold low) (red) 6. Analog set point (hold low with 11 for position c ontrol) (black) 7. Close valve (low) (white) 8. Softstart (hold low) (orange) 9. Reserved (N.C.) 10. Analog scaling control (low 10% FS, high 100%) (blue) 11. Analog setpoint control (hold 11 lo w only for pressure control) (redblack stripe) 12. Setpoint E (low) (black-red-stripe) 13. Setpoint D (low) (white-black stripe) 14. Setpoint C (low) (green-black stripe) 15. Setpoint B (low) (orange-black stripe) 16. Setpoint A (low) (blue-black stripe) 17. Reserved (N.C.) 18. Reserved (N.C.) 19. Valve open status return (hi gh = open) (red-white stripe) 20. Process limit relay PLO relay #1 common 21. Process limit relay PLO relay #2 common 22. Process limit relay PLO relay #2 NO contact 23. Valve closed status return (hi gh = closed) (black-white stripe) 24. Reserved (N.C.) 25. Remote zero (low) (white-red stripe) 26. Stop valve (low) (green-white stripe) 27. Open valve (low) (orange-red stripe) 28. Process limit relay PLO #2 status (low = out of limit) (blue white stripe) 29. Process limit relay PLO #1 status (low = out of limit) (red-black white stripe) 30. +15 VDC output 31. -15 VDC output 32. Power ground 33. + Analog set point input (red-green stripe) 34. Analog set point input (black-red stripe) 35. Analog ground (green-white stripe) 36. Pressure output voltage (white-black stripe) 37. Position output voltage (orange-green stripe)

PAGE 120

108 Appendix B: (Continued) Diaphragm valves Upper reactor gauge connection 8(9)-pin 1. Signal (brown) 2. N.C. (black) 3. N.C. (red) 4. Gain res. (green) 5. Signal (yellow) 6. Power + (blue) 7. Power return (orange) 8. Gain res. (white) 9. N.C. (RF shield) Lower reactor gauge (15-pin male to contact block) 1. Signal (brown) 2. N.C. (black) 3. N.C. (red) 4. Gain res. (green) 5. Signal (yellow) 6. Power + (blue) 7. Power return (orange) 8. Gain res. (white) 9. N.A. (RF shield) 10. N.A. 11. N.A. 12. N.A. 13. N.A. 14. N.A. 15. N.A. Analog output 2-pin DC jack (back of gauge indicator) 1. Positive signal 0-1.5 VDC (red) / (red-black white stripes) 2. Signal ground (black) T.C. gauge 1. Power (red) 2. Return (black) 3. Ground (green) 4. N.C. 5. N.C.

PAGE 121

109 Appendix B: (Continued) 6. N.C. 7. N.C. 8. N.C. 9. N.C. 25-pin digital output transport to relay box 25-pin male breakout board (contact 1) 1. 5 VDC power from linear power supply (red) 2. Power ground from linear power supply (black) 3. Signal ground from PCI6509 digital board (white) 4. H2 safety digital output 5 VDC (green) 5. C3H8 safety digital output 5 VDC (orange) 6. SiH4 safety digital output 5 VDC (blue) 7. N2 safety digital output 5 VDC (red-black stripe) 8. HCl safety digital output 5 VDC (black white stripe) 9. CH3Cl safety digital output 5 VDC (white-black stripe) 10. Auxiliary safety digital out put 5 VDC (green-black stripe) 11. H2 process digital output 5 VDC (orange-black stripe) 12. Ar process digital output 5 VDC (blue-black stripe) 13. C3H8 process digital output 5 VDC (red-black white stripes) 14. SiH4 process digital output 5 VDC (black-white red stripes) 15. N2 process digital output 5 DC (white-black red stripes) 16. HCl process digital output 5 DC (green-black white stripes) 17. CH3Cl process digital output 5 VDC (orange-red stripe) 18. Auxiliary process digital output 5 VDC (blue-white stripe) 19. H2 vent digital output 5 DC (red-white stripe) 20. Ar vent digital output 5 VDC (black-red stripe) 21. C3H8 vent digital output 5 VDC (white-red stripe) 22. SiH4 vent digital output 5 VDC (green-white stripe) 23. N2 vent digital output 5 VDC (orange-green stripe) 24. HCl vent digital out put 5 VDC (blue-red stripe) 25. CH3Cl vent digital output 5 VDC (red-green stripe) 25-pin digital output transport to relay box 25-pin male breakout board (contact 2) 1. Auxiliary vent digital output 5 VDC (red) 2. H2 purge (black) 3. Ar purge (white) 4. Main process valve (green) 5. Top Ar (orange) 6. Main bypass valve (blue) 7. Poppet valve (red-black stripe) 8. T.C. gauge (black-white stripe) 9. Poppet bypass valve (white-black stripe)

