3

If there already is a current flowing from the collector to the emitter, can’t the electrons just keep jumping over the PN junction?
Why does killing the base current make that impossible?

brhans
  • 14,373
  • 3
  • 34
  • 49
Vambyte
  • 31
  • 2
  • 10
    Try to think of a BJT as a bad delivery service: you are in base and ask for a carrier; the emitter ships one to you but it gets snatched by the evil collector. You are still waiting for your carrier, so the emitter ships another one, and this one too is snatched by the collector. Out of 100 carriers, only one makes it to your contact in base. But in order to get that one carrier you have to ask for it, by means of the base current. Stop the base current and the emitter will stop the deliveries. – Sredni Vashtar Feb 13 '23 at 22:07
  • 3
    I assume that your term "open" for the BJT means that it is switched ON and conducting. "Open" implies an open switch which does not allow current flow, as opposed to one that is closed and conducting. – PStechPaul Feb 13 '23 at 22:20
  • 1
    If you want a device that continues to conduct without further input on the control electrode, google "thyristor" or "SCR". – The Photon Feb 13 '23 at 22:21
  • 1
    @SredniVashtar what a great analogy! – Designalog Feb 13 '23 at 22:37
  • 5
    To reiterate the whole "open means conducting" thing. In English electronics usage the words "open" and "closed" or "open" and "shorted" with respect to a transistor is by analogy to a _switch_. So "open" means "off" ("closed" isn't used much, but it _would_ mean "off"). I understand there's a lot of languages that use the analogy to a valve ("open" means "on"), but English needs to uphold it's reputation for being weird. I've been working on _not saying "open", "closed" or "shorted"_, and instead saying "on", "off", "conducting", "not conducting". It just avoids confusion. – TimWescott Feb 13 '23 at 22:38
  • 2
    ... or activated/deactivated. Please don't use hydraulic analogies for electronics. – Andy aka Feb 13 '23 at 22:55
  • Aha ok thanks @PStechPaul – Vambyte Feb 14 '23 at 00:51
  • Fixed it ...... – brhans Feb 14 '23 at 02:13
  • 1
    `can’t the electrons just keep jumping over the PN junction?` question of the century! – Rohat Kılıç Feb 14 '23 at 08:51
  • @Sredni Vashtar In your "analogy" you were using the term "ask for a carrier". I think, this touches the most important point of the whole story - and, therefore, my question: What does this means (ask for...) in electronic terms? I think, to "ask" for something is not enough to understand whats going on.... – LvW Feb 14 '23 at 11:09
  • @LvW fair enough, but this will bring us back to the voltage vs current dilemma. I don't think I have anything more to say of what I wrote here: https://electronics.stackexchange.com/questions/547625/working-of-bipolar-junction-transistor-with-electron-flow/548094#548094 . If I find the time tho, I will make an answer elaborating a bit more. – Sredni Vashtar Feb 14 '23 at 15:57
  • I remember one of my trainees in the past was confused with the water/tap analogy and asked a similar question: `So the flow rate changes as I turn the tap in one direction or another, but if I release the tap or stop it turning the water keeps flowing. So, does this mean that applying a base current only once is enough to keep the collector current flowing?`. I hope yours is not something like this because sometimes mechanical analogies are not enough to explain, or they can be confusing. – Rohat Kılıç Feb 14 '23 at 16:17
  • @Sredni Vashtar Do you really see a "dilemma"? I am sure that you will have an answer for yourself? – LvW Feb 14 '23 at 16:18
  • I must admit that I do not like the "water analogy" at all. This mechanical "model" does not work. It violates mechanical laws. – LvW Feb 14 '23 at 16:20
  • Correction: In this thread, Sredni Vashtar has shown a "water model" which could be used as a working analogy: A pressure (voltage) together with a corresponding mechanical "theshold" can alter the water flow in a main channel. Practically the same model can be used as an analogy to the FET principle - the only (but important) difference is that in the BJT case, both channels are not fully isolated (as far as water flow is concerned) - and there is kind of unwanted but unavoidable "leakage" (current Ib). – LvW Feb 16 '23 at 10:51

