 So, what I will do is next 10 minutes, I could accept may be a few questions and what I request is we may not be able to get across to all of you since there are 38 centers, I could maybe take at the most one out of two questions and my request is please use model and please put your questions there, we will get back to you and as promised by professor Fatak this model would be available for at least one year and we would be looking at your questions and let us make it a interactive session and a long term relationship. So, please upload your questions, if anything was not clear in my lecture please put up your questions and I would get back to you. Thank you. So, we will we will wait for any questions which you would have. We have Tandei Periyar Vellur. So, can you ask a question please Hello, my question is should BJT be called a current amplifier or a voltage amplifier just to be random sir. So, what do you sir? Yeah. So, we have not talked about amplifier so far, now when you talk about when my next lecture I will talk about it. Let me give you a quick answer to your question, the question is is a BJT a current amplifier or a voltage amplifier. Now, mind you BJT is not an amplifier BJT is a device, you need to make an amplifier using BJT. So, you can have 4 types of amplifiers a voltage amplifier, a current amplifier, a trans contactance amplifier and a trans resistance amplifier. We will talk about that in the next lecture. So, you can have both current amplifier and voltage amplifier using BJT. Thank you sir, over and out. Mofakam Jya College Hyderabad ask you a question please. I would like to ask a question, you mentioned about early effect, what are the other consequences of early effect? Yeah. So, the question is about early effect, what are the other effects of early effect? Now, early effect comes into picture when the VCE voltage is large. So, in a normal amplifier where in a small signal amplifier, the basic early effects what it says is, if your VCE increases collector current increases. So, think of any scenario where your collector voltage is changing, then the current will change that is the issue. So, in a small signal amplifier this is not an issue, because the collector voltage will not change beyond the DC value, it will only change by very small values. But if you have an application where you have a large variation in the collector voltage, then early effect comes into picture. But this is actually used to determine the exact value. In most of the biasing circuits, we neglect early effect, but if you want to be correct, you would use it. Another place you would, whether you would use early effect or current sources. So, in another few days, you would be hearing lectures on current sources, IC current sources. There you would see the expressions would use early effect also, but otherwise most of the biasing circuits, we would ignore early effect. So, as a first order approximation, in most calculations we will ignore, but if you want to be more precise, then we have to use early effect. Sir, one more question sir, what are the parameters which affect alpha and beta values, over to you sir? Yeah, what are the parameters that affect alpha and beta? Now, my suggestion to you would be, see in one lecture you cannot cover everything from device to circuits. So, I just limited, but I would ask you to refer to the book which I have given there. It depends on many parameters. For example, the concentration, doping, many, many, many parameters. So, if you look at any textbook on any modern textbook on devices, you would see this. So, it is a parameter alpha and beta. Let us take just beta, because anyway from that you can get alpha. Beta is a function of base weight, the doping and many, many, many parameters, at least 4, 5 parameters of a transistor, device parameter. Yeah, over. So, maybe I can take another question and it will stop. VIT Weller, we have selected you. If you have any questions, please ask it now, over. Sir, can you please distinguish source bias as a universal bias from cell bias? We will come to this, but right now, can we stick to questions which are related to what I did in the class? Since you have asked this question, I will answer. The bias which is generated externally, which is independent of the operating circuit, that has to be distinguished from a bias which generates because of the flow of the current in the transistor itself. So, for example, if you have a transistor which is drawing current and you put a resistor in the emitter lead, then the emitter reaches a particular voltage because of the amount of current which is flowing in the transistor. So, the emitter is in this case self bias, but sometimes you create a reference by an external potential divider or power supply or whatever and that leads to the reference bias. But we will come to these when we have done the transistors. What I would like to do today is to concentrate on things we have done in the session that we have just completed. So, are the universal bias and cell bias same? No, no, I gave you the difference. No, it depends on self bias circuits are those in which because of the operation of the transistor itself, voltages and so on are established which causes this transistor to work properly. So, for example, suppose you have a J FET, this J FET is on you put a resistor in the source lead. As a result the source reaches some positive voltage and if the gate is at ground potential, there is a built in negative bias. This is a self bias simply because you put the resistor in the source. On the other hand, if you had created a voltage source by a potential divider and so on, then that is not dependent on the operating point of the transistor. So, that is the other kind of bias. So, normally whether