PAGE 122

110 Appendix B: (Continued) 10. Vent (green-black stripe) 11. Purge flow valve (orange-black stripe) 12. RF on (blue-black stripe) 13. 1.5 VDC rear diaphragm valve signal (red-black white stripes) 14. Pump N2 (black-white red stripes) 15. Fan power on (white-black red stripes) 16. Aux 1 pneumatic (green-black white stripes) 17. Aux 2 pneumatic (orange-red stripe) 18. Aux 3 pneumatic (blue-white stripe) 19. Aux 4 pneumatic (red-white stripe) 20. Aux 5 pneumatic (black-red stripe) 21. Aux 6 pneumatic (white-red stripe) 22. Aux 7 pneumatic (green-white stripes) 23. Aux 8 pneumatic (orange-green stripe) 24. Aux 9 pneumatic (blue-red stripe) 25. Aux 10 pneumatic (red-green stripe) 15-pin digital input transport from relay box 15-pin male breakout board (contact 1) 1. H2O flow meter interlock relay return (red) 2. Door interlock relay return (black) 3. Gas interlock relay return (white) 4. Pressure interlock relay return (green) 5. RF K1 status relay return (orange) 6. RF K2 status relay return (blue) 7. Generator control voltage positive (red-black stripe) 8. Generator control voltage negative (black-white stripe) 9. RF voltage return positive (white-black stripe) 10. RF voltage return ne gative (green-black stripe) 11. RF power return positive (orange-black stripe) 12. RF power return negative (blue-black stripe) 13. 10 VDC rear diaphragm va lve signal (red-white stripe) 14. Power sense 5 VDC (green-white stripe) 15. Aux. digital inpu t 1(blue-white stripe) 15-pin digital output transport to relay box 15-pin male breakout board (contact 2) 1. Aux 11 pneumatic (red) 2. Aux 12 pneumatic (black) 3. Aux 13 pneumatic (white) 4. Aux 14 pneumatic (green) 5. Aux 15 pneumatic (orange) 6. Aux 16 pneumatic (blue) 7. N.C. (red-black stripe)

PAGE 123

111 Appendix B: (Continued) 8. N.C. (black-white stripe) 9. N.C (white-black stripe) 10. N.C. (green-black stripe) 11. N.C. (orange-black stripe) 12. N.C. (blue-black stripe) 13. +15 VDC from linear power supply (green-white stripe) 14. Power ground (red-white stripe) 15. -15 VDC from linear power supply (blue-white stripe) 15-pin (extra contacts for digital I/O) (contact 3) 1. N.C. (red) 2. N.C. (black) 3. N.C. (white) 4. N.C. (green) 5. N.C. (orange) 6. N.C. (blue) 7. N.C. (red-black stripe) 8. N.C. (black-white) 9. N.C. (white-black stripe) 10. N.C. (green-black stripe) 11. N.C. (orange-black stripe) 12. N.C. (blue-black stripe) 13. 120VAC Power + (red-white stripe) 14. Power Ground (green-white stripe) 15. 120VAC Power (blue-white stripe) 25-pin digital output transport from relay box 25-pin female breakout board to Pneumatic plate 25-pin male breakout board 1. Power ground from linear power supply (red) 2. H2 safety digital output 5 VDC (black) 3. C3H8 safety digital output 5 VDC (white) 4. SiH4 safety digital output 5 VDC (green) 5. N2 safety digital output 5 VDC (orange) 6. HCl safety digital output 5 VDC (blue) 7. CH3Cl safety digital output 5 VDC (red-black stripe) 8. Auxiliary safety digital output 5 VDC (black-white stripe) 9. H2 process digital output 5 VDC (white-black stripe) 10. Ar process digital output 5 VDC (green-black stripe) 11. C3H8 process digital output 5 VDC (orange-black stripe) 12. SiH4 process digital output 5 VDC (blue-black stripe) 13. N2 process digital output 5 VDC (red-black white stripes) 14. HCl process digital output 5 VDC (black-white red stipes)

PAGE 124

112 Appendix B: (Continued) 15. CH3Cl process digital output 5 VDC (white-black red stripes) 16. Auxiliary process di gital output 5 VDC (green-black white stripes) 17. H2 vent digital output 5 VDC (orange-red stripe) 18. Ar vent digital output 5 VDC (blue-white stripe) 19. C3H8 vent digital output 5 VDC (red-white stripe) 20. SiH4 vent digital output 5 VDC (black-red stripe) 21. N2 vent digital output 5 VDC (white-red stripe) 22. HCl vent digital output 5 VDC (green-white stripe) 23. CH3Cl vent digital output 5 VDC (orange-green stripe) 24. Auxiliary vent digital ou tput 5VDC (blue-red stripe) 25. H2 purge (red-green stripe) 25-pin digital output transport from rela y box 25-pin female breakout board to pneumatic plate 25-pin female breakout board 1. Ar purge (red) 2. Main process valve (black) 3. Top Ar (white) 4. Main bypass valve (green) 5. Poppet valve (orange) 6. T.C. gauge (blue) 7. Poppet bypass valve (red-black stripe) 8. Vent (black-white stripe) 9. Purge flow valve (white-black stripe) 10. Pump N2 (green-black stripe) 11. Reserved for pneumatics (orange-black stripe) 12. Reserved for pneumatics (blue-black stripe) 13. Reserved for pneumatics (red-black white stripes) 14. Reserved for pneumatics (black-white red stripes) 15. Reserved for pneumatics (white-black red stripes) 16. Reserved for pneumatics (green-black white stripes) 17. Reserved for pneumatics (orange-red stripe) 18. Reserved for pneumatics (blue-red stripe) 19. Reserved for pneumatics (red-white stripe) 20. Reserved for pneumatics (black-red white) 21. Reserved for pneumatics (white-red stripe) 22. Reserved for pneumatics (green-white stripe) 23. Reserved for pneumatics (orange-green stripe) 24. Reserved for pneumatics (blue-red stripe) 25. Reserved for pneumatics (red-green stirpe)

