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

Collector current of an NPN common emitter circuit

Discussion in 'General Electronics Chat' started by MrAl, May 29, 2013.

Thread Status:
Not open for further replies.
  1. ()blivion

    ()blivion Active Member

    Joined:
    Nov 24, 2010
    Messages:
    857
    Likes:
    210
    Location:
    Level 5
    I understand that, and I know it to be 0.98<α<0.998. I also have never needed to know it's precise part specific value though. I know it as essentially unity, and that has served me fine.

    I'm sorry, that was very clumsily spoken wasn't it. Not unusual for me honestly. That statement should really have been...

    Though, it seems like you may have understood what I was implying without me giving a perfect description. So, +1 for not playing petty word games with me here.

    Speaking of petty word games...

    I did specifically say that "I could easily be slightly off", though I may have been more than "slightly". Also, I am actually reading up on the subject vigorously, if nothing more than to sort the matter out for myself as I'm clearly not 100% certain what goes on in a BJT at the subatomic level.

    But that is all pretty irrelevant.

    I strongly doubt that a vastly deeper insight into the exact quantum mechanics of the BJT is going to significantly change the overview of the rudimentary physics and operation of the device. BJT's themselves work very much in accordance with I[SUB]c[/SUB]=β*I[SUB]B[/SUB], even if internally the charge carriers are actually making conga lines from conductor to conductor.

    Again, I was poorly spoken here. And I was expecting someone to play that card because of this.

    The word "load" is commonly understood to mean "reasonably low impedance". Although all things have some conductivity, so with potential some current will always exist, an insulator does not constitute a load in the above context, as it does not continually draw reasonable currents. Neither does a capacitor. And in both cases, a current had to have existed in the first place to charge it up, so both reduce to "chicken or the egg" arguments.

    A forward biased semiconductor junction on the other hand has a relatively low impedance, which will constantly draw a current. Any such impedance obviously can not develop a voltage across it unless there is a continuous current through it that is also in alignment with Ohm's/Diode law. Thus continuing to support my position that you can't have one without the other.

    Another typographical blunder...

    The equation I[SUB]c[/SUB]=I[SUB]S[/SUB]*(e^(V[SUB]BE[/SUB]/VT)-1), (which from what I can tell is just a simplified form of Ebers–Moll model) is not from first principle because some of the constants and relationships in it are derived from empirical models. This therefore invalidates it as a "physical fact". I[SUB]C[/SUB]=β*I[SUB]B[/SUB] is no different really. Which all makes perfect sense considering both are known to just be close approximations and not exact solutions for the transistor.

    As far as I know, the closest you can get to "physical fact" with a BJT is the Gummel–Poon (charge-control) model. And wouldn't you know it, the Gummel–Poon model is neither voltage nor current, but a combination of both... sound somewhat familiar?

    Man this is terrible. Just know that I am aware I imply much of what I mean when I really should clearly say it. I am introspectively "working on it".

    If that suffices, I'll be moving on...

    No... you really don't. But I suspect I know what you're getting at...

    Ohm's law is the LINEAR relationship between current, voltage, and resistance, with the formulas I=V/R, V=I*R, R=V/I. Diode law is similar in nature, but is an exponential relationship. These are not the same thing. I was never unaware of this.

    My focus with the text you quote was that SOME KIND of interdependent, coexisting, mutually inclusive, relationship exists between voltage and current when dealing with finite loads. Whether the relationship be linear, exponential or X=Δfeng shui^(groundhog shadows/sqrt(moon phases)) doesn't particularly matter to the heart of my argument.

    Yes of course it is, it is a circuit design strategy. And just as it is not any of those other things, it also does not prove that a transistor internally works from voltage control. Nor does it disprove that a BJT is not internally current controlled.

    But you are clearly focusing on practical concerns, so let's be practical.

    The mathematical model that design philosophy is centered around uses many approximations, [an] irrational number, exponential math, and still requires part specific parameters. This way is more complicated and more susceptible to error.

    I[SUB]C[/SUB]=β*I[SUB]B[/SUB] on the other hand, uses less approximations, no irrational constants, no exponential math, is less complicated, and uses values that are published in every single BJT technical paper that has ever been made. And if you can't find β, you can measure it. (Which is also just as possible for the above methods required parameters.)

    Now, given the two, I don't see the sense in going out of your way to use a more complex and obscure formula, that also requires more obscure data, to try and design a real and practical circuit, in situations where the other one is just as acceptable.