3 Answers3

2

Assuming you're talking about the forward-active region, the base-emitter junction must be forward biased to allow charge carriers into the base (and ultimately into the collector). Without wading into the current-controlled vs voltage-controlled argument, one way of looking at it is that, in forward bias, the applied voltage across the base-emitter junction weakens the built-in junction voltage which disturbs the equilibrium between drift and diffusion currents, allowing diffusion to dominate and for majority carriers to move from the emitter into the base. Removing the externally applied base voltage reestablishes equilibrium and the net flow of charge carriers across the BE junction returns to zero.

vir
  • 14,718
  • 13
  • 28
  • I see. As a follow up: Can there be a forward bias without a closed circuit? For example, does the BE junction collapse if I only connect a negative pole to the emitter, but don’t complete the circuit? – Vambyte Feb 14 '23 at 00:49
  • That answer is something simpler: does the lamp in your room turn on if you don't close the circuit with a switch? – LuC Feb 14 '23 at 15:31
  • The BE depletion region is controlled by the BE voltage. Without connecting the voltage source between the two terminals, there is no influence. – vir Feb 14 '23 at 18:23
  • @vir Ok. I think that is where my confusion laid. I thought that the connected negative pole would repell the electrons in the emitter and collapse the junction, but forgot that a voltage had to exist. So, if we remove the base voltage then the BE depletion zone re-establishes because the main current can’t get through the BC junction to establish a voltage due to the reverse bias. Is think thinking correct? – Vambyte Feb 14 '23 at 23:19
  • The depletion zone would return to normal (it shrinks but doesn't disappear entirely under forward bias). Charge carriers would stop being injected into the base from the emitter, so then essentially no charge carriers would cross into the collector and no current would flow. – vir Feb 15 '23 at 17:25
1

Just two sentences:

The base current Ib (which is nearly a fixed portion of the emitter current Ie) as well as the collector current Ic are both - in paralllel and simultaneously - controlled by the base-emitter voltage Vbe (exponential relationship).

Hence, when there is no base current there is no driving voltage Vbe and, of course, no emitter and no collector current.

To answer the question (Why does killing the base current....):

The base current is the result of (and an indication for) a driving voltage Vbe. A current can only be "killed" by opening the path or by setting the driving voltage to zero.

LvW
  • 24,857
  • 2
  • 23
  • 52
  • There is nothing wrong with voltage control, as you describe it. But please tell me: can you change the height of the potential barrier without changing the thickness of the depletion zone? – Sredni Vashtar Feb 14 '23 at 15:51
  • No - of course not. But please tell me: Does a change in Vbe not influence the thickness of the depletion zone? – LvW Feb 14 '23 at 16:22
  • Not until charge has been added/removed in order to cover or uncover ions. It's still the egg and chicken problem: voltage and current are inextricably linked much in the same way the electric field and the magnetic field are but two aspect of the electromagnetic field. – Sredni Vashtar Feb 14 '23 at 16:45
  • OK - of course, both are linked. But one of the two quantities is the cause for the existence of the other, isn't it? We can have V without I but not vice versa. – LvW Feb 14 '23 at 16:56
  • Start from a neutral universe with all charges 'paired'. In order to have V you need to separate charge, by putting positive charge here and negative charge there (a current is needed). But more importantly, in order to keep that voltage when a current is flowing you need to replenish the charge distributions. We are biased by the fact the nearly perfect insulators are easier to come by than perfect conductors. *Once charged*, a capacitor in a vacuum gives you voltage without current. *Once the current flows*, a superconducting loop gives you current without voltage. – Sredni Vashtar Feb 14 '23 at 17:17
  • I like to come back to the real electronic world (instaed of discussing superconducting loops). We have started with the properties of the potential barrier within the pn junction. And in this context, I like to refer to the simple pn diode. Here, the thickness of the depletion area is determined by the applied voltage only, right? Why do you think that this would be not true for the B-E pn junction of the BJT? – LvW Feb 15 '23 at 08:42
  • How do you change the voltage across the diode? You need to change the density of charge at the diode's boundaries, so you need to add or remove charge from the interface (you need a concomitant current). And even then, that won't instantaneously change the potential barrier. The electric field perturbation needs to travel at the speed of light through the diode's body and the charges in its path will react to it. Oversimplifying, we can imagine a blanket of charges being pushed or pulled, uncovering more ions near the border of the DZ in the middle of the diode... – Sredni Vashtar Feb 15 '23 at 16:47
  • ...and trying to neutralize the change in charge at the diodes interfaces . And if you want to sustain the new depletion region size you need to keep supplying or removing charges from the interfaces. Your voltage change won't happen and won't persist without a concomitant current. This applies to the BJT, as well. In the case of the MOSFET, you need a temporary current to bring or remove charges from the gate (still no voltage without a concomitant current), but after that you no longer need to maintain it. (I will add this to my answer, along with the hydraulic model of BJT and MOS). – Sredni Vashtar Feb 15 '23 at 16:49
  • In my post, I tried to answer the original question, which is about the collector resp. the base current. Surely, this question is about "conventional" current, which - driven by a voltage source - flows in an electrical conductor. I am not sure if it makes sense in the context of the question to point out here that any charge displacement (caused by chemical, mechanical or other forces) can also be called electric "current". Of course, I can't and don't want to contradict your explanations of the phenomenon "current", but I think that you go far beyond the original problem. – LvW Feb 15 '23 at 18:09
  • I have edited my answer to add the hydraulic current generator. It seems you think that pressure can not be a consequence of flow. Moving from hydraulics to fluid theory let me ask you: how does a glider stay up in the air? Is it not due to the difference in pressure 'caused' by the different rate of flow under and above the wings? Where is vertical pressure generator? – Sredni Vashtar Feb 17 '23 at 02:46
0

