Emitter resistor with BJT, how does it provide feedback?

[matrixofdynamism]

Use of emitter resistor when using BJT in its various configurations seems to be important. Lets take example of BJT common emitter amplifier.

The emitter resistor is used supposedly to make the bias point more stable, this refers to the Q-point. But how? What would happen if this emitter resistor was not used?

 

[dick_freebird]

If you are driving the base with a stiff voltage source, the
emitter resistor provides direct local negative feedback of
the collector current by raising the emitter voltage and
reducing Vbe.

The resistor particularly stabilizes against self-heating,
which greatly affects Ib for a given Vbe and will cause
thermal current hogging / runaway in paralleled BJTs
as the hottest one will take all the base current. A
properly chosen Re will fight this substantially.

 

[Vbase]

We call it feedback but it isn't feedback.
The resistance of the base is the emitter resistor X current gain (hfe). The bigger the emitter resistor the higher the voltage you need to maintain the base current.
With 0 emitter resistor the base resistance is very low and it is harder to stabilize the base current.

 

[LvW]

We call it feedback but it isn't feedback.
.

Why not?
Of course, the emitter resistor Re provides negative feedback - it is the classical case of current-controlled voltage feedback.
And as a result - the voltage gain follows the classical feedback formula : G=-gmRc/(1+gmRe)with the loop gain (-gmRe).
(Rc: collector resistor, gm: transconductance).

Description of the feedback effect in words: A temperatur-caused unwanted increase of the collector current is limited (reduced) by a corresponding increase of the voltage drop Ve=Ie*Re which reduces the driving voltage Vbe=Vb-Ve. This feedback effect serves as a proof that the transistor works as a voltage controlled current source.

 

[Vbase]

Why not?

Feedback is feeding back to the input part of the output, this is not what that happens here.
About temperature; if the input current is (input voltage - Vbe) divided by the input resistance, then the higher the resistance the less the effect of Vbe changing in temperature.

"This feedback effect serves as a proof that the transistor works as a voltage controlled current source."
The transistor works as a voltage controlled current source even without the resistor.
Where is the proof that it's a feedback? Is it only "feedback effect"?

Your explanation is comfortable to you, and also another person who finds your explanation helpful. For you it is easier to understand math, even when you say: "Description of the feedback effect in words:" you use math mixed with words.
I'm more comfortable explaining things using Ohms Law.

 

[FvM]

The voltage drop over the resistor is substracted from the input voltage, reducing the transistor Vbe, in so far it's clearly feedback to the amplifier input.

The so called "emitter degeneration" resistor is described as feedback in many text books.

- - - Updated - - -

See e.g. Analysis and Design of Analog Integrated Circuits, Gray, Hurst, Lewis, Meyer, 4th edition, paragraph 8.6.1 Local series-series feedback

 

[LvW]

Feedback is feeding back to the input part of the output, this is not what that happens here.

The output of a BJT is the collector current Ic, which causes a voltage Ve=Re*Ie (remember: Ie=Ic+Ib) that acts as a feedback signal because it reduces the input signal.

About temperature; if the input current is (input voltage - Vbe) divided by the input resistance, then the higher the resistance the less the effect of Vbe changing in temperature.

This is not a correct description (input current divided by the input resistance?)

"This feedback effect serves as a proof that the transistor works as a voltage controlled current source."
The transistor works as a voltage controlled current source even without the resistor.
Where is the proof that it's a feedback? Is it only "feedback effect"?
.

Yes - the BJT works as VCCS even without Re. However, some people don`t believe this because they have learned years ago that the collector current would be controlled by Ib (and they stick to this false explanation).
Now - the effect of an emitter resistor Re upon the signal gain can serve as a PROOF that the BJT is voltage-controlled (because of the gain formula following the general feedback analysis from Black).
Everybody who is familiar with this formula and the definition of the loop gain can see that Re provides negative feedback. More than that, of course it is also possible to scetch a model for the feedback loop.

In summary: Negative current-controlled voltage feedback always reduces gain and increases the input resistance - in both cases by a factor equal to (1-loop gain). Exactly this effect can be observed in an Re-stabilized gain stage.

As another proof, it is well known that negative feedback reduces the phase margin (reduced dynamic stability). For this reason, all BJT based voltage follower (buffer, common collector stage) with heavy Re-feedback are sensitive to parasitic effects and tend to be oscillatory.

