What does that mean?
The base current needs to be above a certain value to achieve saturation.
How is that nothing to do with current control?

I think, I have explained that the large base current (compared with linear operation) results from the forward biased B-C junction.
This happens because the product Ic*Rc makes Vce<Vbe.
Hence, the Ib increase is the RESULT of saturation (and it is not the current to "achieve saturation").
For me, this is one of several indications that the BJT is not current-controlled.
Don`t you agree to this description?

So how do you know what is a "low-resistance" without knowing approximately what the base current will be?

Did I ever deny the existence of a base current? Why such a question?
I think you know me well enough from other discussions to know what I mean.
Of course I can - even without a calculator - estimate the expected base current Ib for a selected Ic - and then select resistors that allow a current through the divder chain of at least 10*Ib.

For AC consideration the voltage-control transconductance model is fine, but for biasing and switching applications the current-controlled black-box model is useful.

I disagree. Here are the steps for designing a simple switch:
* Select Vcc, load Rc and a current Ic to make Vce<Vbe=(0.7...0.8) volts
* Calculate the value for a resistor Rs=Vs/Is between the base node and the switching source Vs. For this purpose, use a current Is which is large enough to be on the "safe side" because one (additional) portion of this current Is results from the forward biased B-C juncton. There are some "rule of thumbs" to find a suitable Is value, much larger than Ib=Ic/B.
This is the last step of the design process - don`t you agree?
Where is any indication for Ib-control?
Nevertheless, this is the last step

Why do you think they put the value of Beta and/or Hfe in the data sheets?

Is that really a serious question?
Is the appearance of B for you an indication for current-control?
Beta (resp hfe or h21) is one of various parameters describing the BJT.
It is used (as I have mentioned above) for finding a good guess for Ib=Ic/B (iib=ic/hfe) - however with extremely large tolerance for B (resp hfe). In any case, normally the design starts with selecting a suitable vaue for Ic (which determins gm).
And, finally, the base current determines the input resistance which is a very important parameter for amplfier stages.

Finally, may I ask you something? Just one short and simple question:
What do you think is the justification (reason) for using a darlington combination instead of a single BJT?
Can we expect a large voltage gain due to the extremely large current gain (same or similar collector current assumed)?

Thank you and Merry Christmas
LvW

 

Is that really a serious question?

Yes.
Why do you think it isn't?

Is the appearance of B for you an indication for current-control?

It's an indication that the base current has a relation to the collector current by some ratio, no?

What do you think is the justification (reason) for using a darlington combination instead of a single BJT?

So is that really a serious question?
A Darlington configuration is to significantly reduce the base current (increase the apparent current gain) for a given collector current.
That's why it's used in such applications, as in an audio speaker-amplifier output stage.

Can we expect a large voltage gain due to the extremely large current gain (same or similar collector current assumed)?

Why the straw-man questions?
No, it obviously has only a small effect on voltage gain.

I never said the BJT is a current-controlled device from a physics point-of-view, only that it acts like one from a black-box point-of-view.

I believe this discussion is done, as you seem to be arguing semantics and I'm not interested in that.

 

I believe this discussion is done, as you seem to be arguing semantics and I'm not interested in that.

No - it seems that you have misunderstood my position and my motivation.
I am just an engineer who not only follows generally accepted design rules, but always asks: Why?
And you can only find the answer if you are also interested in the physical background. That´s all.

(By the way: With the same collector current, the voltage gain in the Darlington stage is reduced to 50% - compared to a single transistor).

 

I think, I have explained that the large base current (compared with linear operation) results from the forward biased B-C junction.
This happens because the product Ic*Rc makes Vce<Vbe.
Hence, the Ib increase is the RESULT of saturation (and it is not the current to "achieve saturation").
For me, this is one of several indications that the BJT is not current-controlled.
Don`t you agree to this description?

Thank you and Merry Christmas
LvW

Hey there,

I see you are arguing voltage control over current control for the BJT again, is that right?

"For me, this is one of several indications that the BJT is not current-controlled."
To make any general statement like that without context is just a waste of time. It's equally a waste of time to say just the opposite. It's only proper when we see what application we are dealing with.

