How a BJT Transistor works (base current version)

[(*steve*)]

Final question to Ratch and Steve (just for my understanding what we are talking about

Hi LvW. As you are aware, it is my assertion that BJT's are voltage controlled, but that treating them as current controlled (with the caveat that it is only a rough approximation) is a useful first step in understanding how they are used.

I am following this thread with interest, and my question to Ratch was nothing more than it appeared. I was wondering if he was asserting correlation or causation. If he had answered "causation" I would need to have gone back and read it again because I didn't get that on the first reading and he was not specific in the post.

 

[LvW]

Hi LvW. As you are aware, it is my assertion that BJT's are voltage controlled, but that treating them as current controlled (with the caveat that it is only a rough approximation) is a useful first step in understanding how they are used.

Hi Steve,
your answer reminds me again on one of your posts (post#16) which was a reply to my assertion that - for my opinion - a beginner needs the voltage-control approach in order to understand RE feedback. The contents of your reply was a short comment („That's absolute rubbish“) and a calculation. Of course, it is not a problem to recalculate any feedback system without using the term „feedback“ at all. But this explains nothing. Remember: My emphasis was on „understanding“, which means: Explaining an observable effect using only words.
Do you now understand my position? (By the way: The result of your calculation looks a bit „weird“; I`ll send you a PM on this).

LvW

Remark: Instead of a "PM" I`ve sent the text as a "conversation".

 

[Ratch]

I uploaded the file with the common emitter bjt amp stage referenced in an earlier post so that members can refer to it. Questions/comments are welcome.

Claude

I looked over your calculations and found only a minor arithmetic error, in that ic should be 0.00600785, not 0.006078. That makes the amplification range from -10.8338 to -12.4135. I plotted the gain vs beta below.


As you can see the gain does not change much after beta > 200.

I also calculated the gain disregarding Vbe. The gains are -10.974 to -12.5857 when beta ranges from 50 to 500. In this particular circuit, it hardly seems worth the effort to worry about Vbe.

Ratch

 

[cabraham]

"I've responded plenty. You ask for proof, I give it, ask for examples, I give them."

I cannot find in all your post´s any example or observation which can be explained with current control only - or which cannot be explained with Vbe-control. Please, help me to find it.

Claude, continuously you are speaking of Ie as a controlling quantity - but you totally ignore the fact that there must be voltage source driving such a current.
Or do you deny the old rule "No current without charge separation/charge imbalance (voltage)" ?

The voltage source you mention is NOT Vbe, it is the battery, antenna, mic, etc. Vbe is not the driver, it absorbs energy, not provide it. Besides, Ie can be driven by a current source. By definition in order to provide current, the CCS must output a voltage unless driving a SC.

Your claim is that in every branch of every circuit, barring shorts and opens, wherever there is a current, there has to be a voltage "driving it". Classic prejudice. There certainly has to be a voltage, but the voltage does not necessarily "drive" the current. You invoke the old rule about charge separation/imbalance. But please note that charges don't separate unless they are moved. Moving a charge is by definition a current. So we need voltage to get current, but at the same time we need current to get voltage as well. They are correlated but you blindly assume that voltage is the **cause**.

Correlation is not causation. A forward diode current through the p-n junction obviously is driven by a source either CCS or CVS. But the forward drop Vd is NOT the driver of its own current. That is the confusion the "voltage party" cannot emancipate their minds from. If the overall network is driven by a 12 volt battery, with a 1.0 kohm resistor and diode, the battery redox action imparts energy to charges, 12V = 12 joule/coulomb. The resistor and diode absorb energy, 11.3 joule/coulomb are converted to heat by the resistor, and 0.70 J/V by the diode. Vd, the diode forward drop, is NOT the "driver" of Id. Until that is settled, we will have endless contention.

Your side just keeps insisting that Ie cannot happen w/o Vbe, and that Vbe controls Ie. In steady state analysis, sims, lab tests, it is a stalemate. A change in input signal, mic/antenna, whatever, produces a change in both Ie and Vbe. Ic increases as a result. Ie changing inevitably forced Vbe to change as well. A scope plot affirms that Vbe changes due to increased carrier density in the b-e junction. More charges drifting through the junction results in increased recombination after crossing the junction, and enhancing charge density at edges of depletion zone.

Of course, as soon as Ie increased, more electrons are emitted, so collector current increases immediately since electrons collected must increase. Vbe settles after both Ie and Ic plateau. I provided plots. The fact that there is an emitter resistor does not "spoil" the result. If Vbe was truly what controls Ie, placing an Re in series w/ the emitter does not alter that. A capacitor is still subject to Eli the ice man regardless of emitter resistor.

Lastly, you keep insisting that placing a CVS right across the b-e junction "proves" Ie/Ic dependence on Vbe, but that is well known. We can place an inductor in series with the emitter, energize it, open the path across the inductor so that its only remaining path is through the emitter-base junction. Then the emitter current will rise, followed by Vbe rising. Ic will rise and settle immediately, Vbe a little later. I attached switching converter sims and catch diode waveforms, clearly illustrating that the current enters the diode, giving risr to its voltage.

This is hard to accept because some believe that every current has a voltage driving it. But one can easily demonstrate otherwise. I will post more sims this weekend. I mentioned I would do so last weekend, but I had a business trip which took up the time. BR.

Claude

 

[Arouse1973]

What drives current if its not voltage. Please define what voltage is. Please again answer my questions. I will post my findings this weekend regardless.
Adam

 

[Ratch]

The voltage source you mention is NOT Vbe, it is the battery, antenna, mic, etc. Vbe is not the driver, it absorbs energy, not provide it. Besides, Ie can be driven by a current source. By definition in order to provide current, the CCS must output a voltage unless driving a SC.

