How does Emitter bypass capacitor not affect the amplifier's working?

[Ratchit]

To the Ineffable All,

Could any of you comment on the voltage bias method promoted by Winfield Hill in the link I posted in post #6 of this thread?

Ratch

 

[Winterstone]

When the cap comes into play, we might design this stage such that the cap reactance looks somewhat like a resistor at a particular max design target frequency. If we wanted a gain of 10 at the max frequency, then we would have made the cap value appropriate to make the total resistance look like 1000 ohms. So now we have two more gains to look at:
1. 10000/1000 (my approximation)
2. 10000/1010 (your approximation)

So my approximation yields a gain of 10 and yours a gain of 9.9, not a heck of a lot of difference there.

MrAl,

in reply to your very long contribution I like to concentrate on the technical part only - and I have to apologize for the following sentence:
Are you joking? The originator of this thread must be confused about your "example". Thus, I feel the necessity to clarify something.

*To make it clear: We speak about the role of Re and a shunting capacitor Ce in the emitter path.
The complex impedance is called Xe=Re||(1/wCe) and the "rough approximation" as proposed by you in post#5 is A=Rc/Xe.

*It is common knowledge that it is the purpose of Re to provide DC feedback to stabilize the bias point - and the only purpose of Ce is to "short" this resistor in order NOT to have gain reducing signal feedback.
That means: The value of Ce is selected aiming at a very low Xe value for the operating frequency range ("shorting" Re).
If we want to have some signal feedback the resistor Re is divided in to parts (as I have mentioned in my previous post).

*Now - what are you doing in your quoted ac example above? You select a single operating frequency that corresponds with a fixed value of Xe=1000 ohms.
Each reader of this thread may decide if this a realistic and a fair example to proof your approximation A=Rc/Xe.
Up to now, I never have seen any common emitter amplifier designed for one single frequency with a gain determining capacitor in the emitter path. Do you???
Did you ever think about the influence of the phase shift caused by Ce in your example (stability aspects)? This is another reason for designing 1/wCe very small for the operating frequency range.

You're approximation is better than my "very rough" approximation. Your's just requires more basic knowledge of the workings of the transistor to be able to apply than mine does.

It is not "my approximation" but a gain formula, which is given in all relevant textbooks.
Is there any serious book or article that uses "your" approximation Rc/Xe without pointing to the condition: Only for frequencies below the operating range with Xe>>1/g ?
I was afraid, that a beginner could be "lost" using you formula. What happens for frequencies equivalent to 1/wC= 1 ohm (not very unusual) ?
With your formula (Rc=1E4, re=10) : A=1E4/1=1E4 ; classical formula: A=1E4/(10+1)=909.
By the way - this was my only point of critics: Not to mention the restricted application range of your approximation outside the normal operating frequency range.

Regarding your remark concerning "basic knowledge": I don`t know if or how you ever have explained the transistor basics to students.
Before mentioning any approximate/simplified gain formula (like Rc/Re) I try to teach the basic task of a transistor: To act as a voltage-controlled current source.
In this context, it is obvious to present the transfer characteristic Ic=f(Vbe) and the corresponding slope g=d(Ic)/d(Vbe).=ic/vbe
That is the "basic" knowledge " that is required to derive "my gain formula" (no h parameters necessary as claimed by you).
______________________
I am sorry. This answer happens to be longer than anticipated. But it was my last contribution in this thread.

Thank you
Winterstone


Final remark (added later):
The question of the thread originator (see post#13 and my response in post#17) is a good example, which clearly shows that - before applying any gain formula - it is necessary to know the basic transistor properties: Voltage-controlled-current source with the main parameter "transconductance g". Otherwise, one is "lost" knowing just Rc/Re or even Rc/Xe.

 

[MrAl]

Hello again Winterstone,


Apparently you just cant accept anyone else's approach to this kind if analysis, so what i will do is modify my formula and if you arent happy then that's not my problem anymore


For Winterstone's benefit and hopefully for the OP's benefit, i'll modify my formula to this:
Gac=Rc/(re+Xe)

where
Gac is the AC gain,
Rc is the collector resistance,
re is the internal transistor emitter resistance,
Xe is the combined impedance of the resistance in parallel with the capacitance.

In addition, we might also want to see the additional high frequency limiting resistor R2 in series with the R and C that are in parallel, because we can not depend on the internal re to limit the gain because it is often too low. So that leaves us with this:
Gac=Rc/(re+Xe+R2)

and note that there is also a phase reversal in either of these.

So ok there Winterstone, does that make you feel more comfortable now?

 

[Winterstone]

..............
For Winterstone's benefit and hopefully for the OP's benefit, i'll modify my formula
..........
So ok there Winterstone, does that make you feel more comfortable now?

The only purpose of my replies was not to feel "more comfortable" or to gain some "benefits".
It was just an attempt to help the questioner and to tell him the truth about a gain formula approximation, which is not applicable for the operating frequency range.
But - nevertheless - I appreciate your last reply (forgetting the first line).
Thank you and regards
W.

