transitor: the base pin.

[Zero999]

Imagine the following situation: Introducing the BJT I have stated that Ic is determined by Ib using the simple relation Ic=B*Ib.
Now - after some additional lessons - I explain the temperature dependence of the curent Ic and tell them that - in order to keep Ic constant - I must reduce the voltage Vbe by -2mV per degree temp. change.
Of course, some students will ask: Huhhh? We thougt it is the current Ib that controls Ic. Suddenly it is the voltage Vbe? (That`s what I have experienced often.) 
And why this value of -2mV/K ?
And I must answer and explain that this value is not only measured but that this value was calculated based on charged carrier physics in the pn region of the transistor. 
Do you understand the dilemma? It is really not sufficient to say "doesn`t matter , current or voltage" .
I have to make a decision - and (as I have demonstrated with various examples): A current-control mechanmism cannot explain how BJT circuits really work.   

Of course V_BE decreases by 2mV per degree temperature change because it is a diode junction. You could take an ideal current controlled current source and connect a diode in series with the input and its input will behave like a BJT's base emitter junction.

This debate is silly. I've been talked both models in college. Most of the time I used the current controlled approach but it isn't perfect as it doesn't take everything into account but neither does the voltage control model. If it did, then I wouldn't have to worry about the base current or blowing a BJT up when driving it without a resistor.

Let me guess it's most likely the older people here who argue one way or the other?

With age brings wisdom but it also reduces ones ability to accept new ideas.

[LvW]

This is really a matter of teaching and order of presentation. You have first presented a simple model of the operation of a transistor, of the form  f(Ib, Ic, B) = 0, where B is a parameter and Ib, Ic are model variables. Given a value of B, then for any value of Ib you can fix Ic, or for any value of Ic you can fix Ib. But it is important to explain that this is a simplified model that abstracts away many details, and is only valid in a narrow region of operation.

Later, it will become necessary to introduce more elaborate models of operation, models that include variables Vbe, Vce, Ib, Ic and perhaps the operating temperature, T.

Question 1: Why should I first present a model which is false and which is not able to explain/verify various effects? I would agree if the "more elaborate" model (to be presented later) simply would be an extension of the first model. But that is not the case. 

Questioin 2: Can you, please, show me one single application which is "valid in a narrow region of operation" and where it is advantageous to use the current-control approach?

Question 3: I am sure that in each electronic course the pn diode is explained prior to the BJT. That means: The students are familiar with Shockley`s equation.
Is there any reason to believe that students would not accept that the same equation applies for the pn junction within the BJT? Is there any good reason to deviate from this knowledge and - suddenly - teach something which is simply false? I do not understand such an attitude.

[LvW]

This debate is silly. I've been talked both models in college. Most of the time I used the current controlled approach but it isn't perfect as it doesn't take everything into account but neither does the voltage control model. If it did, then I wouldn't have to worry about the base current or blowing a BJT up when driving it without a resistor.

Hero999, there is a misunderstanding - perhaps I am responsible, but I am not sure.

"but neither does the voltage control model. If it did, then I wouldn't have to worry about the base current or blowing a BJT up when driving it without a resistor."

I think and I hope that I have mentioned several times that - of course - there is base current I have to "worry about"  during design of a BJT stage.
But this simply is a mathematical matter - nothing else. In practice, nobody connects a battery with 0.65 volts across the B-E path.
Of course, I take into account that the top resistor of a base voltage divider carries a current (I1+Ib) and the lower resistor only the current I1.
But - what has this calculation to do with the "model-question" or the "control-question" ?

Please, can you answer this last question? I am really open to learn.
I really have problems to understand the way of thinking of all persons who vote for the "current-control" option.
Where is the advantage? Show me one single application where this approach has - at least - advantages.   

Because you have used the term "silly": The classical BJT gain stage with a low-resistive voltage divider and Re-feedback is designed (after fixing Ic, Rc and Re) starting with VOLTAGES (at the emitter and the base node).  Then, as a next  step,  we calculate the base resistors - of course taking Ib into account.
All of us follow this sequence - am I wrong? - nevertheless, some people think that they have established a current-control mechanism.
In my view: This is silly.
If there would not exist some external requirements (power consumption, input resistance) we could make the divider so low-resistive (I1=50*Ib or more) that we even could NEGLECT the base current.   

