When we come up with conflicting theories for explaining a phenomenon, generally the "better" one wins ("wins" meaning "is accepted as the physical truth"). What I am saying is in this case neither theory is better, so there can be no one winner among them. With this in mind, when you claim "It is NOT the question if one or another principle is "better" or not." you are simply splitting hairs. The truth is as I said, neither principles are better, so neither (or both) can be physically correct as well.Winterstone said:It is NOT the question if one or another principle is "better" or not. The question is simply: What is the physical truth? And there can be only one answer.
Of course I do, this is exactly why I say the two control methods simplify to the same thing. You cannot have a current flow through a load without voltage across it. Conversely, a load can not have a voltage across it without having a current going through it.Winterstone said:[Do] you think that a current can exist only if there is a driving voltage?
When we come up with conflicting theories for explaining a phenomenon, generally the "better" one wins ("wins" meaning "is accepted as the physical truth"). What I am saying is in this case neither theory is better, so there can be no one winner among them. With this in mind, when you claim "It is NOT the question if one or another principle is "better" or not." you are simply splitting hairs. The truth is as I said, neither principles are better, so neither (or both) can be physically correct as well.
So then, how do BJT's physically work?
As far as I know, (I could easily be slightly off), physically the volume{current} of minority charge carriers (which each have a potential{voltage}) moving through the E-B junction, induce an electric field{voltage} on the C-B junction.
It isn't rocket science, this is basic Ohm's law we are talking about.
Ahh - I see. I am "splitting hairs" because I don`t follow your claim that neither (or both) principles can be physically correct (by the way: any proof for that?). Please, be calm.
Current can "induce an electric field" ? Can you give a corresponding reference (please, not wikipedia) ?
And Ohm`s law leads to an exponential expression (Shockley)?
I really don`t know how to interpret your contribution.
W.
Do you consider this really as a fair and objective argument? I am afraid that such a crude logic cannot convince anybody. What creates the field (primarily)?Of course current can produce an E field. An uncharged cap has no E field. A current into the cap will result in charge polarization and an E field.
You keep denying what people present and demand proof, yet you provide no proof except your dogma that V causes I.
Equations like Shockley and Ohm only relate V to I, they do not imply which is in control of the other.
audioguru said:I think with a higher hFE the voltage gain increased because the transistor was biased away from cutoff.
When near cutoff (low hFE) the exponential gm causes the gain to be reduced. With a high hFE then the transistor is operating away from cutoff and is more linear with a higher gain.
I show the transistor without feedback and with slightly different biasing causing it to be near cutoff and away from cutoff.
No... you are splitting hairs between "better/best theory" and "physically true".Winterstone said:Ahh - I see. I am "splitting hairs" because I don`t follow your claim that neither (or both) principles can be physically correct
I'm not uncalm.Winterstone said:Please, be calm.
There is no need for a reference, it's common sense enough that it should be explainable.Winterstone said:Current can "induce an electric field" ? Can you give a corresponding reference
It doesn't have to be Ohm's law, use Shockley/diode law if it makes you happy, this changes nothing about my argument.Winterstone said:And Ohm`s law leads to an exponential expression (Shockley)?
Hi.
I see a circuit, and I see an analysis, but what is it you're trying to claim exactly? Higher hFE should not increase gain, perhaps?
No... you are splitting hairs between "better/best theory" and "physically true".
We can never really know what physically true is as we can only rely on the 6 limited human senses and our imagination. And in the scientific world, proof from unfalsifiability is not proof at all. We can only ever have the best of the most reasonable theories, which is what we are discussing.
I'm not uncalm.
There is no need for a reference, it's common sense enough that it should be explainable.
It's as Claude said, capacitors, for one, are examples of current that has accumulated an electric field. The earths magnetosphere, for two. Current through a resistor, for three... I could cite hundreds of examples honestly. I can also give just as many examples for an electric field creating a current. You are apparently forgetting that any electric system is composed of charged particles, that when in motion, ARE ALSO and electric current.