PAGE 125

113 Appendix B: (Continued) 15-pin transport from relay box 15-pin male breakout board to pneumatic plate 15-pin male breakout board 1. H2O flow meter 5 VDC power (red) 2. H2O flow meter relay return (black) 3. Door 5 VDC power (white) 4. Door relay return (green) 5. Gas sensors 5 VDC power (orange) 6. Gas sensors relay return (blue) 7. N.C. (red-black stripe) 8. N.C. (black-white stipe) 9. N.C. (white-black stripe) 10. N.C. (green-black stripe) 11. N.C. (orange-black stripe) 12. N.C. (blue-black stripe) 13. N.C. (red-white stripe) 14. N.C. (green-white stripe) 15. N.C. (blue-white stripe) 15-pin transport from relay box 15-pin male breakout board to RF generator interior board 1. RF K1 status re lay 5 VDC power (red) 2. RF K1 status relay return (black) 3. RF K2 status relay 5 VDC power (white) 4. RF K2 status relay return (green) 5. Generator control voltage positive (orange) 6. Generator control voltage negative (blue) 7. RF voltage return positive (red-black stripe) 8. RF voltage return negative (black-white stripe) 9. RF power return positive (white-black stripe) 10. RF power return negative (green-black stripe) 11. RF on relay positive conn ection (orange-black stripe) 12. RF on relay return (blue-black stripe) 13. N.C. (red-white stripe) 14. N.C. (green-white stripe) 15. N.C. (blue-white stripe) RF generator interior board 1. 0-20 mA RF voltage return ne gative P 1-1 (black-white stripe) 2. 0-20 mA RF voltage return positive P 1-2 (red -black stripe) 3. 0-20 mA RF power return negative P 1-3(green-black stripe) 4. 0-20 mA RF power return posi tive P 1-4 (white-black stripe)

PAGE 126

114 Appendix B: (Continued) 5. 0-20 mA generator control signal positive P 2-1(orange) 6. 0-20 mA generator control signal negative P 2-2 (blue) 7. Enable input RF on P 2-3(orange-black stripe) 8. Enable reference RF on P 2-4 (blue-black stripe) 9. RF K1 status relay return P 4-1 (black) 10. RF K1 status rela y 5 VDC power P 4-2 (red) 11. RF K2 status rela y return P 4-4 (green) 12. RF K2 status relay 5 VDC power P 4-5(white) 9-pin transport from relay box 9-pin ma le breakout board to pump control 1. 120 VAC to 5 VDC pressure status relay positive (red) 2. 120 VAC to 5 VDC pressure status relay re turn (black) 3. 5 DC to 120 VAC moment ary start bypass positive (white) 4. 5 VDC to 120 VAC momentar y start bypass return (green) 5. 5 VDC to 120 VAC N.C. stop em ergency relay positive (purple) 6. 5 VDC to 120 VAC N.C. stop emergency relay positive (brown) 7. 120 VAC to 120 VAC H2O sensor bypass positive (blue) 8. 120 VAC to 120 VAC H2O sensor bypass return (yellow) 9. 12 VDC power for stop switch LED (orange)

PAGE 127

115 Appendix C: Relay box drawings and electronic IC board design The relay box is first stop fo r most of the control system signals after leaving the breakout box, with the exception of the v acuum gauges and MFC cables. The major purpose of the box is to provide a separati on of the computer power from the power required to manipulate the control system. El ectronics for the RF generator voltage to current conversion are located on the boards as well. The drawings for the box itself were drafted with AutoCAD-autodesk2005 by I. Haselbarth from sketches the author created. Only the sides needed to be cut, a nd the rest of the box was solid plates of Al. The P.C. board design was done with Expre ssPCB software from designs created and simulated in OrCad Capture. Figure C.1: Mechanical drawings displaying the specifications for the front and back plates of the relay box.

PAGE 128

116 Appendix C: (Continued) Figure C.2: The PC style electronic board 1 for the mounting of the relays needed to activate various controls on the reactor. The color green denotes the back side copper, the red denotes front side copper connections, and the white is the silkscreen for part descriptions.

PAGE 129

117 Appendix C: (Continued) Figure C.3: The PC style electronic board 2 for the mounting of the relays and various electronics needed to activate various controls on the reactor. The color green denotes the back side copper, the red denotes front side copper connections, and the white is the silkscreen for part descriptions.

PAGE 130

118 Appendix C: (Continued) Figure C.4: A picture showing electronic board 2 with mounted relays, breakout wiring, and electronics on the right hand side.

PAGE 131

119 Appendix D: Pump starter motor wiring schematic This section contains a sketch of the sa fety system and motor starting system for the Edwards DP40 dry pump. The pump receives three-phase power through 120VAC coil relay, and the coil current is activat ed by one of three switches; a momentary physical on switch, a momentar ily activated computer signal relay, or the low current sensor relays. The computer a nd physical switches are wired in parallel and are used to bypass the low current sensors which will only operate when the pump is activated and consuming power. The safety relays, the wate r sensor, thermal snap switch, and current overload sensor, are located afte r the parallel starting system and are wired in series. A fault in any of these sensors will open th e current path, and deactivate the 120 VAC coil relay. Two final switches, a physical sw itch and a computer si gnal relay switch are placed in the series line to enable the cessati on of current to stop the pump when it is not in use and enable the pump to be disengaged for emergencies. A 120VAC relay is placed in the series line to communicate pump act ivation to the computer for the safety interlocks. When the 120VAC current can flow through this path, the coil magnetically closes the 3-phase relay and allows the pump to activate. The three-phase line has two safety switches in line, one knife switch with fuses for overload protection located on the wall of the cleanroom, and a emergency cut-o ff switch located above the starter to meet OSHA safety requirements.

PAGE 132

120 Appendix D: (Continued) Figure D.1: A hand drawn sketch of the starting syst em for the vacuum pump motor. Safety wiring in series ensures that any fault will deactivate the system.

