DESIGN OF A HIGH-SPEED CMOS COMPARATOR

2y ago
21 Views
2 Downloads
400.39 KB
45 Pages
Last View : 11d ago
Last Download : 3m ago
Upload by : Kaydence Vann
Transcription

DESIGN OF A HIGH-SPEED CMOSCOMPARATORMaster Thesis in Electronics System atLinköping Institute of TechnologybyAhmad SharLiTH-ISY-EX--07/4121--SELinköping 2007-11-07

DESIGN OF A HIGH-SPEED CMOSCOMPARATORMaster Thesis in Electronics System atLinköping Institute of TechnologybyAhmad SharLiTH-ISY-EX--07/4121--SESupervisor:Erik SällISY, Linköping UniversityExaminer:Mark VesterbackaISY, Linköping UniversityLinköping2007-11-07

Presentation Date2007-11-07Publishing Date (Electronic version)Department and DivisionDivision of electronics systemDepartment of Electrical EngineeringLinköpings university Linköpings Sweden2007-12-07LanguageType of PublicationEnglishOther (specify below)Number of Pages30Licentiate thesisDegree thesisThesis C-levelThesis D-levelReportOther (specify below)ISBN Master ThesisISRN: LiTH-ISY-EX--07/4121--SETitle of series (Licentiate thesis)Series number/ISSN (Licentiate thesis)URL, Electronic Versionhttp://www.ep.liu.sePublication TitleDesign of a high-speed CMOS comparator.Author(s)Ahmad SharA bstractT his m aster thesis describ es the d esign of high-speed latched com p ara tor w ith 6-bit resolution , full scale voltageof 1 .6 V and the sa m plin g frequ ency of 25 0 M H z. T he com p arato r is d esigne d in a 0.3 5 9 m C M O S process w itha sup ply voltage of 3.3 V .T he com parator is designed for tim e-in terleaved bandp ass sigm a-delta A D C . D u e to the nature of the targetapplication, it shou ld b e possible to turn o ff the co m po nents to avoid th e static pow er con sum p tion . T hecom p arator o f this d esign im plem ents the turn off tech nique w h en it is not in use. T he settling tim e of thecom p arator is less than half the clo ck cycle w h ich m eans it does not effect th e function ality of the bandpasssigm a -d elta A D C in term s of speed .T he sim u lation results are derived usin g C adence environm ent. T he results show that the com p ara tor has 6-bitresolu tion and pow er consu m ptio n of 4.13 m W for the w orst-ca se freq uenc y of 250 M H z. It fu lfills all theperform an ce re quirem en ts, m ost of them w ith large m a rgins.KeywordsComparator, CMOS comparator, Sigma-delta ADC, Low power design, High-speed.

ABSTRACTThis master thesis describes the design of high-speed latched comparatorwith 6-bit resolution, full scale voltage of 1.6 V and the sampling frequencyof 250 MHz. The comparator is designed in a 0.35 9m CMOS process with asupply voltage of 3.3 V.The comparator is designed for time-interleaved bandpass sigma-delta ADC.Due to the nature of the target application, it should be possible to turn off thecomponents to avoid the static power consumption. The comparator of thisdesign implements the turn off technique when it is not in use. The settlingtime of the comparator is less than half the clock cycle which means it doesnot effect the functionality of the bandpass sigma-delta ADC in terms ofspeed.The simulation results are derived using Cadence environment. The resultsshow that the comparator has 6-bit resolution and power consumption of 4.13mW for the worst-case frequency of 250 MHz. It fulfills all the performancerequirements, most of them with large margins.vii

viii

ACKNOWLEDGMENTSI am extremely grateful to my supervisor Dr. Erik Säll for his support andguidance throughout my thesis work. Dr. Erik Säll’s help at any crucialmoments of my thesis work gave me high motivations and self-confidence.Special thanks to Erik for his patience and kindness. I would like to thankProf. Mark Vesterbacka for his help and being the examiner for this thesiswork.I thank my uncle Dr. Riaz Tarar for his support, guidance and kindnessthroughout all my education carrier. He has been the basic source of myinspiration throughout my life. This is he who makes me all the way to comehere at this point. Thanks uncle.I thank all the staff at ES/ISY for friendly and learning environment. Specialthanks to Greger Karlströms for his help in lab to setup the environment andnecessary installation for this thesis work.I would like to thanks all my friends especially Mr. M. Wasim Sultan and Mr.M. Ali Malik for their moral and technical support, not only for this thesiswork but for throughout the graduation at Linköping University.ix