Your claim is that in every branch of every circuit, barring shorts and opens, wherever there is a current, there has to be a voltage "driving it". Classic prejudice. There certainly has to be a voltage, but the voltage does not necessarily "drive" the current. You invoke the old rule about charge separation/imbalance. But please note that charges don't separate unless they are moved. Moving a charge is by definition a current. So we need voltage to get current, but at the same time we need current to get voltage as well. They are correlated but you blindly assume that voltage is the **cause**.

Correlation is not causation. A forward diode current through the p-n junction obviously is driven by a source either CCS or CVS. But the forward drop Vd is NOT the driver of its own current. That is the confusion the "voltage party" cannot emancipate their minds from. If the overall network is driven by a 12 volt battery, with a 1.0 kohm resistor and diode, the battery redox action imparts energy to charges, 12V = 12 joule/coulomb. The resistor and diode absorb energy, 11.3 joule/coulomb are converted to heat by the resistor, and 0.70 J/V by the diode. Vd, the diode forward drop, is NOT the "driver" of Id. Until that is settled, we will have endless contention.

Your side just keeps insisting that Ie cannot happen w/o Vbe, and that Vbe controls Ie. In steady state analysis, sims, lab tests, it is a stalemate. A change in input signal, mic/antenna, whatever, produces a change in both Ie and Vbe. Ic increases as a result. Ie changing inevitably forced Vbe to change as well. A scope plot affirms that Vbe changes due to increased carrier density in the b-e junction. More charges drifting through the junction results in increased recombination after crossing the junction, and enhancing charge density at edges of depletion zone.

Of course, as soon as Ie increased, more electrons are emitted, so collector current increases immediately since electrons collected must increase. Vbe settles after both Ie and Ic plateau. I provided plots. The fact that there is an emitter resistor does not "spoil" the result. If Vbe was truly what controls Ie, placing an Re in series w/ the emitter does not alter that. A capacitor is still subject to Eli the ice man regardless of emitter resistor.

Lastly, you keep insisting that placing a CVS right across the b-e junction "proves" Ie/Ic dependence on Vbe, but that is well known. We can place an inductor in series with the emitter, energize it, open the path across the inductor so that its only remaining path is through the emitter-base junction. Then the emitter current will rise, followed by Vbe rising. Ic will rise and settle immediately, Vbe a little later. I attached switching converter sims and catch diode waveforms, clearly illustrating that the current enters the diode, giving risr to its voltage.

This is hard to accept because some believe that every current has a voltage driving it. But one can easily demonstrate otherwise. I will post more sims this weekend. I mentioned I would do so last weekend, but I had a business trip which took up the time. BR.

Claude

In a diffusion diode, current is controlled a little differently than in a resistor. The diode current is controlled by diffusion. This is caused by uneven and different doping between the P-N slabs. The voltage across the diode controls the diffusion by opposing the back-voltage caused by the uncovered charges. This is why a diode has a exponential voltage-current curve instead of a linear curve that a resistor has. The physics of the diode show that the diode responds to voltage, not current. You cannot put a current into a diode without first applying a voltage. This is true of any conductive circuit. An electron gun like a CRT or the Sun can shoot electrons into a vacuum, but the diode does not work that way. Now the physics of a magnetic amplifier (MA) works differently. The saturable core of the MA is dependent only on the current, not the voltage. This means the MA is a current controlled device, and not a voltage controlled device like a diode or a BJT or a FET is. If a tube, BJT, or FET responded to current the was a MA does, then we would agree that they are current controlled.

To determine what controls what, you have to look at the device's physics and determine to what it is responding. Stored energy from junction capacitances that skew the phase of voltage-current response do not matter. Neither do external components attached to a device which turn it into a circuit have any validity. You have to follow the physics.

Ratch

 

[Arouse1973]

So the original argument was that Ie controls Ic and not Vbe because in the simulation it shows Vbe lagging Ie. I know the reason for this but I wanted to show that with an ideal VCCS the same result can be obtained.

So in the simulation there is no way for the current in the source to be effected other than the input voltage. In this simulation there is no transistor model at all, no base current and no emitter current influence on collector current.

Now I know that a simulation of this type doesn't show the working physics of a component, that's not the point of this.

Original Circuit and waveforms




So it can be seen that Vbe lags Ie, yep can't deny that.

VCCS circuit and waveforms





And the same can be seen from the VCCS, how is this possible?



The Ie in the first plot is the current leaving the emitter terminal not the emitter junction current, you can't measure that in that simulation. This is effectively the emitter junction current minus Ib obviously. I have added a slight offset so you can see the two together. This shows Vbe tracking effectively Ie with no lag.



Thanks
Adam

 

[LvW]

Lastly, you keep insisting that placing a CVS right across the b-e junction "proves" Ie/Ic dependence on Vbe, but that is well known.
Claude

Claude - in all technical discussions, I try to be as objective as possible without any personal remarks - however, I cannot resist to ask you:
Are you an engineer or a physician or a philosopher or something else?

To me, it is really surprising to see how somebody is trying to complicate things and, at the same time, avoiding clear answers. In your latest answer you mention antennas, batteries, microphones, energy absorptions, conversion to heat, inductors in the emitter path as well as „ELI the ICE man“ (in earlier posts it was „Sue the singer..).

