1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

Understanding Transistor

Discussion in 'General Electronics Chat' started by sdmuashr, Mar 24, 2012.

  1. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    1.) Common base or emitter, e- from the emitter are collected by collector. Emitter follower works the same way. The origin of Ic is the same for all 3 configurations.

    2.) Why is it not convincing? FETs have current gain. But in a FET controlling Ig (gate current) is a poor way of controlling Id. A CCS (constant current source) across g-s of a FET results in a voltage ramp which will inevitably exceed the breakdown voltage of the g-s oxide. A FET cannot be current driven & must be voltage driven. I always regarded FETs as voltage driven because the device does not lend itself to current control/drive. Now regarding current gain of FET, it is different at every value of frequency. Unlike a bjt, the gate current in a FET, Ig, is all displacement current, not conduction current. The Ig is needed to charge/discharge g-s capacitance. At 2.0 kHz it requires twice the Ig value compared to 1.0 kHz.

    Hence for FETs, we do not define a current gain, because it varies w/ frequency. Instead we define a transition frequency, f[sub]t[/sub], where current gain is unity. At lower frequencies, current gain is just f/ f[sub]t[/sub]. So a small signal FET may have a 300 MHz f[sub]t[/sub] value. When frequency is 300 MHz, current gain is 1, or Ig = Id. To implement a 1.0 mA current swing in the drain requires a full 1.0 mA current swing in the gate.

    At a frequency of 3 MHz, the current gain is 100 (300/3). So for Id = 1.0 mA, Ig = 10 uA. Down at a mere 3.0 kHz frequency (human voice for example), the current gain is 100,000. If Id = 1.0 mA, Ig = 10 nA. Pretty straightforward if you ask me. FETs have current gain, but that does not make them "current controlled". Again, if we examine the semiconductor physics of a FET, it is charge controlled just like the bjt. FET is a majority carrier device, bjt is minority carrier device. Voltage control for FET, & current control for bjt work great until the speed approaches the internal time delays, or for switching mode operation. Then the charge control model is best for both devices.

    I design many motor drivers, LED drivers, power supplies, etc., using FETs. Believe me a FET has a finite current gain. It requires gate current to switch states. To change Id we must change Vgs, but to do that we must first change Ig. Ig is necessary but it is not the parameter we directly control, Vgs is directly controlled, hence the VC label. Am I clear?
     
  2. Winterstone

    Winterstone Banned

    Joined:
    Mar 24, 2012
    Messages:
    536
    Likes:
    15
    Location:
    Europe
    No, not at all.

    to 1.) Excuse me, but I am not a beginner. I did already know that e- from the emitter are collected by the collector. Are you joking with me? Is this your answer to my request?

    to 2.) My definition of gain: An input signal controls an output signal and the ratio of both signals is called "gain" (VCVS, VCCS, CCCS, CCVS). It seems you have another definition. Why not?

    Thank you very much for answering my questions.
     
  3. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    1) I thought we were discussing the raw bjt modus operandi, not a composite amp stage.
    2) Again, if "gain" implies input to output of an entire stage, then of course, said stage may be CCCS, CCVS, VCVS, or VCCS. No argument. But I thought this discussion was to examine the raw bjt nature.

    I need to know if you wish to discuss the composite bjt amp stage, bias network & all, or just the raw bjt. When parts are added, then the network can be made to perform as VCCS, CCCS, etc. I was discussing just the bjt.

    Not to be derogatory, but what is your career experience? Practitioner, instructor, HS/college, application engr, R&D? What is your academic coursework in said subject? E/m fields, electronic theory, solid-state physics, etc.? Also 4 more simple questions.

    Is voltage in general the cause of current? Why do power companies produce constant voltage instead of constant current? I believe the answer to these 2 questions is the sticking point.

    You & Ratchit have been presenting arguments that Vbe is the main player. I acknowledge Vbe playing a role, but is not what we use to control Ic. Every device in the world needs current & voltage under dynamic conditions. It's impossible to eliminate either one. Does an incandescent light bulb optical output depend on V or I? Is a light bulb a CC light source or VC light source? How about an LED? These answers will explain to all reading this, why some people cannot accept current control for any device. Please answer my questions as they are simple & should not be misconstrued. I will answer any question you put forth & elaborate further on earlier questions. BR.
     