But then also, how do you explain that the spice model for BJT's has a current controlled current source as one of the main components. What are your thoughts on this?

 

it seems that you have misunderstood my position and my motivation.

I don't believe I have, as they seem perfectly clear, based upon your answers to my comments.

I am just an engineer who not only follows generally accepted design rules, but always asks: Why?

So do I.

And you can only find the answer if you are also interested in the physical background. That´s all.

I had no questions that I did not already know the answer to, none of which need to be corrected.

 

Hello! so I came across this question today , it's simply asking for Vab and I can not understand why my approach is wrong. I think I'm misunderstanding a basic rule here. Any explanation or advice on how to approach these kind of problems will be highly appreciated!

please check the attached files

Hi,

The simplest way to think about this is that when you have equal components that are in parallel, they always share the current between them. If you have two equal components in parallel then the current splits two ways, if you have three then the current splits three ways.
If these were resistors and not diodes and they were all the same value (Ohms), then each of the three parallel resistors would have just one-third (1/3) of the current that the top diode had going through it. That means the voltage across the three would have to be less than the voltage across the single top 'resistor' because three resistors in parallel have total resistance 1/3 of each equal value resistor, and with the top 'resistor' still having the full value, that means the voltage across the top resistor will be higher than that across the three parallel resistors.
What this means of course is that the network does not act like two equal 'resistors' in series like you wanted to assume, rather it's more like one resistor in series with a resistor that has 1/3 of the value of that one top resistor. If you calculate the voltage across both top and bottom sections then, the top 'resistor' will have 2/3 of the total voltage across it while the bottom three will have just 1/3 of the total voltage across them.

This is the same with diodes although their characteristic is nonlinear. It's nonlinear, but it's still monotonic, which means the voltages will not be exactly 2/3 and 1/3 as with the true resistors, but it will still have to be lower across the three in parallel.

For example, if we use this equation for the one top diode:
v=log(i/Is+1)*K*N*T/q

and the current through one of the diodes on the bottom is 1/3 of that at the top, then the equation for one lower diode is:
v=log((i/3)/Is+1)*K*N*T/q

and all we did here was divide the current 'i' by 3 because that is the current that will flow through one of the bottom three diodes.

See if you can use this information to do a new calculation.

 

I don't believe I have, as they seem perfectly clear.
So do I.
I had no questions that I did not already know the answer to, non of which are incorrect.

Hello,

It seems that LvW likes to use the voltage controlled model a lot more than the current controlled model.
I asked him why they use a current controlled current source as a main component in the spice model of BJT's. I'd really like to hear what he has to say on this subject. I don't remember that much about the previous discussions on this.

 

Hello,

It seems that LvW likes to use the voltage controlled model a lot more than the current controlled model.
I asked him why they use a current controlled current source as a main component in the spice model of BJT's. I'd really like to hear what he has to say on this subject. I don't remember that much about the previous discussions on this.

What is used in a particular mathematical model has no bearing, one way or another, on what is actually happening in the real world. It is just that -- a mathematical model the produces an acceptable approximation to the relationships between the voltages and currents associated with a device.

Having said that, I'm not sure what you mean by the SPICE model of the BJT using a current-controlled current source. My understanding (and it's been a long time since I looked into the internals of SPICE and I've never delved very deep) is that the early BJT SPICE models revolved around the Ebers-Moll model or the more sophisticated Gummel-Poon model, both of which are primarily controlled by the base-emitter voltage and not the base current.

 

What is used in a particular mathematical model has no bearing, one way or another, on what is actually happening in the real world. It is just that -- a mathematical model the produces an acceptable approximation to the relationships between the voltages and currents associated with a device.

Having said that, I'm not sure what you mean by the SPICE model of the BJT using a current-controlled current source. My understanding (and it's been a long time since I looked into the internals of SPICE and I've never delved very deep) is that the early BJT SPICE models revolved around the Ebers-Moll model or the more sophisticated Gummel-Poon model, both of which are primarily controlled by the base-emitter voltage and not the base current.