I don`t know if you were (or still are) engaged in teaching electronics. But my position and my arguments result from corresponding experiences.
Do you think that students are satisfied with two different explanations ?  Hence, to me this is by far not a "silly debate".
Why do you think I am so engaged in this discussion (which sometimes merges into a religious matter)?

[LvW]

This debate is silly.

It was silly the first two threads now its gone far beyond silly. LvW seems a little obsessed with the matter.
Nothing like a good old fashioned nerd war.

I suppose you are not engaged in teaching electronics, correct? So you can say "I don`t care about it". Be lucky.

[dannyf]

Imagine the following situation:

Wow, I knew that your understanding of device physics is superficial but I never thought it was this bad, .

The tempco is no repudiation of the bjt's current controlled mechanism. In this context, it is purely the side effect of your driving the b-e junction with a (constant) voltage source.

Ic = B * Ib still holds, except that if you hold the b-e junction's forward voltage constant, Ib will vary by temperature.

Alternatively, you can drive the b-e junction via a (constant) current source, and the tempco's impact on Ic is completely eliminated.

Quite frankly, you can do yourself a huge favor by trying more to understand the basics, and trying less to use terminologies that you have little understanding of - it doesn't help you much (ie. at all).

Why should I first present a model which is false and which is not able to explain/verify various effects?

All models are false / wrong, and that's why models are useful - they help you isolate the important factors so you can focus on what makes sense for your particular study / applications.

Different models can be proposed to explain (different) aspects of the same phenomenon. That doesn't mean that one is correct and others wrong.

If someone has to break that to you, you are in deep @#$@, I mean, trouble.

[amyk]

Different models can be proposed to explain (different) aspects of the same phenomenon. That doesn't mean that one is correct and others wrong.

...just like wave-particle duality. Fortunately, the majority of physicists have accepted that there can be different ways of looking at the same thing.

[IanB]

You are so not understanding what the other poster was saying. Lift doesn't have to be generate by pushing down air. And rotorcraft doesn't work the way you think it does.

Well it does, actually. Heavier than air craft need an upward force to keep them in the air. This upward force is generated by accelerating some of the surrounding air downwards to generate a reaction force (F = ma). Where airplanes differ from rockets in this regard is that airplanes use the surrounding air for their reaction force while rockets generate their own exhaust gases.

[LvW]

Imagine the following situation:

Wow, I knew that your understanding of device physics is superficial but I never thought it was this bad, .
................
Quite frankly, you can do yourself a huge favor by trying more to understand the basics, and trying less to use terminologies that you have little understanding of - it doesn't help you much (ie. at all).

Dannyf, I know that you have some problems to handle some new - and for you perhaps "uncommon" - information. This may explain your polemic attitude.
Nevertheless, as long as you do not answer my technical questions (my former posts) it is perhaps better to avoid such personal attacks.

Alternatively, you can drive the b-e junction via a (constant) current source, and the tempco's impact on Ic is completely eliminated.

Back to the value of -2mV/K:  Your reply is correct, but the information content is zero. 
Nobody has denied that in the ideal and unrealistic case (constant current Ib) the current Ic would remain roughly constant. In this case, the temperature dependence of Ic from Vbe has no effect on Ic. That`s logical, is it not? It is a truism - nothing else. In this case, Vbe is floating and is adapting itself corresponding to the tempco.

However, may I remind you (as I have stated in my recent post) that this value of d(Vbe)/d(T)=-2mV/K was calculated based on the carrier distribution in the pn region? And the result of this calculation is that a VOLTAGE change of 2mV/K can compensate this temperature caused change in Ic.
Do you understand the meaning of this figure? A reduction in Vbe brings the current Ic back to its original value! 
That is the main point which - among others - prooves that Vbe is the quantity which influences Ic. In this calculation, Ib does not appear at all .
 

[miguelvp]

You are so not understanding what the other poster was saying. Lift doesn't have to be generate by pushing down air. And rotorcraft doesn't work the way you think it does.