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. Winterstone

    Winterstone Banned

    Joined:
    Mar 24, 2012
    Messages:
    536
    Likes:
    15
    Location:
    Europe

    Claude, one hour ago I was sure not to respond again. Because I've got the impression that you are not willing to discuss with me on a serious basis.
    OK - I changed my mind.
    And I will, of course, answer your questions because - as I have indicated - a fair, serious and objective discussion between engineers should consist of question and answers.
    However, please understand that at first I would like to have answered at least some of MY questions - in particular question Nr. 1 and Nr. 2 of my last post.

    OK - there is one question in your last post I can answer now:
    I need to know if you wish to discuss the composite bjt amp stage, bias network & all, or just the raw bjt.

    When we speak about the mechanism how the current Ic is controlled we have to assume that the BJT is equipped with a voltage between C and E and a kind of bias network which allows the BJT to act as a controlled current source. Is this a new information for you? This was the basis for all of the contributions in this thread up to today.
    So - what do you mean with "composite bjt amp stage" in contrast to what you call "raw bjt"? Does the "raw bjt" behaves different when it is the heart of a complete gain stage?
     
  6. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    So here are your 2 questions which I will respond to:


    Quote: WHAT is WRONG in my explanation (Question 1)?

    You certainly will agree that a technical discussion (in case of different views) should mainly consist of question and answers.
    I have asked some specific questions - no answer up to now.
    I have asked if your explanations apply to the common base configuration only and to show or describe a corresponding circuit - no response.

    What was the reason for you to speak about the the common base configuration instead of the common emitter (which was the basis of the discussion in this thread up to now)?

    Now I kindly ask you: How do you explain the control function of the common emitter configuration - based on your theory that the emitter current is the first cause of controlling Ic (Question 2).
    I am sure that this would clarify some of the possible misunderstandings.

    Can you please respond to both of my questions 1 and 2. End Quote.

    Actually I did not say that Ie is the 1st control chronologically. Sue is what ultimately controls Ie. But you asked about a CE configuration. If there is no emitter degeneration, i.e. no negative local feedback, then an increase in mic output V & I results in increase in ib/ie. Note - Ib/Ie are fixed dc bias, ib/ic are small signal ac variations. In CE mode signal goes to base side & returns through ground which is emitter side since this example is un-degenerated. The increased I transits through b-e junction & increased V appears across base resistance and/or coupling cap. Until charges transit into b-e junction Vbe cannot immediately change as Cdiff requires charging current to change its voltage.

    As ib/ie increase more holes are injected from base into emitter. As a result the depletion zone incurs more holes on the emitter side as they require a little time to recombine w/ e-. Likewise the original Ie was due to bias network, primarily supplied by Vcc (collector side supply). The signal component of Ie, herein ie, adds to Ic, & more e- are injected from emitter into base. A small fraction of these e- recombine in base, while the rest transit into collector. The base region side of the depletion side incurs an increase in e-. Thus the depletion zone has had its charge increased on both the base & emitter sides, due to increased mic output. The additional charges due to small signal from mic add to the charges from the bias network resulting in increased charge on each side of the depletion zone.

    Thus the b-e electric field is modified since these additional charges have their own E field. By definition, Vbe is the integral of the E field times dl, the path from emitter to base. I cannot accept your model that an increase in Ib/Ie requires that Vbe change first because that cannot happen. The b-e junction is capacitive, & Ib/Ie changes have to precede Vbe changes.

    As The increased number of charge carriers transit through the junction (b-e), Ic must increase since more e- emitted results in more e- collected. In addition, Vbe incurs an increase. It is impossible to raise Ic w/o raising Ib/Ie/Vbe all 3 in unison, but Vbe lags slightly behind. Vbe is not the controlling factor for Ic, rather Ie is. This is hard for some to accept because some cannot conceptualize anything happening unless a voltage controls or drives things.

    I've used the CE topology & explained bjt action re sequence of events, changes in Ie/Ib/Vbe. Let me just clarify that w/o emitter degeneration the network I just described has an undesirable property known as "beta dependent". W/o Re gain varies w/ beta, which for many applications is not desirable. With Re sufficient in value to adequately degenerate the network, beta variations are almost nulled out. The price is lower overall gain. If you wish we can discuss the degenerated amp stage, that's up to you. I hope I've answered your question & I await your response for my questions. Again I will elaborate if you feel my answer needs more explaining. BR.
     