Charges in motion ARE an electric current.
Moving charges to different locations IS an electric field.
For the last time, these things are peas in a pod, you can't have one without having(or having had) the other.
It doesn't have to be Ohm's law, use Shockley/diode law if it makes you happy, this changes nothing about my argument.
It's exactly as Claude has just said, "Equations like Shockley and Ohm only relate V to I, they do not imply which is in control of the other." I only take this a step farther and say that point of fact, these laws really show that they don't just relate to each other, but are always coexistent. Therefore neither can be in control of the other, and to argue otherwise is just a different kind of "chicken or the egg" argument. Who creates a marriage? The man? or the woman?...... Neither/both, it takes both to make the marriage in the first place, and either could break it.
You can't have Ohm/Shockley law without some kind of marriage between volts and current. It's that simple.
We have a 2N22222A transistor and 1k collector resistor (Rc) and a base emitter voltage (Vbe) of 0.65 volts. This is powered by a 10v supply source Vs. What is the collector current Ic ?
What Eric did was he found spec's on the data sheet that specify the collector current with Vbe.
Interesting question arise. For example if we use a Shockley equation (Ic=Is*(e^(Vbe/VT)-1)) to find ( determined) the collector current for a constant Vbe = 0.65V the collector current should also be constant? Or maybe I missing something in Shockley equation?
OK so for RB = 518KΩ and Vbe = 0.65V and I assume Is = 10x10^-15 and VT = 26mV and β = 50.
Also, that formula was in the post that Ratchit linked to in the other thread and he and the original poster seem to think it's the greatest formula.
No Vbe is not a constant. The Vbe varies with Ib current.
It's unfortunate that Ratchit is not here to reply because he has always shown himself to be a strong advocate of the voltage control method.
We have a two equation that we can use to determine the Ic. First equations is
Ic = Hfe *Ib and the second one is Ic = Is*(e^(Vbe/Vt) - 1).
And I think that MrAl want to compare this two equation from engineering point of view.
Which of this two equations gives "better" result in determining Ic current.
And our task is to find Ib and Ic current. Next we build this circuit and we check our calculation on the bench.
We need to add 3) Alpha (current control). Ic = alpha*Ie, is the best control method ever devised. Alpha is very predictable and consistent, unlike beta.
It seems to me everyone opposing MrAl's view just wants to focus on why the circuit sucks, and not ways we can analyze why the circuit sucks.
The transistor alpha certainly is a real parameter of a transistor, to be sure. And it is very "progressive" or "out of the box" thinking to try and use it for something meaningful. But as far as I know, transistor alpha only plays an important role with common-base configurations, and can almost always be considered unity. It's not really a gain factor, as much as it is attenuation.
But other than that, alpha is rather useless for analyzing a transistor from what I know.
We have bjt parts w/ beta=5,000. If Ib is useless undesirable side-effect, why does the semicon OEM make parts w/ beta=50, 100, 250, etc. when they already know how to get beta=5,000. If Ib is useless every bjt would be designed as superbeta.
Reducing Ib incurs other performance tradeoffs that are downright ugly. BR.
Perhaps Ratchit can add to this thread also as he likes the voltage control method most.
I've attached a 4 page calculation sheet I just generated for a generic bjt single stage amp. This was done to demonstrate the interactive nature of bjt parameters. The relations between Ic, hfe, gm, etc., is detailed in this analysis. Feel free to comment or ask for clarity. Best regards.
As far as I know, (I could easily be slightly off), physically the volume{current} of minority charge carriers (which each have a potential{voltage}) moving through the E-B junction, induce an electric field{voltage} on the C-B junction. This induced field weakens what should be an impassable C-B depletion layer. This weakening allows majority charge carriers to pass from the emitter, past the base, into the collector... even though there is a reverse biased junction in that path.
Again...