PAGE 133

121 Appendix E: Labview VI programs for the primary control system This section provides the documentati on and code from the Labview VIs written to facilitate the c ontrol of the LPCVD reactor. MainCVDControlVI.vi Connector Pane Front Panel Controls and Indicators

PAGE 134

122 Appendix E: (Continued) Serial Port A string that uniquely identifies the re source to be opened and written to as well as read from. Th e grammar for the resource name is shown below. Optional string segments are shown in square brackets ([]) Interface Syntax VXI VXI[boa rd]::VXI logical address[::INSTR] GPIB-VXI GPIB-VXI[board] ::VXI logical address[::INSTR] GPIB GPIB[board]::primary address[::secondary address][::INSTR] Serial ASRL[board][::INSTR] The following table shows the default value for optional string segments. Optional String Segments Default Value board 0 secondary address none The following table shows exam ples of address strings. Address String Description VXI0::1 A VXI de vice at logical address 1 in VXI interface VXI0. GPIB-VXI::9 A VXI device at logical address 9 in a GPIB-VXI controlled system. GPIB::5 A GPIB device at primary address 5. ASRL1 A seri al device attached to interface ASRL1. AO Array 1 Cluster Setpnt switch Setpoint 2 Setpoint 1 String Vent Process Safety

PAGE 135

123 Appendix E: (Continued) Automatic Ramp control Start Finish Rate [%/min] Manual RF Board Control Max RF Board Voltage Temp Setpoint Selector Temp. Setpoint 1 Temp. Setpoint 2 Auto/Manual Ridiculous Temperature stop Interlock Reset Emergency Shut Interlock Downstream Pressure Setpoint Torr Purge H2 Ar Main Bypass Pump Inlet Pressure Poppet Valve

PAGE 136

124 Appendix E: (Continued) TOP Ar Purge Process Enable RF RUN AO Array 2 Cluster Setpnt switch Setpoint 2 Setpoint 1 String Vent Process Safety AO 1 Set Var. Cluster Scaling Constant Analog Channel Input Max Voltage Analog Channel Output Process Channel Dig Out Safety Channel Dig Out Vent Channel Dig Out

PAGE 137

125 Appendix E: (Continued) AO 2 Set Var. Cluster Scaling Constant Analog Channel Input Max Voltage Analog Channel Output Process Channel Dig Out Safety Channel Dig Out Vent Channel Dig Out AO Out Tab AO Out Tab 2 Main Vent Auto Purge Poppet Bypass Purge Setpoint Analog Setpoint Throttle Pres./Pos. Throttle Stop Open Close Throttle Softstart Throttle Preset Setopints

PAGE 138

126 Appendix E: (Continued) Pump Stop Pump Start SLM file path file path is the path name of the file. If file path is empty (default) or is , the VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box. SCCM file path file path is the path name of the file. If file path is empty (default) or is , the VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box. Temp file path file path is the path name of the file. If file path is empty (default) or is , the VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box. Learn AI Array 1 Cluster 2 Operation Operation 2 Rate (sl/ccm) Gas Flow String Safety Process Vent Gas Spike Manual Mode Status Hold Mode Status

PAGE 139

127 Appendix E: (Continued) Temp. Spike Board Voltage To RF RF Output Voltage Pyrometer Reading Temperature SCCM Gas Level Water Interlock Door Interlock Gas Interlock Pressure Interlock RF Interlock Safe Interlock Main Gas Valve Pump Inlet mTorr Downstream Pressure Torr Upstream Pressure Torr Main Vent Line Bypass VacLine Main VacLine Process Gas AI Array 2 Cluster 2

PAGE 140

128 Appendix E: (Continued) String Operation Operation 2 Rate (sl/ccm) Gas Flow Safety Process Vent 2 SLM Gas Level RF KW Power K1 Status K2 Status Purge Flow Valve Position % Pump N2 On Pressure Output Torr Power Loss Top Ar Auto Counter

PAGE 141

129 Appendix E: (Continued) Block Diagram

PAGE 142

130 Appendix E: (Continued)

PAGE 143

131 Appendix E: (Continued)

PAGE 144

132 Appendix E: (Continued)

PAGE 145

133 Appendix E: (Continued) List of SubVIs and Express VIs with Configuration Information VISA Configure Serial Port C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Instr\_visa.llb\VISA Configure Serial Port Record VI.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Record VI.vi

PAGE 146

134 Appendix E: (Continued) Emergency Safety.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Emergency Safety\Emergency Safety.vi Gas Control.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Gas Control\Gas Control.vi Temperature Control.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Temperature Control\Temperature Control.vi Digital Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Output Sample Channel.vi Digital Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Input Sample Channel.vi Timer.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Timer.vi Analog Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Output Sample Channel.vi Analog Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Input Sample Channel.vi MKS600Remote2.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Pressure Control\MKS600Remote2.vi Time Delay Time Delay Inserts a time delay in the Express VI. ------------------This Express VI is configured as follows: Delay Time: 0.5 s

PAGE 147

135 Appendix E: (Continued) Time Delay2 Time Delay Inserts a time delay in the Express VI. ------------------This Express VI is configured as follows: Delay Time: 1 s Purge Gas.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Gas Control\Purge Gas.vi Time Delay5 Time Delay Inserts a time delay in the Express VI. ------------------This Express VI is configured as follows: Delay Time: 0.5 s Time Delay6 Time Delay Inserts a time delay in the Express VI. ------------------This Express VI is configured as follows: Delay Time: 1 s VISA Configure Serial Port (Instr).vi C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Instr\_visa.llb\VISA Configure Serial Port (Instr).vi Record VI.vi Connector Pane

PAGE 148

136 Appendix E: (Continued) Front Panel Controls and Indicators SLM Array Array 2[4] SCCM Array Array 2[2] Temperature Array millisecond timer value SCCM file path (dialog if empty) file pathis the path name of the file. If file path is empty (default) or is , the VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box. SLM file path (dialog if empty) file pathis the path name of the file. If file path is empty (default) or is , th e VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box. Temp file path (dialog if empty) file path is the path name of the file. If file path is empty (default) or is , the VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box.

