 Good morning, I will start my second lecture, but before that let me recapitulate the important concept that I discussed in my last lecture. After the brief introduction, I discussed diodes, the very important parameter in high frequency conversion is, one is the reverse recovery current and the reverse recovery time. These two are very important. How these parameters are used in design process? I will discuss at the appropriate time. Let me cover DC to DC conversion, let me and DC to AC conversion and when we are doing design of these converters towards the end, I will tell you how to take care of these parameters. Then I discussed about a BJT. Generally, NPN types are popular. Though the popularity of, though this BJTs are not very popular, they were popular in the 80s, I wanted you to appreciate, in order to appreciate MOSFET and IGBT, one need to know BJT. The problem with the BJT is, we found that as the power rating increases, the gain decreases. Basically, BJT is a current fed device. So, that if the gain comes down, the base current required to drive the BJT increases. So, therefore, in my control circuit or in my base drive circuit, I may require, I am using the word may, I may require a small power transistor to drive another power transistor. Basically, BJT is a current fed device and gain falls as the power rating increases. So, generally the BJT is operated in quasi saturation region. See, BJT has an excellent output characteristics. Output characteristics in the sense, during conduction, I can reduce or I can control the on-state voltage drop. I can reduce it. I can bring down to as low as VCE sat. Fine, my switching frequency may come down. That is the second issue. Then, it is possible to reduce the on-state conduction, on-state losses. The problem with BJT is, it is a current fed device and I require a large gate drive current. Now, let us see another device which is very popular is, especially in DC to DC conversion is MOSFET, power MOSFET. So, in 1978, 25 ampere power MOSFET invented. As of now, see the current rating, 200 volts, 500 ampere, semi-chron MOSFETs are available, 60 volts, 1000 ampere by semi-chron. These are available. See, low voltage, high current. Generally, MOSFETs are used in relatively low voltage at the high current application. Basically, they are very fast devices. Very popular in DC to DC conversion, MOSFET stands for metal oxide semiconductor field, in fact transistor, very fast device. It is a majority carrier device. It is a majority carrier device. Yeah, like BJT is again a non-latching device. I need to have a continuous gate drive. But what sort of a gate drive? We will see. The three terminals that we use in a BJT is base emitter collector, whereas in MOSFET, drain, source and gate. And gate signal is generally applied between gate and source. See the structure. We have an N plus, highly doped, N minus, lightly doped, P junction and N plus junction, gate, source and drain. Now, let us see, we will start from, current flow is from drain to source. We will start from drain to source. Let us see what happens. There is a N, P. So, see here D terminal is here N, cathode of a diode. There is anode here. Then again P N, again N, again N, P N and we reach the source. So, if you see two diodes, if I just see the structure, there are two diodes are connected back to back. So, looks like there cannot be a current flow. If I just see the structure, because two diodes are connected back to back, then, but then MOSFETs are do used in DC-DC conversion. Therefore, it has to conduct. So, how is that possible? Now, let us see what happens when I apply a suitable biasing voltage. See, gate is insulated from the rest of the device. I am telling, gate is insulated from the rest of the device. So, therefore, there is no steady state current, something equivalent to a parallel plate capacitor. So, it is only a displacement current can flow like a parallel plate capacitor. So, MOSFET is in cutoff when gate to source voltage is less than the threshold value. Again, this threshold value one need to see from the data sheet depending upon the voltage and power level. This value does change, but then when the applied voltage V G is greater than the threshold value, what happens is this. A small channel is formed between, I will tell you, a channel is formed between it and source. A channel, once the channel is formed, the electron starts flowing and therefore, the current. When the applied voltage is higher than the gate to threshold value, a channel is formed, electron starts flowing and therefore, the current. So, on conduction or during conduction, how does the MOSFET look like? It is like a flow of water through pipe, it faces only the resistance. So, therefore, during conduction MOSFET appears as if it is a resistive element. So, on state losses are I D square, I D is the drain current into R D S on drain to source resistance R D S on. So, if I compare BJT and MOSFET during conduction mode, on state losses in a MOSFET are slightly higher because I told you it is appears as a purely resistive element R D S on. So, the if I draw the omic, the V I care to say they look something like this, something similar to a BJT active and this region is known as the omic region. V and I is approximately a straight line, approximately a light line. The slope of this is the R D S on, R D S on. So, generally