6:53 PM | 0.6KB/s B ALL ABOUT : CIRCUITS When discussing MOS transistors, there are basically six types of leakage current components in short channel devices: Reverse bias-pn junction leakage current ¢ Subthreshold leakage current ¢ Drain-induced barrier lowering © Vin roll off * Effect of operating temperature ¢ Tunneling into and through gate oxide leakage current ¢ Leakage current due to hot carrier injection from the substrate to gate oxide ¢ Leakage current due to gate-induced drain lowering (GIDL) Before continuing, be sure you're familiar with the basic concepts of MOS transistors that will prepare you for the following information. 1. Reverse-Bias pn Junction Leakage Current Scanned with CamScanner6:53 PM | 3.3KB/s AB (CO |circuits Qeaes 1. Reverse-Bias pn Junction Leakage Current The drain/source and substrate junctions in a MOS transistor are reverse biased during transistor operation. This results in reverse- biased leakage current in the device. This leakage current can be due to drift/diffusion of minority carriers in the reverse-biased region and electron-hole pair generation due to the avalanche effect. The pn junction reverse-biased leakage current depends on doping concentration and junction area. For heavily doped pn junction of drain/source and substrate regions, the band-to-band tunneling (BTBT) effect dominates the reverse bias leakage current. In band-to-band tunneling, electrons tunnel directly from the valence band of the p region to the conduction band of the n region. BTBT is visible for electric fields greater than 10° V/cm. Scanned with CamScanner6:54 PM | 22.4KB/s Z Biul @D OO |circtits = n-side Figure 1. Band-to-band tunneling in reverse-biased pn junction of a MOS transistor. All images used courtesy of K.Roy, et al, “Leakage current mechanisms and leakage reduction techniques in deep-submicrometer CMOS circuits"; Proc. of IEEE, Vol. 91, No. 2, Feb 2003. Note that, in the context of this article, we define tunneling phenomena as taking place even when the energy of the electron is much lesser than the potential barrier. 2. Subthreshold Leakage Current When the gate voltage is less than the Scanned with CamScanner6:54 PM | 2.8KB/s Zo ul Go OO lcirctits Bi 2. Subthreshold Leakage Current When the gate voltage is less than the threshold voltage (Vj,) but greater than zero, the transistor is said to be biased in the subthreshold or weak inversion region. In weak inversion, the concentration of minority carriers is small but not zero. In such a case, for typical values of |Vps| > 0.1V and the entire voltage drop takes place across the drain- substrate pn junction. The electric field component between the drain and source, parallel to the Si-Si02 interface, is small. Due to this negligible electric field, the drift current is negligible and the subthreshold current mainly consists of diffusion current. Drain-Induced Barrier Lowering (DIBL) Subthreshold leakage current is mainly due to Scanned with CamScanner6:54 PM | 4.4KB/s B Clcirctirs Drain-Induced Barrier Lowering (DIBL) Subthreshold leakage current is mainly due to drain-induced barrier lowering or DIBL. In short channel devices, the depletion region of drain and source interact with each other and reduce the potential barrier at the source. The source is then able to inject charge carriers into the surface of the channel resulting in subthreshold leakage current. DIBL is pronounced in high drain voltages and short channel devices. Veh Roll Off The threshold voltage of MOS devices reduces due to channel length reduction. This phenomenon is called Vip roll-off (or threshold voltage roll-off). In short channel devices, the drain and source depletion region enter further into the channel length, depleting a part of the channel. Due to this, a lesser gate voltage is required to invert the channel reducing the threshold Scanned with CamScanner6:54 PM | 0.0KB/s B (CF | circuits Ven Roll Off The threshold voltage of MOS devices reduces due to channel length reduction. This phenomenon is called Vip roll-off (or threshold voltage roll-off). In short channel devices, the drain and source depletion region enter further into the channel length, depleting a part of the channel. Due to this, a lesser gate voltage is required to invert the channel reducing the threshold voltage. This phenomenon is pronounced for higher drain voltages. The reduction in threshold voltages increases the subthreshold leakage current as the subthreshold current is inversely proportional to the threshold voltage. Effect of Operating Temperature Temperature also plays a part in leakage current. Threshold voltage decreases with increasing temperature. Or, in other words, subthreshold current increases with Scanned with CamScanner6:54 PM | 0.0KB/s B © |circuits Effect of Operating Temperature Temperature also plays a part in leakage current. Threshold voltage decreases with increasing temperature. Or, in other words, subthreshold current increases with increasing temperature. 