x

TABLE OF CONTENTS1 Introduction11.1 General information .11.2 Background .21.2.11.2.21.2.31.2.4A high-speed CMOS comparator with 8-bit resolution.2A 6-bit 1 GHz acquisition speed CMOS flash ADC with digital error correction.3A 0.35 9m CMOS comparator circuit for high-speed ADC applications .5Performance analysis of optimized CMOS comparator .61.3 Scope of the work .71.4 Outline .72 Theory92.1 Pre-amplifier .91.2 Comparator offset.92.3 Kickback .102.3.12.3.22.3.32.3.4Sampling switches .10Isolation transistors .11Pre-amplifier .11Neutralization technique .112.4 Parasitics .112.5 Metastability .123 Design details153.1 CMOS latch circuit .153.1.1 Comparator Optimization .16a) Transistors M1-M3 .16b) Transistors M4 & M5 .17c) Transistors M6 & M7 .17d) Transistors M8 & M9 .173.23.33.43.53.63.7SR Latch circuit .17Two phase operation .18Gain and bandwidth of the comparator .19Comparator turn off technique.19Settling time of the comparator .20Kickback noise .214 Simulation results and discussion234.1 Final simulations .234.2 The performance and design parameters.254.3 Conclusion and discussion .255 Abbreviations276 References29xi

xii

LIST OF .64.1Comparator symbol.1A high-speed CMOS comparator with 8-bit resolution.3A 6-bit 1 GHz acquisition speed CMOS flash ADC slice schematic.4The schematic of pre-amplifier of CMOS comparator circuit for high speedADC applications. 5The schematic diagram of the latch part of CMOS comparator circuit for highspeed ADC applications.6Performance analysis of optimized CMOS comparator.7Differential pair with offset measured at the output.10MOSFET parasitics capacitances.12VTC of back-to-back connected two inverters.13Comparator schematic.15CMOS latch schematic.16Schematic diagram of SR latch .18The comparator with turn off switches.19Settling time of the compartor.20Neutralization technique.21Comparator inputs model.23xiii

xiv

1INTRODUCTIONThis thesis presents the basic topologies, design decision and thetheory needed to understand the latched comparator designissues and considerations. The purpose of this project is todesign a latched comparator for single bit time-interleavedbandpass sigma-delta ADC in 0.35 9m process.The main design consideration was the comparator speed andturn off technique when it is not in use. This is because the targetapplication is supposed to be used for a band of differentfrequencies at a time and the components which are not in useshould be turned off to avoid the static power consumption.1.1 General informationA comparator, by definition is “a circuit that compares the two analog inputsignals and decodes the difference into a single digital output signal”. Figure1.1 shows the comparator symbol, where out is the single digital output as aresult of comparison of two analog inputs in1 and in2.In1outIn2Figure 1.1: Comparator symbol.1