    Last edited: Aug 13, 2012
  7. Winterstone

    Winterstone Banned

    Joined:
    Mar 24, 2012
    Messages:
    536
    Likes:
    15
    Location:
    Europe
    Claude, your long explanation cannot satisfy me because it starts with the statement "... results in increase in Ib. At first, this is just a claim. And - as far as could understand - you are trying to justify this claim with a capacitor model: V lags I. Is this interpretation correct?
    In this context - I see a kind of contradiction (as follows):

    Quotation#106: Actually I did not say that Ie is the 1st control chronologically.
    Quotation 96: The increase in Ie, which is caused by Sue & mic, accounts for Ie change, Ib change, Vbe change, & ultimately Ic change. Vbe is merely incidental

    In the following I give you an excerpt from an internet contribution. I do not blindly trust all articles which are distributed by internet - however, it is interesting and - of course (!) - it supports my view.
    The full article is available via the link given in the first line.

    http://amasci.com/amateur/transis.html

    Excerpt:
    PPPPPPPPPPPS
    This article apparently has triggered extensive debates if not flamewars on multiple hobbyist forums, newsgroups, and WP. It's as if many people see Ic=hfe*Ib as holy, while Ic=Is(e^Vbe/Vt) is dark blasphemy which must be kept from the delicate ears of children. The cause of controversy is fairly obvious: at early stages we're all taught that BJTs are current-controlled devices, and only in later engineering physics courses is this claim held up to questioning. Also, the current-control viewpoint works just fine as long as we give it lip service and then turn around and use Spice programs, or as long as we never look too closely at details of the inner workings of the physics. This situation leads most people to firmly decide that Ic really is affected by Ib and not by Vbe. (Or perhaps they believe that, in diodes, the Vf diode drop is caused by the current.) I note that these debates all seem to feature typical flaws:

    Primary is a sort of backwards reasoning: first we take a stance for (or against) current control. Then we hotly defend that stance against all comers while cherry-picking the supporting evidence and ridiculing all contrary evidence. But that's not reason. That's religion or politics. It's how pseudoscientists operate. Science is the very opposite: in science, first we try like hell to avoid rigid preconceptions and emotional biases. We take no stand for or against. Then we honestly ask which side is actually right: ask whether transistors are controlled by voltage or by current. And then we take the answers seriously, without desperately twisting facts to avoid losing face in public, without breaking sweat while having steam shoot out our ears, and without descending into mild insanity triggered by psychological denial that we're genuinely on the 'losing' side. :)

    Second problem: is the BJT current-control viewpoint really held by all scientists and engineers everywhere, while voltage control is terribly wrong? WRONG. :) Look at the Ebers-Moll section of Sedra/Smith, Horowitz/Hill, or most any engineering text. Ask some engineering authors (many are online!) Ask semiconductor physicists. Ask professional engineers. Their answers will surprise you. And don't ask them about abstract models of black-box transistors, ask about the topic of this article: the internal physics: ask whether Vbe or Ib determines Ic, ask what is the origin of the Shockley equation, and what role does Ib and hfe actually play?)

    Third problem: transient Ib current causes confusion. When Vbe changes value, charges must move during the changing profile of the depletion layer, and this requires a momentary charge flow in the base lead. I've seen several people proclaiming that this proves that Ib "causes" Vbe. No, that's a clear attempt to twist facts. The voltage across a capacitor sitting on a shelf isn't being "caused" by any continuing current. In truth it just means that a changing capacitor voltage always requires a momentary current. To explain the high frequency behavior, you need Gummel-Poon and not just Ebers-Moll. This issue doesn't apply to the low-freq or DC case of Ic=Is(e^Vbe/Vt) where Vbe is constant, yet the values of Vbe, Ic, and Ib are all connected together. The Ebers-Moll model shows that Ic is proportional to Vbe, but in order to see this, we must ignore the Emitter-Base capacitance and the transient currents which charge or discharge the EB capacitor during changing conditions.
     
  8. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    That site is heresy. Bill Beaty & I discussed this at length on forums & in private email, & I cannot fathom how he believes this stuff. First of all, I have the Ebers-Moll paper & gave it to Bill. He then conceded that Drs. Ebers & Moll flat out model the bjt as current controlled! His response was "Just because Ebers & Moll think a bjt is current controlled doesn't make it so"!!! But his site still relies on E-M paper as his "proof", a paper he has already dismissed as not reliable.

    Regarding his cited equation:

    Ic = Is*exp((Vbe/Vt)-1), I don't have heartburn as long as alpha is included. Why do these voltage people always omit alpha? Also, "Is" is the saturation current for a diode p-n junction. A bjt has 2 junctions, b-c & b-e. Due to differing doping densities their sat currents differ. For the b-c junction it is "Ics", for b-e junction it is "Ies". The correct equation should read as follows:

    Ic = alpha*Ies*exp((Vbe/Vt)-1).