3. Tunneling into and Through Gate Oxide Leakage Current In short channel devices, a thin gate oxide results in high electric fields across the SiOz layer. Low oxide thickness with high electric fields results in electrons tunneling from the substrate to the gate and from the gate to the substrate through the gate oxide, resulting in gate oxide tunneling current. Consider the energy band diagrams as shown. Scanned with CamScanner6:54 PM | 0.0KB/s |circuits 3. Tunneling into and Through Gate Oxide Leakage Current In short channel devices, a thin gate oxide results in high electric fields across the SiOz layer. Low oxide thickness with high electric fields results in electrons tunneling from the substrate to the gate and from the gate to the substrate through the gate oxide, resulting in gate oxide tunneling current. Consider the energy band diagrams as shown. fi Ty SS @) o © Figure 2. Energy band diagrams of MOS transistors with (a) flat band, (b) a positive gate voltage, and (c) a negative gate voltage Scanned with CamScanner6:54 PM | 6.5KB/s Zo ALL ABOUT — {F | circuits QBs The first diagram, Figure 2(a), is of a flat band MOS transistor, i.e., where there is no charge present in it. When the gate terminal is positively biased, the energy band diagram changes as shown in the second diagram, Figure 2(b). The electrons at the strongly inverted surface tunnel into or through the SiOz layer giving rise to gate current. On the other hand, when a negative gate voltage is applied, electrons from the n+ polysilicon gate tunnel into or through the SiOz layer giving rise to gate current, as shown in Figure 2(c). Fowler-Nordheim Tunneling and Direct Tunneling There are primarily two types of tunneling mechanisms between the gate and the substrate. They are: * Fowler-Nordheim tunneling, where the electrons tunnel through a triangular Scanned with CamScanner6:54 PM | 51.4KB/s Zi © lcircuits Bi Fowler-Nordheim Tunneling and Direct Tunneling There are primarily two types of tunneling mechanisms between the gate and the substrate. They are: * Fowler-Nordheim tunneling, where the electrons tunnel through a triangular potential barrier * Direct tunneling, where the electrons tunnel through a trapezoidal potential barrier “f je “ET : Ey & substrate E p substrate Ey ca nt poly-silicon n+ poly-silicon (a) (b) Figure 3. Energy band diagrams showing (a) Fowler- Nordheim tunneling through triangular potential barrier of the oxide and (b) Direct tunneling through trapezoidal potential barrier of the oxide You can see the energy band diagrams for Scanned with CamScanner6:54 PM | 2.3KB/s BH Bou ( lcirctirs Ai You can see the energy band diagrams for both tunneling mechanisms in Figures 3(a) and 3(b) above. 4, Leakage Current Due to Hot Carrier Injection from the Substrate to Gate Oxide In short channel devices, the high electric field near the substrate-oxide interface energizes the electrons or holes and they cross the substrate-oxide interface to enter the oxide layer. This phenomenon is known as hot carrier injection. p-substrate Scanned with CamScanner6:54 PM | 43.0KB/s B Bi OO lcircuirs Bi p-substrate n+ poly-silicon Figure 4. Energy band diagram depicting electrons gaining sufficient energy due to high electric field and crossing over the oxide barrier potential (hot carrier injection effect) This phenomenon is more likely to affect electrons than holes. This is because electrons have a lesser effective mass and a lesser barrier height as compared to holes. 5. Leakage Current Due to Gate-Induced Drain Lowering (GIDL) Scanned with CamScanner6:54 PM | 0.2KB/s ALL ABOUT = (CO |circuits Qeas 5. Leakage Current Due q to Gate-Induced Drain Lowering (GIDL) Consider an NMOS transistor with a p-type substrate. When there is a negative voltage at the gate terminal, positive charges accumulate just at the oxide-substrate interface. Due to the accumulated holes at the substrate, the surface behaves as a p-region more heavily doped than the substrate. This results in a thinner depletion region at the surface along the drain-substrate interface (when compared to the thickness of the depletion region in the bulk). Va>0 Vsg0 Ves0 Figure 5. (a) Formation of thin depletion region at the dlenin mithatentn intarénan alanc tha jcanned with CamScanner6:54 PM | 0.0KB/s OUT = ( |circtits Qas Figure 5. (a) Formation of thin depletion region at the drain-substrate interface along the surface and (b) flow of GIDL current due to carriers generated by avalanche effect and BTBT Due to a thin depletion region and higher electric fields, the avalanche effect and band- to-band tunneling (as discussed in the first section of this article) take place. Thus, minority carriers in the drain region underneath the gate are generated and are pushed into the substrate by the negative gate voltage. This adds to the leakage current. 