Bipolar junction transistors

Thread Starter

jut

Joined Aug 25, 2007
224
http://www.allaboutcircuits.com/vol_3/chpt_2/8.html

I have two questions about this section.

1. why is it that a common-base amplifier is being used as a introduction to amplifiers when it seems rarely used and hard to understand? Here are a few snippets from the section dedicated to common-base amps:

"This configuration is more complex than the other two, and is less common due to its strange operating characteristics."

"Predicting voltage gain for the common-base amplifier configuration is quite difficult, and involves approximations of transistor behavior that are difficult to measure directly."

"The combination of low current gain (always less than 1) and somewhat unpredictable voltage gain conspire against the common-base design, relegating it to few practical applications."
2. While the explanation of the 'fate' of electrons that enter the base was good and made sense, I didn't see an explanation of why more base current equates to more collector current. Why doesn't more base current mean less collector current? I understand transistors have a certain current gain. But why is this? Why wouldn't the collector current be exponentially related to the base current?
 

PRS

Joined Aug 24, 2008
989
Jut, I see no common base amplifier on the page you linked. There is nothing more complicated about a common-base amplifier from a common emitter amplifier. Both have the same current gain, which is a parameter of the particular transistor. It's about 100 A/A for a typical 2n2222 transistor. I'm not sure what the quotation refered to. I think you took it out of context. The voltage gain of a CB is the same as that of a CE, but the first has a low input resistance compared to the latter.

As for more base current allowing more collector current go back to those depletion diagrams. What happens when such a region is widened or contracted? Another think to consider, if you've actually gotten to the BJT amplifier sections is this. Ic = Ib*B where B is beta and is the current gain of the particular device, approximately 100 for the 2n2222. So if you have .01mA base current, you have 1mA collector current.
 
Last edited:

studiot

Joined Nov 9, 2007
4,998
Well spotted Paul.

The page is all about biasing (correctly) the transistor, which has to be done whatever configuration you use for signal purposes.

It is true that the pictures show the transistor ' on its back' the way common base is often drawn in circuit diagrams.

Try not to be confused by this and remember that there are two separate processes to analysing a transistor circuit.

The DC analysis (biasing) is much simpler and can be completed using only Ohms and Kirchoff's laws.

The AC or signal analysis is more complicated and requires what is known as a transistor model to apply circuit equations to.
The designation common base/emiter/collector constitute such a model and are additional to the analysis shown on the linked to page.
 

Tim Tylor

Joined Aug 24, 2009
2
http://www.allaboutcircuits.com/vol_3/chpt_2/8.html
2. While the explanation of the 'fate' of electrons that enter the base was good and made sense, I didn't see an explanation of why more base current equates to more collector current. Why doesn't more base current mean less collector current? I understand transistors have a certain current gain. But why is this? Why wouldn't the collector current be exponentially related to the base current?
I'm confused on this point too. The section does a good job describing the anatomy of the transistor and the destinations of electrons passing through it, but it doesn't say exactly how the base current controls the collector current. The only explanation I can think of is that the voltage from the base terminal boosts the forward bias of the emitter, pushing it over the 0.6V barrier, but that's just guessing.
 

Ratch

Joined Mar 20, 2007
1,070
Tim Tylor,

I'm confused on this point too. The section does a good job describing the anatomy of the transistor and the destinations of electrons passing through it, but it doesn't say exactly how the base current controls the collector current.
You have a right to be confused. The base current does not control the collector current. It is an indicator of the collector current. Here's how. Both the base current and collector are exponentially related to Vbe. Ic=Is*exp(Vbe/Vt) and Ib=Is*exp(Vbe/Vt)/β . Therefore they are proportional to each other by the semi-constant β over a wide range of currents. Vbe is really the dog that wags the two tails (Ic and Ib). Since Ib and Ic are somewhat proportional to each other via their exponential relationship to Vbe, this proportionality can be used in modeling and calculations. However, in the quantum world, it is really Vbe that is calling the shots. So a BJT is masquerading as a current controlled device, but physically, it a voltage controlled device.

Ratch
 

beenthere

Joined Apr 20, 2004
15,819
How interesting that this has lurched back to life - briefly.

The topic has been covered in the past, such as this discussion and here.

