Discussion in 'General Electronics Discussion' started by BlackMelon, Apr 17, 2016.

1. ### Ratch

1,094
334
Mar 10, 2013
I believe this argument in the past has reached a conclusion. Describing the operation of a BJT with respect to its physics stops the chicken and egg argument dead in its tracks.

Just because a BJT is a voltage controlled current source (transconductance amplifier) does not mean that circuit design should be based on manipulating the small, highly nonlinear Vbe. Why do folks deduce that? In fact, the strategy appears to be to swamp out the Vbe with lots of resistance in the base and emitter circuits (current driven). This assures that the higher bias voltage needed to overcome this resistance will minimize the effect of Vbe. That is well and good, but folks should realize that Vbe is really what is really controlling Ic, and not aver that Ib controls Ic.

Ratch

2. ### Ratch

1,094
334
Mar 10, 2013
R = V/I and its variants is not Ohm's law. It is the definition of resistance. Ohm's law is concerned with the linearity of the resistance. Components like junction diodes do not follow Ohm's law. Other components like most resistors do (within a wide range of values). So Ohm's law is a property of a material, not a calculation of its resistance. Most good physics textbooks explain this.

You would not say "The wind is blowing", or "The stream is flowing", would you. There would not be a wind unless the air is moving, and there would not be a flow unless the water is moving. So why say "current flow", which is the same as saying charge flow flow. Say instead "charge flow", "current is present", or "current exists".

Ratch

3. ### LvW

604
146
Apr 12, 2014
Yes - that`s true.
On the other hand, I think the real working principle of a BJT should be clear as it was discussed also in this forum already some time ago (although some persons are still denying physical facts). For participating in THIS discussion, my only intention is to reveal the surprising (or even "funny") contradiction as described in one of my former answers (post#12):
While discussing the methods for designing a simple BJT-based amplifier stage it was claimed that the application of a "current control view" would be simpler and easier. Hence, "curent-control defenders" believe to apply current-control, but in reality they don`t. They apply exactly the same rules and equations as everybody.
I am really curious to learn in which design step and how the "current-control" feature is applied.
However, up to now I didn`t get any answer - just a statement like Ic=B*Ib, which is not very convincing.

4. ### Ratch

1,094
334
Mar 10, 2013
Everyone should read this link. http://cr4.globalspec.com/thread/68055/voltage-vs-current Especially posts #9 and #10. There is a lot of good info in that thread. By the way, Winfield Hill is the author of "The Art of Electronics". Observe the way he designs circuits. Interesting.

Ratch

5. ### duke37

5,364
772
Jan 9, 2011
[QUOTE="
This is, certainly, not a chicken-egg question.
The voltage is always first; no current without driving voltage.[/QUOTE]
And no voltage without a current.

6. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,497
2,838
Jan 21, 2010
excellent. Then you agree that Vbe does not control Ic

7. ### Ratch

1,094
334
Mar 10, 2013
And no voltage without a current.[/QUOTE]

Yes, current and voltage go together. But that does not explain why a BJT, FET, and vacuum tubes are voltage controlled devices and a magnetic amplifier is a current controlled device. Unless you know the operational physics of these devices, you cannot tell what controls them even if you can do a rubber stamp design with them.

Ratch

8. ### Ratch

1,094
334
Mar 10, 2013
How do you conclude that. Without Vbe, there would not be any control voltage at all across the emitter-base junction. Just because part of the Vbe voltage is lost across the bulk resistance, does not mean that Vbe is completely ineffective.

Ratch

9. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,497
2,838
Jan 21, 2010
My apology, I misquoted LvW.

And the wind certainly is blowing hard right now. And the stream formed by the runoff from the rain is flowing rapidly.

I would hazard a guess that the rain is making the current in our river flow significantly faster. Indeed, I expect the current is flowing fast enough that the white water enthusiasts will probably take a sickie tomorrow so as to take advantage of it.

10. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,497
2,838
Jan 21, 2010
Because you said it.

steve: "So you're saying that some part of the Vbe is not controlling Ic?"

ratch: "Correct"

11. ### Ratch

1,094
334
Mar 10, 2013
The above two sentences contain redundant words. They should say something like "There is a (high, strong, fast) wind here now. And "The stream from the rain runoff is rapid".

I would hazard a guess that the rain is making the current in our river flow significantly faster. Indeed, I expect the current is flowing fast enough that the white water enthusiasts will probably take a sickie tomorrow so as to take advantage of it.[/QUOTE]

Remove the redundant words "flow" and "flowing" in the above sentences.

Ratch

12. ### Ratch

1,094
334
Mar 10, 2013
False conclusion. Just because part of Vbe does not control the emitter-base junction does not mean all of Vbe is ineffective in controlling the Ic .

Ratch

13. ### BlackMelon

188
5
Aug 7, 2012
Guys, I appreciate your answers very much. I asked you this because , in the first place, I was thinking about a circuit capable of checking my current sensor whether it was broken or ok. I came up with an 555 timer, driving a base of a darlington pair. And pair is used to drive a 12V lamp. The sensor monitors current through the lamp, so I'll be able to check the sensor's output signal. (Seems that I've to be careful about the phrase 'current flowing' from this point on right? ) I asked you all this because I needed to do some calculation stuff about the darlington pair.

Your replies are very beneficial to me.

Thank you.
BlackMelon

14. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,497
2,838
Jan 21, 2010
Because whenever a practical alternative is suggested some people jump on it and claim it just ain't so.

just as most of us accept that there is a relationship between voltage and current outside a base-emitter junction which allows us to use one as a predictor of the other (with various constraints depending on the nature of that relationship, most of us also accept that in many circuits the easily calculated Ib can be used in place of a very non linear, difficult to measure Vbe.