6. Leakage Current Due to Punch-Through Effect Scanned with CamScanner6:54 PM | 0.0KB/s K x @ 5 Causes of MOS Trans... ie allaboutcircuits.com ( lcirctits A: the substrate by the negative gate voltage. This adds to the leakage current. 6. Leakage Current Due to Punch-Through Effect In short channel devices, due to the proximity of drain and source terminals, the depletion region of both the terminals come together and eventually merge. In such a condition, “punch-through’ is said to have taken place. The punch-through effect lowers the potential barrier for the majority of carriers from the source. This increases the number of carriers entering into the substrate. Some of these carriers are collected by the drain and the rest contribute to leakage current. You should now be familiar with six types of leakage current associated with MOS transistors. If you have additional questions nhautthaaaaanan inte nlananlacuaa Scanned with CamScanner‘What is Pipelining? Pipelining is the proces of accumulating instracton from the provessor through & pipeline. Ieallows storing ad executing instructions in an onrly process 1s also known as pipeline processing Pipelining technique where mulkipl instructions are overipp ding execution, Pipeline i divided int stages and these stages ate connected with one another to forma pipe like structure. Tnstuctons enter from one end and exit from nother ee Pipcining increases the overall iastue throughput In pipaine system, each segment consists of an input register followed by & ‘combinational ecu. The register is wed to old data aed combinational circuit performs operations oni. The output of combinational criti applied to the {pu egistr of the next segment Pipeline system s lik the moder day assembly ine setup in factories. For example ina car manufacturing indasr. ge assembly Tines ae setup ad at each pin, there are robotic arms to perform a contain task, andthe tho ear moves om Sead tothe next arm This divided into 2 categories 1, Aithmeti Pipeline 2 Instruction Pipeline Avitheti Pipeline ‘Arithmetic pipeline are usually found in most of he computer. They ae wed or foating post operations, mulplication of fixed point numbers et, For example The input othe Foating Point Ader pipeline i ‘ete Aand B are mantsss (significant digit of oating point numbers), ‘while and bare exponents, cating point ation and subtraction is done in 4 pas 1, Compare the exponents 2 Align the mantissa. 3. Add or subract mantssas 44 Produce the resul Registers are wed for storing the intermediate results between th operations Scanned with CamScannerTypes of Pipeline It is divided into 2 categories: 1. Arithmetic Pipeline 2. Instruction Pipeline Arithmetic Pipeline Arithmetic pipelines are usually found in most of the computers. They are used for floating point operations, multiplication of fixed point numbers etc. For example: The input to the Floating Point Adder pipeline is: Here A and B are mantissas (significant digit of floating point numbers), while a and b are exponents The floating point addition and subtraction is done in 4 parts: 1. Compare the exponents Align the mantissas Add or subtract mantissas 4, Produce the result. Registers are used for storing the intermediate results between the above operations. Instruction Pipeline In this a stream of instructions can be executed by overlapping fetch, decode and execute phases of an instruction cycle. This type of technique is used to increase the throughput of the computer system. Scanned with CamScannerScanned with CamScannerScanned with CamScannerScanned with CamScannerrat Scanned with CamScannerrear < Scanned with CamScannerSherali Thre, ou wpe wel cmos ee ed rts pre. “Anata aise ow kowerccan (2) choy, hoe te ee ulema on pet of ht he wi 12 ar dy of Tow: vies assay power common of hgh devs, 2 laced kof bth tne and hig W devices maybe ol. Te following te reaches cane wed ce eahage power dspace andy mode 83, 7 VTCMOS Approach 1 VTCMOS cic mike we of the boy effect reve the subiesd leakage cure, when the cic sin al ae ¥ sereMos Appr [> Inthiapoch MOSFETs wih wo Siforethewold ogee sd in a single cp © Wtses to operon! mades—aene and slep for fit power ‘management Des Vt Assignmese App (DTCMOS) ‘VTCMOS and MTCMOS Approach ln cae of VICMOS, tsk pipe is Wo at Urea voage by chaning brat bie Tratr ily have Tow ih ing ner operation and ‘Sst bs alee wig subst as com ict The veo eee iy sing evens body Nts whe Gh crcl oct tn we. Elicit, ‘Ealge power dsipaton in stndy pode andi valved aditonal wea and higher ‘rt amples. Sata pont con speach. On te