The comparator is a critical part of almost all kind of analog-to-digital (ADC)converters. Depending on the type and architecture of the comparator, thecomparator can have significant impact on the performance of the targetapplication. The speed and resolution of an ADC is directly affected by thecomparator input offset voltage, the delay and input signal range [4].Depending on the nature, functionality and inputs, comparators are classifiedinto different types i.e. voltage and current comparators, continuous anddiscrete time comparators etc. Some basic applications of comparators areanalog-to-digital conversion, function generation, signal detection and neuralnetworks etc.1.2 BackgroundDue to the many comparator applications, researchers have designed andpresented different architectures to fulfill the requirements. The followingstudy gives an overview of some of the different comparator topologiesexamined during the pre-study. The outcome of the pre-study yieldsinformation, which topology to use.1.2.1 A high-speed CMOS comparator with 8-bit resolutionA high-speed CMOS comparator is shown in figure 1.2. The comparatorconsists of three blocks, an input stage, a flip-flop and SR latch. Thearchitecture uses two non-overlapping clocks ( 1 and 2). The circuitoperates in two modes, reset mode during 2 and regeneration mode during1. During reset mode the inputs voltage difference is established at node Aand B [1]. The regeneration happens during a short time when 1 is rising and2 is falling. At the end of regeneration process the SR latch is driven to thedigital output levels.The design was implemented in a 1.5 9m CMOS technology operating at a 2.5 V power supply with 8-bit of resolution and input range of 2.5 V [1]. Atsampling rate of 65 MHz the chip exhibit offset voltage of 3.3 mV and asensitivity 1.5 mV. At 65 MHz clock rate and input signal of 32.5 MHz, thecomparator has total power consumption of 0.85 mW [1].The design is appealing in context of power consumption, however it doesnot fulfill the requirements for this thesis work in term of speed.2

Chapter 1 – IntroductionVddInput stageM13M3SR latchflip-flops1M15M14M11M10QIBM9M8Vin1M1M2 Vin22Q M12M4M5VssFigure 1.2: A high-speed CMOS comparator with 8-bitresolution.1.2.2 A 6-bit 1 GHz acquisition speed CMOS flash ADC with digital error correctionFigure 1.3 shows the schematic of one of the 63 slices of the flash ADC withdigital error correction. The comparator in figure 1.3 is designed for a 6-bitADC converter. It consists of three parts, pre-amplifier, comparator and thenand-gate. The pre-amplifier at the comparator inputs consists of adifferential amplifier with resistances as load. The pre-amplifier is used toreduce the input offset and kickback noise. The pre-amplifier amplifies thedifference between input voltage and the reference voltage generated by theresistive ladder of the ADC [2].3

Pre-ampC1ComparatorVddM3C lkVref M 1AM 1B VinM1NandM4C lkM 4AM6M 7BM 7ARbRaC2M 3AM 3BM 4BM7S1M 10M 2AM5M 2BM2M 8BM 8AM8M 11M9M 12M 13VssFigure 1.3: A 6-bit 1 GHz acquisition speed CMOS flash ADCslice schematic.After the pre-amplifier the comparator generates the digital logic levels fromthe amplified inputs difference. During one clock phase, the regenerationnodes are charged proportionally to the amplifier outputs. In the next phasethe voltage imbalance is amplified by the regeneration loop of the NMOSsand the PMOSs to digital levels. Using this architecture in an ADC withmoderate resolution, a sample-and-hold circuit is not required [2].The nandgate at the end is used to select one of the 63 ROM-lines.This architecture has been implemented in a 0.35 9m technology with 3.3 Vpower supply. The total power consumption (all the 63 slices of Flash ADC)of the pre-amplifier and comparator is 759 mW and digital part is 165 mW atinput frequency of 141 kHz and 1 GHz sampling speed.The power consumption of this architecture is very high. The single slice hasthe power consumption almost 12 mW which is considerable high for thecomparator design of this thesis work.4

Chapter 1 – Introduction1.2.3 A 0.35 @m CMOS comparator circuit for high-speed ADCapplicationsThe comparator of the topology shown in figure 1.4 is used for high-speedADC applications. It consists of a pre-amplifier, a latch and output sampler.The pre-amplifier shown in figure 1.4 has PMOS transistors as differentialinput pairs. One reason to use the PMOSs here is that the DC input is low.The output of pre-amplifier is mirrored to the latch.VddBiasVin M1M2Vref Vref-M3 M4Vout-Vin-Vout M5M6VssFigure 1.4: The schematic of pre-amplifier of CMOScomparator circuit for high speed ADC applications.The latch consists of the cross coupled inverters connected to the groundthrough the clock as shown in figure 1.5. The latch operates in two modes,reset and evaluation, respectively. During reset phase, the latch output voltageis in the middle of the power supply rail voltages, which give a shortregeneration time. The latch is activated during the evaluation phase (whenthe clock signal is high) and outputs are sampled before it goes to reset modei.e. at the end of the evaluation phase.The last part, the sampler, consists of a transmission gate and inverters asbuffers. The transmission gate samples the latch during the evaluation phaseand then the inverters amplifies the samples and buffers the outputs. Thebuffers keep the samples constant for a whole clock period, which relaxes thetiming requirements, for the following stage (e.g. encoder) [3].The comparator designed was fabricated in a 0.35 9m technology with asupply voltage of 3.3 V. The total power consumption of the comparator is 2mW with 6 bit resolution at 1 GHz sampling frequency [3].5