    [SUB]Honestly, for practical concerns I don't see why one wouldn't just use known circuit building blocks and simulators with already correct models, but... let's just put a pin in that. :)/)[/SUB]

    (Self proclaimed) electronics guru Winfield Hill of harvard might say this, and he may be right. But my texts say that I[SUB]CO[/SUB] increases 6% every °C. Though admittedly my text is just the 1957 Ditto machined lecture notes from some random professor of a "Technical Institute" in Massachusetts. It's Massachusetts, I bet they probably let just anyone teach and learn in their institutes, right?

    Hummm, for a second I was thinking this institute had an acronym, but it escapes me for now. :rolleyes::)

    Speaking of Winfield Hill though...
    I wait to analyze, or see analysis of the circuits and methodologies Winfield Hill himself (or possibly an imposter) posted in the following thread...
    http://cr4.globalspec.com/thread/68055/voltage-vs-current I'm sure a conventional explanation and analysis can be found for those circuits just as well.
     
    Last edited: Jun 13, 2013
  2. anhnha

    anhnha Member

    Joined:
    Feb 4, 2012
    Messages:
    208
    Likes:
    1
    Hi, I need help again!
    We often choose Vbe= 0.65 or 0.7 but this is not exact value. I mean that there are some error. But as you said above "very small change in Vbe results in huge changes in the base current" and Ic also change with the same rate. I see that there is much different in Ib, Ic if we choose Vbe =0.65 and in case Vbe = 0.7.
    Can you tell me why we neglect this?
     
  3. Jony130

    Jony130 Active Member

    Joined:
    Feb 1, 2009
    Messages:
    292
    Likes:
    28
    Because for a given Vcc and RB we don't have a huge error in IB current.
    For example if Vcc = 10V and RB = 100KΩ we have IB current range:

    IBmin = (10V - 0.7V)/100KΩ = 93μA

    IBmax = (10 - 0.6V)/100KΩ = 94μA


    And this is why we don't need to know exact Vba value.
    But because of a huge differences in β (50 ± 300) we will get huge difference in Ic current.
     
  4. dave

    Dave New Member

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


     
  5. anhnha

    anhnha Member

    Joined:
    Feb 4, 2012
    Messages:
    208
    Likes:
    1

    haha, that makes sense, thank you:D
     
  6. Winterstone

    Winterstone Banned

    Joined:
    Mar 24, 2012
    Messages:
    536
    Likes:
    15
    Location:
    Europe
    You can find an answer (in form of a calculated example) in post#16 of this thread.
     
  7. anhnha

    anhnha Member

    Joined:
    Feb 4, 2012
    Messages:
    208
    Likes:
    1
    Thanks, i had the answer.
     
  8. ericgibbs

    ericgibbs Well-Known Member Most Helpful Member

    Joined:
    Jan 4, 2007
    Messages:
    21,189
    Likes:
    644
    Location:
    Ex Yorks' Hants UK
    ONLINE
    hi Ratch,
    As stated earlier here is the image for LTSpice plots of a 2N2222 and a 2N2222A, Vbe values.

    Eric
     

    Attached Files:

  9. Ratchit

    Ratchit Well-Known Member

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

    Easy question to answer. The data sheets and yourself indicate that the 2N2222 is a general purpose low power transistor. The data sheets listed in the link below call it a switching transistor.

    I would appreciate a link to just one site that gives a Vbe vs Ic graph of a 2N2222(A). I checked the link below, and only two or three manufacturers or distributors even had graphs for that transistor.

    http://www.datasheetcatalog.com/datasheets_pdf/2/N/2/2/2N2222A.shtml

    My argument was not against the method used, but Al's description of the method. Any way an Is value can be obtained is fine with me.

    I think he meant to say something like "typical devices within a particular model family".

    And also max and min parameter limits.
     
  10. Ratchit

    Ratchit Well-Known Member

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

    You are correct in that QM is not going to help you much in practical understanding of BJTs. But an understanding of diffusion and electrostatics will do so. For small signals BJTs work with ic = gm*vbe.

    Shockley's diode equation came before Ebers–Moll. The diode equation has both a solid theoretical and experimental foundation.

    Depends on what you define as physical fact. They both work good to describe what a transistor does, but no model tells you how a device works.

    Good, you know that then. I would never call the triplet (I=V/R ...etc) Ohm's law. Just as I would never call distance = velocity*time something like "Newton's law". I usually call it the resistance formula. The correct Ohm's law does not relate voltage to current, the resistance formula does.