One should remember that, as a map is not the territory, so a model is not the device. It is a means of explaining aspects of operation. Several such models exist to explain BJT transistor operation. One looks to base current controlling collector current in a predictable manner when designing with BJT's.
 
Last edited:

ELECTRONERD

Joined May 26, 2009
1,147
Tim Tylor,



You have a right to be confused. The base current does not control the collector current. It is an indicator of the collector current. Here's how. Both the base current and collector are exponentially related to Vbe. Ic=Is*exp(Vbe/Vt) and Ib=Is*exp(Vbe/Vt)/β . Therefore they are proportional to each other by the semi-constant β over a wide range of currents. Vbe is really the dog that wags the two tails (Ic and Ib). Since Ib and Ic are somewhat proportional to each other via their exponential relationship to Vbe, this proportionality can be used in modeling and calculations. However, in the quantum world, it is really Vbe that is calling the shots. So a BJT is masquerading as a current controlled device, but physically, it a voltage controlled device.

Ratch
But it does control the emitter current, correct? Just making sure.
 

Ratch

Joined Mar 20, 2007
1,070
ELECTRONERD,

But it does control the emitter current, correct? Just making sure.
It depends on what "it" is. If you mean Vbe, then yes. Otherwise no. Ib does not control anything. Ib is an indication of Ic and thereby Ie also. Vbe is the controller.

Ratch
 

beenthere

Joined Apr 20, 2004
15,819
Your statement
Ib does not control anything
is odd. If Ib is not controlled, the BJT will turn into a puff of smoke. Try it and see.

Or do you have an illustration of some circuit wherein Ic is the controlling mechanism, thus varying Ib? Why, if Ic is, indeed, the control, why do we need to conside or have any concern for Ib? Didn't you characterize Ib as a kind of transistor by product?
 

steveb

Joined Jul 3, 2008
2,436
ELECTRONERD,



It depends on what "it" is. If you mean Vbe, then yes. Otherwise no. Ib does not control anything. Ib is an indication of Ic and thereby Ie also. Vbe is the controller.

Ratch
As we have discovered in previous discussions on this topic. The main confusion comes from the different possible interpretations of what "control" means in this context. Even if one concedes that the terminal currents and voltages of a transistor are mostly determined by the states of the internal base-emitter and base-collector voltages, as they are in the Ebers-Moll or Gummel-Poon models, this is not an indication that the transistor must be externally controlled, in an engineering sense, by a voltage. A careful look at the equations show that the base emitter voltage applied to the external pins is not the "control" or state voltage used in the model. There is an internal base resistance that depends on base current, and it prevents the applied voltage from being the exact internal voltage relevant for internal control. In fact, it requires knowledge of the base emitter voltage, the base collector voltage and the input base current to fully determine the transistor state.

Based on this, if we define the word control to refer to "engineering control" then the transistor is controlled by both current and voltage, simultaneously. One cannot deny the importance of considering terminal voltages as an important control variables. Even in models/circuits that appear current controlled, there is an implied voltage biasing condition to take the transistor out of the cutoff region. However, current control (in the engineering sense) definitely exists in real applications.

I would recommend that we be careful when we use the word control and try to be clear about what we mean by it. Personally, as an engineer, I tend to think about engineering control. I have a good understanding to various mathematical models of transistors, and I rarely think about what control means which regards to mathematics or physics. Instead I classify variables as inputs, states and outputs. This perspective comes from the well known state space representation (either linear or nonlinear) of physical systems. Engineers should think of the input variables as the controls of the state variables and output variables. In fact, states are a special case of output variable, and can not be considered input (i.e. directly controllable) variables. However, some prefer to consider the internal state variables of the transistor (Vbe and Vbc) the controllers, which is confusing since you can not directly force a state variable to change. It must evolve according to it's own dynamic equations. The confusion comes from the fact that the important output variables are direct functions of the state variables. I can see the temptation to say this means the output is controlled by the state. Anyway, if you choose to define this to mean "control" ... well there you have your source of confusion. Just be clear about what you mean.
 

Ratch

Joined Mar 20, 2007
1,070
beenthere,

is odd. If Ib is not controlled, the BJT will turn into a puff of smoke. Try it and see.
It is the excessive Vbe that burns out the transistor. The high Ib is an indication of excessive Vbe.