In this thread, we are discussing a technical question which is rather unambiguous:
Which quantity controls Ic? And to make it clear again: The BJT is a three-terminal device, and as an engineer I am interested only in the question:
What kind of external signal (voltage or current) at which of these 3 available terminals (pins) is necessary to cause (a) a certain quiescent dc current Ic
and (b) a change of this current.
The answer can be found by studying physical effects on charged carrier level supported by measurements, observations and explanations for the behaviour of real existing circuits. (And perhaps it is not an error to look for corresponding answers from some other information sources like universities with high reputation).

Today we can count already approximately 140 contributions - and from the beginning it was my impression (correct me if I am wrong) that Claude does NOT support the claim that the voltage Vbe would be the quantity which controls Ic.
However, suddenly he succinctly states:
"Lastly, you keep insisting that placing a CVS right across the b-e junction "proves" Ie/Ic dependence on Vbe, but that is well known."

Well known ? Are you joking with us? And what are we discussing here since several days?

May I remind you on some of your earlier earlier posts? Here are examples:

post#23: The theory that vbe is controlling ib & ie, which infers control over ic as well, cannot withstand even a mild scrutiny
post#23: The fact that ic/ie/ib settle to final plateau values while vbe continues to rise is proof positive that vbe is NOT the mechanism that "controls" ie/ib/ic.
post#23 It is ie that controls ic, not ib.
post#25 It is clear that Sue the singer is what "controls ie", and clearly NOT vbe.
post#36 Your plot is proof positive that vbe is not in control of ie
post#50 Even at low speed, Vbe does not control Ie/Ic/Ib
post#110 Because Vbe is the result of Ie! That is why I claim it, because it is so. (My comment: Good justification!)

Finally, I must admit that, to me, a serious and proper discussion looks quite different.

LvW

 

[Arouse1973]

"Are you an engineer or a physician or a philosopher or something else?"
Politician

 

[cabraham]

I am an engineer. I am near completion of my Ph.D. Went back to school in my 50's for the doctorate. I've been employed as an engineer for 36 years, doing hardware R&D, analog, digital, optical, and power. I have 2 published papers, and 5 U.S. patents. More to come. I have another sim/plot for viewers to study. THis is an inductor energized by a CVS plus resistor, then it de-energizes into a diode, akin to a switching power supply. Please pay careful attention to the diode I & V. It is all to obvious that I establishes V, beyond a doubt. In the reverse direction, V will establish I since the supply appears across the diode in reverse. I can control V, or V can control I depending on the boundary conditions. Questions/comments welcome.

Regarding Ic/Ie dependence on Vbe, I am not joking at all. If I were to drive the b-e junction w/ a CCS, a collector from another bjt, an inductor, etc., then I would record a functional dependence of Vbe upon Ie. What is a math "function"? What is a functional relationship?

If we have 2 variables, x & y, y is said to be a function of x if and only if there is but one unique value of y for every value of x. An example is y = mx+b. Here y depends on x. But, sometimes, there can be an inverse relation as well. If y = mx+b, is it not equally true that x = (y-b)/m? Is it? Here for every y value, there is one unique value of x.

Sometimes is is not bi-directional. Take y = cos x, it is safe to say that y is a function of x, since every x value results in a unique y value. But x is NOT a function of y. Since x = arcos y, for any y value, take y = 1, there are infinite numbers of x values. That is if y=1, x can equal 0, 2*pi, -2*pi, etc. So y is a function of x, but x is not likewise a function of y.

Take the log/exponential relations found in p-n junctions. Take y = exp(x). This is single valued, for every x there is but 1 y. Likewise, it is equally true that x = ln(y). It is safe to say that x is also a function of y as well, since any y value corresponds to a unique x value.

This is too well known. SO take the Shockley equation for a diode or any p-n junction, Id = Is*exp((Vd/Vt)-1). Here, for every value of Vd, there can be only 1 value of Id, hence Id is a function of Vd. But we know that Vd = Vt*ln((Id/Is)+1) is also valid. Here, for any value of Id, there is but 1 value of Vd.

Just as Id is a f(Vd), it is also true that Vd is a f(Id). Which is the "cause vs. effect" is an endless chicken-egg vicious circle. In steady state it is impossible to change just 1 of them. If 1 is perturbed we know the other is as well. So how do we determine if Vd changes 1st resulting in Id changing, or vice-versa. We perturb the diode and observe the I & V waveforms. The attached plots do just that.

Current in an inductor continues after the battery voltage switches to zero, it continues into the diode D1. THe plot clearly shows the current pre-existing the voltage. We know that it does because inductor current was established while the diode was reverse biased. Please study the plots and I will clarify if needed.

This is hard to believe. The evidence I provided is overwhelming and all my critics do is deny. The myth needing to be dispensed with is the one that assumes currents are caused by voltage. V does not "drive" I. Who says it does?

Claude

 

[Ratch]

e

I am an engineer. I am near completion of my Ph.D. Went back to school in my 50's for the doctorate. I've been employed as an engineer for 36 years, doing hardware R&D, analog, digital, optical, and power. I have 2 published papers, and 5 U.S. patents. More to come. I have another sim/plot for viewers to study. THis is an inductor energized by a CVS plus resistor, then it de-energizes into a diode, akin to a switching power supply. Please pay careful attention to the diode I & V. It is all to obvious that I establishes V, beyond a doubt. In the reverse direction, V will establish I since the supply appears across the diode in reverse. I can control V, or V can control I depending on the boundary conditions. Questions/comments welcome.

It doesn't matter how many degrees, schooling, or experience you have. If you ignore the physics of a device, you will never understand how it really works, even though you have a good grasp of how to make a device do something useful.