    Although alpha is typically 0.98 to 0.998, & its omission results in a very small error, it is all important. I will write up a paper & post it later, deriving Ebers-Moll equations from Shockley diode law & transistor equation. Alpha must be included. Vbe is functionally related to Ib & Ie per Shockley. When the e- from the emitter transit through the base into the collector that action is Ic = alpha*Ie. But Ie is Ies*exp((Vbe/Vt)-1). So Ebers-Moll is derived from Shockley diode equation plus transistor action equation.

    Regarding Bill's question "What role does Ib & hfe actually play?" Here it is, hfe = alpha/(1-alpha). If alpha is zero, then hfe is zero. Ic = 0*exp( ). Ic cannot be expressed w/o alpha, which is related to beta per above. Ic can be computed 3 ways:

    1) Ic = beta*Ib

    2) Ic = alpha*Ies*exp((Vbe/Vt)-1)

    3) Ic = alpha*Ie.

    A p-n junction is described as:

    4) Id = Is*exp((Vbe/Vt)-1).

    I will leave it to the reader to verify that equation 2), the darling of the critics, is merely a combination of equations 3) & 4). One cannot express Ic w/o the use of alpha, or its relative, beta. Equations 1) & 3) express Ic w/o Vbe or Vt. They are, of course implied since Ie & Ib are related to Vbe.

    Bill concludes with Ic proportional to Vbe, which we all know already. He then pulls this one: "In order to see this, we must ignore the Emitter-Base capacitance and the transient currents which charge or discharge the EB capacitor during changing conditions. " Indeed!!! Those transients are what conveys to us that Vbe is incidental & not what controls Ic. Ratchit has stipulated that he wants to keep this in the dc domain & avoid transients altogether. Transients easily disprove the Vbe causal hypothesis. Bill loves to point out that "Ib does not cause Ic". That is so well known, the majority of EEs who use current control never state that Ib is the cause of Ic.

    To collect e-, we need a Vcc supply reverse biasing the b-c junction. We need emitter current so that e- can be injected towards base then into collector. Ib cannot do this alone. If Vcc was removed, forward biasing the b-e junction results in Ib/Vbe/Ie with no collector action at all. Transistor action requires all the above. Ib alone can never be the cause of Ic. I never suggested otherwise. Straw men at its best.

    I like Bill, I just don't get how he can believe his own teachings. If you wish to private message me, I can forward to you his email where he concedes that Ebers-Moll using current control model, but it shouldn't matter anyway because Ebers & Moll don't necessarily have it right. I have it in my archives if you're interested. Cheers.
     
  9. Mr RB

    Mr RB Well-Known Member

    Joined:
    Jul 22, 2008
    Messages:
    4,716
    Likes:
    194
    Location:
    Out there
    I think I am going to bow out of this thread as others have done, as this is taking on all the momentum of an argument about religion and people are too busy pedantically defending what happens inside a transistor to understand how to control the transistor. :D

    As a parting shot, I think some people need to have a careful think about what "controlled" means. This is NOT a forum about theoretical semiconductor physics, it is an electronics forum. In electronics we "control" transistors.

    Let's say I have another device (not a transistor), say a serial controlled light dimmer. I apply serial data to its input, and it adjusts phase angle and "controls" the current through the lamp filament. People could put forward all sorts of complex arguments about the inside of the component, that the lamp brightness is actually "controlled" by phase angle, or is actually "controlled" by a combination of phase angle and the PSU volts vs filament ohms.

    But the FACT is that as a component, used by someone in electronics on an electronics forum, it is a serial controlled device. The mechanism of whatever control systems and voltage effects happen inside the device is fairly irrelevant, and only of importance to pedantic nitpickers.
     
  10. Winterstone

    Winterstone Banned

    Joined:
    Mar 24, 2012
    Messages:
    536
    Likes:
    15
    Location:
    Europe
    Hello to Claude A. and Mr Rb,

    yes - I think I can (partly) agree with Mr Rb. The "discussion" (I have explained why I sometimes have problems to call this disput "discussion") has come to a point where it seems impossible to come to common understanding. In this context, I remember (and you all know this) that there are textbooks supporting current control and other supporting voltage resp. charge control.
    The problem of this thread is that each party simply claims something - without being able to deliver a convincing proof. And the reason for this dilemma - for my opinion - is the relationship in general between voltage and current.
    More than that: What is current? Does it really flow? Or is it the charge that flows?. OK - let's stop at this point.

