Justin Bell Perhaps, if I do another video on the differences between enhancement & depletion mode MOSFET's and JFET, then I guess I might as well include IGBT's.
I used to work at IBM making microprocessors. When you described the MOSFET gate operation, I was reminded of how much time our engineers spent perfecting the gate oxide layer. There were 40 PHD level engineers working full time on gate oxide quality. Good oxide equals faster speed. The more expensive fast CPU processor is identical to the cheaper CPU except for the quality of the gate oxide.
@@charlesclements4350 there are a few vid out there. the most interesting is those using a high magnification microscope to look at processor circuitry.
@@charlesclements4350 It's an awesome process. I worked at a company called Atmel that made alot of ASIC parts then mostly EEPROMS and 8bit MCU's along with a host of other stuff. I was the lead operator for the strip process cleaning the silicon in preparation for the next step. The inspection post cleaning was to me the most amazing. Watching the layers build up. Did quite a bit of inspect during the metal layers as well .... DId get to use a Scanning Electron Microscope ( when I became an Eng. Tech ) that was the best. Also got, at various times, to get cross section views when trying to determine what / where contamination was located.
I think what happens is that the electron moves over, and thus the atom left behind is now genuinely positively charged. Obviously the P type silicon wants to have that extra electron to fill its shell, but technically it is neutrally charged, and the N type wants that extra electron gone from its shell, but is also still neutral. When they meet up and the electrons go from the N type to the P type, they happily fix their she'll situation, but the result is that now you literally have a negative charge on the P type and a positive on the N type. It's almost like the "holes" did move over. Technically of course the situation is fundamentally different, but from an engineering point of view you don't need to care about that, you can just say the two materials swapped roles.
This extra electron appearing in doped matrix was brilliantly explained. I've never seen it done graphically, it still makes sens without the picture, but this way it is really easy to understand and remember.
Hi Dave, I may not be your target audience as I have next to no idea about electronics and am a software developer but I am interested in learning about electronics and love your videos! Is there any chance you could possibly do a video or two on designing circuits, using transistors and the like and maybe interfacing with arduino/raspberry pi so I (and others) can get started with this as a hobby? Many thanks and love the teardowns/DaveCAD/whiteboard enthusiastic videos, keep it up!
What physical aspect would change for devices that can handle larger voltages and currents? Would it be the width of the P and N material (current) and the distance between the P and N material (voltage)? Excellent video, thanx.
Do you have any resources or books to recommend for beginners trying to get into amateur electronics? Electronics seems very interesting, but it is quite daunting and I don't really know where to begin.
Why is Conventional Current used in science and engineering? Because of the positive current-carriers common in nature. "Electric Current" has to work correctly for anything explained by science, not just for electronics involving metals. (And then confused tech people pretending that holes aren't real, or that batteries and electroplating and proton flows don't exist.) Positive currents are: - proton-flow in battery acid - Proton flow in fuel cell Proton Conductor membrane - Positive ion flow in human tissues, salt water, ground, ocean etc. (along with negative ions going opposite.) - Positive ion flows in plasmas along with electrons going opposite. - And of course, holes. Holes really are genuine positive charges, same as protons. They aren't just "lack of electrons," since positive charge is when an existing proton is exposed by a missing electron. Holes are real protons, but only having virtual motion caused by electron-transfer. When an electron cancels out a hole, it leaves behind an exposed proton, as if the positive charge had jumped from atom to atom. Also, CC conventional current is great for teachers testing the students, because if a student complains that the currents are "backwards," send that student back to physics class, since they've just admitted that they have little idea what "electric current" actually is. Put it on a test questions. Thinking that Franklin was right about single-fluid electricity, or thinking that current is somehow backwards, that's an automatic zero in physics classroom. For the ones with solid physics background, CC doesn't look backwards at all. Electric currents in salt water have no electron flows, instead they have two opposite ion populations going in opposite directions, two clouds flowing through each other, yet creating just one electric current, the CC current. And besides, DMMs measure conventional current: they don't break currents down into the actual carrier flows: into percent of protons flowing one way and percent of electrons flowing the other. Clamp-on ammeters tell us the Conventional Current even when clamped around an acid-solution hose, or a human arm being electrocuted, or a fluorescent lamp tube.
"Holes really are genuine positive charges, same as protons." No, it is not. And you yourself explained below why it is not so. Holes - just abstraction for explaining. Hole by itself hasn't positive charge, but ion have it. Ions don't mooves in lattice (except thermal vibration and other not important effects), all moovement created by electrons if we talk about Si-based semiconductors doped by B or P.
@EEVblog Hi Dave, maybe you can answer this question of which i don't really know the answer...What is the maximum input voltage or current that can be sent into the base of a transistor (FOR AUDIO USE - PRE AMPS) and where is this info on the specifications data sheet of that particular transistor? For e.g a BC547B signal transistor NPN, what is the maximum voltage into the base allowed before it burns out? I know that it takes about 0.7 Volts to turn on so that's the minimum input into the base, but can i put 2 Volts into the base and will it be okay OR will that transistor be totally overloaded and distort because of the high voltage (IN AUDIO USE) This is what i don't know the answer to,i suppose my question is like overloading the input of a preamplifier,for example putting an electronic guitar output (maybe 1 volt) into the input of a RIAA TURNTABLE pre-amp (which is designed for about 5mv sensitivity) the guitar output would totally overload the input and distort like crazy, so I am trying to find out if on specification sheets transistors have a max limit into the base voltage and if so what that is called. For example the amount of times a transistor can amplify is know as HFE, I am looking for title like that which signifies the maximum voltage or current allowed into that transistor before its too much for it. When i know this info then i can choose specific transistors for audio applications. DOES OVERLOADING THE BASE INPUT OF A TRANSITOR CAUSE IT STRAIN?, WILL IT DESTROY IT OR JUST DISTORT THE SIGNAL THANKS IN ADVANCE - YOUR VIDEOS ARE GREAT - LOTS OF PASSION - VERY INFORMATIVE AND FUN AND EDUCATIONAL OF COURSE - THANKS FOR ALL YOUR EFFORT
Thanks a lot Dave for your work and your time! I deeply respect people like you who take their time to spread their knowledge free. Of course, some "Iknoweverythingandimbetterthanyou" people will criticize your work, but what you do is popularization of electronic sciences (and you never claimed it to be anything else, and... no one's perfect). So thank you again! BTW, if i may, you should do also a thursday, wednesday or whatever day lesson, where you present and explain, in a 5 minutes format, week after week, a simple electronic circuit function: example like limiter current system, protection system, etc... but kind of circuit that you build with less than 10 basic components (resistor, transistor, capacitor, diode, even AOP in the most extreme case), not the one that you build round with a specialized IC.
5:29 Having the same amount of protons and neutrons does not make it have a neutral charge! Neutrons are neutrally charged so they cannot balance the positively charged protons! Was shouting at the screen lol EEVblog
waicool20 Doh, did I really say that? - Yup, I did! Fixed in annotation, thanks. I shouldn't have put the neutron count in the middle, that's a tad confusing.
Nice!! These are my favorite videos of yours. You did one similar with op amps. I absolutely love these. They are the reason I subscribe. Thanks so much!
Awesome video. You explained this better in 20 minutes than my EE professor did in multiple weeks worth of classes. I think university professors need to lay off the math and explain things like this. When they just want to jump into calculations and numbers they fail to explain the greater picture like you do here.
gamingSlasher Ah, yeah, could have made that better. Basically, small number of holes enter from the base, huge number of free electrons get moved up from the emitter due to the collector voltage, and only some of those electrons match up with holes to exit the base. the rest go up to the collector. Hence small base current, large collector current.
Dave, I love ya man, but you just don't know how transistors work. I don't even know where to begin. I suggest you invite a guest, who is an expert, to correct this. Sorry.
Steve Robbins I can only support what you say, but the Video is still okay and good .... there are also minor errors in the FET explanation (e.g. whiteboard drawing --> bigger gate or bigger n+ doped regions) But: I couldn't do better so it's OK. Still impressive explanation. Much much better than most universities .... This is youtube-Content you can't 100% rely on this, like you can't trust wikipedia
+Steve Robbins, I have subscribed to your channel, being your very FIRST SUBSCRIBER. I recommend that you post a video displaying your vast knowledge of the chemistry of Si conductors, but do so in simple terms, so that even grade school children can understand. Eagerly anticipating the vid! It's great to have so many intelligent people willing to go out of their way to make the world a better place & at such a high personal cost. Not just sit there behind a keyboard typing comments that don't contribute to community at all!....
Steve Robbins Perhaps you can begin by making your own video? Or the comment space here is unlimited, go for it. I did the best I could in less than a day from thinking of the topic, to shooting, editing and uploading without final checking or getting anyone else to check it. Yes, I'm no semiconductor physics expert. I think it's a good enough explanation.
