I have no interest at all in debating with a dogmatist who can only refer to authorities ...

Thank you very much for your nice words.
As far as my reference to "authorities" is concerned: May I kindly remind you on your own words: "I am not interested" (as an answer to my question, see post #14, if you are interested to read my technical arguments in favour of voltage control).
Regards
LvW

 

This is a never-ending argument I've seen many times in various electronics forms.
Technically the BJT physics shows that it is a voltage-controlled, current-output device, but that is really only useful for small-signal circuit design.
For large-signal operation (biasing and switching design) a current-controlled, current-output (black box) model is almost always used, thus the reason for the current-gain and saturation parameters in a BJT data sheet.

Yes - agreement. I share your position to see (and make) a difference between physical realities and models which do not necessarily reflect the physical truth.
(There are three basic BJT models for small-signal operation - one is based on current-control and two use the voltage-control feature; large-signal models are all voltage controlled.).

 

large-signal models are all voltage controlled

Do you not mean current-controlled using the current gain (Beta or hFE)?

 

Speaking of "large-signal" models I am referring to the models implemented in simulation packages ("Gummel-Poon").

 

LvW
I have studied the theory of pn junctions and I am willing to accept that the current formulas are derived from the applied voltage. It's kind of like a voltage control. But that doesn't mean that the transistor is necessarily voltage controlled. There are circuits, and I cited them, that are current controlled. It's the current control of the second transistor by the current of the first transistor in a cascade circuit. Also the transistor is controlled by the current of the photodiode. But you were unwilling (or unable) to deal with these cases. My statement now sounds like this - a bipolar transistor can be voltage controlled and can be current controlled, and in most cases there is an intermediate option. For example a voltage source with internal resistance. Or an equivalent current generator with parallel conductivity. And show me a real circuit driven by a perfect voltage source. I don't mean circuits in simulation, which cause outrage to many people who realize that this is a surefire way to destroy a transistor in reality. Don't confuse control voltage and base-emitter junction voltage.

 

LvW
But you were unwilling (or unable) to deal with these cases. My statement now sounds like this - a bipolar transistor can be voltage controlled and can be current controlled, and in most cases there is an intermediate option.

OK - this sounds as we could come back to a technical discussion - however, your first sentence as quoted above......
Nevertheless:

I think, now I am in the position to identify the source of our disagreements:
When we talk about "control of the collector current", we strictly have to distinguish between
(a) the behavior of a "naked" transistor and
(b) the behavior of a BJT-based circuit.

Theoretical example (little practical significance): Transistor stage with signal injection via a relatively large series resistor Rs .
In this case ,we often speak of "current injection" (laboratory jargon) and one can also show that the collector current, indeed, essentially is determined by the current going through Rs into the base node.

But this view (interpretation) applies - as mentioned - to the behavior of the whole circuit.
In fact, in contrast to this view, the transistor reacts to the thereby changed base voltage:
This is because the series resistor Rs and the resistance of the base-emitter path form a simple voltage divider.

Exactly the same applies to the classical voltage-opamp operated as a transimpedance amplifier:
I am sure, nobody will deny that this opamp can do nothing else than to amplify the differential input voltage.
However, tt looks as if the output voltage would be controlled by the input current, but in reality it is, of course, still the voltage difference between the amplifiers input terminals that determines the voltage at the output.

Insofar - and applicable to BJT-based circuits (with external components) - I can agree to your statement: "..a bipolar transistor can be voltage controlled and can be current controlled, and in most cases there is an intermediate option."

Conclusion: It is very important for an engineer to know whether he is talking about the properties of a single component, an associated simulation model, or an entire circuit.

Can you agree to this?

 

Speaking of "large-signal" models I am referring to the models implemented in simulation packages ("Gummel-Poon").

Okay.
I was referring to the large-signal model used when doing design not simulation.

 

Okay.
I was referring to the large-signal model used when doing design not simulation.

OK, I know what you mean - and I am sure that during design you are not really using a "model" (drawn on the paper or on the screen). Instead, you certainly will have all the relevant equations/relations in your mind, right?