The models currently used use a current controlled current source to produce the collector current. It's a completely linear CCCS. The nonlinear behavior comes from the added diodes, usually at least one BE diode and one BC diode. The diodes may have different parameters, but they use the exponential model that we usually see.
Again, how we say that the external control is to be interpreted is probably up to the individual because it takes both current and voltage, but the current going through the BE diode controls the CCCS, so in a common emitter configuration it's hard not to say that the device is current controlled, although you can't have that current without applying some voltage of course. The 'internal' CCCS is controlled by the current without exception, because after all it's a CCCS not a VCCS. It's only by choice that we can view the BJT as a voltage controlled device, but then it's also by choice that we can view it as a current controlled device.
I don't see any way around this but maybe someone else can provide alternate viewpoints.
This is of course when we consider the model used by spice, not the physics, but then even in physics it's hard to apply a voltage without at least some current.
We usually rate something by the relative ratio of the voltage vs current, and even more by the derivative. For example, if the derivative of current vs time di/dt is zero and the dv/dt is nonzero, we are apt to call that voltage controlled, but if the dv/dt is zero and the di/dt is nonzero, then we would be calling that current controlled. The problem comes in with the naming convention when they are both nonzero, because there is no real comparison between current and voltage as they are completely different aspects of energy. For example, if in the first case di/dt is 1ma/s and dv/dt is 1mv per second, or in a second case the di/dt is 1ma/s and dv/dt is 10mv/s does that mean the first is tied and the second is voltage controlled. It's hard to pick a winner in either case.
On the other hand, if the di/dt is 100 amps/s and the dv/dt is 1mv/s we might be tempted to call that current controlled, but would that even be right.

Here is a link see what you think:
The Spice Gummel-Poon Model - Course on modeling and simulation

 

Take a closer look at that Gummel-Poon model.

The current source is dominated by (if - it) which are both functions of voltage, not current.

 

All these discussions are akin to an argument.: which used to be the chicken or the egg. I have drawn a diagram where there is a control of a transistor and a voltage source and a current source. The Q5 transistor is controlled by the collector current of the transistor, which is quite a current generator in this case. I am of the opinion that it is possible to use both cases and use the method with the least effort. And let the teachers use the correct management mechanism. After all, scientific knowledge is important to them, and not the ability to simply calculate electronic circuits.


For fun, look at my transistor model, which takes into account the effect of avalanche multiplication of charge carriers.

 

Hello,
It seems that LvW likes to use the voltage controlled model a lot more than the current controlled model.
I asked him why they use a current controlled current source as a main component in the spice model of BJT's. I'd really like to hear what he has to say on this subject. I don't remember that much about the previous discussions on this.

I must admit that I do not use any model.
And in my contributions I did not speak about models but on transistor physics.
I am aware that models do not necessarily always reflect physical laws (in the sense of cause and effect).
Some models are good for calculation purposes only.
In any case, we should really try to distinguish between models and physical explanations - otherwise misunderstandings and misinterpretations cannot be avoided (as we can see here in this thread).

As I can read - you would "really like to hear what he has to say on this subject."
OK - here is my answer:
You are wrong - all detailed Spice models do NOT use a controlled current source.
If you would look into the model descriptions you will notice that all Spice models, of course, are based on the classical exponenetial relationship Ic=f(Vbe).
This is absolutely necessary because these models must work for small-signal as well as large signal operation.
(Do you know about a current-controlled model for large signal operation?)
I hope I have answered all of your questions.

 

LvW,
in message #31, I provided a diagram. Please tell me what is controlled by Q5 - current or voltage?

 

Hey there,
I see you are arguing voltage control over current control for the BJT again, is that right? "For me, this is one of several indications that the BJT is not current-controlled."
To make any general statement like that without context is just a waste of time. It's equally a waste of time to say just the opposite. It's only proper when we see what application we are dealing with.

From you last sentence I can learn that you think that the physical working principle of the BJT would depend on the application, right? Perhaps no surprise - but I strictly disagree.

Let me repeat what I have stated several times:
There are many proofs, indications and observations which clearly proof that the BJT is voltage-controlled.
And there is not a single proof for current-control.
(Or do you consider an artificial model which is based on the formula Ic=B*Ib (resp. ic=beta*ib) as a proof ?).