 

[Vbase]

The output of a BJT is the collector current Ic, which causes a voltage Ve=Re*Ie (remember: Ie=Ic+Ib) that acts as a feedback signal because it reduces the input signal.

LvW,
A transistor with Re works the way it wants to work and it doesn't care whether we call it feedback of give it another name.
The way you explain it is favorable by most people and I think you should be satisfied with that. It is not important that some old French farmer likes to explain it differently. The English say: you can't teach an old dog new tricks.

 

[LvW]

LvW,
A transistor with Re works the way it wants to work and it doesn't care whether we call it feedback of give it another name.
The way you explain it is favorable by most people and I think you should be satisfied with that. It is not important that some old French farmer likes to explain it differently. The English say: you can't teach an old dog new tricks.

Your answer was "...it isn`t feedback" (post#3) - and it was my only intention to correct this with the aim to provide the questioner with the correct answer. OK?

- - - Updated - - -

 

[crutschow]

I agree that an emitter resistor provides negative feedback.
To not call it feedback is using a too narrow definition of the word.

 

[Vbase]

Yeah, let's widen the definition a little more and call a potato feedback too!

 

[matrixofdynamism]

So inspite of it being such a great thing to have, I hear that emitter resistor reduces the amplifier gain. Because of this people like to put in a capacitor which grounds the emitter.

Why can't we choose such a Re value that we not need to use a capacitor and detrimental effect on gain is minimal while getting an acceptable level of stability in the Q-point from Re?

 

[Vbase]

Long ago we were taught about the common emitter, resistors at the collector and emitter and from the base to grd and from the base to Vcc. We had to calculate the resistors using E24 values for a transistor that had hfe ranging 100 to 400.
It was impossible to do it without the emitter resistor, this is why we need the emitter resistor.
Most teachers didn't know how to explain the need for the emitter resistor so they invented the explanation that it is for negative feedback despite the fact that feedback is only a by-product. Negative feedback sound sophisticated so we all adapted it.
None of us was clever enough then to ask the teacher the question that you ask now or ask the teacher why we put cap across it. Until today most people call it feedback resistor because they don't know that in real life with wide spread of transistor specs you cannot have common emitter amplifier without it.

 

[FvM]

Why can't we choose such a Re value that we not need to use a capacitor and detrimental effect on gain is minimal while getting an acceptable level of stability in the Q-point from Re?

That's a valid option and used in a number of designs. But more often the designer wants to set DC and AC gain differently.

Shorting the emitter resistor with a capacitor and achieving the maximal gain available in common emitter circuit is only a good idea if linearity and gain stability don't matter. But even at moderate output voltages, the exponential transistor characteristic takes effect and causes noticeable signal distortions. Reducing the AC gain by implementing sufficient f******k turns the circuit into an acceptable amplifier.

- - - Updated - - -

Until today most people call it feedback resistor because they don't know that in real life with wide spread of transistor specs you cannot have common emitter amplifier without it.

The sense of the words doesn't disclose itself to me, particularly the word "because". As mentioned before, you can set AC and DC gain differently by implementing different feedback factors. If you short the transistor with a capacitor, you cancel the AC feedback but still have DC feedback.

 

[LvW]

So inspite of it being such a great thing to have, I hear that emitter resistor reduces the amplifier gain. Because of this people like to put in a capacitor which grounds the emitter.
Why can't we choose such a Re value that we not need to use a capacitor and detrimental effect on gain is minimal while getting an acceptable level of stability in the Q-point from Re?

OK - let me try the following answer:
In electronics, every design is - more or less - a trade-off between conflicting effects.
That means: Improving one parameter/property (in our case: Stability against temperature changes and against parts tolerances) will at the same time worsen another parameter or property (in our case: Signal gain).
This leads to a capacitor Ce across the resistor Re - and the gain will rise again to the old/original value (without Re). However - this increased gain applies only to frequencies above the corner frequency of the highpass circuit formed by Ce and the associated resistances.
More than that, using Ce we loose another advantage of signal feedback: The gain is larger but its value depends more on BJT parameters and its large tolerances.
For this reason, the best trade-off between all these effects is (a) to use an emitter resistance Re=Re1+Re2 and (b) to short only a part of this resistor (Re2) with a shunt capacitor.
Thus, we have good/full dc feedback and - at the same time - moderate signal feedback.
As you can see: This design resembles a trade-off between different feedback options (and it does not help too much introducing "potatoes" into the design).