    Ohm's law is a property of a material (specifically linearity), while Shockley's equation describes a relationship between voltage and current in a semiconductor.

    Winfield Hill's method does not use the β relationship. And it seems to product a superior design.

    WH did not say this, I did, although WH would agree. Your text has a "typo" in it. See the link below.

    http://www.premiumorange.com/daniel.robert9/anglais/Semi_conducteurs5_Suite2.html

    I already gave a brief description of the method in post #129 of this thread.

    Ratch
     
  11. ()blivion

    ()blivion Active Member

    Joined:
    Nov 24, 2010
    Messages:
    857
    Likes:
    210
    Location:
    Level 5
    ic = gm*vbe... sure... why not.

    And since Vbe can be computed from Ib, in accordance with diode law, the two formulas are completely interchangeable.

    I didn't say it didn't, and the method in question still uses all the approximations, irrational numbers, exponential math, and STILL requires you to measure exact part values if you want it to reasonably describe a transistor.

    Physical fact, by most physicists, is something that is or can be reduced down to purely "first principles". In mathematics, an "Axiom" is a "physical fact". The formula in question is neither.

    That which we call a rose by any other name would smell as sweet.

    In a room full of 1000 EE types, if I say "Ohm's law", I can rest assured that 995 of the people will know exactly what I am talking about, even if the other 5 know, and are only picking at details. Now, you can go out of your way to be as confusing as possible in the name of political correctness if you really have too, but I try not to do such things as it detracts from the underlying points.

    Ohm's law is what it is because that is what we call it. If it were "bananas law", we would still call it "bananas law", even if it had nothing to do with bananas or laws.

    It uses an equally part specific relationship, and it's superior results are debatably an artifact of a superior circuit, not a superior understanding of the BJT.

    Fair enough.

    (‎ಠ___ಠ)

    You're telling me, I'm supposed to believe some random web page, whose content was obviously translated from another language, whose last update is apparently from the future (06/01/16), whose visit counter is allegedly just over 1.7 billion (AND in degrees north, WTF?), and finally, who never specifically and concisely says (or shows) that he/she actually did any experiments in hardware to back up his claims...

    And I am to believe this source.......... over MIT lecture notes.....?

    Now I'm not saying you're wrong, I don't know how likely there is to be an uncorrected typo in this text. It could have happened, I have seen a few typos in it so far, but all were corrected. But even if my source is marginally weak, your source is far far worse. You are essentially asking me to ignore an outdated textbook to look at some torn crackerjack prize that, when folded just right, has a resemblance to the fact you say are true. And... LMAO... no... that's not going to happen. Get a better source and I will believe you.

    Incidentally, I have a microammeter, bunches of transistors, a calibrated thermocouple, and maybe some free time and interest. I could probably get to the bottom of this with actual evidence if I wanted.

    I wanted a current control analysis, to see how it stacks up against the voltage control design/analysis. Though your input is still valued.



    EDIT A friend informed me that other countries use a different date format than US. So... statement about that websites date retracted.
    Edit2 And apparently it's MILLION, not billion. (I just woke up BTW)
     
    Last edited: Jun 13, 2013
  12. Ratchit

    Ratchit Well-Known Member

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

    No they are not. One is for DC, and the other is for small signals. Vbe is the 0.6 or 0.7 volts we associate with the forward voltage drop of the diode. vbe is the small signal applied to the base-emitter terminal of the BJT while in the active region. vbe should not be more than 10 mv.

    You said "The equation Ic=IS*(e^(VBE/VT)-1), (which from what I can tell is just a simplified form of model) is not from first principle because some of the constants and relationships in it are derived from empirical models."

    The equation was derived before any model existed. The variables like Is are not derived from any model. The manufacturer publishes a data sheet giving you most of what you need to know about a transistor.

    No, a physical fact is something that can be proven true. An axiom is a self-evident or universally recognized truth. Mathematical truths are not usually referred to as physical facts. The resistance formula can be proven true, so it is a physical fact.

    But we don't call it a rose or banana, so that statement does not apply.

    They would also understand if NASA told them their astronauts "walk" in space. That still does not make it the correct terminology.

    No, Ohm's law is what it is because the physics dictate the properties.

    What specific part relationship is that?

    Yes, I expect you to do so. I have seen this fact in numerous text books and it is a well known fact. Your notes have a typo.

    Feel free to do so.