    To Claude:

    To be honest, nearly everything in your last reply is a repetition of commonly known facts. Nobody will deny that there is an alpha which must be included in the equations.
    I don't know if alpha was forgotten in the cited equation or if the meaning of the symbol for the saturation current is Is=alpha*Ies.
    But I do not consider this as important.

    And also the second part of your reply starting with 4 basic and well-known equations does not help to bring the discussion one step further.
    Why do you stress the importance of Vcc? Was there any disagreement about Vcc?
    As far as I understand you claim that Bill Beaty is wrong. OK - but why? Where is his error?

    This question touches the problem of transients and high-frequency signals.
    And your comment is: Transients easily disprove the Vbe causal hypothesis
    You will not be surprised that this argument cannot convince me.

    It happens very often in electronics that we have
    * to idealize some characteristics, or
    * to neglect parasitics, or
    * to suppress second-order effects

    with the aim to isolate and identify the desired properties of a DUT.
    As an example I mention the fact, that NO formula (gain, input/output resistance, ...) is correct by 100%. We restrict ourselves on the main parameters and accept errors.
    Thus, I think during this discussion we really should have only rather low frequencies in mind. High-frequency effects can conceal the phenomena we are interested in.

    As far as I remember, it was your key argument pro current control that Ib would lead Vbe (because of the diffusion capacitance).
    After corresponding simulations I cannot confirm this observation (I have asked you already: Simulation or measurement?).
    For frequencies up to the lower kHz range Vbe and Ib are in-phase. (BJT Spice model in active region, Ic=1.5 mA)
    (By the way: this is a good point to demonstrate that high-frequency effects influence/disturb the effects we want to study).

    Finally, at the moment I tend to follow Mr RB's suggestion to leave the thread.

    W.
     
  11. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    If you're old enough to remember the 1st time O.J. Simpson was tried, double homicide charges, each side claimed the other just didn't prove their case. I do not wish to reargue the Simpson case, but I felt that the prosecution had a case at least as strong as that which many other DA's had & won a conviction. The mere fact that my writings can be disputed does not make me wrong unless one can show where my info is flawed. I've shown in spads Bill Beaty's error. Reread my post. He assumes way too much. He assumes Vbe is causal, & if one tracks the transit of charge through the be junction, one can see that Ie & Ib can change due to external stimulus, with Vbe eventually catching up. I'll submit a diagram tonight if time permits.

    My parting though, when Faraday induction comes up, the same people who insist bjt is voltage controlled also insist that induction results in constant voltage generation, current determined by load resistance. They will not hear me out when I tell them that a very low load resistance results in constant current operation. These same people do not accept the OEM op amp mode known as "current feedback". They insist it is really voltage feedback modified. In general, these critics cannot accept anything being controlled by a current. An error signal has to be a voltage. A control signal has to be a voltage. They even deny the existance of current sources! Attempts to explain things to them are met with stubborn opposition & they counter with dogma.

    But am I viewing things from my narrow current control viewpoint> Do I have a tendency to define all things w/ current? People on my side who model the bjt as CC, also model FETs as VC, IGBT as VC, vacuum tube as VC, high-Z voltage feedback op amps as VC. I don't force everything to fit into a current control model. My opponents on the other hand, believe that nothing changes unless some voltage changes first. That is not true. Any examination of charge motion, fields, boundary conditions affirms my point.

    As far as Ib/Vbe/Ie being in phase as observed on a scope, that is expected at the low frequencies (lower kHz) that you mentioned. But even being in phase is ample proof that when external stimulus increases Ib/Ie, that Vbe is increasing in unison. The increase in Ib/Ie is not a result of Vbe increasing, rather all 3 are a result of external stimulus. Increase the frequency & see what happens.

    Why does On Semi, Fairchild, TI, etc. model the bjt as CC when viewed externally, & charge controlled internally? I have stated repeatedly that CC is a good fit only externally. When considering internal storage, delays & such, the charge control model works best. BR.
     
  12. Winterstone

    Winterstone Banned

    Joined:
    Mar 24, 2012
    Messages:
    536
    Likes:
    15
    Location:
    Europe
    Hi Claude,

    Quote: I have stated repeatedly that CC is a good fit only externally. When considering internal storage, delays & such, the charge control model works best.

    May be I couldn't find this view behind the lines written by you in former posts (due to my limited knowledge of your language), but for me this is a new "sound".
    Perhaps we can meet each other using this formulation.
    But - don't you also think it is surprising that experienced engineers discuss about the physical working principle of a device that was invented 60 years ago?

    Quote: These same people do not accept the OEM op amp mode known as "current feedback". They insist it is really voltage feedback modified. In general, these critics cannot accept anything being controlled by a current. An error signal has to be a voltage. A control signal has to be a voltage.