 

[MrAl]

Hi again Winterstone, and also reply to Ratchit,

Why forget the first line? You didnt accept it right? But that's ok anyway, as long as we agree now



WHEN GOOD FORMULAS GO BAD

Ratchit:
I take it you agree with that post's formulation which i will repeat here in the hopes it may help:
Ic=Is*(e^(Vbe/VT)-1)

So since you think it is that good and apparently agree with the original poster, then you can easily answer this question:

Given a 2N2222A transistor, a 10v DC supply source, a 1k collector resistor, and a 1 ohm external emitter resistor, and a base emitter voltage of 0.65 volts, what is the collector current? You may assume that VT=0.0257 volts due to the temperature. Since we want to use that formula we dont allow a circuit simulation for now.

[note we might be able to ignore the 1 ohm resistor, making it 0 ohms, which makes this even simpler]

 

[Ratchit]

MrAl,

Ratchit:
I take it you agree with that post's formulation which i will repeat here in the hopes it may help:
Ic=Is*(e^(Vbe/VT)-1) Given a 2N2222A transistor, a 10v DC supply source, a 1k collector resistor, and a 1 ohm external emitter resistor, and a base emitter voltage of 0.65 volts, what is the collector current? You may assume that VT=0.0257 volts due to the temperature. Since we want to use that formula we dont allow a circuit simulation for now.

The post promotes the voltage-centric method instead of the β-centric method of biasing. That formula you quoted does not show up in post #10 of the thread. What does show up is the formula re=0.0257/Ic. The illustration shows an Ic of 0.1 ma, which will result in a re = 257 ohms. So, for an Ic of 0.1 ma, we calculate (Vb-0.65)/257 = 0.1 ma ===>Vb=0.68 volts. That's pretty close to 0.65 volts, so we can add a 1k emitter resistor which will boost Vb up to 0.78 volts. So stabilizing Vb will assure a constant Ic, and the method is independent of β. We decide the current first, then select the components to make it happen.

Ratch

https://cr4.globalspec.com/thread/68055/voltage-vs-current

 

[MrAl]

Hi Ratchit,


Yes but it shows up in the same thread (post 2) by the same guy and that's where the whole voltage control idea stems from.
So if you want to agree with him then do you agree with that formula, and what is the collector current in my previous post?

 

[Ratchit]

MrAl,

So if you want to agree with him then do you agree with that formula, and what is the collector current in my previous post?

Certainly I agree with him on Shockley's formula. I certainly do not want to disagree with Shockley, do I? If you want me to calculate Ic, then you have to supply me with an Is value.

Ratch

 

[MrAl]

MrAl,



Certainly I agree with him on Shockley's formula. I certainly do not want to disagree with Shockley, do I? If you want me to calculate Ic, then you have to supply me with an Is value.

Ratch

Hi,

Exactly the reply i expected. I gave you the whole circuit including the transistor part number, but you still cant do it. Why? Because NOW we need an Is value

Let me come clean here though. What i expected was that you (or anyone else) would not know what Is was, or you (or other) might dig up a value for Is, then do the formula, and still come up with the wrong result. And the whole point of his post was that the voltage control method was somehow better than the Beta method. Sure, if you know all the parameters you're ok, but im not sure if you can find Is in the data sheet while we can usually find Beta and we already know how to handle that.
But i guess we can steal Is from one of the spice models we down load right? That might be acceptable, although a measurement of Is might be better if we can do it. We can set up a couple equations and solve for Is i suppose. But unfortunately it still wont work in that formula because something is missing.

But again the whole point was why abandon 'Beta' if 'Is' is going to be unknown? Maybe you can shed some light on this through your own past experience since you always imply that you like the voltage control method over the current control method (and im not saying that it is totally off the wall either

So tell me how you would go about solving that circuit without using a circuit simulator. Im pretty sure we know how to approach this with the Beta method but the voltage control method still needs to be clarified i think. Of course we can check our results in a circuit simulator later, but i'd like to wait until we have the whole story here first.

Perhaps we could look at ways to measure Is too while we are at it. I intended to do an Article on this but didnt get around to it yet.

Is the Shockley equation wrong? I dont think it is, but then neither is Pavg=I*E. But that doesnt mean that we can apply them to EVERY circuit under the sun. They have to be applied as appropriate. The Shockley equation is for an IDEAL diode. I gave a REAL non ideal part number 2N2222A. That's the difference. So the error is in the application not the formula itself.

 

[Ratchit]

MrAl,

Well, let's look at the Shockley equation. i=Is(e^v/0.0257-1) . When v=100 mv or greater, then the equation becomes approximately i=Is(e^v/0.0257). Differentiating, we get di/dv = 1/re=Is(e^v/0.0257)/0.0257=i/0.0257=====>re=0.0257/i .

Therefore, the formula re=0.0257/i does come from Shockley's equation. The collector current will be approximately the same as the emitter current, which will be (Vb-0.65)/Re . The value of Is is irrelevant because it is swallowed up by the high value of the e^(v/0.0257) term.