EEVblog I've never made a video before. I'm sure it's very hard. And BTW, I give thumbs-up to at least 95% of your videos. You're a very talented engineer. But with all due respect, If I was going to make a video about how transistors work, I'd probably get out the text books and review first. Anyway, I'd start with the PN junction. Doping silicon, adds charges to the matrix. But a key point is these charges can't move, they are fixed. It's like they're glued in place. Mobile carriers are attracted to the fixed charges of opposite polarity, essentially covering them, and making everything charge-neutral. When the PN junction is reverse biased, the applied voltage attracts the mobile charges away from the junction, leaving a "depletion" region, that is void of mobile carriers. Note that the fixed charges are still in the depletion region, because they can't move. Think of a tide going out and uncovering a bunch of sea shells that are stuck in the sand. When the mobile carriers (water) drift away, then uncover the fixed charges (sea shells). And the fixed charges, since they are no longer cancelled by the cloud of mobile charges that were covering them, make an electric field across the junction that opposes the applied voltage. Equilibrium is reached when the mobile charges drift far enough from the junction that they uncover enough fixed charges to make a big enough opposing voltage to equal the applied voltage. That is why the width of the depletion region increases as you increase the applied voltage. Also, when you decrease the doping concentration (the number of fixed charges per unit volume) it takes a wider depletion region to uncover enough charges to oppose the applied voltage. Simple, right? A bipolar transistor consists of three layers, making two back-to-back diodes. The middle layer is the base, and must be very thin (that's a critical fact that I think Dave forgot to mention. The collector is lightly doped to give it a higher breaking down voltage. (Lighter doping = wider depletion region = higher breakdown voltage.) When the base-emitter junction is forward biased, current flows across it, just like it would through any forward-biased diode. But because the base is so thin, many of the mobile carriers from the emitter have enough momentum to pass right through the base, into the collector. That's why it's so important that the base layer is thin. The thinner it is, the higher the BETA. Also, it is commonly said that the collector "sweeps up" the carriers that flow into the base from the emitter, like a vacuum cleaner sucking up dust; this is because the VCE is usually a lot bigger than VBE, so the carriers are much more attracted to the collector. FETs work on a completely different principle. It's true that depletion regions form around the drain and source, like Dave said. But you don't normally get a depletion layer under the gate. In fact, if you increase the drain-to-source voltage enough, making the depletion regions big enough to touch each other, that's called "punch through" and it's bad. What you're supposed to get under the gate is an inversion layer, not a depletion layer. Inversion means that the gate potential has attracted enough mobile carrier to outnumber the fixed carriers. When the fixed carriers were the majority, the channel region (the space under the gate) was an insulator, but when mobile charges become the majority, the channel becomes a resistor. Now, it's not like a normal resistor, because there is a kind of feedback mechanism called "pinch-off" that regulates the drain-to-source current. I don't think I could explain that very well, so I'll just quit here. I hope people find this helpful. And once again, Dave, I'm one of your biggest fans. My criticism was totally meant to be constructive.
There's this book "Horowitz and Hill." Not familiar? You do realize that it contradicts this above BJT explanation. They repeatedly show that BJTs are voltage driven. H&H's book AOA steers students away from "current driven" by giving many examples of broken circuits which were created with the "current input" BJT misconception. The lab book for Horowitz and Hill goes even more into this. They show that slight temperature changes screw up the hfe-based circuits, and totally halt function because of device-to-device variations. hfe-based transistor stages each need a trimpot and a thermistor. Vbe-based designs do not. Then, AOA points out the many circuits which are impossible if transistors are current amplifiers, but which are easily explained if they're voltage amps: impossible things like BJT op-amp front ends! Why aren't BJT op amps based on current-input? Because BJTs are inherently voltage in, current out, and their base terminals need no resistor in series. Also: the bjt CB junction is not thin, that's wrong. That junction always has a relatively thick depletion layer, and it never gets thin. In fact, the higher the supply-voltage, the wider is that depletion zone. But electrons easily flow across it because they're being dumped on the wrong side! They see the wide empty zone as a vacuum, and easily cross it, driven by the strong e-fields in that insulator region. The wide BC junction acts a lot like the vacuum in a vacuum tube; it's insulating unless an electron cloud is supplied.
+wbeaty There is no misconception other than to say that in the absence of voltage there is an absence of current. There is also a book written by Bill Shockley who was probably a lot smarter than Horowitz and Hill put together. I should probably say that either Horowitz or Hill individually is a lot smarter than I, but we have to pick our gurus based on something.
+Richard Gray Of course there's a misconception. This error is exposed in any engineering classroom, especially in the semiconductor physics classes. The error isn't only debunked in AOA book, but everywhere. AOA is just an easily-recognizable example. The above video about BJT operation is oversimplified and wrong, using concepts typically aimed at low-level techs who don't need to do actual BJT designs. If you have no desire to understand the silicon itself, a wrong explanation won't hurt you. But the wrong explanation isn't necessary, since this "analog engineers' secret" isn't that complicated. But it does require that we confront and defeat the wrong concepts we've been taught. Many people won't do that ever. They prefer to believe that grade-school textbooks tell the truth, just like George Washington chopping down the cherry tree, and heroic Columbus who would never think of murdering or enslaving the natives. So, what's the "Lies to Children" part of transistor explanation? It's the mistaken idea that base current Ib can control collector current Ic. To understand transistors' internal function, we must note that the base current has *no direct effect* on collector current. One current cannot affect another. Instead, the entire base circuit sets the value for base voltage Vbe (via the Shockley equation of course.) Then the Vbe potential-barrier determines both currents: Ib and Ic. We can view this as Ib affects Vbe which then affects Ic ...and then we can go further and form a simplified 'black box', where Ib apparently determines Ic directly. But, if we open the hood and look inside, we find that BJTs are voltage-input devices, transconductance devices same as FETs, where the ruling equation is Ie=Is*e^Vbe/nkt -Is. This isn't that complicated! It's just transconductance: volts in and current out. Same as FETs and vacuum tubes. But we do need to have high-school algebra, so we're not scared off by nasty math like the diode exponential equation. And we must face the terrible horrible fact that Vbe is not actually fixed at 0.7v, and it never was. (Oh no!) :) All of BJT-engineering is based on this Shockley equation, the Ebers-Moll version, where base voltage determines Ie and Ic. It's why BJT op-amps have voltage input rather than Ib current input. In op amps, the differential-pair front end is a voltage-input transconductance circuit. (For those who believe in BJT current-input, the long-tailed pair circuit remains a deep and confounding mystery!) The breakthrough to understanding op-amp internals is easy: BJTs are voltage input animals, and your earlier teachers were wrong about the nature of hfe. Also, base-voltage is how current-mirrors work, and half the stuff inside any analog IC is the current-mirrors replacing the large resistors. Try explaining current mirrors or Cascode circuit based on hfe, where Vbe is fixed. Not possible. Of course it's fine to use Ic=hfe*Ib in many situations. It's an extremely useful simplified rule of thumb, like Ohm's law. Just don't try to use it to explain the internal physics of components, or everything gets loopy and distorted. Yes, Shockley himself was beating the drum for "BJTs are current input." He was wrong, but had VERY good reason. His statement appears to be Bell Tel business-hype, done to make certain that USPTO wouldn't reject BJT patent applications, since voltage-input Lilienfeld transistors had already been patented twenty years earlier. Bell Labs had to distance themselves somehow from existing FETs. A good way to do this is to pretend that BJTs are current-input while FETs are voltage input. But it's not honest science (though it may be honest cutthroat business practice.) BJTs use voltage-input and use the (ahem!) Shockley Equation. Shockley worked out the whole semiconductor theory behind them. So, why would he claim that BJTs are current-input when he very well knew that they weren't? If he hadn't, there might not be any Bell Tel transistors; no patent, no profit, no aerospace contracts. Yet his small businessmans-lie is confusing-the-noobs even today. They just have to take some college undergrad courses to get the "hfe stuff" beaten out of them, and learn to slap on that Vbe Shockley equation just as easily as using Ohm's law.
+wbeaty Puhleeze. There is nothing wrong-headed about modeling a bipolar transistor using collector current as a function of base current. But, if you KNOW better than that, then I'm happy for you.
+Richard Gray Yes, that's a lot like modeling diodes as having fixed 0.7V voltage: often useful, even though certain designs would fail. And of course it tells us little about how diodes work inside. The "current amplifier" model of transistors is fine as long as we know that it's just an ultra-simplified black-box model. Major trouble arises whenever people think it's really true, and then they try to base designs on it. The designs don't work (their resistors have to be hand-adjusted for each transistor used, and temperature effects are enormous, requiring thermistor compensation circuits.) Also, classroom trouble crops up when we try to explain such things as Op-amp basics. The op-amp diff-input front-end relies upon base voltage signals, not on hfe, and it cannot be explained in terms of current-amplifier transistors. --- I encountered this problem myself, on a very first professional analog design for an industrial opto-sensor. I was forced to abandon the current-based "hfe-think"; and instead spreadsheeted everything as voltage-input (transconductance) circuits where hfe is ignored. Worked fine. That sensor design is in coke machines everywhere. Win Hill, one of the authors of "Art of Electronics" tells exactly the same story here: cr4.globalspec.com/comment/720374/Re-Voltage-vs-Current Before he could succeed at circuit design, first a wise old engineer had to take him aside and teach voltage-centric BJT philosophy, as opposed to the "current amplifier" stuff he'd learned in early schooling. So, instead of first being taught that BJTs use current input, why don't we just do an end-run instead? Never learn hfe stuff at all, but directly jump into teaching that BJTs are voltage-input devices like FETs? After all, that's what they really are. But then we'll have to face the fact that our earliest teachers were lying when they taught us that George Washington chopped down a cherry tree, or that Columbus was a saintlike hero, or that BJT transistors are current amplifiers. Or that William Schockley cannot tell a lie! :)
+wbeaty Well, I knew as a second-grader that it was bullshit when my teacher told us that George Washington never told a lie or as a seventh-grader when we were told that Isaac Newton watched an apple fall and thus invented gravity. But, later, I never got that feeling about hfe being a useful device parameter.
EEVblog Please do :) This is very interesting. Maybe you could also follow up on the fet, and explain what determines which of the N-doped regions becomes source, and which becomes drain, as well as why the parasitic diode forms.
I've just started studying electrical engineering and I don't get the textbook and I never understand anything on the lectures, so each time i have a test or a laboration I just look at your videos and suddenly I'm the *smartest* one in the room, despite knowing nothing and feeling devastated just hours before. You've literally saved me from failing and even helped me get better grades than I thought was possible. You are the best teacher and I wish the best for you.