we operate the MOSFET in this zone. See here on on state, the channel of the device behaves like a resistance is R D S on, it is the slope of V D S divided by delta V D S divided by delta I D conduction. During conduction, conduction power loss is given by I D square into R D S on. So, outputs like for a for a for a given power level, if I use a BJT, on state power losses are slightly less. So, in other words, output state of a BJT is slightly superior compared to that of a MOSFET. Now, what happens to the input stage? In other words, in other words, the driver circuit. I told you here the structure looks gate is gate and source are isolated. There is gate is insulated from the rest of the device insulated. So, if I apply a potential across it, only the displacement current can flow, something similar to a parallel place capacitor. So, how do I turn it on? By suitably applying a voltage across a gate to source. In other words, I need to charge, I have to charge the gate to source capacitor. I will repeat, I have to charge the gate to source capacitor. So, input impedance at the gate source is tends to infinity. So, in other words, I require a very small steady state current, very small, practically negligible. It is a voltage driven device. It is not a voltage fed device. It is not a current fed device unlike a BJT. So, how do I increase the turn on time of a BJT? I said turn on time of a BJT can be decreased by supplying a high pulse of base current during turn on. So, during turn on period you supply a slightly high pulse current, whereas in a MOSFET, faster that I charge the capacitor, the faster I charge the capacitor between gate to source, faster is the turn on. So, by the way, how do I charge the capacitor at a very fast rate? So, maybe I need to supply initially a pulse current. That current has to flow from the control circuit. Fortunately, that is only during starting this, the charging of the capacitor that is very small. Of course, one way how do I charge this capacitor? Suppose I have, see here I have a problem. If I take gate and source, if this is a voltage source, the rate at which I charge this capacitor will determine the T on. So, if I, of course, I need to connect a very small resistance to limit the starting current. Otherwise, current is limited by the so called the internal resistance of this voltage source. So, the value of this R, external R will determine the charging rate. Smaller the resistance, faster is the turn on, but the smaller the resistance, higher is the peak pulse current that has to flow through this source and this may be a switch. Of course, this switch is generally realized by gates. So, you need to take care of while selecting this R. Maybe sometime later I will discuss the gate drive circuit and how to choose this value of R. So, gate drive requirement for a MOSFET is very minimal compared to or the power rating of a gate of the gate drive circuit for a MOSFET is very small compared to that of a BJT. Basically so happens that it is a voltage fed device and BJT happens like a current fed device. By the way, if you see the structure here, structure here, there is a inherent body diode between drain and source. Something like this, if you see the structure, the MOSFET has a built in body diode, but then what is being done is the speeds are not very comparable with that of a MOSFET. So, generally an external body diode during fabrication, during fabrication, during fabrication this diode is incorporated there. So, how is the characteristic similar to that of a MOSFET? Safe operating area of a MOSFET is a majority carrier device. So, therefore, it has a positive resistance coefficient, positive resistance coefficient. In other words, as the temperature increases, resistance increases. So, therefore, paralleling is easy compared to that of a BJT. BJT paralleling is difficult because it is a minority carrier device and it has a negative resistance coefficient. And if I see the safe operating area of a BJT, there are four zones. They are current limit, voltage limit, power limit and the fourth one is the secondary breakdown. Secondary breakdown is due to the negative resistance coefficient. Whereas, the secondary breakdown is absent in MOS. There are only three zones, current limit, voltage limit and this is the junction temperature limit. Just show you some of the parameters of a MOSFET IRF 640. See, for a drain current of 18 amperes, RDS on is of the order of 0.18 ohms. So, on state power loss is assuming that drain current is 18 amperes. Of course, 18 is the rated current. We may not be able to supply that it may not be able to carry 18 amperes during normal condition. So, I DS square 18 square into RDS on. So, that is the on state power loss and one has to choose the heat sink accordingly. Now, see here the gate to threshold voltage to turn it on is of the order of plus or minus 20 volts. Plus or minus 20 volts to turn it on and to turn it off. See, here the timings turn on delay 14 nanoseconds, rise time 51 nanoseconds, turn off delay 45 nanoseconds and fall time is of the order of 36 nanoseconds. You just compare yesterday I did I think I gave you the characteristics of a 8 ampere BJT and today you compare I am discussing the 18 ampere MOSFET and compare the timings turn on delay. They are in terms of nanoseconds and see here gate to source charge is 13 nanoculums