    I am 100% with you that - of course - current can serve as an error signal. The current mirror which is fed with the current through the inverting input is a clear indication of this principle.

    Perhaps we meet again in another thread.
    Regards
    Winterstone
     
  13. ericgibbs

    ericgibbs Well-Known Member Most Helpful Member

    Joined:
    Jan 4, 2007
    Messages:
    21,187
    Likes:
    644
    Location:
    Ex Yorks' Hants UK
    hi Claude,
    I agree with your definition, its the explanation I was taught and its the principle I use and it has always worked for me in practice.

    Eric
     
  14. Ratchit

    Ratchit Well-Known Member

    Joined:
    Mar 12, 2008
    Messages:
    1,934
    Likes:
    78
    Mr RB,

    As is your right to do so.

    There is no religious aspect to this discussion. Religious beliefs are partly based on faith. The participants in this thread have beliefs based on what they think are scientific facts. The fact that the folks here are headstrong, and argue them passionately does not make it a religious argument.

    Whether right or wrong, isn't that what we are doing?

    Theory is the basis of practice. Doesn't every transistor circuit control a transistor?

    What they are saying is that control inputs can come from multiple sources. You already found out that both Vbe and temperature can control Ic.

    Not so. Treating a system or component as a "black box" has advantages, but also disadvantages. Knowing what goes on inside can be a great help to improving the circuit or application.

    Ratch
     
  15. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    A search of my post history will reveal that my CC model is strictly limited to "black box" viewpoint. The CC model works great when the bjt is operating in active mode, i.e. signal amplifier, at low frequencies, or current mirror, current source, Vbe multiplier, etc. in low speed. If the bjt device is to be operated at high speed, or toggled in the switching modes, cutoff & saturated, then the CC model does not produce good results. Again, the CC model works best when internal charge storage, delay times, rise/fall times, is small compared to the signal time constants, & this can be ignored. In these cases, only the charge control model, QC, is reliable.

    So we seem to have a partial consensus that the CC model is reliable for "black box" viewing of a bjt, when operating in active region at low speeds. CC is fine as long as we do not have to concern ourselves with the device "innards", i.e. stored charge, delays. For those applications, high speed active, and/or saturated switch mode, we must use QC model, as it accounts for internal physics.

    That has been my position since day 1. But notice that I make no mention of the VC model. At high speeds, the Ebers-Moll equation does not include storage, delays, etc. At high speed or switch mode use, VC provides no usefulness. Only QC works. Of course, no silicon crystal is flawless, there are lattice breaks, impurities included, etc. Even the QC model provides only limited accuracy. I'm glad we at least concur on that point.
    Let me emphasize that I do not consider this "religion". Also, it does indeed matter what goes on inside the device if you are at high speed or in the switch mode. When I design a discrete bjt gate driver for fast FET switching, I rely on my knowledge of the *innards* of bjts & FETs to successfully make a good driver. For low speed active mode, it doesn't matter. But I have at times been forced to rely on my knowledge of innards. Best regards to all.
     
    • Like Like x 1
  16. Ratchit

    Ratchit Well-Known Member

    Joined:
    Mar 12, 2008
    Messages:
    1,934
    Likes:
    78
    To the Ineffable All,

    Now I would like to say something about BJT control. Whenever I study BJTs, I read that they are manufactured with two types of materials, each of which have a predominace of different charge carriers. That is why they call a BJT a bipolar device. Then I read that the two different charge carriers diffuse into the opposite slab of material, and either annihilate each other, or continue on to the next slab, or exit a contact. Continuing on, I observe that different amounts of doping are used to control the charge concentrations, which in turn controls the diffusion within the BJT. Then I learn that there are depletion regions in the boundary locations of the NP slabs where diffusion cannot take place due to the absence of charge carriers. I also come across terms like minority carrier diffusion currents. It does not take long to realize that unlike a FET or a vacuum tube, the mechanism for BJT operation is diffusion. So what controls diffusion? All the texts agree that charges from the emitter diffuse to the base and are, for the most part, accelerated on to the collecter by the collector voltage. There is no disagreement that this process cannot continue by itself due to the back voltage, also called barrier voltage. This barrier voltage can be neutralized by the application of Vbe, and the diffusion process can continue. That is why the equations for Ic show up in Sedra and Smith as being dependent only on Vbe and not Ib. This control by Vbe does not mean that the relationship between Ib and Ic should not be used for calculation and design. It only means that Ib is an indicator of Ic and not a control.