EEVblog Sorry, but the physics is really messed up. Can't blame you; nothing else on UA-cam gets it right, but I was hoping. I've been meaning to make a vid, but we all know that'll never happen. Maybe if I go to Australia for vacation...? In the meantime, Pierret's Semiconductor Device Fundamentals is a good resource. BTW: All transistors (MOSFETs, JFETs, BJTs, Vacuum Tubes...) are voltage-controlled. BJT base current is merely coincidental. That is to say: base current in a BJT flows due to the same process that causes collector current, but to claim that base current *causes* collector current would be to confuse correlation with causation. The ideal BJT wouldn't have any base current, if that were possible (they can get close to this ideal with superbeta and heterojunction BJTs).
sorry bro, you can't take out a fire with a mere spit. If nothing else on UA-cam gets it right, then enlighten us with your knowledge, you don't have to go to australia just to make a reaction video about this right?
"If, in discussing a semiconductor problem, you cannot draw an energy band diagram, this shows that you don't know what you are talking about. with th corollary: If you can draw one, but don't, then your audience won't know what you are talking about." Herbert Kroemer
From what I know those symbols were created because of conventional current flow. Anyway, nevermind that. I don't understand why does it make more sense... I'm really curious.
ronettreker Water runs downhill, from the higher up potential to the lower potential for example. Thing like that make conventional current more relate-able and more "obvious" that current flows from positive to negative. That's why they thought that was the case to begin with, it wasn't an arbitrary choice.
EEVblog Isn't the number of electrons in a medium that define it's potential? I thought that electrons flow from a medium wich contains more free electrons to a medium which contain less, in order to achieve equilibrium. I thought the reason why the higher potential medium is negative is because of the higher number of negatively charged electrons. If so, then the negatively charged medium is the one with higher potential, not the positively charged one. Did I got it right?
ronettreker i think you may be missing the point I was trying to make. I was trying to show how (unless you know the details) it's more obvious to the lay person that something (current) would flow from something higher (potential) to lower (potential).
@@Bob-zg2zf I am left with the same question. Its literally the "Chicken or the Egg" paradox! Its a cascade of switches but what takes the softwares logical 1 and converts it to a physical flow of current!??! Ugh.....I must know!
Thanks Dave for the great video. Not sure if it's been covered or not yet, but like Intel & AMD years ago when clock speeds basically topped out, but they kept naming their CPUs with "virtual" clock speeds representing the effective performance improvements due to other innovations like multi-threading and multi-core, the semiconductor industry has been doing the same for the past several generations. So the 14nm node does not literally have a channel length of 14nm, the transistor performance just behaves as if it did. The really impressive dimension though is the gate oxide thickness, which actually is only a few atoms thick, and controlling this thickness (much less reducing it) is one of the main motivating factors behind going to the FINFET non-planar transistor design for 14nm.
cipmars Yes that is true but everything stems from somewhere. Prior to the modern transistor was the vacuum tube triode which was much bigger and used different principles. Even the transistor has changed so much since it was created. THAT's why electronics is so awesome!
Hana Zener Diode That is why it is a pain in the a** after almost 60 year of doing this. I started a boy. :) Now learning SDR. Not sure why, but it is fun.
my teachers could never explain a transistor this clearly. as a result ive been afraid to design circuits with them because of the PFM (pure f__ing magic) associated with their operation. thank you
Wrong. The space charge region between Collector and Base gets BIGGER, which "absorbs" electrons diffusing across the Base region, transporting them automatically over to the collector.
Thanks for the video, but there is one misunderstanding. 15:50. The depletion region is formed by the negative ions in the p-type/substrate side and the positive ions in the n-type (source) side, not electrons and holes like what you mentioned. The negative ions (acceptors) are created in the p-type/substrate side due to the gain of electrons diffused from n-type, whereas the positive ions (donors) are created in the n-type/source side due to the loss of the electrons which diffuses to the p-type/substrate side. Hope you can add notes there (15:50) to not let people misunderstand this. Thanks.
I LOVE! all your videos. Can you do one on how to find total impedance of a parallel RCL circuit im so lost and cant seem to find a good site that explains it.
I would LOVE to see a nice video about how to drive MOSFETS and BJT, including high currents, and inductive loads with "soft off" AWESOME video!!! Learned a lot! Thanks!
WRONG, WRONG, WRONG FOR THE BJT !!! BJT NOT only is NOT an amplifier, but also it's exactly the oposite: a strangler ! BJT does NOT amplify, it simply controls. Amplification rate ? WRONG ! Should be called sensitivity rate. Cheers !
Hey, you don't want to say "BJT transistor" because that's like saying "bipolar junction transistor transistor." It's the same kind of error as saying "PIN number." The same goes for "FET transistor," of course. So will you please just say "BJT" or "FET"?
Kevin Sirois (can't reply) MOSFETs use much less power (high input impedance; at low frequencies, essentially infinite). Easier to scale down (photolithographic). Easier to produce. Less power makes for less heat.
A point I never understood. Why do we call it amplifying? A transistor does not generate (or, amplify in your terms) amper out of nothing, correct? It all depends on the max amper that supply provides (collector lead), after all. It cannot exceed the total amperage of supply, right? Or, it is called amplifying in relation to base amper level/ratio? Could we rephrase that it adjusts the amper at collector to emitter, based on the amper level on BASE up to the max of suppply. Thanks,,.
The atomic section has a few huge gaps. -why can they join ? Electrons required in each shell. -why can electrons move ? You mentioned "electrons moving freely" but an atom missing a stable amount of electrons will always be attracted to electrons in/for the amount of electrons it's missing. -a way to look at N vs P. Stable amount of electrons is neutral. Electrons are negatives. Extra electrons make it more (N)egative. Missing electron(s) make(s) it less negative, thus (P)ositive. (Although, I believe in actually chemistry, it's the opposite.) Also, if it's possible, can you dumb down "how a transistor works" even further. I've watched many videos, read two things, and had someone explain it to me, but everytime I think I understand something, the rules seem to change on me. So, after years (3+) of halfassly investigating, I'm still at a lost. Like, maybe include some sort of mnemonic (letter-word-concept association) for each gate/end, and maybe have some more (numbered) action lines as well. I know there are many types of transistors, so, explain in a way that can be adapted to most (if not all) types of transistors, if you can. And thanks for providing an educational option (ie; teaching from a different perspective - perspective, video, text, animation, et cetera).
An intresting thing is that FETs were basically invented (and patented) quite a couple of years before BJTs but due manufacturing difficulties could not take off. What would the world look like if in the 50s and 60s not much progress had been done on BJTs but on FETs instead? Anyways, I don't think that PNP and PMOS as their opposites would be very intresting to see too, but as a good addition the JFET would be intresting, especially since many people love to use them for amplifiers, and how they can (or can not) be replaced by depletion mode mosfets.
16:37 about explanation of MOS FET. You are wrong. Field effect between substrate and gate, not source and gate. The substrate and source are internally connected in the device.
Thankfully Tek is advertising $100,000 oscilloscopes 10 times an hour, I’m sure they are flying off the shelves... I guess that’s better than having the sales weasel waiting in the lobby...
The next generation will be Sphere - Kinetic with moving atoms inside, not the static, here is the beginning : > KL03: Kinetis KL03 - 48 MHz, Small Form Factor, High Integration, Ultra Low Power Microcontrollers (MCUs)
THIS is educational!!! When you break things down to their simplest form and begin rebuilding them until you have a component, the principles of electronics really become evident! Tired issues, like which type of capacitor is "best," become rubbish. Does that cap allow the electron to move from one atom to another? More efficiently than another? I spend WAY to much time listening to the information Dave shares!!! Back to the money square, I go!!!
On second viewing I just noticed something that bugged me for years. Starting around 2000, the Sandhill Road companies between San Pablo and the 205 were over run with special visa people (H12) or something. Can't remember. What I do remember is that there where truck loads of "EE" guys who had Ph.Ds but had never heard of the partial doping region. It was just nuts. We would ask them in interviews how a BJP amplified a signal and they had no answers at all. It was not on the third world Ph.D. script I guess. We complained to management and were sent to HR with a racism write up or something. I actually quit the field entirely and went to law school at night and rebuilt my career. I could not personally live with the loss of basics that had become the standard. Years maybe decades later I was at a round table dinner with Gordon Moore. I asked him about those days and he laughed at me. Walking out later that night he said - this I never will forget - "now you know why we located Copper Mine (a critical Intel R&D lab) in the middle of Oregon." My interpretation was that he meant that the fundamentals do matter and that the insane Sandhill Road Venture Capital driven model was just filling slots with bodies, not people. Not sure of course that he would say it like that. I was in no position to press the point. Funny that you nailed the key issue right up front. After 20 min. of UA-cam searching on how BJPs work found me not one that even mentions this critical region of the junction. I wonder why.
3 years on and very very informative. Ironically, 14nm processes by Samsung Alliance like 14LPP have a gate length of 30nm. I think finfet doesn't follow the old tradition of decreasing the gate length/pitch (looking at intel's early 22 node too gives some insights) and... Well, Intel is struggling with 10/7nm now. Moore's law, created by an Intel employee and now Intel say they are unable to adhere to it anymore because they are struggling with EUV.
Hi, thanks for your videos which are really helpful, I have a question if you please help me in as I spent quiet time in searching and experimenting about (actually I'm a civil engineer, not electronics) the question is: I have a very weak signal 27MHz to amplify and I tried several ways to amplify using a transistor (not an IC) but the first stage transistor circuit is not giving the expected amplification, the last info I got was that the resistors selection needs to allow such week signal to trigger (turn on) the transistor (impedance issue) for which: - can you please advise me what would be the best range of transistor (lower - like hundreds of ohms - or higher - kilos or megas of ohm-)? - is there a better way to amplify weak signals? thanks a lot in advance.
About the MOSFET: you made a wrong drawing. The source is connected to a positively doped area that contacts the substrate and it forms a body diode along with the drain... And a MOSFET consumes current through the gate actually. Make a small design with a light bulb in the drain of a MOSFET and measure the gate current. It will be constant as long as the bulb is lit. And now explain that!
