BJT's don't work like that

Status
Not open for further replies.

beenthere

Joined Apr 20, 2004
15,819
circuitashes,



No, Vbe controls the charge carriers from the emitter. Ideally they should all pass through the base and continue on to the collector. Unfortunately some are inevitably caught by the base to become Ib. Ib current is not needed to control Ic, but it is proportional to Ic. Although this relationship is exploited to bias the transistor, saying it controls Ic it is like saying that the ammeter in a series circuit controls the current instead of the voltage source. Ratch
Have you really got an understanding of things? The voltage source has nothing to do with current, any more than does the ammeter. The nature of the path from the voltage source negative terminal to the positive has everything to do with the current.

The voltage has to be present to enable current, of course. But, assuming there is a circuit, is is the elements in the circuit that determine current. It is proportional to the voltage, and cannot exist without the potential difference to keep the electrons moving. Asserting that current is byproduct of voltage is correct, but the two are too intimately related to treat one in the absence of the other.

Of course there is no carrier flow in a PN junction in the absence of a voltage difference across it. once the potential has got large enough, the junction (if the voltage is properly applied) goes into conduction. We know that Vbe is essential to the operation of the BE junction in a BJT.

Where the problem we have appears is the assertion that Vbe is the controlling influence. That implies that management of Vbe can make the transistor operate in a completely predictable fashion.

The reality we all deal with is that once Vbe has put the transistor into conduction, it will be destroyed by current unless some element is present to limit - read control - that current. Some level of Vbe must be present for conduction in the junction, but that voltage does not control conduction - it enables it. Managing current through the BE junction is the only control for BJT operation.

By now you must be aware that these forums do not function as a condensed matter seminar. The questions and aid are much more practical. A knowledge of Vbe is quite useful - one may check for proper transistor operation by metering base to emitter and observing something on the order of .7 volts. At that point, though, Vbe has lost its usefulness in circuit operation. As long as it's maintained, it's all current control after that.

If you wish, we can simply agree to disagree on this point and drop this thread. Continuing it in a purely contrarian spirit serves little purpose. Our objections arise from the fact that we can demonstrate that our model functions in the real world, and we have no such demonstration of yours. That doesn't necessarily mean we win, but we lack evidence of the utility or application of your model of transistor operation.
 

studiot

Joined Nov 9, 2007
4,998
I'm still waiting for you to provide a relationship between Vbe and the frequency response of a transistor, that I can use to control said transistor.
 

Attachments

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
beenthere,

Have you really got an understanding of things? The voltage source has nothing to do with current, any more than does the ammeter. The nature of the path from the voltage source negative terminal to the positive has everything to do with the current.
Of course I do. Of course it does. The voltage source supplies the energy to pump the charges around the circuit. Moving charges are current, so the voltage is controlling the current.

The voltage has to be present to enable current, of course. But, assuming there is a circuit, is is the elements in the circuit that determine current. It is proportional to the voltage, and cannot exist without the potential difference to keep the electrons moving. Asserting that current is byproduct of voltage is correct, but the two are too intimately related to treat one in the absence of the other.
You are trying to pick apart what I gave as an analogy. I was trying to illustrate to Ron H that just because you can set up a parameter, it does not mean that the parameter is the controlling the device.

Of course there is no carrier flow in a PN junction in the absence of a voltage difference across it. once the potential has got large enough, the junction (if the voltage is properly applied) goes into conduction. We know that Vbe is essential to the operation of the BE junction in a BJT.
Essential and controlling.

Where the problem we have appears is the assertion that Vbe is the controlling influence. That implies that management of Vbe can make the transistor operate in a completely predictable fashion.
Yes, it does. However, it is difficult to do because of Ic's exponential response to Vbe. So circuits are designed and biased with current methods because of the proportionality of Ib with Ic. That does not negate the fact, however, that the physics of the BJT show Vbe is the control for Ic.

The reality we all deal with is that once Vbe has put the transistor into conduction, it will be destroyed by current unless some element is present to limit - read control - that current. Some level of Vbe must be present for conduction in the junction, but that voltage does not control conduction - it enables it. Managing current through the BE junction is the only control for BJT operation.
The fact that one can burn out a junction does not prove what you are saying. And by control, I do not mean limit. You are on the wrong side of the street to say that Vbe does not control Ic in an exponential manner, as Sedra and Smith show it does.

By now you must be aware that these forums do not function as a condensed matter seminar. The questions and aid are much more practical. A knowledge of Vbe is quite useful - one may check for proper transistor operation by metering base to emitter and observing something on the order of .7 volts. At that point, though, Vbe has lost its usefulness in circuit operation. As long as it's maintained, it's all current control after that.
Current manipulation for how to design circuits to make BJTs work. Vbe for understanding how BJTs work. Vbe never loses its functionality when the BJT is in its active region. It always controls the Ic with Vbe values between 0 and approximately 1 volt.