Or do you have an illustration of some circuit wherein Ic is the controlling mechanism, thus varying Ib? Why, if Ic is, indeed, the control, why do we need to conside or have any concern for Ib? Didn't you characterize Ib as a kind of transistor by product?
Nope, Ic never controls Ib. At least never in the active region. Vbe controls Ic and Ib. So the fact that Ib is a nuisance by-product is irrelevant to your second false supposition.

Ratch
 

Ratch

Joined Mar 20, 2007
1,070
steveb,

As we have discovered in previous discussions on this topic. The main confusion comes from the different possible interpretations of what "control" means in this context. Even if one concedes that the terminal currents and voltages of a transistor are mostly determined by the states of the internal base-emitter and base-collector voltages, as they are in the Ebers-Moll or Gummel-Poon models, this is not an indication that the transistor must be externally controlled, in an engineering sense, by a voltage.
No confusion on my part. I maintain that whether you run a current through the base lead or apply a voltage across the base and emitter, it is the Vbe that controls the Ic. You can put a adjective on the word "control", and generate fancy models. But no matter how you model it, it is Vbe that controls the ejection of electrons or holes into the collector.

However, some prefer to consider the internal state variables of the transistor (Vbe and Vbc) the controllers
Vbc hardly has any effect on Ic as long as we are in the active region and disregard the Early effect.

I can see the temptation to say this means the output is controlled by the state.
It is true that the Ic is affected by the Early effect, the Ico, the Vbe change with temperature, the beta changes, and the bulk resistance of the semiconductor. But they only affect the operating point. No one modulates a transistor by those parameters, so they cannot be called controls. It is only Vbe that is used to control a BJT. Please do not point to phototransistors or radiation effects.

Anyway, if you choose to define this to mean "control" ... well there you have your source of confusion. Just be clear about what you mean.
Well, I hope I have.

Ratch
 

steveb

Joined Jul 3, 2008
2,436
Well, I hope I have.

Ratch
In my opinion, you haven't been clear. Just look two excerpts from your post.

I maintain that whether you run a current through the base lead or apply a voltage across the base and emitter, it is the Vbe that controls the Ic. You can put a adjective on the word "control", and generate fancy models. But no matter how you model it, it is Vbe that controls the ejection of electrons or holes into the collector.
Here, you are NOT referring to engineering control. You speak of a person running either current or applying a voltage, but say that it is Vbe and not the input voltage or current that does the control. OK, fine. I'll accept that and acknowledge that your definition of control is different than mine. By the way, I'm assuming that you realize that it is the internal base emitter voltage Vb'e' that is "controlling", not the external Vbe at the transistor terminals.

It is only Vbe that is used to control a BJT.
Here, you ARE referring to engineering control. You are stating that the Vbe is USED to control a BJT. This is a different meaning of the word control. It's also a false statement because, one can inject a base current with a current source. Whether we apply an external Vbe or an external Ib, we can not directly control the internal Vb'e' and Vb'c' states which affect the transistor voltage and current responses. These state variables obey dynamic equations and evolve. We can only influence them with input variables. Remember that the Vb'e' and Vb'c' junctions include voltage dependent capacitors, and these voltages can not instantaneously change. Hence, the transistor is controlled (internally) by the evolution of these internal state variables, which I ackowledge Vb'e' is the (typically) more important one, but Vb'c' can't be ignored. However, the transistor is controlled (externally) by input variables, of which there are three (Vbc, Vbe and Ib). With three inputs, there are multiple options for applying engineering control to the system.

Take your pick of whether you want to define control to mean physical control by internal state variables (Vb'c' and Vb'e') or engineering control by external input variables (Vbc, Vbe and Ib). However, for the benefit of those trying to learn, be clear and don't mix them up.

There is another issue I have with your post in that you seem to want to restrict the disscussion to a transistor biased in the linear region, and to ignore secondary effects.

Vbc hardly has any effect on Ic as long as we are in the active region and disregard the Early effect.
You forget that getting the transistor to operate in that range is part of the engineering control. You also forget that we sometimes deliberately operate (and hence control) the transistor in saturation. Hence, your previous statement
It is only Vbe that is used to control a BJT.
is doubly false, since Vbc (or equivalently Vce) has a big effect on the transistor in general.
 