Dave, Great simplified explanation of how transistors work. But I think you are pronouncing "planar" incorrectly, and I'm Australian too. In my opinion it should be pronounced like "play-nah" rather than "plan-ah" so that it rhymes with trainer, explainer, and entertainer. source: en.m.wiktionary.org/wiki/Rhymes:English/e%C9%AAn%C9%99(r)
Hi Dave, can you please make another video about transistors (BJT and MOSFET) where you can explain on how to actually correctly read a datasheet? There are so many notations and values and I just don't know where to look to check if the transistor meets my requirements. Thank you!
This is the best explanation of transistors I have ever heard/read/seen. Only question... When you say the holes/electrons "switch" at a P/N junction. Is it more the case that the electrons just jump across from the N type and fill the hole in the P type material. Effectively blocking the holes. I cant conceive of the holes "moving" An electron shell cant gain an extra hole can it? Thanks Dave :)
When I trained with the P.M.G. in the sixties we were taught that the old conventions were wrong, that current flowed from -ve to +ve and that electron flow was the new future in electronics. Easily understood and remembered when you study the operation of a vacuum tube amplifier. At the same time, Cycles per Second (CPS) was being dispensed with and replaced with Hz (in honour of Hertz). Voltage was to be represented by E (for electromotive force) in lieu of V. So, I = E / R. Out with the old and in with the new. Hole movement was theorectical, the electrons moved into the holes leaving holes behind them. In a theatre, there's an empty seat (hole) at the far end of the row. Instead of the new patron (electron) stumbing all the way down the row, it's easier for everyone in the row to move to an adjacent seat. The empty seat (hole) effectively moves to the other end of the row but it's really the patrons (electrons) that are moving. Without a discussion of manufacture and the actual fusing together of the PN layers in an oven, there is no clear understanding here of how the depletion layer forms. These aren't pieces of silicon that are just placed next to each other, they are FUSED. The depletion layer creates an actual potential difference that must be overcome by external forces before current can begin to flow. When and why did the electronics industry revert from the NEW sixties conventions and decide to fall back to confusing ancient conventional current flow and also, bring back "V"?
So, are FETs symmetric or not? The schematic symbol suggests that it is not, but from the drawing it looks pretty much that it is. Dave, can you please briefly explain where symmetry is broken in physical implementation of FETs?
With some luck moors law will still hold for the next 2 decades. Some of the professors at my local technical university have been successfully working at the Karslruhe University on way smaller kinda mechanical switches. Collector and Emitter are made from a silver-base, and the actual switching happens with silver-atom bridging the tiny gap. Not entirely sure how it works, but those switches are insanely efficient - and they can carry a mindbogglingly huge current in the range of micro-amps over a single atom. Would be nice if they gave some more coverage on that.. and they didn't mention it this year sadly.
I still don't see (for the BJT) how the base emitter current is amplified into a higher collector current...? Somehow, what, 5 electrons entering the emitter results in, what, 100 electrons going through the base region and causing 100 holes to move into the collector region...? What accounts for that?
Just want to point out that the doping on the collector side is light, not heavy. Normally, the doping level is as such: Emitter >> Base > Collector. This is so that when in reverse bias, there won't be a large current flow. If you make a symmetrical BJT, then forward and reverse bias will cause same current flow magnitude.
EEVblog thank you for the explanation , but i have a question about the threshold voltage of the CB junction , practically I notice that collector voltage isn't responsible for overcoming the CB depletion region but depends on the Ib current , ( I mean that even if the collector voltage is small , the transistor will act as a short circuit if we applied large Ib .... and with a very small Ib - after Vbe > 0.7 - even with a large collector voltage the transistor will act as a high resistance because of the CB depletion region ) so instead I think it depends on Ib
Please do a followup video about popular transistor configurations, basic building blocks. Many people struggle with that. I've done several commercial designs and still could use a revision.
So I have TWO questions: 1. What happens to the characteristic behaviour of the circuit with increased temperature (i.e. does it become more or less efficient or something else?), and 2. What is the nature of either transistor type with aging (after being used for many years - i.e. - what is the mode of failure or problem due to aging?? thanks
Obviously BJTs and FETs don't work *exactly* the same way, or they wouldn't have made the two types (and you already said that one was for current vs. the other for voltage).
Dave wisely avoided the confusion surrounding the direction of the arrow. Scientists at Bell Labs incorrectly depicted "conventional flow" (plus to minus) when everyone who has ever worked with vacuum tubes knows that electron flow is from minus to plus.
Heh great video Dave! What I would do in the past is just match up the numbers and not really think of how the things work. Today I wondered how leg transistors work and this video was on the first page of youtube. You certainly explain stuff so even this dummy can understand it. :) I am learning about how stuff works here recently as I am working on a custom pinball machine and no not one of those toy kinds. A full sized 52" long pinball machine. ;) I think the mechanical and circuit boards will be the hardest. And I would send it off to you for Mailbag day but that would be costly and right now still in development as well.
You get 300USD if you build 100A 100v dc dc buck. I finished electronics engineering school and noone in class could build one, but claim to know how a transistor work.
I would love to see a "fundamental" about how to use FET transistors, what they are good or bad for and "traps for young players". Knowing how they are made is good I suppose but there is more to it than that. Why is the arrow pointing in the wrong direction for example (as compared to BJTs)?
I have worked in electronics, both as a career and as a hobby for more than 40 years. Even after reading several articles and books about how transistors work (remember the old RCA transistor manual?), I never really understood it as clearly as your explanation here. Thanks.
Sorry to say but very disappointed by your explanation of the BJT in this video of all others which have been awesome. many instances you kept contradicting yourself.
so is it true that between gate and source on the mosfet it is like ceramic capacitor and does the voltage on the gate regulate the resistance between drain and source or what? and also why is that parasitic diode between source and drain
If this was taught at my university like this, I would surely be more curious back then when I was younger. We were all about smoking resistors and blowing caps.
I want to try and build semiconductor chips at home. Yeah, the performance will be horrible (a piece of oxidized aluminum will work for the gate and insulator, right?) but I want to see if I can do it.
i hate it when physicists talk about "holes moving around". holes don't move, electrons move. when an electron move from an full orbital to a partially full orbital, it leaves a hole behind and fill a previously-unoccupied hole. it confused people when they talk about moving holes, when the actual explanation is much simpler and much more sensible.
I think it was a great idea to take on this task but suggest you consider re-doing it after cleaning up some of the issues. In particular I suggest you re-read the section in your physics book on PN junctions and structure the BJT discussion terms of Base *current* flow. In particular, "Vbe" only matters as the forward voltage of the transistor's PN junction, and that what *does* matter is that you structure current into the base such that that current times the transistor's gain (Hfe) result in the transistor being fully 'on'. From a physics perspective you are pulling enough electrons into P layer from the collector side, that there is always a path for an electron to go from the emitter to the collector. The P layer is essentially not there (this is called the saturation current, and for a switching application its your target current) Also suggest that you replace silicon 'matrix' with silicon crystal. That would align better with other sources people might go to after watching this for more information.
Laws of physics & can't get much smaller but one could alloy the substrate with dna to make a wireless field type connection, another type of nano induction. How? It's like growing a garden.
Literally the best lecture on semiconductors I've ever seen.
Afrotechmods Yeah, I like your channel too, Please make more vids :-).
Afrotechmods Thanks! I'm sure there are much better technical explanations out there though.
EEVblog Do you plan to do a video on IGBTs?
Justin Bell Perhaps, if I do another video on the differences between enhancement & depletion mode MOSFET's and JFET, then I guess I might as well include IGBT's.
Agreed. I need more information that, "here's this and here's how you use it." I want to know why it works. This is the perfect deep-dive.
I used to work at IBM making microprocessors. When you described the MOSFET gate operation, I was reminded of how much time our engineers spent perfecting the gate oxide layer. There were 40 PHD level engineers working full time on gate oxide quality. Good oxide equals faster speed. The more expensive fast CPU processor is identical to the cheaper CPU except for the quality of the gate oxide.
Now that you mentioned it, Jeff, just how are those things built? Did any of you make a video of what goes on in a plant where those things are made?
@@charlesclements4350 there are a few vid out there. the most interesting is those using a high magnification microscope to look at processor circuitry.
Wow ! Thanks, tbled52, I'll keep my eyes open for them.
Respect
@@charlesclements4350 It's an awesome process. I worked at a company called Atmel that made alot of ASIC parts then mostly EEPROMS and 8bit MCU's along with a host of other stuff.
I was the lead operator for the strip process cleaning the silicon in preparation for the next step. The inspection post cleaning was to me the most amazing. Watching the layers build up. Did quite a bit of inspect during the metal layers as well .... DId get to use a Scanning Electron Microscope ( when I became an Eng. Tech ) that was the best.
Also got, at various times, to get cross section views when trying to determine what / where contamination was located.
Thanks for taking the time to make this video. I have learned something new today and you're responsible for that.
This is really useful. Even for people who work with this regularly to be able to explain it to other people. Thanks!
I think what happens is that the electron moves over, and thus the atom left behind is now genuinely positively charged. Obviously the P type silicon wants to have that extra electron to fill its shell, but technically it is neutrally charged, and the N type wants that extra electron gone from its shell, but is also still neutral. When they meet up and the electrons go from the N type to the P type, they happily fix their she'll situation, but the result is that now you literally have a negative charge on the P type and a positive on the N type. It's almost like the "holes" did move over. Technically of course the situation is fundamentally different, but from an engineering point of view you don't need to care about that, you can just say the two materials swapped roles.
This extra electron appearing in doped matrix was brilliantly explained. I've never seen it done graphically, it still makes sens without the picture, but this way it is really easy to understand and remember.
One of your best videos Dave. First thing that comes time mind is you should write books but these videos are far more effective.
Many thanks for another trip back to basic EE stuff. Actually, a bit of physics as well.
Excellent.