PAGE 149

137 Appendix E: (Continued) Block Diagram List of SubVIs and Express VIs with Configuration Information Write To Spreadsheet File.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Write To Spreadsheet File.vi Write To Spreadsheet File.vi Converts a 2D or 1D array of single-precision numbers to a text stri ng and writes the string to a new byte stream file or appe nds the string to an existing file. Connector Pane

PAGE 150

138 Appendix E: (Continued) Front Panel Controls and Indicators file path (dialog if empty) file path is the path name of the file. If file path is empty (default) or is , th e VI displays a dialog box from which you can select a file. Error 43 occu rs if you cancel the dialog box. 1D data 1D data contains the single-precision numbers the VI writes to the file if this input is not empty. 1D data contains the sing le-precision numbers the VI writes to the file if this input is not empty. append to file? (new file:F) Append to F ile? indicates whether to append the data to an existing file. 2D data 2D data contains the single-precision numbers the VI writes to the file if 1D data is not wired or is empty. 2D data contains the sing le-precision numbers the VI writes to the file if 1D data is not wired or is empty. format (%.3f) format specifies how to convert the numbers to characters. transpose? (no:F) If transpose? is TRUE, the VI transposes the data after converting it from a string.

PAGE 151

139 Appendix E: (Continued) delimiter (Tab) delimiter is the character or string of characters, such as tabs, commas, and so on, to use to delimit fields in the spreadsheet file. The default is a single tab character. prompt new file path (Not A Path if cancelled) Ne w File Path returns the path to the file. Block Diagram List of SubVIs and Express VIs with Configuration Information General Error Handler.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\Utility\error.ll b\General Error Handler.vi Write File+ (string).vi C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Utility\file.llb\Write File+ (string).vi Close File+.vi C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Utility\file.llb\Close File+.vi

PAGE 152

140 Appendix E: (Continued) Open/Create/Replace File.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\Utility\file.llb\ Open/Create/Replace File.vi Emergency Safety.vi Connector Pane Front Panel Controls and Indicators Shut Interlock Interlock Reset RF ON

PAGE 153

141 Appendix E: (Continued) Run Mode K1 Status K2 Status Safe Interlock Emergency RF Interlock Pressure Interlock Gas Interlock Door Interlock Water Interlock Power Loss Block Diagram

PAGE 154

142 Appendix E: (Continued)

PAGE 155

143 Appendix E: (Continued)

PAGE 156

144 Appendix E: (Continued)

PAGE 157

145 Appendix E: (Continued) List of SubVIs and Express VIs with Configuration Information Digital Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Input Sample Channel.vi Timer.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Timer.vi Digital Input Sample Channel.vi Connector Pane Front Panel Controls and Indicators DAQmx Physical Channel Boolean

PAGE 158

146 Appendix E: (Continued) Block Diagram List of SubVIs and Express VIs with Configuration Information DAQmx Create Virtual Channel.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Virtual Channel.vi DAQmx Start Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Start Task.vi DAQmx Clear Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Clear Task.vi DAQmx Create Channel (DI-Digital Input).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Channel (DI-Digital Input).vi DAQmx Read.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\read.llb\DAQmx Read.vi DAQmx Read (Digital B ool 1Line 1Point).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\read.llb\DAQmx Read (Digital Bool 1Line 1Point).vi Timer.vi Connector Pane

PAGE 159

147 Appendix E: (Continued) Front Panel Controls and Indicators Boolean 6 wait Specifies the target number of sec onds the VI waits after the start time. When the VI reaches the Time Target (s), Time has Elapsed is TRUE. RF Interlock Elapsed Time (s) Displays the amount of time in seconds that has elapsed since the start time and the Present (s) time. Block Diagram

PAGE 160

148 Appendix E: (Continued) List of SubVIs and Express VIs with Configuration Information Elapsed Time Elapsed Time Keeps track of time by indicating when a cer tain amount of time has elapsed. The elapsed time is the present time minus the start time that you specify. ------------------This Express VI is configured as follows: Time Target: 1 s Auto Reset: Off Gas Control.vi Connector Pane Front Panel

PAGE 161

149 Appendix E: (Continued) Controls and Indicators millisecond timer value Process Gas AO Array 1 Cluster Setpnt switch Setpoint 2 Setpoint 1 String Vent Process Safety AO 1 Set Var. Cluster Scaling Constant Analog Channel Input Max Voltage Analog Channel Output Process Channel Dig Out Safety Channel Dig Out Vent Channel Dig Out AO Out Tab

PAGE 162

150 Appendix E: (Continued) AO Array 2 Cluster Setpnt switch Setpoint 2 Setpoint 1 String Vent Process Safety AO 2 Set Var. Cluster Scaling Constant Analog Channel Input Max Voltage Analog Channel Output Process Channel Dig Out Safety Channel Dig Out Vent Channel Dig Out AO Out Tab 2 Gas Spike array sccm Array 2[4]

PAGE 163

151 Appendix E: (Continued) array slm Array 2[2] AI Array 1 Cluster 2 Operation Operation 2 Rate (sl/ccm) Gas Flow String Safety Process Vent AI Array 2 Cluster 2 String Operation Operation 2 Rate (sl/ccm) Gas Flow Safety Process Vent 2

PAGE 164

152 Appendix E: (Continued) SLM output cluster Array 2[1] Array 2[0] SCCM output cluster Array 2[2] Array 2[3] Array 2[0] Array 2[1] Array 2[2] Block Diagram List of SubVIs and Express VIs with Configuration Information 2D to 1D Array.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\2D to 1D Array.vi Gas Handler.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Gas Control\Gas Handler.vi

PAGE 165

153 Appendix E: (Continued) Gas Handler 2.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Gas Control\Gas Handler 2.vi 2D to 1D Array.vi Connector Pane Front Panel Controls and Indicators Array Array 2 element Block Diagram Gas Handler.vi Connector Pane