 

What is used in a particular mathematical model has no bearing, one way or another, on what is actually happening in the real world. It is just that -- a mathematical model the produces an acceptable approximation to the relationships between the voltages and currents associated with a device.

Yes - it is a pity and it creates many misunderstandings if no clear distinction is made between the models and the physical working principles.

* Simple example: When calculating a two-resistor voltage divider, it is common practice to calculate the output voltage as U2=I*R2 (current generates voltage).
However, we know (should know) that this does not correspond to the physical reality, because a flowing current is always the result of a driving voltage. Nevertheless, the result of the calculation will be correct.
Some may consider such a consideration to be "semantics" - but I think a practicing engineer should know and take into account the difference between model and reality.

 

How is the bipolar transistor Q1 controlled by current or voltage in the above cascaded (cascode) electronic circuit?

 

LvW,
in message #31, I provided a diagram. Please tell me what is controlled by Q5 - current or voltage?

Thank you for your question - it may help to clarify things and reveal misunderstandings.
My anwser to your question:
From the diagram, I can`t tell you. I see that the collector of Q2 is connected to the base of Q5.
So I can imagine that (a) there will be a current into the base of Q5 - and (b) I know that this current will correspond to a certain voltage between base and emitter of Q5.
So - how can I (can you?) deduce from the figure if the base current Ib(5) or Vbe(5) will control the collector curent Ic(5) ?
But as you know - the BJT model within the simulation program will react upon the voltage Vbe(5) and find the corresponding Ic(5).

I have no problems to admit that, perhaps, it would be best to use the relation Ic=Ib*B for calculating by hand the collector current Ic(5).(provided I have a good guess for B).
A similar consideration applies to the Darlington transistor.
But this approach does not automatically imply that I consider the BJT (as an isolated component) to be current-controlled.
It is the circuit consisting of BJT and resistors that I use with a valid formula for the calculation.
(This is not semantics, as I have been accused of, but only the clear distinction between physical effect and practical formulas.
In this context, please take another look at my example (voltage divider) in post#35).
A simulator does not care if I=f(V) or V=f(I).

This is the "crux" with the BJT (and the source of many misinterpretations): Some calculations are more simple (and correct) when we consider it as current-controlled (with attention to extremely large B tolerances).
But - on the other hand - we should know that this is just a model because there are many proofs/indications for voltage-control .
And - to know about this working principle is very important. Otherwise, nobody can understand how circuits work or even invent new circuits.
More than that, there are some circuit techniques (negative feedback) which are commonly in use to counteract the uncertainties connected with the exponential relationship Is*[(exp(Vbe/Vt)-1].
(In this context, I remind you on the well known tempco d(Vbe)/d(T)=-2mV/K.).

I hope that you now understand the primary goal of my contributions.

Greetings
LvW

 

After all of the hundreds of lines about so very much,, I would point out that the reason for the much more detailed explanations we often get are to help a newcomer understand that the grossly simplified models that are used in some of the marginally accurate explanations are incomplete, and promote an incorrect understanding. Neither a diode nor a transistor base start conducting from zero forward volts, nor do they instantly start full conduction at some point just above zero.
Just as LEDs do not suddenly switch on and light up fully at some magic forward voltage and current. The fact that so much is rather non-linear is ignored in a lot of explanations, and that is a problem.

 

You shouldn't have given the formula d(Vbe)/d(T)=-2mV/K. It's not right! This coefficient depends on the current and temperature! Here's the calculation:

 

You shouldn't have given the formula d(Vbe)/d(T)=-2mV/K. It's not right! This coefficient depends on the current and temperature! Here's the calculation:

Can you really not imagine why I have mentioned this formula in the context of our discussion?
Surely not for some special calculations - but to show how and why the collector curent Ic depends on temperature.
Hence, the real and exact number is of less imporance here.
Remember the background of the subject :
When the current Ic (resp. the factor Is in the exponential formula Ic=f(Vbe)) rises due to a temperatur increase we must reduce Vbe in order to bring Ic back to its former value. And the amount of Vbe reduction is given by this tempco (-2mV or whatever the actual value is).
(There is no expression which gives a corresponding Ib reduction)