It is the volume{current} of carriers, combined with the resistance to this carrier flow, that creates the accumulation of potential{voltage}. Thus the internal workings of the bjt transistor function on both voltage and current at the same time, as I have already said.
Conversely, a load can not have a voltage across it without having a current going through it.
Equations like Shockley and Ohm only relate V to I, they do not imply which is in control of the other.
The transient behavior of p-n junctions is proof positive that V does NOT control I. I changes w/o V changing, then V catches up. If V controls I the opposite would be observed. Can't refute that.
We can never really know what physically true is as we can only rely on the 5 limited human senses and our imagination.
Therefore neither can be in control of the other, and to argue otherwise is just a different kind of "chicken or the egg" argument.
You can't have Ohm/Shockley law without some kind of marriage between volts and current.
No he didn't. He found the point on the β vs Ic curve that matched the product of β and the Ib he calculated. There is no Vbe vs Ic anywhere on the data sheets.
Eric,
I agree, but not the 2N2222.
Ratch
They are for "typical" devices. But there is a wide range of Vbe that is listed in the text of the datasheet since each transistor is different.Many transistor datasheets do show Vbe versus Ic.
They are for "typical" devices. But there is a wide range of Vbe that is listed in the text of the datasheet since each transistor is different.
The datasheet for the MPS2222 shows that Vbe is a minimum of 0.6V and a maximum of 1.2V.
You cannot buy a "typical" transistor, you get whatever they have.
Eric,
I did a reasonable search on the net for a Vbe curve of the 2N2222, but not an exhaustive one. Anyhow, Is is easy to calculate by taking two simple easy measurements.
Ratch
To the Ineffable All,
I have been reading this thread for a couple of days now, and I think the top of my head is going to blow off.
I won't be able to comment on everything, so I will just get the highlights.
MrAl,
You have to find the saturation current Is, and it is not in the data sheets. However, you can apply a voltage at the base terminal, say 0.2 volts, and read the current of the collector. Then apply the formula iC = Is*exp(VBE/VT) and get a value for Is.
No he didn't. He found the point on the β vs Ic curve that matched the product of β and the Ib he calculated. There is no Vbe vs Ic anywhere on the data sheets.
Jony130,
No, you are correct. The trick is to keep Vbe constant and compensate for temperature changes.
Is can vary between 10^-15 to 10^-12 amps. That is a 1 to 1000 range. How can you assume an Is unless you measure it like I outlined above?
MrAl,
I assume you are referring to Shockley's formula. It is a good one.
Jony130,
And also with Ic.
MrAl,
I have returned, and will expound on the method later.
Jony130,
Both equations are valid and give equally good results.
Better get started.
Claude Abraham,
Alpha and beta have a one to one relationship, i.e. and alpha can be converted to a unique beta value and vice versa.
Oblivion,
I will have something to say about this later.
Alpha in known in electronics as the common base current gain. Beta is the common emitter current gain.
Its concept is useful.
Claude Abraham,
I think you answered your own statement. If they could eliminate or greatly reduce Ib without bad side effects, that would be great.
MrAl,
Indeed I can.
Claude Abraham,
I will look it over when I have more time.
Oblivion,
Back to the textbooks for you. A BJT is a diffusion driven device. That is why its current is exponential with respect to voltage. The highly concentrated charge carriers in the emitter diffuse into the base and are swept away into the collector by the reverse bias voltage. The mobile charge carriers leave behind ions whose charges cause a voltage that opposes further diffusion. The base voltage cancels this voltage and keeps the diffusion current going. So the movement of charge in the emitter-base region is not directly from the voltage, but from diffusion.
An insulator can have a voltage without any current existing through it. A fully energized capacitor can have a voltage across it with any current existing in the the cap's branch.
Claude Abraham,
I agree with you on that point. The physics of the device determine what controls what.