Hi Dave,
I may not be your target audience as I have next to no idea about electronics and am a software developer but I am interested in learning about electronics and love your videos! Is there any chance you could possibly do a video or two on designing circuits, using transistors and the like and maybe interfacing with arduino/raspberry pi so I (and others) can get started with this as a hobby?
Many thanks and love the teardowns/DaveCAD/whiteboard enthusiastic videos, keep it up!
What physical aspect would change for devices that can handle larger voltages and currents? Would it be the width of the P and N material (current) and the distance between the P and N material (voltage)? Excellent video, thanx.
Dave, any chance of the FET one promised in this video? Really enjoyed this one, but it is FETs I l know little about :(
Do you have any resources or books to recommend for beginners trying to get into amateur electronics? Electronics seems very interesting, but it is quite daunting and I don't really know where to begin.
Why is Conventional Current used in science and engineering?
Because of the positive current-carriers common in nature. "Electric Current" has to work correctly for anything explained by science, not just for electronics involving metals. (And then confused tech people pretending that holes aren't real, or that batteries and electroplating and proton flows don't exist.)
Positive currents are:
- proton-flow in battery acid
- Proton flow in fuel cell Proton Conductor membrane
- Positive ion flow in human tissues, salt water, ground, ocean etc. (along with negative ions going opposite.)
- Positive ion flows in plasmas along with electrons going opposite.
- And of course, holes.
Holes really are genuine positive charges, same as protons. They aren't just "lack of electrons," since positive charge is when an existing proton is exposed by a missing electron. Holes are real protons, but only having virtual motion caused by electron-transfer. When an electron cancels out a hole, it leaves behind an exposed proton, as if the positive charge had jumped from atom to atom.
Also, CC conventional current is great for teachers testing the students, because if a student complains that the currents are "backwards," send that student back to physics class, since they've just admitted that they have little idea what "electric current" actually is. Put it on a test questions. Thinking that Franklin was right about single-fluid electricity, or thinking that current is somehow backwards, that's an automatic zero in physics classroom.
For the ones with solid physics background, CC doesn't look backwards at all. Electric currents in salt water have no electron flows, instead they have two opposite ion populations going in opposite directions, two clouds flowing through each other, yet creating just one electric current, the CC current.
And besides, DMMs measure conventional current: they don't break currents down into the actual carrier flows: into percent of protons flowing one way and percent of electrons flowing the other. Clamp-on ammeters tell us the Conventional Current even when clamped around an acid-solution hose, or a human arm being electrocuted, or a fluorescent lamp tube.
"Holes really are genuine positive charges, same as protons."
No, it is not. And you yourself explained below why it is not so. Holes - just abstraction for explaining. Hole by itself hasn't positive charge, but ion have it. Ions don't mooves in lattice (except thermal vibration and other not important effects), all moovement created by electrons if we talk about Si-based semiconductors doped by B or P.
@EEVblog Hi Dave, maybe you can answer this question of which i don't really know the answer...What is the maximum input voltage or current that can be sent into the base of a transistor (FOR AUDIO USE - PRE AMPS) and where is this info on the specifications data sheet of that particular transistor?
For e.g a BC547B signal transistor NPN, what is the maximum voltage into the base allowed before it burns out? I know that it takes about 0.7 Volts to turn on so that's the minimum input into the base, but can i put 2 Volts into the base and will it be okay OR will that transistor be totally overloaded and distort because of the high voltage (IN AUDIO USE)
This is what i don't know the answer to,i suppose my question is like overloading the input of a preamplifier,for example putting an electronic guitar output (maybe 1 volt) into the input of a RIAA TURNTABLE pre-amp (which is designed for about 5mv sensitivity) the guitar output would totally overload the input and distort like crazy, so I am trying to find out if on specification sheets transistors have a max limit into the base voltage and if so what that is called.
For example the amount of times a transistor can amplify is know as HFE, I am looking for title like that which signifies the maximum voltage or current allowed into that transistor before its too much for it.
When i know this info then i can choose specific transistors for audio applications.
DOES OVERLOADING THE BASE INPUT OF A TRANSITOR CAUSE IT STRAIN?, WILL IT DESTROY IT OR JUST DISTORT THE SIGNAL
THANKS IN ADVANCE - YOUR VIDEOS ARE GREAT - LOTS OF PASSION - VERY INFORMATIVE AND FUN AND EDUCATIONAL OF COURSE - THANKS FOR ALL YOUR EFFORT
SP330Y when in doubt, Google for datasheets.
Moore's law is nonsense.
I put a transistor into my drawer 10 years ago and it did not become smaller at all.
The Kaiser Hahaha. You are hilarious.
The Kaiser Hahaha. You are hilarious.
The Kaiser I remember putting a transistor in a drawer 10 years ago. I just went to look and I had 32 transistors, just what I expected.
mart fart It is about getting smaller, NOT more and I bet you left 2 transistors of opposite sex in the drawer, so that was to be expected.
+The Kaiser mine became mosfets, as expected. Measure emitter-base, and its open in both directions.
The clearest explanation of how transistors work I have ever seen. Thanks Dave!
dogastus Thanks, glad it was understandable for you.
Agreed, and I did 2 years of micro-electronics specialization, and spent countless hours making diodes and transistors and logic gate in the lab
dogastus Entirely agreed. This is fantastic.
good job
Thanks a lot Dave for your work and your time! I deeply respect people like you who take their time to spread their knowledge free. Of course, some "Iknoweverythingandimbetterthanyou" people will criticize your work, but what you do is popularization of electronic sciences (and you never claimed it to be anything else, and... no one's perfect). So thank you again!
BTW, if i may, you should do also a thursday, wednesday or whatever day lesson, where you present and explain, in a 5 minutes format, week after week, a simple electronic circuit function: example like limiter current system, protection system, etc... but kind of circuit that you build with less than 10 basic components (resistor, transistor, capacitor, diode, even AOP in the most extreme case), not the one that you build round with a specialized IC.
5:29 Having the same amount of protons and neutrons does not make it have a neutral charge! Neutrons are neutrally charged so they cannot balance the positively charged protons! Was shouting at the screen lol EEVblog
waicool20 Doh, did I really say that? - Yup, I did! Fixed in annotation, thanks. I shouldn't have put the neutron count in the middle, that's a tad confusing.
Nice!! These are my favorite videos of yours. You did one similar with op amps.
I absolutely love these. They are the reason I subscribe.
Thanks so much!
+EEVBlog, that's the best explanation ever, textbook authors should learn how to do this from this one! Massive thumbs-up! Thx!
I learned more in this 23 minute video than I did during an entire semester of my semiconductors course in college. Well done!
This was awesome. Thank you very much! I had that "Ah-Ha!" moment here where it all clicked.
you had an A-Ha Moment and you decided to TAKE HIM ON huh ? LOL
Awesome video. You explained this better in 20 minutes than my EE professor did in multiple weeks worth of classes. I think university professors need to lay off the math and explain things like this. When they just want to jump into calculations and numbers they fail to explain the greater picture like you do here.
I think they might first explain it like dave and then go to the maths.
This is very exact problem I have in my studies.
Omg YES!
Also try having professors who barely speak English.
Very good presentation. The only difficult thing was the explanation of the current amplifying physics. That was unclear. The mosfet was very clear.
gamingSlasher Ah, yeah, could have made that better. Basically, small number of holes enter from the base, huge number of free electrons get moved up from the emitter due to the collector voltage, and only some of those electrons match up with holes to exit the base. the rest go up to the collector. Hence small base current, large collector current.
Dave, I love ya man, but you just don't know how transistors work. I don't even know where to begin. I suggest you invite a guest, who is an expert, to correct this. Sorry.
Steve Robbins I can only support what you say, but the Video is still okay and good .... there are also minor errors in the FET explanation (e.g. whiteboard drawing --> bigger gate or bigger n+ doped regions)
But: I couldn't do better so it's OK. Still impressive explanation. Much much better than most universities ....
This is youtube-Content you can't 100% rely on this, like you can't trust wikipedia
You should do your own video and post a link back on this thread.
+Steve Robbins, I have subscribed to your channel, being your very FIRST SUBSCRIBER. I recommend that you post a video displaying your vast knowledge of the chemistry of Si conductors, but do so in simple terms, so that even grade school children can understand.
Eagerly anticipating the vid!
It's great to have so many intelligent people willing to go out of their way to make the world a better place & at such a high personal cost. Not just sit there behind a keyboard typing comments that don't contribute to community at all!....
Steve Robbins Perhaps you can begin by making your own video? Or the comment space here is unlimited, go for it. I did the best I could in less than a day from thinking of the topic, to shooting, editing and uploading without final checking or getting anyone else to check it. Yes, I'm no semiconductor physics expert. I think it's a good enough explanation.
EEVblog I've never made a video before. I'm sure it's very hard. And BTW, I give thumbs-up to at least 95% of your videos. You're a very talented engineer. But with all due respect, If I was going to make a video about how transistors work, I'd probably get out the text books and review first.
Anyway, I'd start with the PN junction. Doping silicon, adds charges to the matrix. But a key point is these charges can't move, they are fixed. It's like they're glued in place. Mobile carriers are attracted to the fixed charges of opposite polarity, essentially covering them, and making everything charge-neutral. When the PN junction is reverse biased, the applied voltage attracts the mobile charges away from the junction, leaving a "depletion" region, that is void of mobile carriers. Note that the fixed charges are still in the depletion region, because they can't move. Think of a tide going out and uncovering a bunch of sea shells that are stuck in the sand. When the mobile carriers (water) drift away, then uncover the fixed charges (sea shells). And the fixed charges, since they are no longer cancelled by the cloud of mobile charges that were covering them, make an electric field across the junction that opposes the applied voltage. Equilibrium is reached when the mobile charges drift far enough from the junction that they uncover enough fixed charges to make a big enough opposing voltage to equal the applied voltage. That is why the width of the depletion region increases as you increase the applied voltage. Also, when you decrease the doping concentration (the number of fixed charges per unit volume) it takes a wider depletion region to uncover enough charges to oppose the applied voltage. Simple, right?