eer hand, tase MEMOS aypoach MOS transistors of wap retold voles faked in which power gg wai ser the Sack beeen the loge transistors ander power or rou, ths ceting ial Spy ral ora val ground a, pec. The lope Hock enti al ew tears for fast siching speeds whe the sich Vnssrs eat a ots fre bal ang hgh Vin wane oes the lenge per Gon Sa, ita preslcon aprox Standby and Runtime Leakage Power ‘Stanby nage power distin lkes pice when the cc sw ia we, be ings do ot change an clock me pled. ‘On thee and une Tlekage power dsption aks pace when the cu is ting wed ‘SAAdlabatic Logi Circuits Sie complenestry oul exe semiconduce (CMOS) cies ar extenely soxcenfal nverm of mur re Denon of wy vais ach ower power ‘sipion, real peraon and avalaisy of compatraded design (CAD) ‘ote to + Werhve sen th al he cic vs make a io (an VM nso fe ‘xhsiching even an th supply vgs Vi remains conta As a coneaeee he ouput hoge makes tanston fom Ot Vd with «oad ‘Spactance Cent of C's des rm th power pp) 6/9 Out ofthis, 24. 8 estore in the capi andthe rsinng hal is dsp Inthe pope meta-onkscmaconducor MOS) ive + Sobeqacty non the opt nde sche rem Vil to Othe ery tha wat Soren the apicor dsm fe iype met-oxie Semon (MOS) “The power dition hat takes place ects ofthese ning event converted to bea which is ulna rlased fo the ertonmee. This as fa teaching ‘Sasoyemcat Iie gatal werting. To sole prvor iain Seah digas “cede the soppy vlage dere the me Capaciane cr mii the aber ‘stung event tae receling cater we ave dnc these pens ¥ ‘A-ovel apc Yo achive ener snption lw thi lowe it aC * ‘Tb he reed ina ew clan Sf creats how os“ eel’ Adabate ‘sichin sac ve!aproch tht has made posi vo elie he lrom= fpover compoting applications Wabout sag the supply ollage The tm Stab” retro he thrmedmamic pers ta extagen Hel with te cvrnnent The cece charge sr between yas eat noes can be ‘Sianisred a he press Ih adaue CMOS ici, he energy cmsompon it ‘minimied ty’'slowing Gown the charge wnsprtfetcn the dai ad sure termina of the mal -osde semicon lft rane (MOSFET) sith sd mcoverng th energy witout sping es Conyenionalcargag ofa opaciter C trough ressor "7" Conveaonal charging oe cepa Ia he dition ofan eery of 1aav'ss 0 Ww R Scanned with CamScanner54 Sic complementary me-ondesemicondut (CMOS) ccs are extenely cou terme of market sue boaav of many avantagos ack wer wet ‘sipation.rlble operon and valbty of computraded dst (CAD) ‘oan 17 Wetuve scent ll te cc aes make exo (and Vd sin fo ‘xh swiching even andthe up voltage eld emis const sa consequence, the oor tae kes a tain from Oto Vd wih 9 toad ‘ipcince Cam nergy of CVs dn ot he poe Su ¥ Ont of tis, V2C.V 7A sored inthe apace he emiing al ssid Inthe pype meta onesemicendacter @MOS) setwork Sane ene cpa node sch rom Va 0, the energy tht was red nh apart them type metal oxie-smcondtr (MOS) \ Thepomer dspnton hl hes lace Hea of hee sithing event convened {erat which tiny rleanod wo the exnoneat. This fs fa-raching “Snsequcnes ke patel warning Te rluse power spain the cca desis ‘Simrahice the vopply vlige daca he ode caste or aime The mb ‘tomiching eer othe rceing capers, we hve cued tee spreac ‘7 Naunel approach achieve every sition Plow ths lowe i of C= ‘Tash mead ia nw ie fie kena ci Abate richie! approach ht ae possible elie the lw Ppt computing aplkatos witht sclg the apy voltage The tem aa” relr oe temdynaic process at exchag> no eat wih the honest. "The elecie charge inser between vais cca nodes Can De ‘Sneed the proces In adabe CMOS cul, the eat comin is ‘aimed by shoving. down the charge Wanspor tween the att and Source ‘Some of the tl-onde semiconductor Bl-etettassor (MIOSFET) switch Sitecrwering the energy without siping as eat ‘Conventional charging of a apart C through a resistor 7 Comentonalcirig capac Teds oe sition of an every of races feceereeel wo Y R vad 8 Adiabatic charsing of capacitor S| i 1h dis Scanned with CamScannerAdiabatic Amplification if i A ~< ~ characteristics The following battery characteristics must be taken into consideration when selecting a battery: * Type Voltage Discharge curve Capacity Energy density Specific energy density Power density Temperature dependence Service life Physical requirements Charge/discharge cycle Cycle life Ost bility to deep discharge * Application requirements . lo | . [> | 1) Type See primary and secondary batteries page. 2 Maltana Scanned with CamScanner7:16 PM | 6.6KB/s Zi all es CD 1)Type See primary and secondary batteries page. 2) Voltage The theoretical standard cell voltage can be determined from the electrochemical series using E° values: E° (cathodic) — E° (anodic) = E° (cell) This is the standard theoretical voltage. The theoretical cell voltage is modified by the Nernst equation, which takes into account the non-standard state of the reacting component. The Nernstian potential