If you wish, we can simply agree to disagree on this point and drop this thread. Continuing it in a purely contrarian spirit serves little purpose. Our objections arise from the fact that we can demonstrate that our model functions in the real world, and we have no such demonstration of yours. That doesn't necessarily mean we win, but we lack evidence of the utility or application of your model of transistor operation.
I think that we are already disagreeing and probably won't convince each other. Were Galileo and Copernicus contrarians too? I cannot help it if my view is unpopular. You can demonstrate how to make BJTs work. How can I demonstate how they are working? Only by pointing out relationships and device behaviors others have discovered. If others reject this evidence or ignore it, I cannot do anything more. Ratch
 

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
studiot,

I'm still waiting for you to provide a relationship between Vbe and the frequency response of a transistor, that I can use to control said transistor.
What makes you think there is one? I have a hard time understanding your question. Could you elucidate why you think there is such a relationship? How does the relationship between Vbe and Ic relate to frequency response? I said many times previously that knowing Vbe controls Ic at the quantum level is for understanding, not necessary for designing practical circuits. Ratch
 

studiot

Joined Nov 9, 2007
4,998
Let us say I want to control the current supplied to the antenna of transmitter operating at 2400 Mhz. This would be a collector load to the output transistor, at this frequency.
What relationship should I employ to calculate what Vbe I would need to do this?
 

studiot

Joined Nov 9, 2007
4,998
Or perhaps we should try it another way.

Let us suppose we have a properly biased transistor set up to observe Vbe and Ic.
Let us further supply a voltage signal to the base from a suitably amplitude stable signal generator.
This constant amplitude alternating voltage varies Vbe by some amount \(\Delta\)V.
As you rightly say we can observe a corresponding change in Ic.
However this change in Ic will itself vary as we change the frequency of the signal generator, despite the fact that Vbe + \(\Delta\)V is the same at all frequencies.

This observational fact proves that there are components of Ic that do not depend upon Vbe, since we observe different effects on Ic at different frequencies.

I am simply asking for a voltage explanation of these facts since you contend that Ic depends solely upon Vbe and nothing else.

Engineers like myself are, of course, interested in an equation that allows me to calculate this effect as such theory is pretty esoteric without.

I am sure you are aware there is such an equation using base current.
 
Last edited:

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
studiot,

Or perhaps we should try it another way.

Let us suppose we have a properly biased transistor set up to observe Vbe and Ic.
Let us further supply a voltage signal to the base from a suitably amplitude stable signal generator.
This constant amplitude alternating voltage varies Vbe by some amount V.
As you rightly say we can observe a corresponding change in Ic.
However this change in Ic will itself vary as we change the frequency of the signal generator, despite the fact that Vbe + V is the same at all frequencies.

This observational fact proves that there are components of Ic that do not depend upon Vbe, since we observe different effects on Ic at different frequencies.
I think you are trying to say that the gain and other characteristics of a BJT change with frequency. That perhaps Vbe is not as effective at a higher frequency as it was at DC. So what? I am sure that FETs and OP AMPs, which you will no doubt agree are voltage controlled creatures, also have different high frequency characteristics.

I am simply asking for a voltage explanation of these facts since you contend that Ic depends solely upon Vbe and nothing else.
I said that the Ic of a BJT responds to the Vbe and that any change to Ib is going to change Vbe. The fact that temperature, junction capacitance, doping, and manufacturing methods can also influence Ic does not mean that those influences are considered to be controlling factors, i.e., nobody tries to control a BJT with temperature or junction capacitance. The voltage explanation was given because it controls the charges expelled from the emitter which is the primary control. You keep trying to say that because Ic responds to Vbe, Vbe should be used to design circuits to do what is wanted. Because Vbe is an exponential relationship to Ic, it is difficult to design a circuit that way. Instead Vbe should be used to understand what is going on. It is much better to use the proportional relationship of Vbe to Ic to make a BJT to do something, knowing that you are really controlling the BJT with Vbe.

I am sure you are aware there is such an equation using base current.
And does not the Ic also change with Ib at a different proportion depending on frequency? Ratch
 

beenthere

Joined Apr 20, 2004
15,819
studiot,



I think you are trying to say that the gain and other characteristics of a BJT change with frequency. That perhaps Vbe is not as effective at a higher frequency as it was at DC. So what? I am sure that FETs and OP AMPs, which you will no doubt agree are voltage controlled creatures, also have different high frequency characteristics.
So what? So what? Either Vbe controls, or it does not - that's what. You just agreed that your assertion that Vbe is the actual control of BJT conduction is not the case.