eventually the circuit configuration is the same or not is irrelevant. The intention of providing the bias through the operation of transistor itself is called self biasing and otherwise you have the whatever term these are not standard terminology universal bias or voltage reference biasing. By looking at the characteristics curve, how can we know that which configuration has a high impedance and a low output impedance input impedance? Which configuration common bias, common emitter, common collector? How can we know that? This is not a characteristic of the transistor, it is a characteristic of the circuit. So, you have to analyze the circuit of what kind of feedback it has, what kind of connection it has. I will leave it for Professor John to answer the details. See, this is again a topic I will take up in the next lecture. Now, as Professor Shama rightly said, you cannot talk about a device and impedance. You have to talk about a circuit and the moment you talk about a circuit, you have to look at it in detail. So, the concept of an input impedance, output impedance I will be covering in the next class and that time please wait till then. And still I think in two lectures we cannot cover everything. So, we can only finally, maybe give you a few things. So, I think that is the correct answer. So, you need to look at the whole circuit and then decide the impedance. Over. K. J. Somya, Mumbai, we have selected you. If you have any questions, please go ahead and ask. Over. When I going for the fabrication of silicon diode or germanium diode, the cut-in voltage is 0.7 for silicon and 0.3 for germanium. Whereas, this cut-in voltage of barrier potential or you can say it is a built-in potential, you can vary by changing the doping. But, while solving the numerical in any book if you refer, always you are getting for germanium it is approximately equal to 0.7 and for germanium it is approximately equal to 0.3. Why it is so? What is the reason to stick to 0.7 for silicon and 0.3 for germanium? Actually, in my view these are terrible approximations. It is neither 0.7 nor 0.3, these voltages vary with current. The reason why they work at all is because the dependence on voltage on current is logarithmic. That means, the current has to change by several orders of magnitude for the voltage to change. So, for example, in case of silicon actually in case of any diode, the value of the voltage across the diode is k t by q log i by i 0. Therefore, this voltage is a function of i 0. Now, if i 0 is like say 10 minus 14 amperes or so, then for a milli ampere of current you will have log of say 10 minus 3 divided by 10 minus 14 which is 10 to the power 11. On the other hand, if you had 10 times as much current even then the voltage will change by a few k t by q. So, log of 10 will be to the base e will be of the order of 3 point something. So, about 3 k t by q a few millivolts. So, as a result the total voltage change for a very large change in current is quite small and it is decided actually by i 0 which can then be changed by choice of material doping etcetera, etcetera. But the voltage across a diode is not constant. It is logarithmic and because the log changes very slowly in these ranges for large values. Therefore, we take it as a quasi constant and taking of 0.7 and 0.3 are really gross approximation. They work reasonably well in circuits which have reasonable amount of negative feedback, but we should not take these as overall accurate figures. And the leakage current is much higher in case of germanium because it has a lower band gap and that is why the corresponding voltage is lower. i 0 is much higher. Over. So, the second question is in case of diode if you try to measure the barrier potential that is 0.7 for silicon. So, it is not obvious the answer is not possible. So, I read one answer from a statement is. So, if you connect the probe across the diode. So, that contact potential of the probe will cancel out the built in potential of the diode. But the second view in my mind is in case of depletion layer you are having immobile charges that is the heavy atoms. So, if the atoms are not moving there is no flow of current. So, it is not possible to measure the voltage across the diode over. I think in this case Mr. Streetman is right and you are wrong. For example, in case of a charged capacitor you do not have any mobile charges and still you can measure the voltage. If you have a polarized dielectric you can still measure the potential you are not drawing any current from it. So, it is not necessary to have current flowing in order to measure emf. For example, if you have a null measurement in which no current is drawn, but the voltages have to be equal for the bridge to balance. Then it is not necessary for something to have a mobile charges in order for the potential to be measured. Remember we are talking of an emf and not a voltage. The reason for this is somewhat quantitative, but it is in fact related to what I talked about this morning. That means the Fermi level and the Fermi level difference decides the overall measured voltage that you measure and the point is that if the p and n sides have Fermi levels which are different then so does the metal and if you go around the full circle then it has to add up to 0. Otherwise you will be producing energy without doing any work on it. So, it is because of that reason that the local potentials are just so that you cannot measure any voltage. It is in fact I can give you counter examples in which you do not have any mobile charges. For example, a piece of wood charged to some very high voltage and now you want to measure without drawing any current from it you can in fact measure the emf. So, it is not necessary to have mobile charges for an emf to be there. Depletion