VddM7SampBM6SampBVout Vout-ClkBoutputSamplerSampM8SampM2M1Vin M3ClkoutputSamplerM4 Vin-M5VssFigure 1.5: The schematic diagram of the latch part ofCMOS comparator circuit for high speed ADCapplications.1.2.4 Performance analysis of optimized CMOS comparatorThe comparator topology shown in figure 1.6 consists of two parts, CMOSlatch circuit and S-R latch circuit. This design does not use a separateamplifier but the CMOS latch circuit does some amplification. I have selectedthis topology for my thesis work. One reason to select this topology is that thecomparator does not use any separate amplifier and so the powerconsumption is less. I will discuss the design in more detail in chapter 3.6

Chapter 1 – IntroductionVddVbias t2M8 M6M14M16M7 M9M17M15VssFigure 1.6: Performance analysis of optimized CMOS comparator.The design operates at 2.5 V power supply at frequency of 500 MHz. Atmaximum sampling frequency, 500 MHz, the comparator achieves 10-bitresolution for 1 V differential with power consumption of 272 9W.1.3 Scope of the workThe scope of this work is the design of a comparator for a time-interleavedbandpass Sigma-Delta ADC. The design is implemented in a 0.35 9mtechnology with 6-bits of resolution at a sampling frequency of 250 MHz.The main consideration is to minimize the power consumption and avoid thestatic power consumption by switching it off when it is not in use.1.4 OutlineThis thesis document is organized as in chapter 2 comparator related theory ispresented along with some design issues used for this thesis work. Designdetails of the comparator and some performance measurements are discussedin chapter 3. The simulation results, conclusion and discussion are presentedin chapter 4.7

8

Chapter 2 – Theory2THEORYThis chapter gives the basic theory needed to understand thefundamental parts in comparator design. The purpose of thischapter is also to see how different components can contribute incomparator performance.2.1 Pre-amplifierThe pre-amplifier is a circuit which is used to amplify the signal so that it caneasily drive the load. In most latch comparator designs pre-amplifiers are alsoused to avoid the kickback effect from the latch and input referred offset [2].The comparator design in this thesis work does not use a separate preamplifier but the CMOS latch performs the amplification. We will discussmore about the CMOS latch in chapter 3.2.2 Comparator offsetDue to the mismatch between input transistors, the circuit exhibits a dc offsetof different values. This value of dc offset depends on the mismatch of inputand output voltages. The figure 2.1 shows a differential pair with perfectsymmetry of input and output nodes, i.e. Vin 0 as well as Vout 0, hence thecircuit has no offset error. On the other hand if the input is zero and output isnot equal to zero, the circuit exhibits mismatch and suffers a dc offset. This dcoffset is equal to the value of Vout when the input voltage (Vin) is zero, and iscalled the output referred offset.The input-referred offset voltage can be defined as the input level whichforces the output voltage to go to zero [6]. The offset can limit theperformance of comparator and can make the system nonlinear. The precision9

VddRARBVoutVoutVinVinFigure 2.1: Differential pair with offset measured at theoutput.of the comparator is also affected by the offset.2.3 KickbackDuring the regeneration process the latched comparator uses the positivefeedback mechanism to scale the digital level. The voltage variations at theregeneration nodes are coupled to the inputs and disturb the input voltages.This disturbance is called the kickback noise. There are many solutions to thisproblem [7], a few techniques are discussed in chapter 2.3.1 through 2.4.4.2.3.1 Sampling switchesA sampling circuit consists of a capacitor and switch (a MOS transistor)controlled by the clock. The sampling switch is placed before the comparatorinputs. During the regeneration phase these switches are opened anddisconnect the inputs from rest of the circuit [7]. The switches should be sizedas small as possible compared to the total capacitances at the inputs tominimize the effect of charge injection [1].10