Well it does, actually. Heavier than air craft need an upward force to keep them in the air. This upward force is generated by accelerating some of the surrounding air downwards to generate a reaction force (F = ma). Where airplanes differ from rockets in this regard is that airplanes use the surrounding air for their reaction force while rockets generate their own exhaust gases.

Although some air pushes down due to the angle of attack, most of the lift happens when air particles are accelerated on the top of the air foil causing a difference in pressure from the bottom foil.

At least that is my understanding, but I haven't done any airmodels for a long time since I was a kid but I did get my parents to buy me subscription to a airmodel course by post. So I'm rusty since that was a long time ago.

[c4757p]

Almost everybody in this thread needs to be slapped on the head with the words correlation does not imply causation.

[dannyf]

Although some air pushes down due to the angle of attack,

Pitch can create lift, but it also increases drag thus slows down the airplane which in turn decreases lift.

If anything, the up side of a wing's leading edge does push up the air a little.

most of the lift happens when air particles are accelerated on the top of the air foil causing a difference in pressure from the bottom foil.

Yes, you can have zero pitch (or even slightly negative pitch) and generate lift - that pretty much says it all about where lift comes from.

The same with two boats in parallel motion get sucked together.

[IanB]

Although some air pushes down due to the angle of attack, most of the lift happens when air particles are accelerated on the top of the air foil causing a difference in pressure from the bottom foil.

Again, the angle of attack. What is this fixation with the angle of attack?

[miguelvp]

Although some air pushes down due to the angle of attack, most of the lift happens when air particles are accelerated on the top of the air foil causing a difference in pressure from the bottom foil.

Again, the angle of attack. What is this fixation with the angle of attack?

It affects the lift coefficient:


source: http://en.wikipedia.org/wiki/Lift_coefficient

[IanB]

Again, the angle of attack. What is this fixation with the angle of attack?

It affects the lift coefficient:

Well, doh!

But why does it keep getting introduced into this thread?

No wonder so many have been having trouble with transistors. There is an almost complete inability to marshal facts, to separate the relevant from irrelevant, and to sort out the relationships between the facts that matter.

Is it not obvious that there is a parallel between current and voltage in transistors, and between pressure differentials and downward air deflection in wings?

[T3sl4co1l]

The fundamental problem is that people follow blindly rather than reasoning it out themselves.  Practical experiments can be designed to positively or negatively prove different aspects of working hypotheses.  This isn't sociology here, it can be tested and repeated with complete certainty.

I presented exactly such an experiment earlier in the thread!  Amazingly enough, it was summarily ignored by all!

The other fundamental problem is, the people who teach those who follow blindly, also follow blindly.  School curricula NEVER say "oh by the way there's more, but we're just teaching you THE SIMPLIFICATION NOW because we don't want to overwhelm you".  They just say, "oh, there's no such thing as sqrt(-1), if you get that, you've done it wrong."  "OH BY THE WAY THERE'S COMPLEX NUMBERS"    Shit like this and it's no wonder everyone seems to unanimously have problems with math, science, any technical subject!

Tim

[Electro Fan]

Thanks for the informative and entertaining thread. 

I happened to see the article below.

The article might just be presenting semantics or it might offer some meaningful distinctions or it might be wrong - I really don't know - in my mind voltage is needed to enable current (so even after reading this thread and the article I don't get how something can be current driven without also being voltage driven) - but I can't claim to have seen any electrons, holes, or other such phenomenon.  I just try to observe results, learn from others, study, and hopefully understand - especially when the proverbial light bulb goes on, or off    (And after reading this thread and the article I'd say my light bulb is flickering )

Seriously, I don't want to heat up the debate but rather just contribute info that might possibly help explain something or at least help a friendly discussion. 

Happy New Year EEVers!  EF

---

Difference Between BJT and MOSFET

http://www.differencebetween.net/technology/difference-between-bjt-and-mosfet/

The transistors BJT and MOSFET are both useful for amplification and switching applications. Yet, they have significantly different characteristics.