Regarding Ic/Ie dependence on Vbe, I am not joking at all. If I were to drive the b-e junction w/ a CCS, a collector from another bjt, an inductor, etc., then I would record a functional dependence of Vbe upon Ie.

That is where you are going off track. Just because two quantities have a one to one corresponding relationship over a range of values does not mean that their cause and effect is ambiguous. You can find the inverse of the relationship of Ie and Veb, and determine what Vbe would be at a particular value of Ie, but that does not prove which quantity is dependent on the other. You have to understand the physics of the device before you can determine what the dependency is.

What is a math "function"?

A functional relationship expressed as a mathematical equation.

What is a functional relationship?

A correspondence between the values of two or more quantities.

If we have 2 variables, x & y, y is said to be a function of x if and only if there is but one unique value of y for every value of x. An example is y = mx+b. Here y depends on x. But, sometimes, there can be an inverse relation as well. If y = mx+b, is it not equally true that x = (y-b)/m? Is it? Here for every y value, there is one unique value of x.

Yes, mathematically speaking, any of the variables can be defined to be dependent or independent.

Sometimes is is not bi-directional. Take y = cos x, it is safe to say that y is a function of x, since every x value results in a unique y value. But x is NOT a function of y. Since x = arcos y, for any y value, take y = 1, there are infinite numbers of x values. That is if y=1, x can equal 0, 2*pi, -2*pi, etc. So y is a function of x, but x is not likewise a function of y.

That is why mathematicians define a range where the one to one correspondence exists. That range is called the "principal value" of the function.

Take the log/exponential relations found in p-n junctions. Take y = exp(x). This is single valued, for every x there is but 1 y. Likewise, it is equally true that x = ln(y). It is safe to say that x is also a function of y as well, since any y value corresponds to a unique x value.

Yes, and the point is?

This is too well known. SO take the Shockley equation for a diode or any p-n junction, Id = Is*exp((Vd/Vt)-1). Here, for every value of Vd, there can be only 1 value of Id, hence Id is a function of Vd. But we know that Vd = Vt*ln((Id/Is)+1) is also valid. Here, for any value of Id, there is but 1 value of Vd.

Yes, and the point is?

Just as Id is a f(Vd), it is also true that Vd is a f(Id). Which is the "cause vs. effect" is an endless chicken-egg vicious circle. In steady state it is impossible to change just 1 of them. If 1 is perturbed we know the other is as well. So how do we determine if Vd changes 1st resulting in Id changing, or vice-versa. We perturb the diode and observe the I & V waveforms.

The physics of the diffusion diode put an end to the "chicken and egg" question. Physics shows that the diode current is dependent of the voltage across the diode. Sure, you can calculate what the voltage would be if such and so current were present in the diode, but that does not show that current is the independent quantity. If you send a current into the diode with a CCS, the CCS will put a voltage across the diode to according to Schockley's equation with voltage as the independent variable.

The attached plots do just that.

Fine, you can drive a diode with a CCS, but then you are evaluating a circuit with a high value resistor, not just a diode by itself.

Current in an inductor continues after the battery voltage switches to zero, it continues into the diode D1. THe plot clearly shows the current pre-existing the voltage. We know that it does because inductor current was established while the diode was reverse biased. Please study the plots and I will clarify if needed.

Yes, caps and inductors are energy storage elements. The energy stored in their electromagnetic fields move charges and do work when there is no counter-acting voltage or current, and a conduction path exists. When does that prove?

This is hard to believe. The evidence I provided is overwhelming and all my critics do is deny.

Your evidence is falsely interpreted and much of the time is irrelevant. You appear to extrapolate mathematical relationships and circuit based voltage-current plots into physical principles which are just not true. Finally, you do not seem to understand the fact that if you study a circuit into which a device is embedded, you are not studying only the device itself.

The myth needing to be dispensed with is the one that assumes currents are caused by voltage. V does not "drive" I. Who says it does?

Physics does. A charged particle is not going to move unless it is shot from a CRT gun, ejected from the Sun, attracted or repelled by a electric field, or undergoes a direction change from a magnetic field if the charge is moving with respect to right angles of the magnetic field. In a straight wire, none of the above exists except an electric field caused by applying a voltage across the wire. Perhaps you can supply an example circuit devoid of any electrical energy storage elements where electrons move without voltage.

Ratch

 

[LvW]

Hi Ratch - I was prepared to answer Claude, however I can spend my time for some more important actions.
The reason is because you already have done the job.
It is really surprising (funny?) that somebody with his background ("near completion of Ph.D.") constantly is repeating the story about y=exp(x) and x=ln(y).
He doesn`t understand that physics does not allow all mathematical manipulations.
(Electrical wattage V*I causes a temperature increase proportional to the thermal resistance: delta(T)=Rth*V*I) ; what about the recverse effect?)
And - instead of explaining the appearent inconsistency as mentioned in my former post - he is repeating (at least) the third time in a row all his assertions.
And he doesn`t understand that simulations cannot proove everything - in particular not questions connected with cause and effect.

LvW

 

[cabraham]

e

It doesn't matter how many degrees, schooling, or experience you have. If you ignore the physics of a device, you will never understand how it really works, even though you have a good grasp of how to make a device do something useful.



That is where you are going off track. Just because two quantities have a one to one corresponding relationship over a range of values does not mean that their cause and effect is ambiguous. You can find the inverse of the relationship of Ie and Veb, and determine what Vbe would be at a particular value of Ie, but that does not prove which quantity is dependent on the other. You have to understand the physics of the device before you can determine what the dependency is.

