2

Is the following a correct understanding of an NPN transistor?

enter image description here Basically, a positive voltage source needs to flow through the Base of an NPN transistor and out through the Emitter, so it needs to flow in (via the base) from a more positive source than it flows out (via the emitter).

Is this a correct understanding of what makes a transistor turns on? Is there anything else that might be added to it, such as how much current/voltage is required to turn the transistor 'on' or if there's a minimum voltage/transistor that needs to flow from the through the Base to the Collector for it to 'continue' to flow through.

David542
  • 669
  • 2
  • 7
  • 15
  • 1
    Keep in mind that BJT transistors are _current_ controlled devices. The statement "positive voltage source needs to flow" does not make sense. It is current that "flows", by applied force of voltage, so to speak. – Maple Jan 18 '20 at 07:44
  • Maple...are you really sure? Can you proove it or is it just a believe? In fact, BJTs are voltage controlled!! Several explanations are proofs are available!! – LvW Jan 18 '20 at 09:30
  • https://www.researchgate.net/publication/338670662_Tim_Tom_and_Ben_try_to_unlock_the_secrets_of_the_transistor – LvW Jan 18 '20 at 09:44
  • @LvW Well... the Ic=β*Ib still works in most cases, does it not? besides, my comment was mostly about "voltage source needs to flow" which I am _really_ sure is nonsense. – Maple Jan 18 '20 at 09:53
  • @maple....You should strictly distinguish between (a) practical rules which can be applied and which will "still work in most cases" and (b) the theoretical question if the BJT is (physically spoken) voltage or current controlled. And in your comment you spoke about the transistor as a "current controlled device" - and this is simply wrong! – LvW Jan 18 '20 at 11:53
  • 1
    @LvW Yes, from a semiconductor device physics point of view the transistor is controlled by voltage. But engineers usually use a higher level abstraction of the transistor, where it is useful to view the current as the controlling factor, since the collector current is proportional to the base current. I think you are being needlessly pedantic. – Elliot Alderson Jan 18 '20 at 14:11
  • @Elliot Alderson, whether you call it "pedantic" or not - when a statement is wrong, I am so free to call it wrong! I am a senior engineer (and I have worked for 25 years as a prof) and I have found that - in particular for engineers - it is very important to distinguish between (a) useful formulas and rule of thumbs and (b) theoretical laws and principles. But it is up to you to disagree, On the other hand, of course, you are right when somebody is only following cook book recepies during design of BJT stages... – LvW Jan 18 '20 at 15:03
  • 1
    @LvW Wouldn't it be more correct to say that the transistor is controlled by semiconductor energy levels? Saying that it is controlled by **voltage** makes it hard to explain how phototransistors work. – Elliot Alderson Jan 18 '20 at 15:46
  • No - I don`t think so. The BJT is a device with three terminals. Everybody who wants to use the BJT must not know details about internal energy levels or quantum physics - but it is important to know (in particular when inventing new circuits) which signal types between the two steering terminals (B and E) will do the job of varying the output current: Voltage or current? Why should we stick on current-control (without a single proof !) when at the same time there are several indications (proofs) for voltage control? – LvW Jan 18 '20 at 16:44