BJT, as in Bipolar Junction Transistor, is a semiconductor device that replaced the vacuum tubes of the old days. The contraption is a current-controlled device where the collector or emitter output is a function of the current in the base. Basically, the mode of operation of a BJT transistor is driven by the current at the base. The three terminals of a BJT transistor are called the Emitter, Collector and Base.

A BJT is actually a piece of silicon with three regions. There are two junctions in them where each region is named differently ‘“ the P and N. There two type of BJTs, the NPN transistor and the PNP transistor. The types differ in their charge carriers, wherein, NPN has holes as its primary carrier, while PNP has electrons.

The operation principles of the two BJT transistors, PNP and NPN, are practically identical; the only difference is in biasing, and the polarity of the power supply for each type. Many prefer BJTs for low current applications, like for switching purposes for instance, simply because they’re cheaper.

Metal Oxide Semiconductor Field-Effect Transistor, or simply MOSFET, and sometimes MOS transistor, is a voltage-controlled device. Unlike the BJT, there is no base current present. However, there’s a field produced by a voltage on the gate. This allows a flow of current between the source and the drain. This current flow may be pinched-off, or opened, by the voltage on the gate.

In this transistor, a voltage on an oxide-insulated gate electrode can generate a channel for conduction between the other contacts ‘“ the source and drain. What’s great about MOSFETs is that they handle power more efficiently. MOSFETs, nowadays, are the most common transistor used in digital and analog circuits, replacing the then very popular BJTs.

Summary:

1. BJT is a Bipolar Junction Transistor, while MOSFET is a Metal Oxide Semiconductor Field-Effect Transistor.

2. A BJT has an emitter, collector and base, while a MOSFET has a gate, source and drain.

3. BJTs are preferred for low current applications, while MOSFETs are for high power functions.

4. In digital and analog circuits, MOSFETs are considered to be more commonly used than BJTs these days.

5. The operation of MOSFET depends on the voltage at the oxide-insulated gate electrode, while the operation of BJT is dependent on the current at the base.

[LvW]

Instead of commenting the last reply I like to enclose a pdf file showing a slide created by one of the world leading semiconductor developers: Barrie Gilbert.

[Zero999]

Instead of commenting the last reply I like to enclose a pdf file showing a slide created by one of the world leading semiconductor developers: Barrie Gilbert.

That may be very helpful for designing an analogue multiplier but it's no good for calculating the base resistor value when one wants to use a transistor as a saturated switch to turn on an LED; in that case it's much better to consider the transistor as a current controlled switch. Trying to use the voltage controlled model for a saturated BJT switch will cause big fuck ups "3.3V should be a high enough voltage to turn on the BJT enough to allow 200mA to flow through that motor but fuck the BJT and MCU are now dead and there's lots of smoke!"

[LvW]

Hero999 - thank you for this remark.
This gives me the opportunity to state that all of my previous contributions, statements and remarks are related to small-signal operation of the BJT (e.g. as an amplifier) only. 

[G0HZU]

Instead of commenting the last reply I like to enclose a pdf file showing a slide created by one of the world leading semiconductor developers: Barrie Gilbert.

Hi LvM

I still don't have a horse in this race. I think 'current control' and voltage control' are both risky labels to attach to a BJT if the winning label comes at the total demise of anything else.

At work I'm happy to design using behavioural models like the Ebers Moll or Gummel Poon (or even small signal s2p models) right up to many GHz and for simple stuff I'll just use a calculator.

If I had to comment on your Barrie Gilbert slide I'd say that the Vbe equation suggests that the BJT can be modelled as a voltage controlled current source based on numbers bashed into the supplied equation. So the equation suggests that Vbe is the daddy.

But under the hood I think there is a lot more to it and the meaning of the word 'control' is open to (mis)interpretation. Maybe a room full of semiconductor physicists could debate this one.

But I doubt their answers will make me any better at designing RF amplifiers or other circuits using BJTs. Although it would be a fascinating debate for an onlooker, I'm not sure I 'need' to know what they might say. Even if I understood it all it won't affect the manufacturer supplied models I use and it probably won't change how I design amplifiers or simple switching/control circuits using BJTs. That's why I don't see much value in it all.