 

[Vbase]

f******k

Your efforts to avoid arguments is clever and amusing.

Do you think you can design a common emitter amplifier without emitter resistor using 5% resistors, BC237 with hfe 200 to 800 ?
If the answer is no then it explain why we need this resistor.

 

[LvW]

....
It was impossible to do it without the emitter resistor, this is why we need the emitter resistor.
Most teachers didn't know how to explain the need for the emitter resistor so they invented the explanation that it is for negative feedback despite the fact that feedback is only a by-product. .

... most teachers? Are you sure?
Each relevant textbook explains the role of Re correctly as "feedback". It is not a "by-product", in contrary - it is the desired design step. Negative feedback is the most important design method used for all kinds of amplifiers!

Vbase, perhaps the following example makes it clear to you:
What is the task of the resistive voltage divider placed between the output of an opamp and the inverting input?
Answer: It fulfills exactly the same task as the emitter resistor Re for a BJT stage: Negative feedback for stabilizing the operating point with the consequence of a reduced signal gain.

Of course, this gain reduction is desired because now - due to the feedback effect - the gain value is nearly determined by the external resistors only.
And now you can compare this effect again with the BJT stage and its gain: G=-gm*Rc/(1+gm*Re)=-Rc/(1/gm + Re)~-Rc/Re (for Re>>1/gm).

 

[Vbase]

The sense of the words doesn't disclose itself to me, particularly the word "because". As mentioned before, you can set AC and DC gain differently by implementing different feedback factors. If you short the transistor with a capacitor, you cancel the AC feedback but still have DC feedback.

Well, it's a shame that some of the words aren't clear but it isn't very important.
You can have different gain for ac and dc by other ways, like series resistor to the base with a cap in parallel, or a cap in parallel to collector resistor, or cap from collector to base.
What you cannot do is have common emitter amp without emitter resistor, the emitter resistor is not an option.

 

[FvM]

Do you think you can design a common emitter amplifier without emitter resistor using 5% resistors, BC237 with hfe 200 to 800 ?
If the answer is no then it explain why we need this resistor.

Of course I can, by using voltage controlled current (or "shunt-shunt") feedback. But if I prefer the discussed emitter resistor for DC stabilization or even consider it nessecary in some situation, does this change the feedback nature of this circuit means?

 

[Vbase]

... most teachers? Are you sure?
Each relevant textbook explains the role of Re correctly as "feedback". It is not a "by-product", in contrary - it is the desired design step. Negative feedback is the most important design method used for all kinds of amplifiers!

Vbase, perhaps the following example makes it clear to you:
What is the task of the resistive voltage divider placed between the output of an opamp and the inverting input?
Answer: It fulfills exactly the same task as the emitter resistor Re for a BJT stage: Negative feedback for stabilizing the operating point with the consequence of a reduced signal gain.

Of course, this gain reduction is desired because now - due to the feedback effect - the gain value is nearly determined by the external resistors only.
And now you can compare this effect again with the BJT stage and its gain: G=-gm*Rc/(1+gm*Re)=-Rc/(1/gm + Re)~-Rc/Re (for Re>>1/gm).

Teachers is including books that teach. So you answered it yourself.

Comparing transistors to op amps isn't that different to comparing transistors to potatoes.

Please read post #16 and answer if you can have a common emitter amp without emitter resistor.

- - - Updated - - -

Of course I can, by using voltage controlled current (or "shunt-shunt") feedback. But if I prefer the discussed emitter resistor for DC stabilization or even consider it nessecary in some situation, does this change the feedback nature of this circuit means?

The feedback is a by-product of the emitter resistor which must be there in our circuit of common emitter. You can utilize the feedback or cut it with a capacitor or do other things with it. The unfortunate thing is that most teachers misled pupils by telling them everything about the feedback but they forgot to tell them that the resistor isn't an option, and most pupils still think that that resistor is for feedback.
When I tell those pupils that the resistor is there because it has to be there they blame me for not knowing what the resistor is for.