I've been saying the same. I showed in a different thread the function of a switching power supply. When FET is off, catch diode is placed across the input supply. When FET is on, inductor in series with diode. In 1st case, V controls I in the diode. In 2nd case, I controls V. I've stated many times that either can be the independent quantity.

A functional relationship expressed as a mathematical equation.



A correspondence between the values of two or more quantities.



Yes, mathematically speaking, any of the variables can be defined to be dependent or independent.



That is why mathematicians define a range where the one to one correspondence exists. That range is called the "principal value" of the function.



Yes, and the point is?



Yes, and the point is?



The physics of the diffusion diode put an end to the "chicken and egg" question. Physics shows that the diode current is dependent of the voltage across the diode. Sure, you can calculate what the voltage would be if such and so current were present in the diode, but that does not show that current is the independent quantity. If you send a current into the diode with a CCS, the CCS will put a voltage across the diode to according to Schockley's equation with voltage as the independent variable.

But your argument does not show that voltage is independent either. I've said the same since day one. Only perturbing and observing can we see which is independent.

Fine, you can drive a diode with a CCS, but then you are evaluating a circuit with a high value resistor, not just a diode by itself.



Yes, caps and inductors are energy storage elements. The energy stored in their electromagnetic fields move charges and do work when there is no counter-acting voltage or current, and a conduction path exists. When does that prove?



Your evidence is falsely interpreted and much of the time is irrelevant. You appear to extrapolate mathematical relationships and circuit based voltage-current plots into physical principles which are just not true. Finally, you do not seem to understand the fact that if you study a circuit into which a device is embedded, you are not studying only the device itself.

Device behavior does not change because it is in a circuit. A bjt is exclusively used in a circuit, not bare. Any study of its bare properties has little relevance in practical real world applications.

Physics does. A charged particle is not going to move unless it is shot from a CRT gun, ejected from the Sun, attracted or repelled by a electric field, or undergoes a direction change from a magnetic field if the charge is moving with respect to right angles of the magnetic field. In a straight wire, none of the above exists except an electric field caused by applying a voltage across the wire. Perhaps you can supply an example circuit devoid of any electrical energy storage elements where electrons move without voltage.

I've been saying ad infinitum that some type of source, CVS or CCS, is needed to move charges. But Vbe/Vd are voltage drops, not emf's. A battery voltage is an emf. The current and voltage in the battery are oriented so that ions inside it are gaining energy. Hence the "12 volts" in a 12 volt battery is the energy of 12 joules GAINED by each coulomb of charge. The 0.10 volts across the wire is the energy per charge LOST. If there is 1 point needing emphasis it is that. THe energy storage/conversion elements provide emf, and the passive elements incur drops. In a diode-resistor-battery circuit, the battery provides all the energy. Vd, the diodes forward drop is not moving charges. You insist that E fields and their associated voltage are what moves charges, but you are mistaken to think that a diode's own forward drop Vd is what moves the charges forming current Id.

Ratch

 

[cabraham]

Another set of sims, these are very helpful. The circuit is a discreet linear voltage regulator I've been using a while. I've designed this low dropout regulator into trucks, busses, cars, and airplanes. It provides protection against over-voltage transients like load dump, inductive switching, lightning strokes etc. It uses 3 active devices, 2 bjt parts, and one FET. They all operate in the active region as amplifiers.

Please note that the pnp bjt exhibits a collector current that closely tracks the emitter current, ref file name "ovp0 q2 vbe-ie-ic zoom.jpg". But vbe continues to settle long after ie/ic settle. It is too obvious that ic is tracking ie, with vbe exerting virtually no influence on ic.

Now we look at the FET, file name "ovp0 q3 vgs-ig-id zoom.jpg". This is interesting. As expected, gate current ig leads gate-source voltage vgs by a lot. No surprise as a FET is almost a pure capacitance, opposed to a bjt where much of the current is conducted, only a fraction displacement to charge diffusion capacitance. But look at id, the drain current. Drain current lags gate current and closely tracks vgs, the gate-source voltage. Id vs. Vgs is closely correlated. Ig is needed to charge up Vgs, but Id does not respond until Vgs is charged. Id is controlled not by Ig, but by Vgs.

In the pnp bjt, as soon as Ie changes, Ic responds immediately, Vbe lagging. After Ie reaches plateau, Ic also plateaus and both settle in unison. But Vbe continues to change before settling at its new plateau.

This really demonstrates the current controlled nature of bjt devices, and the voltage controlled nature of FET devices. Peace to all.

Claude

 

[Ratch]

"I've been saying the same. I showed in a different thread the function of a switching power supply. When FET is off, catch diode is placed across the input supply. When FET is on, inductor in series with diode. In 1st case, V controls I in the diode. In 2nd case, I controls V. I've stated many times that either can be the independent quantity."

That's right, you did say that, but you are wrong. In a junction diode, it is voltage that controls current. Sending current through a diode will apply a voltage across it that will match the Schockley equation. I said that and explained why in terms of physics in my previous replies to you. It does no good to constantly repeat that current controls voltage. You have to explain why that is so in terms of the physics of the device, like I did for voltage controlling current many times before.

"But your argument does not show that voltage is independent either. I've said the same since day one. Only perturbing and observing can we see which is independent."

Yes it does, because my explanation is based on the physics of the device. You can perturb and disturb all day, but if you cannot explain what happens using the physics of the device, then it is all for naught.

"Device behavior does not change because it is in a circuit. A bjt is exclusively used in a circuit, not bare. Any study of its bare properties has little relevance in practical real world applications."