    Some further opinions:

    A model of a device is just that, a model. It may help determine what a device does, but it does not necessarily disclose or prove how the device works.

    A storage charge may cause a current without any external voltage, because it has its own internal voltage.

    Driving a device with a current source is the same as connecting a large external resistor to the device. That is not the same as connecting only a pure voltage source. It takes a voltage to initiate a current, unless there is already stored charge present. If you drive a device with a current, you are effectively applying a voltage to it.

    The order in which events happen has no relevance in determining control. Control is determined by the physical cause of the changes.

    The topology of the external components connected to a device does not change the way the device works, only the way it behaves.

    Ratch
     
  17. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    Again, we do not "apply Vbe", it is a consequence of an external source providing an E field & charges drift as a result. The Ib/Ie currents do not depend on Vbe. Then you state " It takes a voltage to initiate a current," which I respond with - how does the voltage appear? Voltage does not cause current. A power source be it battery, generator, photodiode, etc., generates both. As long as you keep invoking the myth that "voltage initiates current" you will always view every electrical device as voltage driven. What about an LED, universally described as current controlled? We adjust light output by varying current, albeit there has to be a forward voltage. Is an LED VC or CC? Please answer.

    I think you have your own definition of "control". Reread my posts & I've explained in detail why Vbe is not the cause nor the control of the currents in the bjt. You make declarations based on pure imagination. This business that Vbe is what initiates Ib/Ie is flat out wrong. This is your view. Re Sedra & Smith, they acknowledge a relation between Vbe & Ic, but also between Ic & Ie, as well as Ic & Ib. You selectively focus all your attention on Vbe as the cause. I believe for you it is religion. Proof to the contrary is ignored by you.

    Once again, the CC model is not meant to cover internal physics. But the model which extends beyond CC is not VC, but rather QC. When using a bjt as a saturated switch only charge control works. How can VC help with switching operation in saturation? Lastly you say "If you drive a device with a current, you are effectively applying a voltage to it." That proves my point that you view voltage as more basic than current universally. You simply cannot accept that the 2 are inclusive, neither 1 more basic than the other. Here is my response - When you drive a device with voltage, you are effectively applying current to it. Unless the load resistance is 0 or infinite, you need both I & V. Driving a device w/ current only implies that should the resistance change, the source will output a different voltage in order to maintain the fixed current. Likewise, a voltage source does the reverse.

    Another faux pas on your part - "Driving a device with a current source is the same as connecting a large external resistor to the device". Not true. A 1.0 milliamp CCS plus a 1.0 kohm shunt resistor may be electrically equivalent to a 1.0V CVS plus 1.0 kohm series resistor, but not thermally equivalent. If open, the CVS dissipates 0 power, CCS dissipates 1.0 mW. If shorted, CVS dissipates 1.0 mW, CCS disspates 0. The 2 are not completely equivalent. Also, a true CCS does exist.

    Are you aware that the power company can just as well deliver constant current source to our homes? Spinning the turbines at constant torque will provide CCS operation. But they spin them at constant speed for CVS operation, & as a benefit we get constant frequency. You just cannot accept that current & voltage have no pecking order. Either one can be the independent variable, with the other determined by impedance.
     
    Last edited: Aug 14, 2012
  18. Ratchit

    Ratchit Well-Known Member

    Joined:
    Mar 12, 2008
    Messages:
    1,934
    Likes:
    78
    Claude,

    Are saying that "applying a voltage" to something is not correct? Of course a voltage difference causes a electrostatic field to form. How does that disprove that Vbe controls Ic?

    Still don't believe Sedra? Then perhaps Neudeck in attachment #1. Notice how both Ic and Ib both vary with Vbe and temperature.

    Countless ways. I believe we are talking about it being applied as a signal.

    If there is a conduction path, it sure does.

    Power sources may conduct current if there is a conduction path. Otherwise, only voltage is present.

    Voltage has to be present for current to exist, but voltage can be present without any current also.

    The current in a LED is physically controlled by the voltage across it according to Schockley's equation. The fact that the brightness corresponds to current does not make a LED current controlled physically. It still operates by bipolar diffusion, which is controlled by a external voltage.

    Your explanations do not make sense.

    I consider them well thought out.

    If wrong, they why do Sedra and Neudeck publish those equatons? Of course since Ic is almost the same as Ie, there is going to be a relationship. My focus in on Vbe controlling Ib and Ic is because Sedra, Neudeck, and other textbooks say that Vbe controls the diffusion process. This is solid evidence that that proposition is true. I don't believe it on religious faith, I believe it on solid evidence.