Chapter 2 – Theory2.3.2 Isolation transistorsIsolation transistors isolate the input differential pair from the regenerationprocess. Isolation transistors are usually a set of NMOS transistors controlledby the clock and placed between the drain of differential pair andregeneration outputs. During the regeneration phase the isolation transistorsare switched off preventing the charge injection to the differential inputs. Thistechnique results in low kickback noise.2.3.3 Pre-amplifierThe pre-amplifier is the most commonly used solution placed in front of thecomparator to reduce the kickback effect. The pre-amplifier also amplifiesthe input difference and reduce the input-referred offset[2]. The pre-amplifiermay increase the gain and bandwidth of the system but power consumption isalso increased.2.3.4 Neutralization techniqueThis technique is used in the designs where differential inputs are directlyconnected to the regeneration nodes. Due to the non zero impedance of thecircuit preceding the comparator, the inputs of the comparator are disturbedby the drain voltage variations of the differential pair. By adding the twocapacitances between the gate and drain of the differential pair, as shown infigure 3.7, with a value equal to the Cgd of the differential pair will cancel thekickback noise [7]. This technique is further discussed in detail in chapter 3.2.4 ParasiticsThe parasitics play a critical role in analog designs. The ac behavior of theMOSFET is crucially effected of parasitics. The figure 2.2 shows a simplemodel to illustrate the parasitics of a MOSFET.Between every two of four nodes of MOSFET, there exists a capacitance. Thecapacitance depends upon the gate voltage and it changes values according tothe region of operations. The capacitances are; overlap capacitance betweengate and source/drain, depletion capacitance between channel and substrate,oxide capacitance between gate and channel and junction capacitance11

CgdDCdbCgbGCgsBCsbSFigure 2.2: MOSFET parasitics capacitances.between source/drain and substrate.2.5 MetastabilityNormally in all latching comparators metastability is a problem which occurswhen the input is near the comparator decision point [9].Comparator metastability occurs when very small signals appear at the inputof a comparator close to the comparator decision point. Normally all kind oflatching comparators exhibit this problem [9]. In such cases, the comparatoris not able to make a decision, i.e latch its output to the stable point, withinthe allotted time. This metastability delay is random and could switch theoutput to the wrong logical levels which can cause system malfunction orfailure. The figure 2.3 shows the voltage transfer characteristics of two backto-back connected inverters. Each inverter has two stable points; Vdd orground. The mid point where the two curves intercept each other ismetastable point (MSP) as shown in the figure 2.3. Ideally the MSP of aninverter is at half of the input range i.e. Vdd/2. Now, if the input at the firstinverter slightly deviates from Vdd/2, the output at the second inverter goesto one of the stable states. In this band of range the output is unpredictableand can switch to wrong logic level.12

Chapter 2 – Theorystable pointVoutmetastable pointstable point0VinFigure 2.3: VTC of back-to-back connected twoinverters.13

14

Chapter 3 – Design details3DESIGN DETAILSFor this thesis work I have selected one of the topologies presented in chapter1. The comparator design consists of two parts, the CMOS latch and SR latchas shown in the figure 3.1. There is no separate pre-amplifier in this design.VddVbias t2M8 M6M14M16M7 M9M17M15VssFigure 3.1: Comparator schematic3.1 CMOS latch circuitThe CMOS latch circuit includes the biasing part, differential andregeneration part, as shown in figure 3.2, followed by the SR latch. Theamplification is done by the PMOS differential pairs. In the following part wewill discuss the CMOS design in more detail.15