That's right, the device behaves the same, but the external components that make up the circuit have to be taken into account. Not so, studying a device's bare properties is important for understanding. A bare BJT is a transconductance DEVICE. But, if you insert a lot of resistance into the base circuit, then you have a current amplification CIRCUIT. The transistor still behaves like a transconductance device, but the circuit is something else. That is analogous to an OP-amp which is a twin high gain voltage amplifier. But look at all the things you can make by connecting up different components to its leads.

I've been saying ad infinitum that some type of source, CVS or CCS, is needed to move charges. But Vbe/Vd are voltage drops, not emf's. A battery voltage is an emf. The current and voltage in the battery are oriented so that ions inside it are gaining energy. Hence the "12 volts" in a 12 volt battery is the energy of 12 joules GAINED by each coulomb of charge. The 0.10 volts across the wire is the energy per charge LOST. If there is 1 point needing emphasis it is that. THe energy storage/conversion elements provide emf, and the passive elements incur drops. In a diode-resistor-battery circuit, the battery provides all the energy. Vd, the diodes forward drop is not moving charges. You insist that E fields and their associated voltage are what moves charges, but you are mistaken to think that a diode's own forward drop Vd is what moves the charges forming current Id.

I never said that Vd directly moves the charges in a junction diode. The voltage across the bare diode and the voltage source have the same value. If you insist on a diode-resistor-voltage source, then you have a circuit with an external component. Otherwise let's stick to a diode-voltage source. If you are going to get hung up on a battery, then we can substitute an electronic voltage source. The important thing is to put a voltage across the diode. Yes, in a diode, electric fields caused by diffusion propel the charges, and the Vd partly cancels the back-voltage caused by the uncovered charges due to the diffusion. I said that many times before, too.

"This really demonstrates the current controlled nature of bjt devices, and the voltage controlled nature of FET devices. Peace to all."

No it doesn't. Only an explanation using the physics of the device can determine what controls what.

Ratch

 

[cabraham]

"I've been saying the same. I showed in a different thread the function of a switching power supply. When FET is off, catch diode is placed across the input supply. When FET is on, inductor in series with diode. In 1st case, V controls I in the diode. In 2nd case, I controls V. I've stated many times that either can be the independent quantity."

That's right, you did say that, but you are wrong. In a junction diode, it is voltage that controls current. Sending current through a diode will apply a voltage across it that will match the Schockley equation. I said that and explained why in terms of physics in my previous replies to you. It does no good to constantly repeat that current controls voltage. You have to explain why that is so in terms of the physics of the device, like I did for voltage controlling current many times before.

"But your argument does not show that voltage is independent either. I've said the same since day one. Only perturbing and observing can we see which is independent."

Yes it does, because my explanation is based on the physics of the device. You can perturb and disturb all day, but if you cannot explain what happens using the physics of the device, then it is all for naught.

"Device behavior does not change because it is in a circuit. A bjt is exclusively used in a circuit, not bare. Any study of its bare properties has little relevance in practical real world applications."

That's right, the device behaves the same, but the external components that make up the circuit have to be taken into account. Not so, studying a device's bare properties is important for understanding. A bare BJT is a transconductance DEVICE. But, if you insert a lot of resistance into the base circuit, then you have a current amplification CIRCUIT. The transistor still behaves like a transconductance device, but the circuit is something else. That is analogous to an OP-amp which is a twin high gain voltage amplifier. But look at all the things you can make by connecting up different components to its leads.

I've been saying ad infinitum that some type of source, CVS or CCS, is needed to move charges. But Vbe/Vd are voltage drops, not emf's. A battery voltage is an emf. The current and voltage in the battery are oriented so that ions inside it are gaining energy. Hence the "12 volts" in a 12 volt battery is the energy of 12 joules GAINED by each coulomb of charge. The 0.10 volts across the wire is the energy per charge LOST. If there is 1 point needing emphasis it is that. THe energy storage/conversion elements provide emf, and the passive elements incur drops. In a diode-resistor-battery circuit, the battery provides all the energy. Vd, the diodes forward drop is not moving charges. You insist that E fields and their associated voltage are what moves charges, but you are mistaken to think that a diode's own forward drop Vd is what moves the charges forming current Id.

I never said that Vd directly moves the charges in a junction diode. The voltage across the bare diode and the voltage source have the same value. If you insist on a diode-resistor-voltage source, then you have a circuit with an external component. Otherwise let's stick to a diode-voltage source. If you are going to get hung up on a battery, then we can substitute an electronic voltage source. The important thing is to put a voltage across the diode. Yes, in a diode, electric fields caused by diffusion propel the charges, and the Vd partly cancels the back-voltage caused by the uncovered charges due to the diffusion. I said that many times before, too.

"This really demonstrates the current controlled nature of bjt devices, and the voltage controlled nature of FET devices. Peace to all."

No it doesn't. Only an explanation using the physics of the device can determine what controls what.

Ratch

 

[cabraham]

Hi Ratch - I was prepared to answer Claude, however I can spend my time for some more important actions.
The reason is because you already have done the job.
It is really surprising (funny?) that somebody with his background ("near completion of Ph.D.") constantly is repeating the story about y=exp(x) and x=ln(y).
He doesn`t understand that physics does not allow all mathematical manipulations.
(Electrical wattage V*I causes a temperature increase proportional to the thermal resistance: delta(T)=Rth*V*I) ; what about the recverse effect?)
And - instead of explaining the appearent inconsistency as mentioned in my former post - he is repeating (at least) the third time in a row all his assertions.
And he doesn`t understand that simulations cannot proove everything - in particular not questions connected with cause and effect.