from there you can calculate the capricent and therefore, the current that is required to charge that capacitance. So, if I compare the difference between a MOSFET and a BJT current control device is a voltage control device minority carrier device is a majority carrier device. Therefore, negative resistance and a positive resistance coefficient has secondary breakdown no secondary breakdown paralleling is difficult paralleling is easy on state power loss B C sat into I C is low I D squared R D S on is higher than on state loss of BJT turn off time is higher because if I saturate it as what is known as a storage time plus the fall time whereas, it is a very fast device gate draw requirement as minimum and MOSFET and you are much higher compared to in BJT is much higher compared to MOS. So, the problem with the MOS is the on state power loss is there a way out yes what is there is what is known as a cool MOS as the name suggests it has a very low on state resistance conduction losses are very low, but then they are very expensive. So, that is about MOSFET. So, we have discussed BJT and MOSFET BJT seem to be having excellent output characteristics in other or the on state power loss seem to be low, but then gate draw requirement is high whereas, MOS seem to have an excellent input characteristics input impedance at the gate is really tends to infinity. So, gate draw requirement is minimum there, but then during conduction on state power losses are high. So, what could be an ideal device according to me in the ideal device is a device which is having the input characteristics similar to that of a MOS and output characteristics similar to that of a BJT. So, therefore, hence the name insulated gate bipolar transistor and IGBT insulated gate something similar to MOSFET output is a BJT bipolar junction transistor. I said in insulated gate in MOS structure current is only due to the majority carrier device, but then in BJT sorry in IGBT BT stands for bipolar junction transistor. So, there is going to be I need to modify the existing MOS structure. So, what price that I am paying we will see by the way this was invented by Jainth Baliga in 1983. Here is the structure of an IGBT I have added or there is another P structure here P layer n plus n minus P plus n minus this is the MOS structure there is an additional P layer P plus. So, what happens when I apply a suitable gate to source voltage here what will happen now poles start moving towards n electrons start flowing to n minus region that is. So, therefore, n minus layer receives electrons from n plus as well as poles from P plus some of them recombine and remaining electrons are collected at the source. So, junctions of poles are injected in n minus region some electrons combined with the holes remaining holes are collected at the source. Now, because of this P structure we have a P plus and n plus structure. So, junction J 1 now can block a negative voltage come on in the sense this voltage if I apply a negative voltage it can this junction J 1 if I connect drain to a negative of the source and if I make source positive this junction gets reverse biased and therefore, it can block a negative voltage can block a negative voltage. But then if you see the structure I have a P plus and n plus. So, plus indicates is a highly doped. So, therefore, the reverse breakdown voltage of this junction is relatively low relatively low. So, the V I characteristics something similar to that of V J T, but then we had a, but then control parameter is the gate current here it is a V G S gate to source voltage this is for the comparison between I G B T and a MOS see the modified structure see the slope of this line and the slope of this line. So, on straight losses in I G B T are slightly less compared to that of a MOS by the way coming back to the structure do I require this n plus layer what if I or what happens if you remove this n plus layer it will still work like an I G B T same thing P n P n now there is P plus n plus n minus P plus n minus. So, if I if there is no n plus layer I have only P plus and n minus n minus indicates is a lightly doped. So, so if that is the case this junction J 1 can block a higher voltage. Now, due to this n plus layer this J 1 cannot block a higher voltage. So, n plus layer between P plus and n minus drift is not essential for operation of an I G B T some I have just n minus layer. So, those are known as non punch through non punch NPT I G B T. So, if both are present n plus as well as n minus are present they are known as punch through I G B T's punch through I G B T's, but then what are the advantages of NPT NPT. So, non punch through wherein there is no n plus it can block a higher voltage. Now, so it is almost symmetrical both it can block positive as well as negative voltage it is what is known as a symmetrical device. Symmetrical device whereas, whereas if there is a n plus layer it can block a higher voltage in the positive when it is forward biased and it can block a smaller voltage when it is reverse biased. So, it is a non symmetrical device. So, in other words NPT you may be able to use it in AC supply, but then a punch through you cannot use it. So, see the structure here there is no n plus here this is n plus. So, this is a symmetrical device both it can block positive as well as negative voltage this device cannot block negative voltage