VddVbiasVinClkM2M1M4M5M6M7VrefOut Out-M8M3ID3ID2M9VssFigure 3.2: CMOS latch schematic.3.1.1 Comparator Optimizationa) Transistors M1-M3First, we will consider the PMOS differential pair [M1-M3]. As demonstratedin [5], the difference of current between differential and regenerations stagecan be calculated as:4IDkWI I D2 – I D3 --- ----- V ------------- – V 22LWk ----L(3.1)From the equation 3.1, the PMOS transistors sizing can have significanteffect on the comparator performance. Increasing the W/L ratio of PMOStransistors of differential pair{M1-M3} will produce the large I [4]. If Ibetween differential pair and regeneration pair is large, it will cause either M8or M9 to saturate for a small difference of input voltages (Vin and Vref). Inthis way the offset error can be reduced [4]. However if the I is too large,NMOS transistors will not be able to drive the SR latch for noticeable timeand SR latch will be disabled before the regeneration happens.16

Chapter 3 – Design detailsb) Transistors M4 & M5PMOS transistors M4 and M5 are controlled by the clock and act as cascodedevice. Since the design does not use a separate pre-amplifier, these cascodetransistors (also called isolation transistors) help to minimize the kickbackeffect by separating the inputs from the outputs during the regenerationprocess. These switches may limit the voltage swing and over load recovery.c) Transistors M6 & M7NMOS transistors M6 and M7 implement the switching transistors. Theswitching time of switching NMOS transistors is given by [8]V t WLC js1T t --- 2 ---------------------IDft(3.2)The equation 3.2 shows by decreasing the W/L ratio of switching transistorswill increase the switching time and also speed up the regeneration process.d) Transistors M8 & M9NMOS transistors M8 and M9 implement a regeneration circuit [4]. Thedrain current of the two cross coupled NMOS transistors affect directly theSR latch as well as the regeneration process. If the W/L ratio of the NMOStransistors is too large, it will produce more drain current which yields fastregeneration. However too much drain current will discharge both nodes atthe input of SR latch which will increase the offset voltage for properoperation [4].3.2 SR Latch circuitThe SR latch is shown in figure 3.3. The basic function of the latch is to act as17

memory that keeps values for a whole clock period. It may also add somegain to the outputs. The latch provides an interface between analog anddigital levels since the outputs of the comparator are digital. Otherwise ifanalog inputs are connected directly to the digital levels (the comparatoroutputs), the system becomes unstable. The digital levels can change quitemuch and can produce bounces even due to small noise spikes.The ratio of PMOS and NMOS sizes effect the resolution as well as the gainof comparator. The SR latch in this design has optimum transistor values forgood hysteresis M15VssFigure 3.3: Schematic diagram of SR latch3.3 Two phase operationThe comparator works in two phases, charging or amplification phase and regeneration phase during one cycle of comparison. Charging or amplificationhappens when the clock signal is low. During amplification phase the PMOScascode pair or isolation transistors M4 and M5 turn on while the switchingNMOS transistors M6 and M7 are disabled. Inputs are amplified and sampledat intermediate nodes (differential nodes).During the period when clock is high, the NMOS switching transistors M6 &M7 turn on and regeneration occurs. The differential nodes are discharged toGND.18

Chapter 3 – Design details3.4 Gain and bandwidth of the comparatorThe gain and bandwidth of the comparator depends on NMOS and PMOStransistors sizing as well as the biasing current to the comparator. By keepingthe configuration, input DC voltage as 1.65 V and biasing current 200 9A,yield the following results. Gain is 5.9dB, -3dB cut-off frequency 731 MHzand unity gain frequency is 1.0 GHz.3.5 Comparator turn off techniqueSince the comparator is intended to be used in a time-interleaved bandpasssign-delta modulator, it is turned off when not in use to avoid the static powerconsumption. To turn off the comparator, two extra NMOS transistors areintroduced as switches in the design as shown in the following figure 3.5.These NMOS switches are controlled by an external “enable” signal to turn“ON” and “OFF”. If the enable signal is low i.e. 0, both transistor MT1 andMT2 turn off and disconnect the CMOS latch from biasing circuit and SRlatch from GND. The comparator consumes almost no power when it isturned off.VddM1M0Enable ClkOut2M8 M6M17M15M14M16M7 M9VssMT2Figure 3.4: The comparator with turn off switches.19Enable