A bipolar transistor consists of three layers, making two back-to-back diodes. The middle layer is the base, and must be very thin (that's a critical fact that I think Dave forgot to mention. The collector is lightly doped to give it a higher breaking down voltage. (Lighter doping = wider depletion region = higher breakdown voltage.) When the base-emitter junction is forward biased, current flows across it, just like it would through any forward-biased diode. But because the base is so thin, many of the mobile carriers from the emitter have enough momentum to pass right through the base, into the collector. That's why it's so important that the base layer is thin. The thinner it is, the higher the BETA. Also, it is commonly said that the collector "sweeps up" the carriers that flow into the base from the emitter, like a vacuum cleaner sucking up dust; this is because the VCE is usually a lot bigger than VBE, so the carriers are much more attracted to the collector.
FETs work on a completely different principle. It's true that depletion regions form around the drain and source, like Dave said. But you don't normally get a depletion layer under the gate. In fact, if you increase the drain-to-source voltage enough, making the depletion regions big enough to touch each other, that's called "punch through" and it's bad. What you're supposed to get under the gate is an inversion layer, not a depletion layer. Inversion means that the gate potential has attracted enough mobile carrier to outnumber the fixed carriers. When the fixed carriers were the majority, the channel region (the space under the gate) was an insulator, but when mobile charges become the majority, the channel becomes a resistor. Now, it's not like a normal resistor, because there is a kind of feedback mechanism called "pinch-off" that regulates the drain-to-source current. I don't think I could explain that very well, so I'll just quit here.
I hope people find this helpful. And once again, Dave, I'm one of your biggest fans. My criticism was totally meant to be constructive.
Current flow = charge flow flow. Ugh. Other than that, takes me back to my days in chemistry class.
There's this book "Horowitz and Hill." Not familiar? You do realize that it contradicts this above BJT explanation. They repeatedly show that BJTs are voltage driven.
H&H's book AOA steers students away from "current driven" by giving many examples of broken circuits which were created with the "current input" BJT misconception. The lab book for Horowitz and Hill goes even more into this. They show that slight temperature changes screw up the hfe-based circuits, and totally halt function because of device-to-device variations. hfe-based transistor stages each need a trimpot and a thermistor. Vbe-based designs do not. Then, AOA points out the many circuits which are impossible if transistors are current amplifiers, but which are easily explained if they're voltage amps: impossible things like BJT op-amp front ends! Why aren't BJT op amps based on current-input? Because BJTs are inherently voltage in, current out, and their base terminals need no resistor in series.
Also: the bjt CB junction is not thin, that's wrong. That junction always has a relatively thick depletion layer, and it never gets thin. In fact, the higher the supply-voltage, the wider is that depletion zone. But electrons easily flow across it because they're being dumped on the wrong side! They see the wide empty zone as a vacuum, and easily cross it, driven by the strong e-fields in that insulator region. The wide BC junction acts a lot like the vacuum in a vacuum tube; it's insulating unless an electron cloud is supplied.
+wbeaty There is no misconception other than to say that in the absence of voltage there is an absence of current. There is also a book written by Bill Shockley who was probably a lot smarter than Horowitz and Hill put together. I should probably say that either Horowitz or Hill individually is a lot smarter than I, but we have to pick our gurus based on something.
+Richard Gray Of course there's a misconception. This error is exposed in any engineering classroom, especially in the semiconductor physics classes. The error isn't only debunked in AOA book, but everywhere. AOA is just an easily-recognizable example. The above video about BJT operation is oversimplified and wrong, using concepts typically aimed at low-level techs who don't need to do actual BJT designs. If you have no desire to understand the silicon itself, a wrong explanation won't hurt you. But the wrong explanation isn't necessary, since this "analog engineers' secret" isn't that complicated. But it does require that we confront and defeat the wrong concepts we've been taught. Many people won't do that ever. They prefer to believe that grade-school textbooks tell the truth, just like George Washington chopping down the cherry tree, and heroic Columbus who would never think of murdering or enslaving the natives.
So, what's the "Lies to Children" part of transistor explanation?
It's the mistaken idea that base current Ib can control collector current Ic.
To understand transistors' internal function, we must note that the base current has *no direct effect* on collector current. One current cannot affect another. Instead, the entire base circuit sets the value for base voltage Vbe (via the Shockley equation of course.) Then the Vbe potential-barrier determines both currents: Ib and Ic.
We can view this as Ib affects Vbe which then affects Ic ...and then we can go further and form a simplified 'black box', where Ib apparently determines Ic directly. But, if we open the hood and look inside, we find that BJTs are voltage-input devices, transconductance devices same as FETs, where the ruling equation is Ie=Is*e^Vbe/nkt -Is. This isn't that complicated! It's just transconductance: volts in and current out. Same as FETs and vacuum tubes. But we do need to have high-school algebra, so we're not scared off by nasty math like the diode exponential equation. And we must face the terrible horrible fact that Vbe is not actually fixed at 0.7v, and it never was. (Oh no!)
:)
All of BJT-engineering is based on this Shockley equation, the Ebers-Moll version, where base voltage determines Ie and Ic. It's why BJT op-amps have voltage input rather than Ib current input. In op amps, the differential-pair front end is a voltage-input transconductance circuit. (For those who believe in BJT current-input, the long-tailed pair circuit remains a deep and confounding mystery!) The breakthrough to understanding op-amp internals is easy: BJTs are voltage input animals, and your earlier teachers were wrong about the nature of hfe. Also, base-voltage is how current-mirrors work, and half the stuff inside any analog IC is the current-mirrors replacing the large resistors. Try explaining current mirrors or Cascode circuit based on hfe, where Vbe is fixed. Not possible.
Of course it's fine to use Ic=hfe*Ib in many situations. It's an extremely useful simplified rule of thumb, like Ohm's law.
Just don't try to use it to explain the internal physics of components, or everything gets loopy and distorted.
Yes, Shockley himself was beating the drum for "BJTs are current input." He was wrong, but had VERY good reason. His statement appears to be Bell Tel business-hype, done to make certain that USPTO wouldn't reject BJT patent applications, since voltage-input Lilienfeld transistors had already been patented twenty years earlier. Bell Labs had to distance themselves somehow from existing FETs. A good way to do this is to pretend that BJTs are current-input while FETs are voltage input. But it's not honest science (though it may be honest cutthroat business practice.) BJTs use voltage-input and use the (ahem!) Shockley Equation. Shockley worked out the whole semiconductor theory behind them. So, why would he claim that BJTs are current-input when he very well knew that they weren't? If he hadn't, there might not be any Bell Tel transistors; no patent, no profit, no aerospace contracts. Yet his small businessmans-lie is confusing-the-noobs even today. They just have to take some college undergrad courses to get the "hfe stuff" beaten out of them, and learn to slap on that Vbe Shockley equation just as easily as using Ohm's law.
+wbeaty Puhleeze. There is nothing wrong-headed about modeling a bipolar transistor using collector current as a function of base current. But, if you KNOW better than that, then I'm happy for you.
+Richard Gray Yes, that's a lot like modeling diodes as having fixed 0.7V voltage: often useful, even though certain designs would fail. And of course it tells us little about how diodes work inside.
The "current amplifier" model of transistors is fine as long as we know that it's just an ultra-simplified black-box model. Major trouble arises whenever people think it's really true, and then they try to base designs on it. The designs don't work (their resistors have to be hand-adjusted for each transistor used, and temperature effects are enormous, requiring thermistor compensation circuits.)
Also, classroom trouble crops up when we try to explain such things as Op-amp basics. The op-amp diff-input front-end relies upon base voltage signals, not on hfe, and it cannot be explained in terms of current-amplifier transistors.
---
I encountered this problem myself, on a very first professional analog design for an industrial opto-sensor. I was forced to abandon the current-based "hfe-think"; and instead spreadsheeted everything as voltage-input (transconductance) circuits where hfe is ignored. Worked fine. That sensor design is in coke machines everywhere.
Win Hill, one of the authors of "Art of Electronics" tells exactly the same story here: cr4.globalspec.com/comment/720374/Re-Voltage-vs-Current Before he could succeed at circuit design, first a wise old engineer had to take him aside and teach voltage-centric BJT philosophy, as opposed to the "current amplifier" stuff he'd learned in early schooling.
So, instead of first being taught that BJTs use current input, why don't we just do an end-run instead? Never learn hfe stuff at all, but directly jump into teaching that BJTs are voltage-input devices like FETs? After all, that's what they really are. But then we'll have to face the fact that our earliest teachers were lying when they taught us that George Washington chopped down a cherry tree, or that Columbus was a saintlike hero, or that BJT transistors are current amplifiers. Or that William Schockley cannot tell a lie! :)
+wbeaty Well, I knew as a second-grader that it was bullshit when my teacher told us that George Washington never told a lie or as a seventh-grader when we were told that Isaac Newton watched an apple fall and thus invented gravity. But, later, I never got that feeling about hfe being a useful device parameter.
interesting, looking more for things like FET,DIAC
zombielinkinpark Deliberately didn't want to cover JFET's in this one, just briefest of mentions,
EEVblog Oh, i see, can't wait a video about it
zombielinkinpark Thought about another video explaining the differences between enhancement & depletion mode mosfet's, JFET etc.
EEVblog Please do :) This is very interesting. Maybe you could also follow up on the fet, and explain what determines which of the N-doped regions becomes source, and which becomes drain, as well as why the parasitic diode forms.
dumle29 I've done the parasitic diode, somewhere...
I've just started studying electrical engineering and I don't get the textbook and I never understand anything on the lectures, so each time i have a test or a laboration I just look at your videos and suddenly I'm the *smartest* one in the room, despite knowing nothing and feeling devastated just hours before. You've literally saved me from failing and even helped me get better grades than I thought was possible. You are the best teacher and I wish the best for you.
Have you finished your degree?