LvW

I can't believe anybody believes this stuff. Regarding "reverse effect", heating up a resistor then reversing the process, is a bit off topic. I stated that just as I is a function of V, so is V a function of I. Function implies single value of a variable for every value of the other variable. Heating a resistor with I & V, emitting photons as heat, then attempting to "reverse the process" by gathering the photons back and regenerating I & V, is certainly not germane to this discussion.
Simulations can prove or disprove cause and effect sometimes. I submitted the l-r network where a diode is placed directly across an inductor. The inductor gets energized by a battery and resistor, then a switch opens, allowing inductor to de-energize through diode. The diode is shunted by 100 kohm, which is very large, since the diode resistance is much smaller. So we have an inductor with low winding resistance directly across a diode. The diode current leads the diode voltage, clearly visible in my plot.

If you wish, the inductor could be superconducting, so that it possesses no R to "mess up" the test results according to Ratch. I will address other problems in another reply, but for now LvW, please examine the sims above labeled "l-r-d0". They show a lot you are at a loss to explain. Also, please examine the "ovp" sims. A discrete low drop-out voltage regulator (LDO) uses 2 bjt parts, plus 1 FET.

The lower bjt, an npn, has an Re (emitter resistor). You claim that Re "messes up" test results as does Ratch. The FET has gate resistor Rg, and is driven by the upper bjt, a pnp part. So the FET is not "in the raw" either. Please notice that the FET waveforms clearly show that although Ig (gate current) greatly leads Vgs (gate-source voltage), the drain current Id is following Vgs. Vgs slowly decays exponentially toward its plateau and Id follows like a shadow.

But look at Ic/Ie vs. Vbe. Ic is slightly, very slightly lagging Ie, no surprise here. Ie is 1st established, then moments later electrons emitted from emitter get collected by collector. A slight latency is expected. But Vbe is way behind Ie/Ic, lagging drastically. Ie/Ic plateau, then Vbe drags its feet taking a relatively long time to settle. If Vbe change is really what controls Ie/Ic, you have never explained how Ie/Ic quickly attain new values long before Vbe does the same.

This voltage is causal theory is really a cult following. It is a dogma held w/ the force of a religion. No scientific test can affirm your theory. You assert w/o proof. I offer proof, and you claim that my Re is corrupting the test, the bjt must be naked to see it. Yet, the resistor in the FET gate does not seem to corrupt the voltage controlled nature of FETs. The FET plots clearly indicate voltage control, not current.

Timing inside a device cannot be reversed by external parts. Raising an Re value can affect internal bjt timing via time constant needed to charge junction capacitance, but the resistor cannot reverse time. If Vbe is causing Ie/Ic, how is it the Ie/Ic change ahead of Vbe in time? Nobody who rejects current control can answer that so you just discredit my sims based on bjt being inserted in a network instead of naked. Yet FETs can show their voltage controlled nature fine when placed in a circuit having over a dozen parts. Go figure.

Claude

 

[cabraham]

Regarding current sources, there seems to be some weird stuff being claimed. A current source is just that. A source that outputs whatever voltage is needed to maintain constant current in the face of varying impedances. Current sources are just as "real" as voltage sources. Take a bicycle generator. It can function as either, CVS or CCS. An ideal source of either type is impossible to make, but a non-ideal CCS can be made as easily as a non-ideal CVS.

A bike gen can be regulated by the pedaler for CVS or CCS. For CCS, if the handle bars are equipped with ammeter, and field current adjustment (wound rotor field), maintaining CCS function is as follows. Biker pedals until current on ammeter reads target value. Field current can be adjusted so that biker is producing a speed and torque that is comfortable. Say, the target is 1.0 amp, and the load is 10 ohm. Field current is adjusted so that a comfortable pedaling speed gives 1.0 amp. But then load changes from 10 to 20 ohm. To maintain 1.0 amp, power must double. If biker maintains torque, then field current must double. Or biker can maintain field current and maintain torque while doubling speed (rpm). But either approach results in constant current.

I know what critics are thinking - Claude all you describe is a "voltage" generator that allows adjustable voltage. But as we will see, a CVS is analogous to the CCS. To maintain a constant 10 volts, the biker pedals and observes a volt meter. With a 10 ohm load, the field current is adjusted for comfortable pedal speed/torque. Now drop the load to 5 ohms, and maintain 10 volts. What must be done? Increase field current or pedal speed or both. Torque will increase due to increased current. In the CCS case, load changing from 10 to 20 ohms dictates increased voltage.

The CVS generator at power plants operates much like this bike example. Generators do not naturally output constant voltage nor constant current. The stator inductance is very large. An ideal CVS has zero R, and most important zero L, or zero X_L. As frequency increases, this is hard to achieve. In open circuit, no load condition, a CVS shaft speed and field current are adjusted for 10 volts. Assume the generator full power is 100 watt, so that 1.0 ohm is full load at 10 volts.

The inductive reactance for a typical generator, known as synchronous reactance, is between 0.50 and 2.50 ohms reactive (j0.50 to j2.50 ohm). When open, V = 10V, and the emf generated is 10V. Synchronous reactance X_S is 1.50 ohm, typical. But fully load the output with 1.0 ohm resistive. The terminal voltage will drop like a rock. The new voltage at the terminals is 10V(1/(1.0 + j1.50)) = 5.547 volts. That is a 45% drop! To restore the 10 volts, we must increase field current and/or pedal with more torque.

By increasing torque we increased current. By increasing field current we compensated for the voltage dropped across X_S. Generators must be controlled to produce constant voltage, it's not natural. The power company is so good at it, we don't see the effort it takes. Our wall outlet is 120 volts, 60 Hz, and stays that way whether we draw 1.0 milliamp, or 20 amps. It is very easy to slip into the heretical mindset that V causes I, or V is the independent variable.