Do not try to prove your assertion about BJT's is still valid because a device based on different technology exhibits high frequency limitations. The discussion was not about FET's nor about IC's.
 

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
beenthere,

So what? So what? Either Vbe controls, or it does not - that's what. You just agreed that your assertion that Vbe is the actual control of BJT conduction is not the case.
Yes, Vbe controls Ic, but not as effectively at high frequencies due to other factors like junction capacitance.

Where did I agree that actual control of BJT conduction with Vbe is not the case?

Do not try to prove your assertion about BJT's is still valid because a device based on different technology exhibits high frequency limitations. The discussion was not about FET's nor about IC's.
Then studiot should not try to aver that Vbe is not controlling Ic at higher frequencies. I am not comparing the technologies, I am showing that every device changes with frequency. But that does not change the physics of its response. Ratch
 

studiot

Joined Nov 9, 2007
4,998
So let us take my example one stage further.

We have a properly biased transistor supplied with an alternating voltage signal.
Let us now modulate the frequency of this alternating signal with a desired second signal.
This will modulate Ic with the desired signal, without any change in Vbe.

Are you suggesting this is not true?

If so that is the same as saying that the NBFM 2.4Ghz DECT phone and Wifi system here at Studio T does not work and I am not typing this post via my router!

Vbe is constant for the final ouput transistor to the antenna and therefore plays no part in the signal transmission.
 
Last edited:

studiot

Joined Nov 9, 2007
4,998
In fact instead of smart-alec bickering we should all be celebrating the wondrous effect that transistor action brings to circuitry.
The voltage model you propose is not wrong it is incomplete.

The current model, conventionally adopted is not right it is also incomplete.

However when proposing a replacement model it should predict/explain/quantify a greater proportion of observable effects than that which it is to replace.

However you have steadfastly refused to expand your voltage model to include observable effects, easily explainable and calculable by the current model.

Which is a pity because this thread could have acted to expand the knowledge base at AAC.
 

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
studiot,

So let us take my example one stage further.

We have a properly biased transistor supplied with an alternating voltage signal.
Let us now modulate the frequency of this alternating signal with a desired second signal.
This will modulate Ic with the desired signal, without any change in Vbe.

Are you suggesting this is not true?

If so that is the same as saying that the NBFM 2.4Ghz DECT phone and Wifi system here at Studio T does not work and I am not typing this post via my router!

Vbe is constant for the final ouput transistor to the antenna and therefore plays no part in the signal transmission.
It will modulate both Vbe and Ib. Vbe will vary between 0 and approximately 1 volt in the active region and that will directly affect the Ic current in an exponential manner. Ratch
 

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
studiot,

The voltage model you propose is not wrong it is incomplete.
It is not a model. It is an insight of how a how and why a transistor works.

The current model, conventionally adopted is not right it is also incomplete.
The various transistor models used work just fine for determining what a transistor does and how it interacts with other elements within a circuit. They do not provide much insight of why transistors do what they do.

However when proposing a replacement model it should predict/explain/quantify a greater proportion of observable effects than that which it is to replace.
Who said anything about proposing a replacement model?

However you have steadfastly refused to expand your voltage model to include observable effects, easily explainable and calculable by the current model.
Because I am not proposing a new model. I am just explaining why a BJT works the way it does.

Which is a pity because this thread could have acted to expand the knowledge base at AAC.
It did, it does. Ratch
 

studiot

Joined Nov 9, 2007
4,998
Vbe will vary between 0 and approximately 1 volt
Why on earth do you say this?

I specified that Vbe +\(\Delta\)V remains constant throughout.

That is the transistor sees the following voltages at its base

Vbe the DC bias, a fixed direct voltage

+\(\Delta\)V the signal voltage from the signal generator as in post #106, a fixed amplitude alternating voltage of variable frequency.

This voltage is set by the signal generator and made up of the generator signal plus the modulating signal, fed into the modulating input of the generator. This is similar in principal to post #106 except that instead of turning the frequency knob I am now feeding in a modulating signal to the generator to send a wanted signal to the transistor base. As before this will vary Ic without any change to the voltage at the base. My modulating signal could be a ramp which sweeps the input frequency through the entire range of the transistor response spectrum from Ft to 0. This will sweep Ic from

Icstatic + 1xIb (as given by the DC bias) at Ft when the current gain is 1

Through

Icstatic + \(\beta\)max x Ib at some frequency

to

Icstatic + \(\beta\)min x Ib at low frequency


This can be a very considerable range of control, and all with Vbe +\(\Delta\)V unaltered.