EEVblog Sorry, but the physics is really messed up. Can't blame you; nothing else on UA-cam gets it right, but I was hoping. I've been meaning to make a vid, but we all know that'll never happen. Maybe if I go to Australia for vacation...? In the meantime, Pierret's Semiconductor Device Fundamentals is a good resource. BTW: All transistors (MOSFETs, JFETs, BJTs, Vacuum Tubes...) are voltage-controlled. BJT base current is merely coincidental. That is to say: base current in a BJT flows due to the same process that causes collector current, but to claim that base current *causes* collector current would be to confuse correlation with causation. The ideal BJT wouldn't have any base current, if that were possible (they can get close to this ideal with superbeta and heterojunction BJTs).
Hezekiel Randolph i
sorry bro, you can't take out a fire with a mere spit. If nothing else on UA-cam gets it right, then enlighten us with your knowledge, you don't have to go to australia just to make a reaction video about this right?
"If, in discussing a semiconductor problem, you cannot draw an energy band diagram, this shows that you don't know what you are talking about.
with th corollary:
If you can draw one, but don't, then your audience won't know what you are talking about."
Herbert Kroemer
Yes, I like fundamentals in Dave's way.
Clear as mud : )
I'm curious. Do FET's suffer the same fate as some capacitors? Dialectric break-down due to moisture, temperature, etc.
Why do we even use conventional current flow? It appears to only create more confusion.
ronettreker Because it makes more sense and also follows the the arrows in diode and transistor symbols.
From what I know those symbols were created because of conventional current flow. Anyway, nevermind that. I don't understand why does it make more sense... I'm really curious.
ronettreker Water runs downhill, from the higher up potential to the lower potential for example. Thing like that make conventional current more relate-able and more "obvious" that current flows from positive to negative. That's why they thought that was the case to begin with, it wasn't an arbitrary choice.
EEVblog Isn't the number of electrons in a medium that define it's potential? I thought that electrons flow from a medium wich contains more free electrons to a medium which contain less, in order to achieve equilibrium. I thought the reason why the higher potential medium is negative is because of the higher number of negatively charged electrons. If so, then the negatively charged medium is the one with higher potential, not the positively charged one. Did I got it right?
ronettreker i think you may be missing the point I was trying to make. I was trying to show how (unless you know the details) it's more obvious to the lay person that something (current) would flow from something higher (potential) to lower (potential).
I feel slightly DOped after this. In a good way of course :)
This was very helpful, but one question remains for me: How exactly is the positive voltage generated at the gate?
@@Bob-zg2zf I am left with the same question. Its literally the "Chicken or the Egg" paradox! Its a cascade of switches but what takes the softwares logical 1 and converts it to a physical flow of current!??! Ugh.....I must know!
I kinda like fundamentals friday!
Thanks Dave for the great video. Not sure if it's been covered or not yet, but like Intel & AMD years ago when clock speeds basically topped out, but they kept naming their CPUs with "virtual" clock speeds representing the effective performance improvements due to other innovations like multi-threading and multi-core, the semiconductor industry has been doing the same for the past several generations. So the 14nm node does not literally have a channel length of 14nm, the transistor performance just behaves as if it did. The really impressive dimension though is the gate oxide thickness, which actually is only a few atoms thick, and controlling this thickness (much less reducing it) is one of the main motivating factors behind going to the FINFET non-planar transistor design for 14nm.
Imagine inventing something like this and coming up with a way of building them. THAT's what boggles the mind!
cipmars That's why the three physicists which invented the first transistor were awarded with the physics Nobel price.
genkiadrian I didn't know that, but it's 100% desirved.
cipmars Yes that is true but everything stems from somewhere. Prior to the modern transistor was the vacuum tube triode which was much bigger and used different principles. Even the transistor has changed so much since it was created. THAT's why electronics is so awesome!
Hana Zener Diode That is why it is a pain in the a** after almost 60 year of doing this. I started a boy. :) Now learning SDR. Not sure why, but it is fun.
Jerry Hubbard Should say "as a boy".
my teachers could never explain a transistor this clearly. as a result ive been afraid to design circuits with them because of the PFM (pure f__ing magic) associated with their operation. thank you
Can you do more quantum physics please.
Best lesson on semi-conductors I've experienced online or in class. Thanks EEVlog!
Yeah! It seems like they want you to know all of the math, instead of practical theory, and applications.
Wrong. The space charge region between Collector and Base gets BIGGER, which "absorbs" electrons diffusing across the Base region, transporting them automatically over to the collector.
I need more Fundamental Friday in my feed...
holy shit, since when does firefox support 1080p50, i cant really figure out whats clearer... the video framerate or the explaination. thx dave
Absolutely Superb!! Best explanation I've seen.
Moreover Australian accent made it just PERFECT!!
Thank you mate!!
Thanks for the video, but there is one misunderstanding. 15:50. The depletion region is formed by the negative ions in the p-type/substrate side and the positive ions in the n-type (source) side, not electrons and holes like what you mentioned. The negative ions (acceptors) are created in the p-type/substrate side due to the gain of electrons diffused from n-type, whereas the positive ions (donors) are created in the n-type/source side due to the loss of the electrons which diffuses to the p-type/substrate side. Hope you can add notes there (15:50) to not let people misunderstand this. Thanks.
I LOVE! all your videos. Can you do one on how to find total impedance of a parallel RCL circuit im so lost and cant seem to find a good site that explains it.
I would LOVE to see a nice video about how to drive MOSFETS and BJT, including high currents, and inductive loads with "soft off" AWESOME video!!! Learned a lot! Thanks!
WRONG, WRONG, WRONG FOR THE BJT !!!
BJT NOT only is NOT an amplifier, but also it's exactly the oposite: a strangler !
BJT does NOT amplify, it simply controls.
Amplification rate ? WRONG ! Should be called sensitivity rate.
Cheers !
Dave, you are the engineer par excellence, I love fundamental Fridays, I really learned something. thanks you are DA Man!!!
Hey, you don't want to say "BJT transistor" because that's like saying "bipolar junction transistor transistor." It's the same kind of error as saying "PIN number." The same goes for "FET transistor," of course. So will you please just say "BJT" or "FET"?
You explained in 10 minutes what took a whole semester in college... and I actually understand it better. Well done.
Kevin Sirois
(can't reply)
MOSFETs use much less power (high input impedance; at low frequencies, essentially infinite). Easier to scale down (photolithographic). Easier to produce.
Less power makes for less heat.
A point I never understood. Why do we call it amplifying? A transistor does not generate (or, amplify in your terms) amper out of nothing, correct? It all depends on the max amper that supply provides (collector lead), after all. It cannot exceed the total amperage of supply, right? Or, it is called amplifying in relation to base amper level/ratio?
Could we rephrase that it adjusts the amper at collector to emitter, based on the amper level on BASE up to the max of suppply. Thanks,,.
The atomic section has a few huge gaps.
-why can they join ? Electrons required in each shell.
-why can electrons move ? You mentioned "electrons moving freely" but an atom missing a stable amount of electrons will always be attracted to electrons in/for the amount of electrons it's missing.
-a way to look at N vs P. Stable amount of electrons is neutral. Electrons are negatives. Extra electrons make it more (N)egative. Missing electron(s) make(s) it less negative, thus (P)ositive. (Although, I believe in actually chemistry, it's the opposite.)
Also, if it's possible, can you dumb down "how a transistor works" even further. I've watched many videos, read two things, and had someone explain it to me, but everytime I think I understand something, the rules seem to change on me. So, after years (3+) of halfassly investigating, I'm still at a lost. Like, maybe include some sort of mnemonic (letter-word-concept association) for each gate/end, and maybe have some more (numbered) action lines as well. I know there are many types of transistors, so, explain in a way that can be adapted to most (if not all) types of transistors, if you can.
And thanks for providing an educational option (ie; teaching from a different perspective - perspective, video, text, animation, et cetera).
An intresting thing is that FETs were basically invented (and patented) quite a couple of years before BJTs but due manufacturing difficulties could not take off. What would the world look like if in the 50s and 60s not much progress had been done on BJTs but on FETs instead? Anyways, I don't think that PNP and PMOS as their opposites would be very intresting to see too, but as a good addition the JFET would be intresting, especially since many people love to use them for amplifiers, and how they can (or can not) be replaced by depletion mode mosfets.
16:37 about explanation of MOS FET. You are wrong. Field effect between substrate and gate, not source and gate. The substrate and source are internally connected in the device.
Thankfully Tek is advertising $100,000 oscilloscopes 10 times an hour, I’m sure they are flying off the shelves... I guess that’s better than having the sales weasel waiting in the lobby...
The next generation will be Sphere - Kinetic with moving atoms inside, not the static, here is the beginning : > KL03: Kinetis KL03 - 48 MHz, Small Form Factor, High Integration, Ultra Low Power Microcontrollers (MCUs)
Hi Dave,Thanks for the videos.Can you please explain about PVT characterization and also on TT,FF,SS cornors.
THIS is educational!!! When you break things down to their simplest form and begin rebuilding them until you have a component, the principles of electronics really become evident!
Tired issues, like which type of capacitor is "best," become rubbish. Does that cap allow the electron to move from one atom to another? More efficiently than another?
I spend WAY to much time listening to the information Dave shares!!! Back to the money square, I go!!!
On second viewing I just noticed something that bugged me for years.
Starting around 2000, the Sandhill Road companies between San Pablo and the 205 were over run with special visa people (H12) or something. Can't remember.
What I do remember is that there where truck loads of "EE" guys who had Ph.Ds but had never heard of the partial doping region. It was just nuts. We would ask them in interviews how a BJP amplified a signal and they had no answers at all. It was not on the third world Ph.D. script I guess.
We complained to management and were sent to HR with a racism write up or something. I actually quit the field entirely and went to law school at night and rebuilt my career. I could not personally live with the loss of basics that had become the standard.