because of this junction or if at all if it blocks it is a very small voltage. Now, turn on and turn off an I G B T turn off mode turn on something similar to that of a mass now turn off there are there is a mass as well as the BJT in an I G B T both are combined there are two distinct turn off intervals during turn off. See here drain current and drain to source voltage I am we are reducing it. So, first the channel the whether the MOSFET the current due to the MOSFET effect reduces the channel disappears quickly MOSFET blocks, but then the or this is due to the majority carriers, but then minority carriers in n minus layer gradually recombine that is something known as the BJT current or what is known as the tail current. It takes see minority trapped in n minus layer they gradually combine it takes significant time to recombine or this process is relatively slow, but that but that the, but then the voltage across the device starts increasing. So, during this period of BJT current what is known as the tail current voltage across the device as as increased it is in the process of blocking mode voltage is high current is still finite. So, they are going to be of substantial as significant power loss during turn off. So, tail current is a very important parameter and that determines the turn off process. So, I D is relatively slow known as a train current voltage this period should be very small since VDS has attained reasonably a high value losses are high, but then this time punch through I G B T has a a smaller tail time why there is an n plus layer n plus layer n minus layer the minority carriers are trapped it takes time to recombine, but then here the concentration is high. So, therefore, it takes lesser time to recombine. So, tail current tail or this T F fault time is relatively smaller in case of punch through a BJT punch through I G B T just see the time storage time something similar to that of a BJT MOSFET current as reduces faster and this current take its own time. This process is there in almost all the devices even is there in GTO I have not covered GTO because I do not think the GTO is on this way out or it is by the way it is not very popular there is a reason I have not covered safe operating area I G B T that is about it finally, the smart power modules they have the entire bridge driver circuit as well as protection circuit just the power modules if there are only power module I need to design a driver circuit as well as the protection circuit all this protection circuit have to be designed in the gate driver itself. Whereas, a smart power module everything is incorporated or fabric during the fabrication so, size of the entire power electronic equipment reduces significantly. So, one such IPM is all the terminals protection gate drive circuit are brought out these are three output terminals plus minus and the driver terminals they are brought out in part modules. So, that is about devices we started from one device single three leged element two power modules and from there to a spot smart power modules that is a significant reduction in size and therefore, the space that occupied by this devices this about the devices I will take up few questions. Okay regarding IPM modules where we can get the smaller power modules and India? No, no we have to import them there are Indian agents maybe I will I will. Okay sir thanks sir. Amritha Coimbatore please go ahead with your question. Sir, I just want to ask regarding the resistance which is used to limit the starting current when we are trying to charge the gate source capacitance. Yes, I will cover this at the appropriate time see there is a you need to pay a price you have to see the data sheet and you have to find out from there see here I need to charge this now you need to tell me how do I charge this capacitor at a very fast rate gate and source assume that there is a capacitor in fact there is one I need to charge this at a very fast rate ideal condition is hot and this external resistance tends to 0 are you with me? Take for example this is a voltage source I need to charge this capacitor. Okay now smaller the resistance faster the turn on faster the charging time smaller the resistance faster is the charging and first then therefore faster the turn on smaller the resistance higher is the current this switch and the source has to supply this source may be able to supply but then this switch that I have to realize through gates may not be able to supply that much of current. So, we need to see how much is the capacitance here and what sort of a gate or the driver circuit what sort of a buffer that I am using and based on that we can decide. Hang on for some time I did tell you that during design time I will tell you how to choose this value of R I did tell you that yeah I will tell you in the sense if you are using a gate that then you need to find out the fan out or fan in capacity of the of those gates or what you can do is I have a gate buffer then may be a small this is a buffer then may be a small signal transistor a small signal transistor and then connect a small suitable resistor and this I can directly connect a buffer to or through a small resistor here to a capacitor if this is not able to supply this current you use a small signal transistor and thereby put a resistor and you need to choose the first I need to know the type of MOS