will change with time either because of use or self-discharge by which the activity (or concentration) of the electro-active component in the cell is modified. Thus the nominal voltage is determined by the cell chemistry at any given point of time. The actual voltage produce will always be lower than the theoretical voltage due to polarisation and the resistance losses (IR drop) of the battery and is dependent upon the load current and the internal impedance of the cell. These factors are dependent upon electrode kinetics and thus vary with Scanned with CamScanner7:16 PM | 0.3KB/s Ki 2)Voltage ‘The theoretical standard cell voltage can be determined from the electrochemical series using F° values: (cathodic) ~ E° (anodic) = E° (cell) This is the standard theoretical voltage. The theoretical cell voltage is modified by the Nemst equation, which takes into account the non-standard state of the reacting component. ‘The Nemstian potential will change with time either because of use or self-discharge by which the activity (or concentration) of the electro-active component in the cell is ‘modified. Thus the nominal voltage is, determined by the cell chemistry at any given point of time, ‘The actual voltage produce will always be lower than the theoretical voltage due to polarisation and the resistance losses (IR drop) of the battery and is dependent upon the load current and the internal impedance of the cell. These factors are dependent upon electrode kinetics and thus vary with temperature, state of charge, and with the age of the cell, The actual voltage appearing at the terminal needs to be sufficient for the intended application. ‘Typical values of voltage range from 1.2V fora Ni/Cd battery to 3.7 V for a Lion battery. The following graph shows the difference between the theoretical and actual voltages for various battery systems: Scanned with CamScanner7:16 PM | 0.8KB/s all Yes 3) Discharge Curve The discharge curve is a plot of voltage against percentage of capacity discharged. A flat discharge curve is desirable as this means that the voltage remains constant as the battery is Used up, 4) Capacity. The theoretical capacity of a battery is the quantity of electricity involved in the electro- chemical reaction. Itis denoted Q and is given by: Q=anF where x= number of moles of reaction, n= number of electrons transferred per mole of reaction and F = Faraday's constant The capacity is usually given in terms of mass, not the number of moles: jolecular Mass. This gives the capacity in units of Ampere-hours per gram (Ah/g), In practice, the full battery capacity could never be realised, as there is significant weight contribution from non-reactive components such as binders & conducting particles, separators & electrolytes and current collectors & substrates as well as packaging Typical values range from 0.26 Ah/g for Pb to 26.59 Ah/g for Ho. 8) Eneray density, The energy density is the energy that can be derived peer unit volume of the weight of the cell 18) Specific energy density, The specific energy density is the energy that can be derived per unit weight of the cell (or sometimes per unit weight of the active Scanned with CamScanner7:16 PM | 0.1KB/s al Xi GD ‘).Energy density, ‘The energy density is the energy that can be derived peer unit volume of the weight of the cell ‘8) Specific energy density, ‘The specific energy density is the energy that ccan be derived per unit weight of the cell (or ‘sometimes per unit weight of the active electrode material) It is the product of the ‘specific capacity and the operating voltage in cone full discharge cycle. Both the current and the voltage may vary within a discharge cycle land thus the specific energy derived is calculated by integrating the product of current ‘and voltage over time. The discharge time is related to the maximum and minimum voltage threshold and is dependent upon the state of availability of the active materials and/or the avoidance of an irreversible state for a rechargeable battery. D Bower density The power density s the power that can be derived per unit weight of the cell (W/kg) 8) Temperature dependence The rate of the reaction inthe cell will be temperature dependant according to theories of kinetics. The internal resistance also varies with temperature; low temperatures give higher intemal resistance. At very low temperatures the electrolyte may freeze giving a lower voltage as ion movement is impeded. At very high temperatures the chemicals may decompose, or there may be enough energy available to activate unwanted, reversible reactions, reducing the capacity ‘The rate of decrease of voltage with increasing discharge will also be higher at lower temperatures, as will he capacity- this is illustrated by the following graph: EEE a Scanned with CamScanner