I can give you a couple of intuitive hand-wavy descriptions that exemplify the need for a base current in a BJT.

The bad delivery service analogy

Think of a BJT as a bad delivery service.
You are in base and ask for a carrier by placing a carrier of opposite sign in it. The emitter tries to ship you one, but unfortunately it ends up being snatched by the evil collector (this is because the phenomenon of diffusion does not obey the rules of charge imbalance and doesn't care about the electric field). Since the charge imbalance hasn't been remedied, your request still stands, and the emitter ships you another carrier, but this one as well ends up being snatched by the collector. On average, out of 100 (or beta+1) carriers the emitter is shipping, only one makes it to your contact in base. I know, right? Stormtroopers have better aim!

All of this does not actually happen in successive isolated steps, but it's a statistical process on the high number of electrons that constitute the base current. The point is that in order to get that one carrier from the emitter, you have to ask for it by placing a carrier of opposite sign in base by means of the base current. Stop the base current and the emitter will stop the deliveries.

From the point of view of conduction current in the external circuit, you are either pulling or pushing a tiny base current (depending on the BJT type), and that current happens to be associated with much larger currents in the emitter and collector that are proportional to it. When Ib goes to zero, so do Ie and Ic.

The hydraulic model of the BJT

The hydraulic model works best when referred to an NPN transistor: here water flow will be conventional electric current flow and we will now see the base current to actually add to the collector current to produce Ie. Still, we will be able to see how, once the transistor is conducting (i.e. the conduit is open), the presence of the base flow is required to keep it open.

Hydraulic model of a BJT

In the figure above, a small base water flow is required to move a piston that blocks the main flow of water from collector to emitter (remember, we are using conventional current now). Note that we can control the position of the piston by either using flow or pressure, and that you cannot have one without the other.

A gravitational pump (or a constant torque rotating pump) is an example of voltage generator equivalent. If the base conduit is vertical and oriented as in figure, adding water will increase the weight of the column of liquid, thus increasing the pressure in the base circuit to the point where the spring won't be able to keep it in position. As the piston moves downwards, a base current will necessarily flow (and merge with the collector-to-emitter flow).
What about current generators? A peristaltic pump, or a gear pump (or any rotating constant velocity pump) are examples of hydraulic current source equivalent. As you move the liquid at a fixed rate in the base comduit, since water is uncompressible, the piston will necessarily have to move to let the water pass.
Of course since flow and pressure are conjugated variables you can't have one without the other and the movement of water in the base conduit is associated with an increase in pressure against the piston (that's what develops the force that counters the springs attraction).