  • @LvW Oh, come on, you link a document you wrote as if it was some independent confirmation of what you believe? In there you just built and destroyed your strawman to confirm your belief that voltage causes current. Well, that belief is not universally shared. Here https://www.researchgate.net/post/Are_there_causal_relationships_in_Ohms_law_If_so_which_is_the_cause_and_which_is_the_effect , for example, it is argued that you cannot assume a cause-effect relationship between V and I - see the answer by Simone Orcioni. As for BJT charge control, current control or voltage control are all valid. – Sredni Vashtar Jan 19 '20 at 17:31
  • @Sredni Vashtar, I have linked the document hoping to receive fromyou some TECHNICAL comments....have you some counter examples? Anything wrong with my reasoning? I am still waiting for a proof of current-control. I know that the following is not a real proof, but a severe indication for voltage control: I am sure you know the details of the Gummel Poon model for the BJT. Did you realize that the modelig of the base current is INDEPENDENT on the main transistor function (which uses Vbe of course)? So - even in this model the base current is treatet as a kind of unwanted defect. – LvW Jan 19 '20 at 17:46
  • @LvW In our past exchanges I have suggested TWO BOOKS. Apparently you dismissed them as not technical enough and then you produced a... piece of fiction? Is this loosely based on Galileo's Dialogue? I wonder what would your answer be if your strawman were to ask "Suppose I magically materialize a charge in base, what is the probability that an opposite charge supplied by the emitter will neutralize it, versus the probability that such charge will be collected by the collector?". I left the signs out to allow for NPN or PNP reasoning (I prefer PNPs when looking at the principles). – Sredni Vashtar Jan 19 '20 at 17:55
  • @Sredni Vashtar, two comments: (1) I am still waiting for TECHNICAL comments (anything wrong?) instead of polemic words; and (2) regarding the answer from Simone Orcioni (Researchgate), you probably have overlooked that he spoke about s-parameter presentation and " incident and reflected waves (a, b parameter) to describe our resistor". Do you really think, that this is applicable in our case? Regarding your last sentence: To me, it sounds a bit helpless (sorry to say) to state that charge or current or voltage control - "all are valid" – LvW Jan 19 '20 at 18:09
  • @LvW I already answered your 'main question' "Do you really think that two additional charged carriers in the base can release 500 additional carriers arriving at he collector?" here: https://electronics.stackexchange.com/questions/470000/ce-amplifier-non-sinusoidal-signals but you keep ignoring anything that does not conform with your based view. As for my last sentence before, it's the char limits in comments that makes my last sentences so short. What I meant is that there is no "TRUE" or "SACRED" way of operation for a BJT. They are all equally valid, each has its strenghts and shortcom... – Sredni Vashtar Jan 19 '20 at 18:19
  • @Sredni Vashtar, you felt it necessary to label my view as "biased" and as a "misconception"....I can live with that. But do you know that your view is in deep contrast to really reliable sources? I only mention: Berkeley, Stanford, MIT, Barrie Gilbert, Winfield Hill (AoE), Ian Getreu, Spice modeling ... Do you need references? – LvW Jan 19 '20 at 19:52
  • @Sredni Vashtar, perhaps we can agree to the following ? Physically spoken, the BJT is a device having an output current Ic that is determined by the base-emitter voltage Vbe (several proofs, indications, calculations). However, in some cases - and for calculation purposes - we are free to treat the BJT as a current controlled device. – LvW Jan 20 '20 at 08:39