Voltage 'control' or current 'control'? I'm not aiming to pick a winner on this one. I'd be happier betting on both horses to be flawed or limited losers

[free_electron]

Instead of commenting the last reply I like to enclose a pdf file showing a slide created by one of the world leading semiconductor developers: Barrie Gilbert.

Do you have the whole document ? the fact that this was presented at a conference on bandgaps leads me to believe this is part of a simplified explanation how to design band gap circuits. Bandgaps work because they send a constant (temperature compensated) current through a forward junction. for a diode at constant current the forward drop is constant (in this case vbe). Thie voltage across the  first junction junction is then used to control a second junction. a delta amplifier cancels out the error and in doing so rejects the temperature creating an ultra stable voltage equal to the bandgap of silicon ( around 1.2 volts)

in that case the base current is indeed considered a parasitic. you don't care as it is driven from an opamp and set by the emitter resistor in the bandgap circuit. you are after the Vbe in this system. currents in bandgaps are parasites.

[LvW]

I still don't have a horse in this race. I think 'current control' and voltage control' are both risky labels to attach to a BJT if the winning label comes at the total demise of anything else.
..................
Voltage 'control' or current 'control'? I'm not aiming to pick a winner on this one. I'd be happier betting on both horses to be flawed or limited losers

Hi GOHZU,
I agree with you - nearly up to 100% - as far as you are speaking to me as a designer.
 And - that`s what I have mentioned several times in this thread (see my posts#148 adn #153):
Independent on our "religious attitude" (voltage or current control) - we all are using the same formulas and even the same design steps for designing amplifiers.  Hence, it is complete nonsense to say "I am using the current-control or the voltage-control approach". There is only one single method to design a common emitter stage with Re-feedback and a resistive voltage divider.
However -  the situation is completrely different for somebody who has the task to EXPLAIN the working principle of the BJT to students. 
Students are not satisfied to hear: "Doesn`t matter - why and how it works, apply a set of formulas - and be happy if the circuit works..".

I gave some examples (tempco -2mV/K, Early-effect,...) which can be explained with the voltage control principle only (I intentionally avoid the term "model" which may give rise to misunderstandings). Therefore, I see no reason why I should start with the current-control approach (some people believe it is easier - it is NOT).

But at he end of your contribution you have mentioned a very important aspect: ..."the meaning of the word 'control' is open to (mis)interpretation"
Yes - that´s true.
Let´s take a basic example: An opamp wired as a current-to-voltage converter (current source at the inv. input which has a feedback resistor).
My description: We use a voltage controlled opamp to realize a circuit that produces an ouput voltage which is controlled by an input current.
In short: The circuit is current controlled, but the active device (of course!) is still voltage controlled.

And that is the terminology I have used in my previos contributions: The controlling quantity for the BJT (alone!) is that quantity which directly acts upon the electrical field within the depletion area of the BJT, which in turn determines the current Ic.

 

[LvW]

Instead of commenting the last reply I like to enclose a pdf file showing a slide created by one of the world leading semiconductor developers: Barrie Gilbert.

Do you have the whole document ?

No sorry.
But (as given already in my reply#118) here is another statement from B. Gilbert :

The old current-in, current-out seems view simple at first, but that's about as far as it goes.
We clearly agree that the BJT should be seen in the same way as an MOS device, explaining that the DC base current of the BJT
is actually due to a defect (of sorts) and only a nuisance.
At Analog Devices we have made BJTs (under special conditions) having a DC beta of over 25,000.

[Zero999]

V_BE can be a parasitic and just as much of a nuisance when designing an emitter follower. In that case, it would be nice if the BJT had no V_BE and was completely current controlled. That way there wouldn't be that nagging V_BE to limit the maximum output voltage It would make designing a low drop-out regulator much easier, rather than having to have a common emitter on the output which can be unstable, there could be a nice emitter follower which could saturate close to the positive rail, rather than V_BE below it.

[miguelvp]

Re: Barry Gilbert if you notice the slide ruler you will realize that he is using this for analog computing so he only cares about adding logarithmic relations off two BJT transistors.



Interesting one hour long talk:


Check @9:10 for a glimpse of multiplication using bipolar transistors like a slide ruler will.


But with that model you loose linearity but handy for his purpose.