PAGE 166

154 Appendix E: (Continued) Front Panel Controls and Indicators Process Gas AO Array 1 Cluster Setpnt switch Setpoint 2 Setpoint 1 String Vent

PAGE 167

155 Appendix E: (Continued) Process Safety AO 1 Set Var. Cluster Scaling Constant Analog Channel Input Max Voltage Analog Channel Output Process Channel Dig Out Safety Channel Dig Out Vent Channel Dig Out AO Out Tab Gas Spike Array AI Array 1 Cluster 2 Operation Operation 2 Rate (sl/ccm) Gas Flow String

PAGE 168

156 Appendix E: (Continued) Safety Process Vent Block Diagram List of SubVIs and Express VIs with Configuration Information Analog Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Input Sample Channel.vi Digital Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Output Sample Channel.vi Analog Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Output Sample Channel.vi

PAGE 169

157 Appendix E: (Continued) Analog Input Sample Channel.vi Connector Pane Front Panel Controls and Indicators DAQmx Physical Channels physical channe ls specifies the names of the physical channels to use to create virtual ch annels. The DAQmx Physical Channel Constant lists all physical channels on devices and modules installed in the system. Max Voltage maximum value specifies in units the maximum value youexpect to measure. Sample Rate rate specifies the sampling ra te in samples per ch annel per second. I f you use an external source for the Sample Clock, set this input to the maximum expected rate of that clock. mean mean is the average of the values in the input sequence X. Block Diagram

PAGE 170

158 Appendix E: (Continued) List of SubVIs and Express VIs with Configuration Information DAQmx Create Virtual Channel.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Virtual Channel.vi DAQmx Create Channel (AI-Voltage-Basic).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Channel (AI-VoltageBasic).vi DAQmx Timing.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\timing.llb\DAQmx Timing.vi DAQmx Timing (Sample Clock).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\timing.llb \DAQmx Timing (Sample Clock).vi DAQmx Start Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Start Task.vi DAQmx Read.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\read.llb\DAQmx Read.vi DAQmx Read (Analog 1D DBL 1Chan NSamp).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\read.llb\DAQmx Read (Analog 1D DBL 1Chan NSamp).vi Mean.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\analysis\baseanly.llb\Mean.vi DAQmx Clear Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Clear Task.vi Digital Output Sample Channel.vi Connector Pane

PAGE 171

159 Appendix E: (Continued) Front Panel Controls and Indicators DAQmx Physical Channel Boolean Block Diagram List of SubVIs and Express VIs with Configuration Information DAQmx Create Virtual Channel.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Virtual Channel.vi DAQmx Write.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\write.llb\DAQmx Write.vi DAQmx Start Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Start Task.vi DAQmx Clear Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Clear Task.vi DAQmx Create Channel (DO-Digital Output).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Channel (DO-Digital Output).vi

PAGE 172

160 Appendix E: (Continued) DAQmx Write (Digital Bool 1Line 1Point).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\write.llb\DAQmx Write (Digital Bool 1Line 1Point).vi Analog Output Sample Channel.vi Connector Pane Front Panel Controls and Indicators Setpnt switch Setpoint 2 Setpoint 1 Scaling Constant DAQmx Physical Channel Max Voltage maximum value specifies in units the maximum value you expect to measure. Setpoint

PAGE 173

161 Appendix E: (Continued) Block Diagram List of SubVIs and Express VIs with Configuration Information DAQmx Create Virtual Channel.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Virtual Channel.vi DAQmx Create Channel (AO-Voltage-Basic).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Channel (AO-VoltageBasic).vi DAQmx Write.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\write.llb\DAQmx Write.vi DAQmx Write (Analog DBL 1Chan 1Samp).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\write.llb\DAQmx Write (Analog DBL 1Chan 1Samp).vi DAQmx Start Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Start Task.vi DAQmx Is Task Done.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\task .llb\DAQmx Is Task Done.vi DAQmx Clear Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Clear Task.vi

PAGE 174

162 Appendix E: (Continued) Gas Handler 2.vi Connector Pane Front Panel Controls and Indicators AO Array 2 Cluster Setpnt switch Setpoint 2 Setpoint 1 String

PAGE 175

163 Appendix E: (Continued) Vent Process Safety AO 2 Set Var. Cluster Scaling Constant Analog Channel Input Max Voltage Analog Channel Output Process Channel Dig Out Safety Channel Dig Out Vent Channel Dig Out AO Out Tab 2 Process Gas Array AI Array 2 Cluster 2 String Operation Operation 2 Rate (sl/ccm)

PAGE 176

164 Appendix E: (Continued) Gas Flow Safety Process Vent 2 Gas Spike Block Diagram List of SubVIs and Express VIs with Configuration Information Analog Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Output Sample Channel.vi Digital Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Output Sample Channel.vi

PAGE 177

165 Appendix E: (Continued) Analog Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Input Sample Channel.vi Temperature Control.vi Connector Pane Front Panel

PAGE 178

166 Appendix E: (Continued) Controls and Indicators Temp. Setpoint 2 Temp. Setpoint 1 Automatic Ramp control Start Finish Rate [%/min] Manual RF Board Control Max RF Board Voltage Temperature Selector Auto/Manual Ridiculous Temperature Previous Temperature millisecond timer value RF ON Manual Mode Status Hold Mode Status Temp. Spike Pyrometer Reading Temp output cluster T [deg C]

PAGE 179

167 Appendix E: (Continued) Temp array millisecond timer value RF Output Voltage RF KW Power Board Voltage To RF K1 Status K2 Status Block Diagram List of SubVIs and Express VIs with Configuration Information Manual-Ramp Output %.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Temperature Control\ Manual-Ramp Output %.vi

PAGE 180

168 Appendix E: (Continued) PID Engine.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Temperature Control\PID Engine.vi Analog Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Input Sample Channel.vi Digital Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Input Sample Channel.vi Manual-Ramp Output %.vi Connector Pane Front Panel Controls and Indicators Automatic Ramp control Finish Start

PAGE 181

169 Appendix E: (Continued) Rate [%/min] Max RF Voltage Manual RF Voltage Cycle Time output (%) Ramp output in percent (0 to 100). Block Diagram List of SubVIs and Express VIs with Configuration Information Ramp.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Temperature Control\Ramp.vi Ramp.vi Setpoint ramp generator. Out put ramps linearly toward the setpoint at "Rate" percent per minute. Sign (+ or -) of Rate determines polarity of ramp. Setpoint and output are expressed in percent. Output can be forced to a desired value by setting Initialize to TRUE. "Initial Value" is then written into the output. Setting "Run" to FALSE makes the output hold at the last value. Initialize overrides Run.