Bonus track: comparison with the MOSFET
The hydraulic model of the BJT implies both pressure and flow must be simultaneously present to keep the piston down. How does it differ from the hydraulic model of a MOSFET? Have a look at the figure below:

Hydraulic model of a MOSFET

Here we only have the option of applying pressure because the gate hydraulic circuit is separated from the drain-source conduit and only acts on a rubber membrane (the hydraulic equivalent of a capacitor). A little current is necessary at the beginning to fill the volume difference due to the stretched position of the membrane. Once the piston is in the desired position no more fluid will need to flow and all you need to keep the channel open is just the pressure. If you were to use a peristaltic pump, instead, the pressure against the membrane would increase until it blows.

Back to electronics

Let's get back to electrolandia: the need for the current in the BE junction can be explained by analyzing a simple diode: the current that flows in the diode is inextricably related to the voltage across it as you can immediately see from the VI characteristic of the device: you can't have a nonzero voltage without the corresponding current, and you cannot have a nonzero current without the corresponding voltage.

The need for a current when you change the voltage across a diode (like the BE diode in a transistor) can be explained by considering what charges responsible for the voltage across the device and what makes the size of the depletion layer vary.
In order to change the voltage across the diode you need to change the density of charge at the diode's boundaries, and this can only be done by adding or removing charges from the interface (and to move charges you need a concomitant current). This is not a once and for all effort: to begin with, altering the charge at the diode extremes won't instantaneously change the potential barrier height. The electric field perturbation needs to travel at the speed of light through the diode's body and the charges in its path will react to it.
Oversimplifying, we can imagine a blanket of charges being pushed or pulled, un/covering more ions near the border of the depletion layer in the middle of the diode and trying to neutralize the change in charge at the diode's interfaces. If you want to sustain the new depletion region size you need to keep supplying or removing charges from the interfaces. Your voltage change won't happen and won't persist without a simultaneous current.
This applies to the BJT, as well.
In the case of the MOSFET, you need a temporary current to bring or remove charges from the gate (still no voltage without a concomitant current), but after that you no longer need to maintain it.

Links
I am sparing you my 'tooth two-chain' model of the BJT, but if you want to learn about an interesting model that explains the role of base current from the point of view of device physics, have a look at Streetman, "Solid State Electronics Devices". I described the model and referenced the text in the following two answers:

BJT transistor, collector base junction
Working of Bipolar Junction Transistor with electron flow

The last one has many references to the voltage vs current discussion. Later on I will update (and clean up - oh those figures!!!) my answer to this question to show how a diode powered by a solar cell can be considered a good approximation of a junction paired with a current source generator.

Sredni Vashtar
  • 4,008
  • 2
  • 15
  • 25
  • I must admit that - up to now - I have not seen a "water analogy" as shown by you. And I think - it works (in contrast to other water models of a BJT). In particular, the following description is important (quote):"In particular, if we increase the pressure in the base circuit, the piston will move and a base current will flow (and merge with the collector-to-emitter flow).". Here, the "pressure" models the voltage Vbe - hence, this model confirms the voltage-control with the base current as the result. – LvW Feb 16 '23 at 10:16
  • @LvW a peristaltic pump, or a gear pump, or a constant velocity pump will be the hydraulic equivalent of a current source. If you use such a pump, the pressure is the consequence of the fluid flow. I will later edit this in in the answer. – Sredni Vashtar Feb 16 '23 at 13:45
  • To be clear: pressure and flow are still concomitant, but the flow is fixed, while the pressure changes depending on the circuit. – Sredni Vashtar Feb 16 '23 at 13:52
  • What will happen if the upper part of the piston would be conical ? I think, in this case, the flow (base current) would increase with the pressure? – LvW Feb 16 '23 at 14:55
  • @LvW reducing the section of a conduit has the same effect as reducing the section of a conductor withouth changing conductivity: it (geometrically) increases resistance and a pressure/voltage builds-up (again: it's concomitant with the flow/current). A resistor of the same section would be a sand filter. Again, pushing fluid throug it will create a pressure difference. But analogies can only go that far. For example, water is uncompressible, while the electron gas can be compressed at will and this will break the analogy at some point (think of drift velocity and water velocity). – Sredni Vashtar Feb 16 '23 at 18:10