charges can give you voltages and in fact the same diode that you are talking of if you shift the Fermi level from its equilibrium value by shining light on it then you do measure a voltage otherwise your solar cells will never work. So, the idea is that the equilibrium value of the Fermi voltage is such that it adds up to 0. You have to disturb the Fermi level by shining light injecting current or take the carrier concentration away from the equilibrium value then the voltage will move from this value and only then you will see a voltage difference or any emf. We will see a little more of this in my next lecture. M K S is a spoon eh. If you have any questions please go ahead over. Hello sir, good afternoon sir. Sir, these are related to BJT. The most of the circuits they have used the NPN transistor. So, any explanation regarding this or due to this mobility or all these things only the NPN transistor are using because there are mostly high mobility electrons are there in both in both in or. Yeah, when you talk about high frequency circuits yeah NPN is used, but otherwise normal circuits I think there is no reason why should have only NPN, but high frequency circuits yes frequency response the point of view yes. Actually the leakage behavior makes a difference between the NPN and NPNP and the differences are actually more technological than device physics base. So, for example, if you have the diffusion constant of boron is much higher than that of arsenic and so on which is used. So, if you use boron in the emitter is the most heavily doped region. So, if you use boron in the emitter region then it tends to diffuse into the base and that does not give you such a good transistor. On the other hand if you can use very shallow and very heavily doped emitter then you get good transistor characteristics and that is possible only for NPN. So, the difference is not because of electrons and holes or because of device characteristics the differences are technological and the diffusion constant of boron versus N type in theory. Yeah, DOEC Srinagar please ask your question. Over. Effect of temperature on transistor characteristics under that how would we expect that the transistor would constant current source when there is a variation in temperature and of course it is going to affect the collector current over to you sir. The constant current source I mean you see it is true that all the circuits are affected by temperature, but you can design constant current sources. The way like for example, some circuits done is you would make the reference itself you would make it temperature insensitive by having you know senior and diode connected in series 1 with a positive temperature coefficient and a 1 to the negative temperature coefficient. So, it is possible there are I mean ICs and all available there are circuits available. Some of these questions will be coming up. Over. We have another question. Hello sir, my question is for professor John. This is he said in his lecture that VB is decreases by 2 millivolts when temperature is increased by 1 degree Celsius. I want to know sir what is the operating temperature of BJT for proper functioning and what will happen when temperature will decrease it by 1 degree Celsius? Now, I am I do not really understand what exactly the problem see the issue is when you use a BJT let us say in an amplifier like take a common emitter amplifier which we will talk about next lecture see what happens is when the temperature changes the VB will decrease. So, there will be a tendency for the because of the forward bias to current to increase. Now, in a biasing circuit by putting a resistance in the emitter there you would counter that. So, that is the way it is done. So, in a circuit this has to be taking care. So, this is why in a biasing circuit you would put a resistor in the emitter. If you do not put then this effect which we talked about right now will come into picture. So, the way you this is why in a biasing circuit you have to put a resistance in the emitter. So, that you get a negative feedback. So, any change in the temperature and the corresponding increase in the emitter current will be countered by that extra voltage which is developed due to the increase current. So, I hope it is clear over. Sir, I want to know what is the temperature needed for the proper functioning of semiconductor devices? Actually, first of all the semiconductor devices should remain semiconductor. Now, if you recall my lecture what led to the generation of electrons and holes? It was the availability of energy. If we increase the temperature we increase the amount of energy. This exponentially increases the number of electrons and holes available. Now, at some point if the thermally generated electrons and holes become more than those provided by doping then there will be no effective of doping. There will be no n type. There will be no p type and your device will stop working. Now, where this happens depends on which material you choose. For silicon this might occur around say 150 degrees to somewhere between 150 to 200 degrees centigrade. The typical commercial range of operation of transistor is considered to be 0 to 70 and the military range is somewhere from minus 40 to 100. So, this is roughly the range in which silicon devices work conveniently. On the other hand that does not mean that you cannot have devices working at other temperature ranges. At much higher temperatures you obviously need semiconductors whose band gap is more. So, that even with the availability of energy they will produce fewer thermally generated electron hole pair and this will happen with wide gap semiconductors and wide gap semiconductors could be for example, diamond or silicon carbide. So, therefore operation at temperatures