    Ratch
     
  13. Ratchit

    Ratchit Well-Known Member

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

    Thanks for running those Pspice plots.

    Ratch
     
  14. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,031
    Likes:
    952
    Location:
    NJ
    Hello again,

    Yes thanks to Eric for doing the spice simulation and the plots so we can take a look at that too.

    I also found the required data sheet that i am looking at in more detail now, so Ratchit here is the link:

    http://www.electro-tech-online.com/custompdfs/2013/06/PN2222A.pdf

    That's not the same data sheet i have but i think the Vbe vs Ic curves are the same.

    I have also now had time to calculate values for the non ideal transistor constants and am looking at that in more detail too now. I should be able to post equations soon.

    But what i see in the W. Hill post is not quite the same thing. I've done that kind of design for ages now and i have promoted that technique years ago, but that is not the same as using the Shockley equation. The fact that we can force a voltage across the emitter resistor is a much simpler way to bias the transistor. It basically steals the idea from the "constant current" transistor circuit which we all know and love :) In other words, create a constant current circuit and then allow the base current or base emitter voltage to vary with the input signal and we know we get a change in output with amplification because of the Rc/Re ratio. That's not Shockley, that's more toward what Claude had mentioned. It's certainly easy to prove that if anyone is in doubt about this. So basically what this means now is we have another bias method in addition to the Shockley and Beta methods.

    The curve fitting i did for the 2N2222A came out quite well, with a max error of about 5 percent in collector current so im pretty close now but need a little more time due to other things in the house at the moment.

    If Jony is reading this and cares to do a few more measurements, what we need here is a set of measurements of Vbe vs Ic, with collector emitter voltage say at least 5 volts and hopefully not vary too much.
     
  15. Claude Abraham

    Claude Abraham Member

    Joined:
    Mar 17, 2006
    Messages:
    197
    Likes:
    23
    Location:
    NE Ohio, USA
    Has anybody read my 4 page sheet. I detail the basics of a single amp stage, how to compute gain based on parameters. It's all there.
     
  16. ()blivion

    ()blivion Active Member

    Joined:
    Nov 24, 2010
    Messages:
    857
    Likes:
    210
    Location:
    Level 5
    I glanced at it Claude, looks solid. My understanding of the analysis was "In a nutshell... emitter degeneration works" right? :)

    One thing I have seen recently that I was a tad curious about was R1 being attached to the collector not Vcc... I don't want you to make another analysis if you don't want, so don't worry about it. But if you feel like round two, let us know.

    To me it looks like even more stability is gained, caused by significant negative feedback, at the cost of even more gain.


    Ratchit
    I have plenty a good counter to all your points and more, but I'm not going to argue with you over that trivial nonsense. In the end, your continuation to make such infinitesimal distinctions is blatantly petty to the extreme of being malicious without warrant. Even if you genuinely do believe yourself and your mission, repeatedly accosting and arguing with people to the tune of "Ohm's law isn't really Ohm's law... you know" while being fully aware that such comments will, with almost absolute certainly, start a negitive argument, is by any objective standard "trolling". It's common sense the accepted most obvious answers are the correct ones regardless of the technicalities. The only person that would continue to focus on these pointless details are clearly just trying to detract from the main topic or have serious OCD.

    In the end, anyone that knows anything about electronics knows when concerning reasonable load, voltage and current go together like bees and honey. So any argument about "voltage controlled" vs "current controlled" with BJT's is pointless. The parameters are intimately related in such devices, but current control is, has been, and probably will always be the accepted standard. So you can nitpick to detract from this obvious truth all you want, it's not going to make me look bad.
     
  17. MrAl

    MrAl Well-Known Member Most Helpful Member

    Joined:
    Sep 7, 2008
    Messages:
    11,031
    Likes:
    952
    Location:
    NJ
    Hello there Claude,


    I took a look at it and i must say the included portrait of yourself at the very end of the last page does not do you justice :)

    But really i did not go into detail yet because i was focused on the Shockley equation for now, and then intended to look at this better after i was satisfied i got that equation going good enough, so i could compare.
    We are dealing with so many things right now so there are a lot of posts coming in.