No it isn't. Yes, I can. A phase difference is not a proof that one entity does or does not control the other. The physics of the device determines that. A phase difference means that some energy is being stored or released. It has nothing to do with what parameter controls the device.
Oblivion,
Sure we can. Many scientific facts and mathematical relations can be proven beyond a shadow of a doubt.
No it isn't. The physics of the system determine what controls what.
One of these days, I will have to show you what Ohm's law really is.
------------------------------------------------------------------------------------------------
My turn:
There has a lot of talk about biasing, but very little or nothing about stabilization. Biasing sets the Q point and stabilization keeps it there. Winfield Hill's method is a design strategy. It is not an analysis tool, proof of what controls a device, or proof that beta does not influence a BJT circuit. Also, it does not mean that the signal applied to the BJT is going to be directly applied to Vbe.
From what I read in this thread, I don't think too many of you really understand how it works. Look at what he said, "...a voltage-control design will assume a horrible low beta and a high Ib, and choose low values for R1 and R2..." The low values of Rb and high values of bias current will swamp out any changes caused by beta variations. It's not that beta doesn't have an effect, it is just that the effect is very small. So now he has the base locked down at 1.3 volts and that is not going to change no matter what beta is. The emitter current of Q1 is also locked at (1.3-0.7)/Re. That is not going to change much either. Continuing further, we have the signal voltage across 75 ohms of the Q2 emitter. So now re is stable at 1 ma and Re isn't going to change. So the AC gain will be a solid RL/(Re+re). Plus he lists other benefits of this design. You can read in textbooks that a small value of Rb improves beta variations, and it will also help prevent runaway current from Icbo. Icbo, also called Ico, is a permanent thermal current generator that starts up whenever the collector-base is reverse biased. It cannot be turned off, but if it crosses into the base-emitter junction, its effect becomes "betatized". That is why MrAL's first curcuit invites thermal runaway. Ico doubles for every 6°C for silicon transistors. Therefore, a designer will want to keep the emitter resistance high and the base resistor low so that a good part of Ico will harmlessly shunt out to the base circuit.
Another thing I wanted to say is that models are good for determining what a device does, but it does not show what parameter is in control.
Ratch
The Vbe curves are found on the data sheet as Eric pointed out. When i said that is what Eric did i meant that he found the curve on the data sheet and used that to determine Ic.
In your replies you imply that the Shockley equation is a good one. You said:
"You have to find the saturation current Is, and it is not in the data sheets. However, you can apply a voltage at the base terminal, say 0.2 volts, and read the current of the collector. Then apply the formula iC = Is*exp(VBE/VT) and get a value for Is."
However that equation is for an ideal diode. But that's not the real point i wanted to get to. I wanted to see this equation work with ANY values that anyone cares to dig up. The clincher however is that i want to see it work with at least two different levels of Vbe, where in both cases we do have Ic of at least some significant value like 1ma or such. But if we have Ic1=1ma then the other test Vbe should generate Ic2=10ma. So that's probably a good starting point anyway. To show how this equation works when we have Ic1=1ma and Ic2=10ma. Ic1 is the first test where we have Vbe1, and of course Ic2 is the second test where we have Vbe2 a different level of base emitter voltage. I picked the 2N2222A because it is a common transistor that can be obtained everywhere and there are good data sheets out there. So we could read the data sheets and if anyone wants to test this in real life they can do so too. So it should be an interesting thing to do i think.
So part of the story is if Shockleys equation is so good then we should be able to find a way to use it. And since it speaks for the physical transistor it should work for any circuit not just some. If it does not work for some then we should be able to find out why it does not.
The spice model does give a value for Is of 500e-15 if that helps.
So the main point is that we should be able to find a way to actually use this. So far we have only talked about it and stated this and that and how good it is, but if we can use it then it will do us some good.
For the 2N2222? I looked at four or five sheets on the net and did not see any. I guess I am blind in one eye and can't see out the other. Perhaps you can provide a link where it is.
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?