3.6 Settling time of the comparatorThe settling time is the time the comparator takes to settle, after it is turnedon, when the valid inputs are available at the input.Figure 3.5: Settling time of the comparator.The figure 3.6 shows the simulation waveform for the setting time of

1.2.1 A high-speed CMOS comparator with 8-bit resolution A high-speed CMOS comparator is shown in figure 1.2. The comparator consists of three blocks, an input stage, a flip-flop and SR latch. The architecture uses two non-overlapping clocks ( 1and 2). The circuit operates in two m

Related Documents:

A: Yes, High-Speed option may reduce the performance of other network service, but High-Speed option controls QoS(Quality of Service) and protects other important services. Q3: Is every file transfer speed improved with High-Speed option? A: High-Speed option improves only for users who have High-Speed option.

Manual Camry 4-Cyl 6-Speed Automatic LE 4-Cyl 6-Speed Manual LE 4-Cyl 6-Speed Auto SE 4-Cyl 6-Speed Manual SE 4-Cyl 6-Speed Auto LE V6 6-Speed Auto SE V6 6-Speed Auto XLE 4-Cyl 6-Speed Auto Hybrid 4-Cyl ECVT XLE V6 6-Speed Auto MSRP** 19,395 20,445 20,850 21,900 22,165 23,165 24,565 25,840 25,925 26,150 29,045 MPG

high-speed tool steels. The mold parts of both steel types were manufactured in the same way from heat treatment to machining, and the powder high-speed tool steel was 66HRC and the high-speed tool steel was 61HRC. As a result of the experiment, it was observed that the number of punching of powder high-speed tool steel was improved by 40-50%,

A speed study will determine the speed limit. A speed study now is likely to result in a much higher speed limit, based on the high design speed for the street. In order for a speed limit to be effective and promote safe vehicle travel, the Corrine Drive Recommended Design needs to be implemented first. For the speed limit to be lowered, cars

2010 COROLLA Standard Features and Options* Corolla 1.8L 4-cyl 5-Speed Manual Corolla 1.8L 4-cyl 4-Speed Auto S 1.8L 4-cyl 5-Speed Manual LE 1.8L 4-cyl 4-Speed Auto S 1.8L 4-cyl 4-Speed Auto XLE 1.8L 4-cyl 4-Speed Auto XRS 2.4L 4-cyl 5-Speed Manual XRS 2.4L 4-cyl 5-Speed Auto Cup holders (two front and two rear) Fabric-trimmed 4-way adjustable .

6-speed Manual 5C12N2 5C12M2 - 6-speed DSG Auto. 5C12N3 5C12M3 - 2.0T Beetle R-Line 6-speed Manual 5C1RN2 5C1RM2 - 6-speed DSG Auto. 5C1RN3 5C1RM3 - 2.0L Beetle TDI 6-speed Manual 5C119L 23,495 22,555 6-speed DSG Auto. 5C119M 24,595 23,611 2.0L Beetle TDI w/ Sunroof 6-speed Manual 5C129L 25,195 24,187 6-speed DSG Auto. 5C129M .

Transmission Automatic 6-speed DSG Manual 6-speed Automatic 6-speed DSG Automatic 7-speed DSG Automatic 7-speed DSG Manual 6-speed Automatic 7-speed DSG Automatic 7-speed DSG WEIGHT Kerb weight - in standard version with a 75kg driver (kg) 1,561 (1,604) 1,615 (1,658) 1,630 (1,673) 1,695 (1,738) 1,667 (1,710) 1,705 (1,748) 1,740 (1,783) 1,752 .

2.2 Speed limits in New South Wales 11 2.3 key factors in setting of speed limits 15 2.4 Speed review procedures 17 2.5 10 step process: speed zone review 18 SECTION 3: rEfErENCE INfOrmaTION TypES aNd SIgNpOSTINg Of SpEEd lImITS 22 3.1 Casualty concentrations and rates 22 3.2 Types of speed limits 23 3.3 Signposting of speed zones 29