1 Answers1

2

The image of Figure 1 may help a little.

enter image description here

Figure 1. Horowitz and Hill, The Art of Electronics. "Transistor Man" looks at the current at the base, and adjust the current at the collector so as to be a multiple of the base current.

Transistor
  • 168,990
  • 12
  • 186
  • 385
  • Perhaps it is useful (and necessary) to add that the shown figure is an illustration to a simplified explanation of the transistor principle. In a succeeding paragraph of the mentioned book (Art of Electronics) it is explained that the BJT is in fact a device where the collector current is determined and controlled by the base-emitter voltage only !! The base current is nothing else than an unwanted by-product which, however, cannot be avoided. – LvW Jan 18 '20 at 11:59
  • I was waiting for you. :^) Despite my username, transistor theory is not my strong point. – Transistor Jan 18 '20 at 12:16
  • I think your transistor theory is just fine. It is the appropriate model of a transistor for engineering purposes. Arguments about whether the transistor is voltage controlled or current controlled belong on physics.SE – Elliot Alderson Jan 18 '20 at 14:12
  • @Elliot Alderson, I disagree: Each electronic engineer working with transistors must know how the BJT really works. This is the result of my experience!! Do you need some examples ? Look here: https://www.researchgate.net/publication/338670662_Tim_Tom_and_Ben_try_to_unlock_the_secrets_of_the_transistor – LvW Jan 18 '20 at 15:18
  • @LvW Yes, engineers need to know how the transistor **works**. That is, the relationships between voltage and current at the various terminals, and modes of operation. I think you would agree that the relationship between base current and base-emitter voltage is pretty well defined, so whether you say the transistor is **controlled** by the current or the voltage is like arguing whether the car's speed is controlled by the foot pedal or the fuel injectors. – Elliot Alderson Jan 18 '20 at 15:49
  • @Elliot Alderson, your "example" is a misconception. You completely missed the point...sorry to say. Have you really never noticed the contradiction between the (wrong) current-control assertion (it is nothing else!) and several circuits which can be explained with voltage control only? Can you explain the voltage feedback effect of RE using your current-control explanation ? I would be surprised...but I am curious....waiting for your explanation. – LvW Jan 18 '20 at 16:00
  • 1
    @LvW In what way is my example a misconception? Oh, and I can explain how RE provides feedback because I do not ignore the relationship between Vbe and Ib...but I don't need to enforce a strict use of the word **controlled**. In fact, I don't have to use the word **controlled** at all. If you really understand how the BJT works you don't need to rely on such generalizations. Maybe you taught that way, but I don't. – Elliot Alderson Jan 18 '20 at 16:05
  • @Eliot Alderson, do you really completely ignore (a) Shockleys equation and (b) the transconductance gm and (c) the tempco -2mV/K and (d) the physical background for the EARLY effect and (e) the working principle of the current mirror and (f) the principle of the log-domain and (g) all relevant publications from Berkeley, Stanford, MIT,...and (h) finally, the physical logic which tells you that a small quantity can never directly control a larger quantity of the same kind ? It`s really surprising . – LvW Jan 18 '20 at 16:20
  • @LvW you can explain those thing with current control as well, but you will ignore everything that does not conform to your biased view, shifting the burden on other people to disprove your misconception. For example, the Early effect is due to base width modulation and you will be quick in pointing out that such modulation is linked to the applied voltage but --- how do you increase or decrease the depletion region if not by removing or adding charges? Also, I notice you no longer mention Barrie Gilbert after I pointed out that page where he refused to give you a direct answer. It's progress. – Sredni Vashtar Jan 19 '20 at 17:45
  • @Sredni Vashtar: (1) When we can explain "those thing with current-control as well" - where is it? Yes - I am waiting for a disprovement of my examples (which you nicely call "misconceptions"). And - regarding Barrie Gilbert: He has used the words "defect" and "nuisance" in connections with the base current. – LvW Jan 19 '20 at 18:17
  • @LvW your misconception is not in the examples. I use voltage control when it's easier to use voltage control, current control when it's easier to use current control and charge control when it's easier to use charge control. Of course when you are considering voltage feedback it's easier to use the voltage control model - you'd have to jump through hoops, otherwise - but that does not make it the "SACRED" and "ONLY" way to describe a BJT (and THAT is your misconception). As for Gilbert, I have already explained to you that when you work with translinear amplifiers Ib is indeed a nuisance. – Sredni Vashtar Jan 19 '20 at 18:24
  • @Sredni Vashtar, Sorry, but you completely missed (again) the point. I think the working principle of the BJT is independent of whether you like one version or the other better....(when it is easier.....voltage control...). Of course, an engineer can use the formulas for current-control for some applications ...I never have denied this!! But that is not the question!! The question is a physical one - and independent on the user!! Dont you understand the difference? The BJT does not know what you intend to do. It is a device which follows certain laws - thats all. Try to see the difference! – LvW Jan 19 '20 at 20:03
  • @LvW I am not missing the point: what I am trying to tell you is that one thing is a relationship between variables, and another is a cause-effect relationship. You arbitrarily choose to put causality in a V-I diode characteristic and call your view the 'only' correct one. As if a diode could develop a voltage without having a current through it. What I am saying is that your choice of assuming causality is purely arbitrary. And this is why I consider voltage, current and charge models as equivalent. Not because they lead to simpler calculation, but because there is not a "TRUE" one. – Sredni Vashtar Jan 19 '20 at 20:45
  • Guys, can we all look at the question at the top of the page and try and help the OP? Let's have our rants somewhere else! – Transistor Jan 19 '20 at 22:10
  • @tTransistor, OK - you are right - I agree. It does not make any sense to discuss on this level without any good technical arguments, but with polemic words instead. (Example: "As if a diode could develop a voltage without a current through it"). My answer: A diode does not "develop" a voltage. A voltage is developed in a battery or something like that....and can be applied to a diode with a current through it as a result. – LvW Jan 20 '20 at 08:25