Last edited:

Ratch

Joined Mar 20, 2007
1,070
steveb,

In my opinion, you haven't been clear. Just look two excerpts from your post.


Quote:
Originally Posted by Ratch
I maintain that whether you run a current through the base lead or apply a voltage across the base and emitter, it is the Vbe that controls the Ic. You can put a adjective on the word "control", and generate fancy models. But no matter how you model it, it is Vbe that controls the ejection of electrons or holes into the collector.

Here, you are NOT referring to engineering control. You speak of a person running either current or applying a voltage, but say that it is Vbe and not the input voltage or current that does the control. OK, fine. I'll accept that and acknowledge that your definition of control is different than mine. By the way, I'm assuming that you realize that it is the internal base emitter voltage Vb'e' that is "controlling", not the external Vbe at the transistor terminals.
Yes, I realize that bulk resistance and contract resistance reduce the effectiveness of Vbe.

Originally Posted by Ratch
It is only Vbe that is used to control a BJT.

Here, you ARE referring to engineering control. You are stating that the Vbe is USED to control a BJT. This is a different meaning of the word control. It's also a false statement because, one can inject a base current with a current source. Whether we apply an external Vbe or an external Ib, we can not directly control the internal Vb'e' and Vb'c' states which affect the transistor voltage and current responses. These state variables obey dynamic equations and evolve. We can only influence them with input variables. Remember that the Vb'e' and Vb'c' junctions include voltage dependent capacitors, and these voltages can not instantaneously change. Hence, the transistor is controlled (internally) by the evolution of these internal state variables, which I ackowledge Vb'e' is the (typically) more important one, but Vb'c' can't be ignored. However, the transistor is controlled (externally) by input variables, of which there are three (Vbc, Vbe and Ib). With three inputs, there are multiple options for applying engineering control to the system.
It doesn't make any difference whether Vbe is indirectly controlled. It lowers or raises the E-B junction barrier voltage, which in turn controls the injection of charge carriers from the emitter to the collector. There are other parameters that affect the Ic, but they are not used to modulate the transistor. For instance, Vbc is only used to set the operating point, but not modulate the transistor. I believe calling it "engineering control" does not take away from the fact that Vbe rules.

Take your pick of whether you want to define control to mean physical control by internal state variables (Vb'c' and Vb'e') or engineering control by external input variables (Vbc, Vbe and Ib). However, for the benefit of those trying to learn, be clear and don't mix them up.
It is a distinction without much difference. Eventually it boils down to a change of Vbe which controls Ic.

There is another issue I have with your post in that you seem to want to restrict the disscussion to a transistor biased in the linear region, and to ignore secondary effects.
Correct. Vbe rules the roost in the linear region, and secondary effects like Ico are nuisances that are compensated away, and not used to control the transistor.

Quote:
Originally Posted by Ratch
Vbc hardly has any effect on Ic as long as we are in the active region and disregard the Early effect.


You forget that getting the transistor to operate in that range is part of the engineering control. You also forget that we sometimes deliberately operate (and hence control) the transistor in saturation. Hence, your previous statement
I have not forgotten. That is why I exempted the cutoff and saturation region.

Quote:
Originally Posted by Ratch
It is only Vbe that is used to control a BJT.

is doubly false, since Vbc (or equivalently Vce) has a big effect on the transistor in general.
If that were true, it would only be singularly false. But the collector operates as a imperfect current source, so as long as it is operating in the linear region. Therefore, its Ic will be relatively stable even if Vbc changes. This is shown by looking at the Ic curves of a BJT and noting how horizontal they are. Any tilt toward the vertical is caused by the Early effect which I noted earlier.

Ratch
 

beenthere

Joined Apr 20, 2004
15,819
I might sat that definitions of "control" and its application differ. So far, this is the same music in a different dance. I would leave this with a quote from another of your posts:
This topic has been pounded to death
 

steveb

Joined Jul 3, 2008
2,436
I might sat that definitions of "control" and its application differ. So far, this is the same music in a different dance. I would leave this with a quote from another of your posts: "This topic has been pounded to death."
Well then, may it rest in peace.
 
Top