There are many good quality signal generators on the market that can do this as I have specified. The only parameter that varies in the input is the frequency. I cannot stress this too strongly.

I expect you didn't read my post properly, advice given to students every year at this (exam) time - Read the paper properly and answer the question they ask not a different one they haven't asked.
 
Last edited:

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
studiot,

Quote from Ratch
Vbe will vary between 0 and approximately 1 volt

Answer from studiot
Why on earth do you say this?
Because that is the range of Vbe in the active region.

I specified that Vbe +deltaV remains constant throughout.
If that term contains deltaV then it cannot be constant. Perhaps you should have said the voltage range is constant.

That is the transistor sees the following voltages at its base

Vbe the DC bias, a fixed direct voltage

+deltaV the signal voltage from the signal generator as in post #106, a fixed amplitude alternating voltage of variable frequency.
An voltage with a fixed AC amplitude and a constant DC component.

This voltage is set by the signal generator and made up of the generator signal plus the modulating signal, fed into the modulating input of the generator. This is similar in principal to post #106 except that instead of turning the frequency knob I am now feeding in a modulating signal to the generator to send a wanted signal to the transistor base. As before this will vary Ic without any change to the voltage at the base. My modulating signal could be a ramp which sweeps the input frequency through the entire range of the transistor response spectrum from Ft to 0. This will sweep Ic from

Icstatic + 1xIb (as given by the DC bias) at Ft when the current gain is 1

Through

Icstatic + Betamax x Ib at some frequency

to

Icstatic + Betamin x Ib at low frequency


This can be a very considerable range of control, and all with Vbe +deltaV unaltered.

There are many good quality signal generators on the market that can do this as I have specified. The only parameter that varies in the input is the frequency. I cannot stress this too strongly.
All you are showing is that the BJT has a frequency response. You can expect something to change at the higher frequency. Obviously Vbe's effectiveness deteriorates at higher frequencies, even if its voltage range is kept constant. That is to be expected. The lower beta at higher frequency shows that more Ib current is present and its proportionality to Ic is changing. Why are you demonstrating this? I agree with you that things change with frequency.

I expect you didn't read my post properly, advice given to students every year at this (exam) time - Read the paper properly and answer the question they ask not a different one they haven't asked.
I think you did not explain what you mean properly. Ratch
 

studiot

Joined Nov 9, 2007
4,998
I agree with you that things change with frequency.
Go to the top of the class!!

You have just confirmed the first point I made in my first post in this thread.

I said that other parameters can also vary response, besides the base emitter voltage and independently of it.

Hold the input frequency constant.
At a certain Vbe you get a certain Ic.
Change the Vbe and you get a different Ic.

Great and agreed from the start.

Hold voltage conditions at the base emitter constant
At a certain frequency you get a certain Ic
Change the frequency you get a different Ic

So what's the difference?

And why can't your model provide calculations to establish the change in Ic for the two situations?

Cut the cr__p and answer a straight question with a straight answer.
 

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
studiot,

Hold the input frequency constant.
At a certain Vbe you get a certain Ic.
Change the Vbe and you get a different Ic.

Great and agreed from the start.

Hold voltage conditions at the base emitter constant
At a certain frequency you get a certain Ic
Change the frequency you get a different Ic

So what's the difference?
The difference is the parameter you use to vary the output. No one uses frequency degradation to control the Ic of a BJT. The debate was about whether a BJT is voltage or current controlled, not frequency controlled.

And why can't your model provide calculations to establish the change in Ic for the two situations?

Cut the cr__p and answer a straight question with a straight answer.
As I said previously, I am not proposing a new model. I am just explaining why a BJT works the way it does. That is a straight answer. Ratch
 

Ron H

Joined Apr 14, 2005
7,063
I still think that the basic disagreement here revolves around semantic differences regarding the meaning of "control".
BTW, I believe that the Gummel-Poon model deals with the high frequency limitations of the Ebers and Moll model. From what I have read, the Gummel-Poon model, which is used in Spice simulations, simplifies to the Ebers-Moll model when you eliminate some parameters.
 

Thread Starter

Ratch

Joined Mar 20, 2007
1,070
Ron H

I still think that the basic disagreement here revolves around semantic differences regarding the meaning of "control".
I tried to clarify it by using the word "responding" sometimes. I also stated clearly that Vbe is the cause of Ic and Ib is the consequence of Vbe. So I am either right or wrong, but not misunderstood by semantic differences. Ratch
 
Status
Not open for further replies.
Top