PAGE 182

170 Appendix E: (Continued) This VI should be called from inside a while loop with a fixed cycle time. Note that the cycle time (part of the Tuning Parameters) is in seconds and MUST be supplied; control calibration depends on this value. This VI is reentrant, so you can call it from multiple, independent higher-level VI s without interference. Connector Pane Front Panel Controls and Indicators setpoint (%) Ramp setpoint in percent. Range: 0 to 100%. run (T) TRUE puts controller in "hold". Output freezes. rate (%/min) Ramp rate in percent per minute. Positive values ramp up. cycle time (sec) Cycle time of this contro ller, i.e., how often it is called by your process. Note that this is in SECOND S. Value will directly affect all tuning parameters, so don't mess this up! initialize TRUE forces output to Initial Output value. FALSE is normal ramping operation. initial output (%) Force output to this value. Range: 0 to 100%. output (%) Ramp output in percent (0 to 100).

PAGE 183

171 Appendix E: (Continued) Block Diagram PID Engine.vi Connector Pane Front Panel

PAGE 184

172 Appendix E: (Continued) Controls and Indicators Ridiculous Temp Previous Temperature Temperature Setopint Auto/Manual TRUE selects automatic control (default), FALSE selects manual mode. Bumpless transfer is used from manual mode to automatic control. manual control Control output value us ed when auto? is set to FALSE. RF ON Max RF Board Voltage Tuning Params Kc Proportional band in percent. Range : 0 to 1000. Controller GAIN is 1/PB. Ti Integral i minutes per reset. Enter zero to disable reset. Td Rate in minutes per repeat. Ente r zero to disable derivative action. cycle time (-1) Interval (in seconds) at which this VI is called; used in calculations. If less than or equal to zero, an internal timer with 1 millisecond resolution is used. Temperature Spike T [deg C] Returned RF Board Volts Hold Mode Status

PAGE 185

173 Appendix E: (Continued) Block Diagram List of SubVIs and Express VIs with Configuration Information ModbusPyroRead_Logged.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Temperature Control\ModbusPyroRead_Logged.vi

PAGE 186

174 Appendix E: (Continued) Analog Current Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Current Output Sample Channel.vi PID % to EGU.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\addons\control\pid\ pid.llb\PID % to EGU.vi PID.vi C:\SiC CVD\Control System\Contro l System MF1\Control System Programs\PID\PID.vi PID EGU to %.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\addons\control\pid\ pid.llb\PID EGU to %.vi PID EGU to % (DBL).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\addons\control\pid\pi d.llb\PID EGU to % (DBL).vi PID % to EGU (DBL).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\addons\control\pid\pi d.llb\PID % to EGU (DBL).vi ModbusPyroRead_Logged.vi VI let you communicatie via a serial inerface with Modbus instruments. PS. No error checking is done in this version; Error clusters only implemented for future use Users might implement this for their own needs... This VI is referenced to the Modicon M odbus Protocol as desc ribed in the Modicon Modbus Protocol Reference Guide PI-MBUS-300 rev. D This VI is only tested with a Eurotherm T103 unit controller but should alse work with any other Modbus Instrument. Maarten van Bree AIR Technical Automation info@air.nl Connector Pane

PAGE 187

175 Appendix E: (Continued) Front Panel Controls and Indicators Modbus slave address (1) Serial Port A string that uniquely identifies the re source to be opened and written to as well as read from. Th e grammar for the resource name is shown below. Optional string segments are shown in square brackets ([]) Interface Syntax VXI VXI[boa rd]::VXI logical address[::INSTR] GPIB-VXI GPIB-VXI[board] ::VXI logical address[::INSTR] GPIB GPIB[board]::pri mary address[::secondary address][::INSTR] Serial ASRL[board][::INSTR] The following table shows the default value for optional string segments.

PAGE 188

176 Appendix E: (Continued) Optional String Segments Default Value board 0 secondary address none The following table shows exam ples of address strings. Address String Description VXI0::1 A VXI devi ce at logical address 1 in VXI interface VXI0. GPIB-VXI::9 A VXI device at logi cal address 9 in a GPIB-VXI controlled system. GPIB::5 A GPIB device at primary address 5. ASRL1 A seri al device attached to interface ASRL1. Data Logging String send to device String received from device Calculated CRC CRC can be us ed inside the MBmaster VI for enhanced error checking. Chan.1 Current Chan.1 Temperature String String 2

PAGE 189

177 Appendix E: (Continued) Block Diagram

PAGE 190

178 Appendix E: (Continued) List of SubVIs and Express VIs with Configuration Information MBcrc.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\Temperature Control\MBcrc.vi Simple Error Handler.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\Utility\error.ll b\Simple Error Handler.vi Write Characters To File.vi C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Utility\file.llb\Write Characters To File.vi VISA Configure Serial Port C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Instr\_visa.llb\VISA Configure Serial Port VISA Configure Serial Port (Instr).vi C:\Program Files\National Instruments\La bVIEW 7.1\vi.lib\Instr\_visa.llb\VISA Configure Serial Port (Instr).vi MBcrc.vi VI calculates CRC for Modbus protocol Connector Pane Front Panel