    Oblivion:
    It was my idea to look at the different ways of calculating say the bias point and see what happens. If that involves using current control in one circuit and voltage control in the other circuit that's ok with me, as long as we can see some definite results. I dont even care if there is a little error involved (as long as it is not sky high of course) just as long as we can get some usable results and that these results work in a practical circuit.
    The posts i have read say the same thing others have said basically, that the Shockley equation is good. But then in their circuit they use more like what Claude is talking about with the emitter resistor playing the larger role.
    So i decided it may be time to figure out what is going on here and try to use both methods (or more methods as they come up) to do this, and find some definite results. The kind of results i want are like what Eric showed early in the thread, where he looked up the graph on the data sheet and found an answer. Yes the temperature can vary, but we will account for that later too. For now we just need some simple equations that describe each method, then take a look at how each one works and how well it works.
    Some people call the different equations voltage control or current control, but i dont mind whatever they want to call them, as long as we see a result that matches experimental measurements to some degree that seems ok.
    BTW thanks for your contributions to this thread too.

    I also didnt get to thank Jony for the measurements. That's a good idea for sure. If we can get some real world measurements to go on we can then find out if our equations are working, and how good they are working, if they work at all :)
     
  18. ericgibbs

    ericgibbs Well-Known Member Most Helpful Member

    Joined:
    Jan 4, 2007
    Messages:
    21,189
    Likes:
    644
    Location:
    Ex Yorks' Hants UK
    ONLINE
    hi Al,
    Two more plots from LTS for the 2N2222 transistor CE amplifier.

    E
     

    Attached Files:

    Last edited: Jun 14, 2013
  19. Ratchit

    Ratchit Well-Known Member

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

    Thanks for the link. The site I looked at would not download the Fairchild datasheet.

    Ratch
     
  20. Ratchit

    Ratchit Well-Known Member

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

    Well then, don't keep it a secret. Bring it on.

    Which nonsense is that? We talked about several subjects.

    Oh, you mean about what defines Ohm's law. I never said that Ohm's law was not Ohm's law. I said that Ohm's law was not the triad of V=I*R, etc. It was instead the linearity of resistance of a material. The policy of not pointing out a truth because some people might not believe it or would argue about it is silly. It is only "trolling" if you cannot back it up with facts. Your statement of not believing what I said denigrates two good physics books and the professors that wrote them. You should objectively peruse this link. http://www.electro-tech-online.com/threads/buy-dso-oscilloscope.222/

    Just because voltage and current have a relationship to each does not mean either one controls a device. In a BJT, it is the base-emitter voltage that controls the emitter/collector current while in the active region. That is determined by the physics of the BJT. You can derive models that show either current or voltage dependency, but that does not reveal the physics of the device.

    Ratch
     
  21. Winterstone

    Winterstone Banned

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

    Hello Claude Abraham,

    I just went through the 4 pages - and I must confess it`s really a good work. This applies, in particular, to the calculation of the base voltage Vb and the emitter current Ie.
    (The last part of your paper with gain calculation is standard, for my opinion).

    Let me explain (not for you, but for the benefit of some other readers who perhaps are not too familiar with such calculation) - and for comparison purposes I start with one of my posts:.

    1.) In post#16 I have designed a similar BJT stage - however starting with a given collector current Ic. Then, I have applied the standard methods for parts calculation, including selection of a suitable current through the base circuitry.
    The results were not too surprising because we all know that sufficient emitter degeneration (dc feedback with Re) can stabilize the current - which in this case means that the calculated
    bias resistors (calculated based on Vbe=0.65V and a rough assumption regarding hfE) allow a collector current that in reality is only a few percent smaller than the given value (proofed by simulation).

    2.) In your worked-out example you, instead, have started with a given circuit (including all resistor values) - and it was your first goal to find the corresponding value for Ic.
    At first sight it looks as simple as my example, but it isn`t.

    The reason is that we cannot apply the standard calculation methods as contained in all textbooks. That is because we cannot easily isolate the collector current Ic in the set of equations (we have Ic=alpha*Ie and Ic=hfE*Ib).

    You have shown that and how this problem can be circumvented by splitting the base voltage Vb in two parts applying superposition (Vb´ and Vb´´ ). I must confess that this is a very intelligent and elegant method, which leads to the correct result.
    In this context I like to mention that for calculation of Vb´´ you have replaced the base-emitter voltage by an ideal voltage source Vbe=0.65 volts. This is in accordance with the substitution theorem that, however, is not very-well known (can be found only in very few textbooks on circuit theory). Really an interesting approach. Congratulation.

    Winterstone

    PS: I have detected a small calculation error: page 3/4, first line: 0.006078 must be 0.0060078.
     
    Last edited: Jun 14, 2013
Thread Status:
Not open for further replies.

Share This Page