 

I am sure that during design you are not really using a "model" (drawn on the paper or on the screen). Instead, you certainly will have all the relevant equations/relations in your mind, right?

A "model" does not be drawn or on the screen to be used in calculations.
For BJT large signal design I use the current-controlled current-out black-box model.

And the Spice BJT models have a BF parameter(forward beta) which is used the Gummel-Poons model, so it is not just a voltage-control model as you implied.

 

And the Spice BJT models have a BF parameter(forward beta) which is used the Gummel-Poons model, so it is not just a voltage-control model as you implied.

I must admit that I am really surprised about this sentence.
Do you really think that the existence of the BF parameter is an indication of current control?
Of course, this parameter is used because Ic/BF must be used to model the magnitude of the flowing base current. That is the whole secret.
My recommendation: Consult the standard work explaining the various existing transistor models:
"Modeling the Bipolar Transistor", I.E. Getreu (Tektronix), Elsevier Scientific Publishing Company, 1978

 

The bipolar transistor is controlled by the voltage at the pn junction, which is derived from an external circuit. For example a current source. Let's take the Sziklai transistor as an example. We don't consider how the first transistor is controlled - this is determined by the external circuit, but the first transistor is a current generator and this current controls the second transistor. The voltage at the base-emitter pn junction is determined by this current.

 

The bipolar transistor is controlled by the voltage at the pn junction, which is derived from an external circuit. For example a current source. Let's take the Sziklai transistor as an example. We don't consider how the first transistor is controlled - this is determined by the external circuit, but the first transistor is a current generator and this current controls the second transistor. The voltage at the base-emitter pn junction is determined by this current.

I agree to you that an explanation based on the assumption of current-control is in this case (Sziklai circuit) more illustrative (easier) than an explanation using the voltage-control method.
However - can this be a good reason to argue against voltage control?

When I understand your sentences correctly, you think that the 1st BJT would be controlled by a voltage and the 2nd one by a current?
So - the control mechanism depends on our view? (I remember your chicken-egg-phenomenon).
No - I don`t think so. A good engineer should try to UNDERSTAND what really is going on.

Remember the case of an opamp used as a transimpedance amplifier. Of course, it is much easier to use Ohms law and to state that Vout is determined (controlled) by the input current: Vout=Iin*R.
This calculation is correct, however, when we are asking what - physically spoken - really happens, we cannot ignore that the opamp can do nothing else than to amplify the differential voltage actross its inputs.

Therefore, also the Sziklai combination can, of course, be explained using the physically correct explanation Ic=f(Vbe). We have nothing to do than to verify the fact, that the E-B path of the upper (second) transistor and the C-E path of the lower (first) transistor form a non-linear voltage divider with a current driven by the applied voltage.

(PS: I forgot to react upon your statemet: "I'm surprised you don't know how a reverse bias photodiode works. That's a shame."
If you like, I can explain to you why and how the photodiode is able to develop an open-circuit voltage of app. 0.2V across its terminals.)

 

Explain why the photodiode is not connected anywhere. And the voltage can be higher. I see you do not know that the photodiode can operate in two modes. You only know about the photovoltaic mode. This is the basic mode of the solar cell. And the main mode of the photodiode is negative bias operation. You should be ashamed of yourself. I have designed many photodiodes and I know what I am talking about.
https://www.allaboutcircuits.com/te...hotoconductive-modes-of-photodiode-operation/

 

Regarding the Sziklai couple. You can of course calculate the Base-Emitter voltage of the second transistor. You have found some kind of divider. So how do you calculate this Base-Emitter voltage. Demonstrate it without using the current amplifying property of the transistor ? I will laugh. By the way I mentioned the cascode circuit. It has a similar situation, but simpler. You can of course claim that you will not solve my problems. That is, you will just ignore them. Will you deny that the transistor amplifies the current? Or does it amplify the voltage?

 

........................................................................................
You should be ashamed of yourself. .............

Sorry, I was on vacation for two weeks.
Looking now - with some distance - at the discussion so far, I have no desire to continue.
Too bad - you can't do without personal (unobjective) remarks. Does this make you more credible?