    Then why are you presenting it to disavow my claim?

    Models tell you what a device does, but not how it works.

    As I said before, we are firmly in the active region.

    More basic? No, both have the same relevancy. Voltage is necessary for current to exist. If a voltage is present, current may or may not be present depending on a conduction path. If a current is present, an external or internal voltage is present also. Neither one is more basic that the other, and they have nothing to do with whether a BJT is controlled by Vbe or Ib. Driving the base with a current or voltage source is irrelevant, because it only responds physically to the Vbe presented.

    I am looking for the first faux pas. You are right, they are not thermally equivalent. But that does not matter too much as to whether Vbe or Ib controls Ic, does it?

    So does that mean that at constant current, all the outlets will have arcs across them to maintain the constant current until something is plugged into them? The physical relationship determines what is the control, and in a BJT or LED, which depend on diffusion, the voltage is in control because it regulates the diffusion.

    Ratch
     
  19. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    With constant current, the switches are placed across the outlets. A switch closed is the off position. An appliance is plugged into an outlet which is across the switch. An open switch turns the appliance on. All the outlets are in series for a given branch. With CVS, the outlets in a branch are parallelled. If all devices are switched off, all switches are closed. All of these outlets have 0 volts across them, but the current is constant & no power is consumed. You have to reorient your thinking with CCS, as it is the counterpart of CVS.

    LED forward voltage does not control the current. The page you copied from Sedra & Smith makes no reference to such a concept. I cannot convince you that the external source provides the E field & energy to initiate conduction. The junction provides a barrier which the power source must overcome by losing some energy. Just as Id = Is*exp((Vd/Vt)-1), it also holds that Vd = Vt*ln((Id/Is)+1). Physics texts usually use the I-V form, but many refer to the V-I form of the same equation. Your whole crux is that after all diffusion has taken place, strange you never mention drift, the value of Vd attained is what dictates the current. If you insist on ignoring transients & staying in dc, then it is pointless to argue. In dc & low frequency operation, it is impossible to ascertain which mechanism accounts for this or that. Only pulsing quickly & measuring on a scope can give insight into what is happening.

    To summarize, when bjt is working in conditions where CC model is inadequate, such as high speed, or saturated switch, the only model physicists invoke is charge control. You seem to ignore that model & want to use VC for internal physics, which nobody uses. All Sedra & Smith invoke is the functional relation between Vbe & Ic.

    You don't even accept that power distribution can just as well be CC or CV. It isn't done because conduction losses are so much higher than insulator losses. Generating full current/variable voltage all the time results in heavier losses than full voltage/variable current. Also, you mentioned that if current is present there has to be a voltage. But the reverse is true as well. In static conditions a superconducting inductor can sustain a current w/o voltage, & a perfect capacitor can sustain voltage w/o current. But with ac, neither can independently exist. An ac wall outlet in the open state cannot have ac voltage w/o ac current. Just ponder on that.

    You don't consider current as anything more than the result of connecting a voltage across an impedance. In your mind every current is driven by a voltage. With that frame of mind, current controlled devices don't exist. Every LED maker describes LEDs as current driven, & all LED drivers use constant current using switching a FET with an LED, inductor &catch diode to keep the LED at constant current. Increasing brightness is done by increasing current, with an increase in Vd being incidental.

    I cannot keep harping abut this. If you believe voltage is what drives current, my talking is pointless. Good day.
     
  20. Mr RB

    Mr RB Well-Known Member

    Joined:
    Jul 22, 2008
    Messages:
    4,716
    Likes:
    194
    Location:
    Out there
    Darn you beat me to it! The LED example was going to be my next argument against the sillyness of trying to teach the BE diode junction as "voltage controlled".

    I also like your "black box" reference, and feel that is a large part of the argument in teaching what is the most useful model of bipolar transistor controlling.

    We must keep in mind the thread was started by questions from a beginner, in regards to (I quote him); " the pratical use of transistors (BJT, FETs) and how they work in circuit about their biasing. The material should be down to earth."

    That is inline with a description of "controlling" the "black box" transistor using current control.

    Semantic discussion of the physics of what is happening inside the black box is of a lower usefulness and lower relevance in regards to the OP's question, and to what the OP should learn as a starting point on "controlling a transistor".
     
  21. skyhawk

    skyhawk New Member

    Joined:
    Feb 9, 2007
    Messages:
    235
    Likes:
    7
    Location:
    Morgantown, WV
    Last edited: Aug 14, 2012

Share This Page