A CCS is much like a CVS. Constant current from a generator is not natural, it is contrived and artificial, just like CVS. The bike generator has a 1.5 ohm reactance. Per Thevenin/Norton, we have a voltage source in series with 1.5 ohm, or a current source in parallel with 1.5 ohm. Full load is 1.0 ohm. The truth is, the generator makes a lousy CVS, and a lousy CCS. For load normalized to 1.0 ohm, a "good" CCS should have over 100 ohms of shunt impedance. That way a 1.0 ohm load gets almost the entire generator current, and little regulation if any would be needed. Likewise a CVS that is "good, would possess less than 0.010 ohms of series resistance, assuring that little voltage drops in the resistance, almost all reaches load.

But, that doesn't happen, so we artificially manipulate our filed current, speed and torque to maintain either CC or CV. Both are attainable to an extent, but neither ideally. Peace to all.

Claude

 

[cabraham]

One more issue is the inductor. An inductor is NOT a high value voltage source plus a resistor of high value. That is a heresy. When an inductor is energized, then placed across a device, de-energizing into said device, the inductor transfers energy. Let's use electrons as the carriers. Electrons in the conduction band can stay there until knocked down to valence. When an inductor sources current, the conduction electrons need no E field to stay in conduction. But any resistance in the path results in electrons in conduction colliding with lattice ions. This can and does result in enrgy transfer where electrons drop in energy state from conduction down to valence. Hence these electrons are no longer conducting. Energy conservation occurs, a photon is emitted, and we feel the heat radiated from a resistor with current. THe inductor across the dioe in my sims, has current, and sources said current to diode. No E field is driving them, they are already in the highest state, the conduction level. Once they enter the diode, transit through n & p regions, some collide, dropping down to valence, giving off photons. Then they reach the junction, cross and recombine, ionizing and increasing depletion zone E field. The line integral of E field is voltage. The voltage Vd has increased as a result of inductors electrons in conduction band transiting through p & n, then crossing junction, and increasing depletion zone charge build up.

Inductor current produced a change in diode forward voltage drop, period. There is no magic large resistor in series with a large voltage source. If there was, more power would be dissipated. Inductors with current made from superconductors have been retaining their current for years with 0 loss. A voltage source plus resistor can't do that. The junction voltage Vd is clearly determined by the current. No other explanation makes sense. Sims are not infallible. But I've designed hundreds of power supplies, motor drivers, LED drivers, etc. I've seen on a scope in the lab waveforms from inductive energy transfer. Same result all the time. These sims affirm what I've witnessed in the lab. Anybody can set up these circuits and observe for themselves. Peace.

Claude

 

[Ratch]

Claude,

Sorry for the delay, but I was out of town for the last few days.

"I've been saying the same. I showed in a different thread the function of a switching power supply. When FET is off, catch diode is placed across the input supply. When FET is on, inductor in series with diode. In 1st case, V controls I in the diode. In 2nd case, I controls V. I've stated many times that either can be the independent quantity."

Still don't get it, do you? We are discussing whether a BJT device is a voltage or current controlled device. Why do you jump to a discussion of a FET circuit in an attempt to prove otherwise? Your example proves nothing with respect to determining what controls a BJT. So show that a BJT without any external energy components is a current controlled amplifier.

"But your argument does not show that voltage is independent either. I've said the same since day one. Only perturbing and observing can we see which is independent."

Of course it does. I explained in detail using physical principles why voltage is the controlling factor is determining junction diode current. To say other is disingenuous. Give me a good explanation, not an misinterpreted example, of why physics cannot determine what is the independent or dependent parameter of a device.

"Device behavior does not change because it is in a circuit. A bjt is exclusively used in a circuit, not bare. Any study of its bare properties has little relevance in practical real world applications."

That's right, but its behavior within a circuit can be obfuscated and misinterpreted as it is in the examples you have given thus far. You cannot take the behavior of a circuit and match that with the device within. We are not discussing what is real and practical. Instead we are discussing what controls a BJT.

"I've been saying ad infinitum that some type of source, CVS or CCS, is needed to move charges."

So does every textbook and lecture ever written or given.

"But Vbe/Vd are voltage drops, not emf's."

What does that have to do with what is controlling a BJT?

" A battery voltage is an emf. The current and voltage in the battery are oriented so that ions inside it are gaining energy. Hence the "12 volts" in a 12 volt battery is the energy of 12 joules GAINED by each coulomb of charge. The 0.10 volts across the wire is the energy per charge LOST. If there is 1 point needing emphasis it is that. THe energy storage/conversion elements provide emf, and the passive elements incur drops. In a diode-resistor-battery circuit, the battery provides all the energy."

Of course the battery or power supply supplies the energy to move the charge carriers in a circuit. How could it be otherwise? How does a discussion of the energy transfer between the power supply and battery prove your believe that a BJT device can be alternatively thought of as a current controlled device?

" Vd, the diodes forward drop is not moving charges."

I said many times that diffusion moves the charges in a BJT or junction diode. Haven't you read any of my posts?

"You insist that E fields and their associated voltage are what moves charges, but you are mistaken to think that a diode's own forward drop Vd is what moves the charges forming current Id."

Vd is the loss of charge energy density. You cannot have a charge movement across a resistance without incurring this loss. The energy to move the charges comes from the power supply. I never said that Vd is what moves the charges in a BJT or junction diode. Vd is a consequence of charge movement. But, what does that have to do with your proving your believe that a BJT is a current controlled device.?

Ratch