much higher than 150 to 200 degrees centigrade we use silicon carbide or diamond or special semiconductors whose band gap is in fact higher. At very low temperatures there is a problem of freeze out and the freeze out occurs because we have been assuming that a dopant atom the extra electron or hole is so loosely bound that at any reasonable temperature it is free. So, therefore each dopant contributes an electron or a hole, but if the temperature is very low this assumption will not be true and the energy available will be so low that even this loosely bound electron or hole will still stick to the original dopant atom and not become free for conduction. This condition is called freeze out and this typically occurs near liquid nitrogen temperature. So, therefore the there is no definite temperature range which can be given depends on the material. However, convenient ranges are from let us say 0 to 70 for commercial devices about minus 40 to 100 for military grade devices with silicon and much higher for silicon carbide and diamond based semiconductors and all the way down to liquid nitrogen temperatures on the cooled side if you use careful design of enchused materials properly. So, that is roughly the temperature range over. S.R.Ram, Kanjipuram, if you have any question please ask. Actually, behalf of student I am asking this question while we are taking class we are asking a simple basic definition for potential and voltage how we should distinguish potential and voltage. What is the difference between potential and voltage? We are using those two terms how we should distinguish potential and voltage. Voltage is a measurement of potential. So, this is the difference between richness and rupees. You measure how rich somebody is by finding out how many rupees he holds. So, you measure the potential in volts. So, voltage is the measure whereas potential is actually a measure of the potential energy. The word potential comes from the potential energy and therefore, if a charge is at some voltage then it holds the energy equal to q times v and this energy is not because of its motion it is not kinetic energy it is potential energy and therefore, the corresponding voltage is called its potential. In order to measure it you measure it in volts and the number of volts is called the voltage. So, these are these are the difference is that of nuance. You can you can say somebody is rich or you can say somebody has lots of rupees. KK Wag Institute Nasek. So, please ask a question. Can you elaborate on the effects of leakage current on various regions? Well, if I gave you a complete answer to this question you will all miss your lunch. However, I will give you a partial answer. Essentially, the whether a semiconductor is affected more or less by temperature depends on its band gap. So, when we say that there is a certain amount of energy available this energy as a function of temperature is given by KT by q. The question is what fraction is KT by q of the band gap? Now, germanium has a low band gap about 0.2 volts or so. It depends actually, but the sorry 0.7 volts or so, but silicon has a band gap of 1.1 electron volts. Therefore, the same temperature energy KT by q is a different fraction of the band gap. As a result the capability to generate electron hole pair at the same temperature is different in the two materials and indeed as I talked in response to a different question earlier if you had silicon carbide or diamond then the capability of the same temperature to produce electron hole pair would be even lower. Therefore, at the same temperature the number of thermally generated electron hole pair is much smaller in higher band gap material, but it is much larger in germanium. The net result of this is that the number of minority carrier is larger in low band gap semiconductors compared to high band gap semiconductor for the same doping. The majority carrier is fixed by the doping the minority carrier is given by n i square divided by doping. So, if n i is small the minority carrier is small by a squared ratio. So, and the leakage current is caused by the minority carriers because while the device is reverse biased for majority carriers it is actually forward biased for minority carriers to go over. So, if you have lots of minority carriers then the leakage current will be high that is the reason that germanium and other low band gap materials have higher leakage. In fact, the built in voltage is also directly related to it as I said earlier it is given by k t by q log of i divided by i 0. So, if i 0 is high i by i 0 is low its log is low and therefore, the built in voltages. So, one effect of leakage current is the built in voltage higher the leakage current lower the built in voltage. The other thing is that it is more temperature dependent. That means, with small changes in temperature the number of carriers which become available minority carriers which become available thermally generated minority carriers that increases much more for low band gap and high leakage materials. In terms of actual circuit performance leakage current is often detrimental because leakage current simply means that there is a temperature dependent component of the current the bias shift the operating point of the transistor shift with temperature and also the impedances are lower because there is a lot of current flowing and often the leakage current leads to noise because it is random in nature. So, as a result there are many detrimental effects of leakage current and to honestly describe each one of them will involve missing of lunch for I am sure you if your companion if not you will never forgive me for that. So, we will hold that for future for some time. Thank you if there are no further questions we will break for lunch now.