PAGE 191

179 Appendix E: (Continued) Controls and Indicators Data in: Device adress Data out: CRC When reading from a port th is indicator displays the st ates of the lines in the port. The radix of the contro l is set to binary. Leading zeroes are not displayed. The rightmost digit represents the state of the least significant bit of the port. If a digit is set to 1 the state of the lin e is high. A 0 indicates a low state. Block Diagram Analog Current Output Sample Channel.vi Connector Pane Front Panel Controls and Indicators Setpnt switch Setpoint 2 Setpoint 1

PAGE 192

180 Appendix E: (Continued) Scaling Constant DAQmx Physical Channel Max Voltage maximum value specifies in unitsthe maximum value you expect to measure. Setpoint Block Diagram List of SubVIs and Express VIs with Configuration Information DAQmx Create Virtual Channel.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Virtual Channel.vi DAQmx Write.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\write.llb\DAQmx Write.vi DAQmx Write (Analog DBL 1Chan 1Samp).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\write.llb\DAQmx Write (Analog DBL 1Chan 1Samp).vi DAQmx Start Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Start Task.vi DAQmx Is Task Done.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\task .llb\DAQmx Is Task Done.vi

PAGE 193

181 Appendix E: (Continued) DAQmx Clear Task.vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\configure\ta sk.llb\DAQmx Clear Task.vi DAQmx Create Channel (AO-Voltage-Basic).vi C:\Program Files\National Instruments\LabVIEW 7.1\vi.lib\DAQmx\create\channels.llb \DAQmx Create Channel (AO-VoltageBasic).vi PID.vi Revised 7-14-92; Added Cycle Time. A positional-algorithm PID Algorithm as desc ribed on page 164 of "Process Control Systems" by F.G. Shinskey (1988). Rate action is on the process variable only. Anti-reset windup and bumpless transfer are supported. This VI should be called from inside a while loop with a fixed cycle time. If Cycle Time is less than or equal to zero, an internal timer with 1 millisecond resolution is used for calculations. For loops that run faster than about 10 Hz, supply a cycle time in seconds. This VI is reentrant, so you can call it fr om multiple, independent higher-level VIs without interference. Setpoint, process variable, and output are all expressed in percent. Reverse action (also called increase-decrease) is th e "normal" controller mode wh ere the output decreases if the process variable is hi gher than the setpoint. Switching from auto to manual is bumpless; the initial value of the Manual Output control is subtracted out so that is is really a "rel ative" output control. Switching from manual to auto is also bumpless. A bi as tracking technique is used (see Shinskey, p. 138). This method is effective for modest values of de viation, beyond which a proportional kick will occur. Switching to Hold causes the controller to ignor e further changes in the process variable and suspends any integral action. Thus, the output stays constant at the last value. Switching back in to Run may re sult in a proportional kick in the output. No setpoint tracking is performed. Connector Pane

PAGE 194

182 Appendix E: (Continued) Front Panel Controls and Indicators hold (F) TRUE puts controller in "hol d". Reset action stops and the output freezes. Bumpless transfer is used from run to hold. process variable (%) Process variable (t he thing you are trying to control) in percent. Range: 0 to 100%. setpoint (%) Process setpoint in percent. Range: 0 to 100%. Tuning Params Kc Proportional band in percent. Range : 0 to 1000. Controller GAIN is 1/PB. Ti Integral i minutes per reset. Enter zero to disable reset. Td Rate in minutes per repeat. Ente r zero to disable derivative action. auto (T) TRUE selects automatic contro l, FALSE puts controller in manual. Bumpless transfer is used from automatic to manual. reverse acting (T) TRUE selects reverse (increase-decrease) action, the usual mode for controllers where the output goes down if the input is greater than the setpoint. manual out (%) Manual output setting. Th is is a "relative" percentage control.

PAGE 195

183 Appendix E: (Continued) Kd Influences loop stability. Values between 10 and 20 are normal. Lower values (>=1) result in slightly less stable systems, as a rule. cycle time (-1) Interval (in seconds) at which this VI is called; used in calculations. If less than or equal to zero, an internal timer with 1 millisecond resolution is used. output (%) Output of control al gorithm in percent (0-100). Block Diagram

PAGE 196

184 Appendix E: (Continued) MKS600Remote2.vi Connector Pane Front Panel Controls and Indicators Downstream Pressure/Position Setpoint Analog Setpoint Control Analog Setpoint On/Off Softstart Analog Scaling Learn Zero

PAGE 197

185 Appendix E: (Continued) Stop A B C D E Open Close Pump Stop Pump Start Downstream Pressure Torr Pressure Output Torr Position Output % Valve Closed Valve Open PLO#1 Status PLO#2 Status

PAGE 198

186 Appendix E: (Continued) Block Diagram List of SubVIs and Express VIs with Configuration Information Analog Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Output Sample Channel.vi Digital Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Output Sample Channel.vi Analog Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Analog Input Sample Channel.vi

PAGE 199

187 Appendix E: (Continued) Digital Input Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Input Sample Channel.vi Purge Gas.vi Connector Pane Front Panel Controls and Indicators Auto Purge Counter Previous Purge Out H2 Out Ar Out Auto Purge Counter Next Auto Purge Out

PAGE 200

188 Appendix E: (Continued) Block Diagram

PAGE 201

189 Appendix E: (Continued) List of SubVIs and Express VIs with Config Digital Output Sample Channel.vi C:\SiC CVD\Control System\Contro l System MF2\Control Systems Programs\General Use\Digital Output Sample Channel.vi