various capacity current carrying capacity of this and only then we can decide this value of R one have to see the data sheets I will do this I will take a typical example and solve it in the class do not worry. Sir we know very well about BZT's and MOSFET and nowadays light activated SCR is in market if we use that light activated SCR for the reception of the radiation from the solar radiation or is there any plan in future to produce SCR triac those devices for power electronic applications. See I do not know in the sense what happens in the future I do not know but then one thing is for sure SCR is a basically an ideal device I do not require a continuous gate drive my continuous just a pulse I require just a sharp pulse during torno so my entire gate drive circuit the size of the gate drive circuit reduces significantly if I come unfortunately when I converted to AC I need to have a grid there for for commutating the SCR line commutation so otherwise in a DC to DC conversion I cannot use SCR if I use then I have to use force commutation so therefore they are not being used that is besides reason we do not even teach force commutation circuit in the class also I do not know an ideal device could be yes torno of having a device which can be turned on by just by a pulse and turn off using gate I do not know time will tell time will tell whether such devices will come Vijay Vada whereas in schematic you have shown that the drain is connected to the body which one is right which one though there is a no difference between as for the geometries there will be no difference between source and drain normally source will be shorter to see it is a it is a channel no no no listen it is a channel it is a bipolar it can it can it can conduct in both directions you have to suitably apply gate drive signal and therefore MOS can carry current in both directions no issues depending upon the biasing depending upon the biasing current can from source to drain both it is possible but then you need to apply suitable gate to gate signals it is the flow it is the flow of electrons or at the is the direct during conduction I said a channel is formed and the electrons are flowing it is something similar to that of a water flow through a pipe so it is it can current can flow in both the directions no issues. Shivaji university Kolhapur please go ahead with your question will you relate the two diode structure of MOSFET with the diagram what we have shown that is just from the structural point of view once you apply a gate to source signal that structure is not valid I told you a channel is formed and flow of currents starts see here this is the structure start from N to source N P P N so if if I just look at the structure yes there cannot be a power there cannot be a current flow but then things change when I apply a suitable voltage at the gate to gate between gate and source above the threshold a channel from and therefore a current starts flowing if I just look at the structure yes there cannot be a current flow because there are two diodes connected back to back. Sir generally this N plus is used for low resistance connection N plus is a highly doped only the doping level. Yes it is drain where it is connected to the drain but it is used generally used as a low low resistance connection part. Ok so so what. Then that side what we have shown that is C and N plus packet on. How does it matter see I can show only one layer does not matter N minus will determine or the thickness of N minus will determine the voltage rating. Yes KJ Somya please go ahead with your question. There is an inbuilt body drain on the MOSFET. So if reverse current has to flow through what it will flow through the diode or through the structure. That depends that depends on what exactly you want whether you want whether see it depends on the voltage when the diode the voltage drops of the diode or the voltage drop across the MOSFET are you with me diode also can carry current the MOSFET also carry current. Now now it all depends on the supply voltages power loss across a power loss during the conduction period of diode and power loss during the conduction of MOS that will determine whether you want to operate whether you want to make diode to conduct or to make MOSFET to conduct is entirely up to you is entirely up to you. If you do not apply suitable gate signal diode will carry current diode will start carrying current and you apply suitably bias the gate MOSFET MOSFET will start conducting current are you with me you can make diode if there is no gate current and there is a current flow that current diode starts conducting and if you apply in order to reduce the losses if you want to make the BJT to MOSFET to conduct you have to apply suitable gate drive to the MOSFET it is what is known as the synchronous converter. There is a question NPT IGBT or IGBT for solar inverter is my question do not worry let us see whether we will first we will discuss the inverter let us see whether it has to block a negative voltage or not and if it has to block negative voltage then we will choose the appropriate device let us not discuss or let us not try to try to find an answer to this question right now first we will see what sort of a voltage that device has to block in that solar inverter if it has to block a negative voltage then yes you have to choose an appropriate IGBT.