Sure, there are circuits where understanding that Vbe is the controlling factor is the only way to predict circuit operation, but that predictive capability tends to be qualitative rather than quantitative.

As an example, consider a current mirror. Without an understanding of the effect of Bbr the operation cannot be deduced.

However, with knowledge of the effect of Vbe we can predict the qualitative aspects of the operation assuming matched devices in close thermal contact. What it can't do is predict the actual Vbe, or at least in any way that is practically useful.

duke37 likes this.
15. ### Ratch

1,094
334
Mar 10, 2013
Do they give good reasons why?

Who is suggesting that you measure and use the variation of Vbe?

Perhaps qualitative knowledge is all that is needed.

What is Bbr?

If two transistors are made together from the same substrate, then it can be assumed that they are going to be matched and in close thermal contact with each other.

Ratch

16. ### BobK

7,682
1,688
Jan 5, 2010
Once again all of the Vbe fans miss my point. Ib and Ic are related by a simple linear formula. I don't care whether one controls the other. It makes no difference when I design a circuit. In a small signal amplifier circuit, Vbe barely changes while Ib and Ic change over a wider range, due to the exponential relationship of Ib and Ic to Vbe. Thus it is possible to estimate Vbe to be 0.6V and have everything work out close enough for government work.

Your analogy is of controlling speed by looking at the speedometer is perfectly apt, and works against your argument. I don't set the speed of my car by looking at the position of the accelerator pedal, the wind direction and speed, the slope of the road, the friction between the road and the tires, the road conditions etc, and neither do you. We both set the speed of our cars by looking at the speedometer.

And I set Ic by looking at Ib.

Bob

17. ### Ratch

1,094
334
Mar 10, 2013
What is/was your point? I looked back at your previous post and could not find any explanation of what it is.

If that formula includes beta, then the relationship is going to be quite variable, because beta has a wide variance between transistors unless you are willing to pay handsomely to sort and match

One of the questions this thread discussed was whether the collector current was controlled by Vbe or Ib. Your indifference to this factoid is irrelevant to the discussion.

It is the difference between knowing how to drive and knowing how a car works.

How do you figure that?

Which proves my point that those above things you mentioned do not control the car. By the same viewpoint, Ib does not control Ic.

Then Ic will vary due to the vagaries of beta. Better to set Ic by applying a set voltage across the emitter resistor. When you do, be sure to add about 0.7 volts for the base-emitter drop. I believe that controlling the Ic via voltage across Re was what LvW was talking about. If you control Ic this way, then beta can vary all over the map, and Ic will not change.

Ratch

18. ### Laplace

1,252
184
Apr 4, 2010
The part I don't understand is how Vbe controls Ic when circuit analysis technique is to assume Vbe is a constant voltage around 0.7 volts. In reality there is a highly non-linear temperature-dependent relationship between Vbe and Ic that can best be summarized as "Vbe is a constant voltage around 0.7 volts."

In small-signal analysis, the transistor is usually modeled as a transconductance amplifier where the output current is linearly proportional to vbe - the portion of signal voltage appearing between the B and E terminals of the transistor model (not Vbe, the voltage between the B and E terminals of the physical transistor).

So what is actually behind this inappropriate fascination with Vbe?

19. ### LvW

604
146
Apr 12, 2014
You do not care?
Bob - unfortunately, you didn`t respond to my questions to you (see my post#12).
May I repeat one of my question (see my post#12):
* How do you choose the resistors for the resistive voltage divider at the base?
* Don`t you follow the rule "as low as possible with respect to other constraints (input impedance, power consumption)" because you want to produce a VOLTAGE at the base node?
(Therefore, I think you "care" about it - perhaps without realizing this fact?)

* Don`t you never use the relation gm=Ic/Vth which describes the transconductance properties of BJTs ?
* While "looking" at Ib, don`t you look at Vbe at the same time?
Remember: The equation for the corresponding resistor RB (between collector and base) is RB=(Vc-Vbe)/Ib. Perhaps you simply "forget" the real meaning of Vbe because it is used with a - more or less - fixed value? But this fact does not degrade the importance of Vbe!

Last edited: Apr 19, 2016
20. ### LvW

604
146
Apr 12, 2014
Because we do not know the exact Vbe value for the desired current Ic we are applying negative feedback. One of the many advantages of negative feedback is a sensitivity reduction of circuit parameters upon (uncertain) parameters of the active device.
This can be visualized looking at the following equations for the two resistors forming the DC voltage divider at the base node (valid for Re-feedback in the emitter path):
R1=(Vcc-Vb)/(I1+Ib);
R2=Vb/I1 with
Vb=(Vbe+Ie*Re).
As you can see, the voltage Vb at the base node is less sensitive to the (uncertain) voltage Vbe because of the feedback voltage (voltage !!) Ie*Re.
Inappropriate?
May I remind you at the most important property of a gain stage: The VOLTAGE GAIN ?
For each gain stage (common emitter C-E, C-B or C-C) the gain expression contains the transconductance gm=d(Ic)/d(Vbe), which is the slope of the Ic=f(Vbe) function.
And this function is independent on the beta value (unfortunately called "current gain", which is a misnomer). THAT is the "fascination" of Vbe: It determines Ic and the Ic variations and, thus, the voltage gain.

Again: The voltage gain of a stage does NOT depend on the beta value! Hence, I think we should rather ask: "What is actually behind this inappropriate fascination with Ib ?"

Quote Barrie Gilbert (Analog Devices):
The base current is purely incidental (it is best viewed as a „defect“)

Last edited: Apr 19, 2016