Years maybe decades later I was at a round table dinner with Gordon Moore. I asked him about those days and he laughed at me. Walking out later that night he said - this I never will forget - "now you know why we located Copper Mine (a critical Intel R&D lab) in the middle of Oregon."
My interpretation was that he meant that the fundamentals do matter and that the insane Sandhill Road Venture Capital driven model was just filling slots with bodies, not people. Not sure of course that he would say it like that. I was in no position to press the point.
Funny that you nailed the key issue right up front. After 20 min. of UA-cam searching on how BJPs work found me not one that even mentions this critical region of the junction. I wonder why.
3 years on and very very informative. Ironically, 14nm processes by Samsung Alliance like 14LPP have a gate length of 30nm. I think finfet doesn't follow the old tradition of decreasing the gate length/pitch (looking at intel's early 22 node too gives some insights) and... Well, Intel is struggling with 10/7nm now.
Moore's law, created by an Intel employee and now Intel say they are unable to adhere to it anymore because they are struggling with EUV.
Hi,
thanks for your videos which are really helpful,
I have a question if you please help me in as I spent quiet time in searching and experimenting about (actually I'm a civil engineer, not electronics)
the question is:
I have a very weak signal 27MHz to amplify and I tried several ways to amplify using a transistor (not an IC) but the first stage transistor circuit is not giving the expected amplification, the last info I got was that the resistors selection needs to allow such week signal to trigger (turn on) the transistor (impedance issue) for which:
- can you please advise me what would be the best range of transistor (lower - like hundreds of ohms - or higher - kilos or megas of ohm-)?
- is there a better way to amplify weak signals?
thanks a lot in advance.
About the MOSFET: you made a wrong drawing. The source is connected to a positively doped area that contacts the substrate and it forms a body diode along with the drain... And a MOSFET consumes current through the gate actually. Make a small design with a light bulb in the drain of a MOSFET and measure the gate current. It will be constant as long as the bulb is lit. And now explain that!
Dave,
Great simplified explanation of how transistors work. But I think you are pronouncing "planar" incorrectly, and I'm Australian too. In my opinion it should be pronounced like "play-nah" rather than "plan-ah" so that it rhymes with trainer, explainer, and entertainer.
source: en.m.wiktionary.org/wiki/Rhymes:English/e%C9%AAn%C9%99(r)
Hi Dave, can you please make another video about transistors (BJT and MOSFET) where you can explain on how to actually correctly read a datasheet? There are so many notations and values and I just don't know where to look to check if the transistor meets my requirements.
Thank you!
This is the best explanation of transistors I have ever heard/read/seen.
Only question... When you say the holes/electrons "switch" at a P/N junction. Is it more the case that the electrons just jump across from the N type and fill the hole in the P type material. Effectively blocking the holes. I cant conceive of the holes "moving" An electron shell cant gain an extra hole can it?
Thanks Dave :)
When I trained with the P.M.G. in the sixties we were taught that the old conventions were wrong, that current flowed from -ve to +ve and that electron flow was the new future in electronics. Easily understood and remembered when you study the operation of a vacuum tube amplifier. At the same time, Cycles per Second (CPS) was being dispensed with and replaced with Hz (in honour of Hertz). Voltage was to be represented by E (for electromotive force) in lieu of V. So, I = E / R. Out with the old and in with the new. Hole movement was theorectical, the electrons moved into the holes leaving holes behind them. In a theatre, there's an empty seat (hole) at the far end of the row. Instead of the new patron (electron) stumbing all the way down the row, it's easier for everyone in the row to move to an adjacent seat. The empty seat (hole) effectively moves to the other end of the row but it's really the patrons (electrons) that are moving. Without a discussion of manufacture and the actual fusing together of the PN layers in an oven, there is no clear understanding here of how the depletion layer forms. These aren't pieces of silicon that are just placed next to each other, they are FUSED. The depletion layer creates an actual potential difference that must be overcome by external forces before current can begin to flow. When and why did the electronics industry revert from the NEW sixties conventions and decide to fall back to confusing ancient conventional current flow and also, bring back "V"?
Awesome dive into into the physical world of transistors!
So, are FETs symmetric or not? The schematic symbol suggests that it is not, but from the drawing it looks pretty much that it is. Dave, can you please briefly explain where symmetry is broken in physical implementation of FETs?
With some luck moors law will still hold for the next 2 decades.
Some of the professors at my local technical university have been successfully working at the Karslruhe University on way smaller kinda mechanical switches.
Collector and Emitter are made from a silver-base, and the actual switching happens with silver-atom bridging the tiny gap.
Not entirely sure how it works, but those switches are insanely efficient - and they can carry a mindbogglingly huge current in the range of micro-amps over a single atom.
Would be nice if they gave some more coverage on that.. and they didn't mention it this year sadly.
I still don't see (for the BJT) how the base emitter current is amplified into a higher collector current...? Somehow, what, 5 electrons entering the emitter results in, what, 100 electrons going through the base region and causing 100 holes to move into the collector region...? What accounts for that?
Just want to point out that the doping on the collector side is light, not heavy. Normally, the doping level is as such: Emitter >> Base > Collector. This is so that when in reverse bias, there won't be a large current flow. If you make a symmetrical BJT, then forward and reverse bias will cause same current flow magnitude.
EEVblog thank you for the explanation , but i have a question about the threshold voltage of the CB junction , practically I notice that collector voltage isn't responsible for overcoming the CB depletion region but depends on the Ib current , ( I mean that even if the collector voltage is small , the transistor will act as a short circuit if we applied large Ib .... and with a very small Ib - after Vbe > 0.7 - even with a large collector voltage the transistor will act as a high resistance because of the CB depletion region ) so instead I think it depends on Ib
Great, Thanks! I'm stil struggling with the BJT, but I guess I'll just look at it again until I get it.
Please do a followup video about popular transistor configurations, basic building blocks. Many people struggle with that. I've done several commercial designs and still could use a revision.
So I have TWO questions:
1. What happens to the characteristic behaviour of the circuit with increased temperature (i.e. does it become more or less efficient or something else?), and
2. What is the nature of either transistor type with aging (after being used for many years - i.e. - what is the mode of failure or problem due to aging??
thanks
Obviously BJTs and FETs don't work *exactly* the same way, or they wouldn't have made the two types (and you already said that one was for current vs. the other for voltage).
I hope you know what you are talking about Semiconductor Device Physics
Dave wisely avoided the confusion surrounding the direction of the arrow. Scientists at Bell Labs incorrectly depicted "conventional flow" (plus to minus) when everyone who has ever worked with vacuum tubes knows that electron flow is from minus to plus.
Heh great video Dave!
What I would do in the past is just match up the numbers and not really think of how the things work.
Today I wondered how leg transistors work and this video was on the first page of youtube.
You certainly explain stuff so even this dummy can understand it. :)
I am learning about how stuff works here recently as I am working on a custom pinball machine and no not one of those toy kinds.
A full sized 52" long pinball machine. ;)
I think the mechanical and circuit boards will be the hardest.
And I would send it off to you for Mailbag day but that would be costly and right now still in development as well.
You get 300USD if you build 100A 100v dc dc buck.
I finished electronics engineering school and noone in class could build one, but claim to know how a transistor work.
University take MONTHS and fail to explain this subject as clear as Dave did
I would love to see a "fundamental" about how to use FET transistors, what they are good or bad for and "traps for young players". Knowing how they are made is good I suppose but there is more to it than that. Why is the arrow pointing in the wrong direction for example (as compared to BJTs)?
I have worked in electronics, both as a career and as a hobby for more than 40 years. Even after reading several articles and books about how transistors work (remember the old RCA transistor manual?), I never really understood it as clearly as your explanation here. Thanks.
Sorry to say but very disappointed by your explanation of the BJT in this video of all others which have been awesome.
many instances you kept contradicting yourself.
so is it true that between gate and source on the mosfet it is like ceramic capacitor and does the voltage on the gate regulate the resistance between drain and source or what? and also why is that parasitic diode between source and drain
Very good explanation. FETs are new since I studied Physical Electronics in the 60's!
What about The insulated-gate bipolar transistor (IGBT)? Can you make the video on this??
Dave knows alot of chemistry wow, well elaborated usefull video indeed,kudos!
If this was taught at my university like this, I would surely be more curious back then when I was younger. We were all about smoking resistors and blowing caps.
I want to try and build semiconductor chips at home. Yeah, the performance will be horrible (a piece of oxidized aluminum will work for the gate and insulator, right?) but I want to see if I can do it.
i hate it when physicists talk about "holes moving around". holes don't move, electrons move. when an electron move from an full orbital to a partially full orbital, it leaves a hole behind and fill a previously-unoccupied hole. it confused people when they talk about moving holes, when the actual explanation is much simpler and much more sensible.
Dave would make an awesome classroom instructor!
You explained it way better than my Iranian instructor(No Bias), back in the day, Dave. Many thanks!
I think it was a great idea to take on this task but suggest you consider re-doing it after cleaning up some of the issues. In particular I suggest you re-read the section in your physics book on PN junctions and structure the BJT discussion terms of Base *current* flow. In particular, "Vbe" only matters as the forward voltage of the transistor's PN junction, and that what *does* matter is that you structure current into the base such that that current times the transistor's gain (Hfe) result in the transistor being fully 'on'. From a physics perspective you are pulling enough electrons into P layer from the collector side, that there is always a path for an electron to go from the emitter to the collector. The P layer is essentially not there (this is called the saturation current, and for a switching application its your target current)
Also suggest that you replace silicon 'matrix' with silicon crystal. That would align better with other sources people might go to after watching this for more information.
Laws of physics & can't get much smaller but one could alloy the substrate with dna to make a wireless field type connection, another type of nano induction. How? It's like growing a garden.
Интересно , кто-же ставит дислайк ? Спасибо за видео ! I wonder who puts the same dislayk? Thanks for the video!
Far far too much 'chatter and scribbles'. Good sleep medicine!