Ratch

 

[MrAl]

Hi,

Perhaps so, but that's not the question that i asked. This has been your third reply to the post with the question and still no answer to what Ic is.
You have the formula now so do you want to calculate Ic or not?

 

[Ratchit]

MrAl,

Perhaps so, but that's not the question that i asked. This has been your third reply to the post with the question and still no answer to what Ic is.
You have the formula now so do you want to calculate Ic or not?

I would think anyone could calculate it. Using Vb of 0.78 volts and RE of 1k like I suggested, we get Ic =(0.78-0.65)/1000 = 0.13 ma . That is only 0.13 volts across Rc, so maybe we can lower the Vcc voltage, or increase Vb some more to increase Ic. Better yet, increase Rc to 10k so as to get a gain of Rc/(RE +Re) = 10000/(1000+198) = of approximately 8.35 .

Ratch

 

[MrAl]

Hello again,


This has been your fourth reply now to the post with the question and you still have not answered the question. I am not trying to be argumentative just pointing out that the "voltage controlled" approach you seem to like (and i dont argue that it is exceptionally bad) has not yet produced a result for the question, yet it is put forth as if it is exceptionally accurate and easy to use. So where is the accurate and simple answer?

Here is a repeat of the question:

Formula given:
Ic=Is*(e^(Vbe/VT)-1)

Question:

Given a 2N2222A transistor, a 10v DC supply source, a 1k collector resistor, and a 0 ohm external emitter resistor, and a base emitter voltage of 0.65 volts, what is the collector current? You may assume that VT=0.0257 volts due to the temperature.

[note i replaced the 1 ohm resistor with 0 ohms to make this simpler but it would come out very close to the same result with the 1 ohm resistor]

So we already have the operating point picked out for us, as well as the collector resistor and there is no external emitter resistor.

Care to shoot for a fifth post with no direct answer

 

[Winterstone]

May I help? It`s a simple task:
Ic=Is*9.62*1E10 A.

 

[Ratchit]

MrAl,

Question:

Given a 2N2222A transistor, a 10v DC supply source, a 1k collector resistor, and a 0 ohm external emitter resistor, and a base emitter voltage of 0.65 volts, what is the collector current? You may assume that VT=0.0257 volts due to the temperature.

[note i replaced the 1 ohm resistor with 0 ohms to make this simpler but it would come out very close to the same result with the 1 ohm resistor]

So we already have the operating point picked out for us, as well as the collector resistor and there is no external emitter resistor.

Care to shoot for a fifth post with no direct answer

I believe I already answered the question, with the assumptions I made, and assuming that the 10 volts is applied only to the collector. If no voltage is applied to the base, then no current will exist.

Ratch

 

[Winterstone]

Yes but it shows up in the same thread (post 2) by the same guy and that's where the whole voltage control idea stems from.
So if you want to agree with him then do you agree with that formula, and what is the collector current in my previous post?

Hello again MrAl and Ratch,

because we have met already in this thread, I take the opportunity to jump into this new discussion in order to contribute my opinion.

I know, my previous answer in post#34 regarding the resulting Ic value was not (could not be) really satisfying.
However, that`s the only answer that can be given if Is is unknown.
But we have, of course, a similar problem using the relation Ic=B*Ib.
In this case - assuming for example Ib=1uA and B unknown - we have Ic=B*1E-6 A .

However, such an exercise has nothing to do with the physical principle of Ic control. And this was - as far as I have understood - the start of the discussion, was it not? (See the quoted sequence above).

Here we have clearly to distinguish between „to control“ and „to determine“ the Ic value, respectively. And your question to Ratch concerns Ic determination!
According to Shockley`s equation, the amount of collector current is determined by three quantities (Vt, Vbe, Is), however, there is only one external parameter that can control (modify, vary) the current Ic - and that is undoubtly the base-emitter voltage Vbe.

To me, it seems to be logical that a particular value for Ic can be calculated only if all parameters that determine this value are known. And this applies to B as well as to Is.
I am aware that the uncertainty range for Is is much larger than for B - however, this cannot be a criterion for the discussion voltage vs. current control (physically!).
(By the way, the influence of the Is uncertainty range can be reduced drastically using Re feedback).

More than that, we shouldn`t confuse the physical principle of Ic control with the practical implementation. Of course, there are solutions for dc biasing a BJT with a current Ib.
However, a closer look reveals that we have, of course, not a „current source“. Instead, it is a simple voltage divider between the static DC base-emitter resistance Rbe of the BJT and a large external resistor Rb - driven with a voltage.
And one shouldn`t forget: For calculating Rb (based on a given Ib) we need a value vor Vbe ! I repeat: Even the so-called „current biasing“ needs an assumption for Vbe.

Regards
W.

 

[MrAl]

MrAl,



I believe I already answered the question, with the assumptions I made, and assuming that the 10 volts is applied only to the collector. If no voltage is applied to the base, then no current will exist.

Ratch

Ok, so this is your fifth post after the question and you still havent answered it. You're really being confusing now because you said if no voltage is applied to the base then no current will exist, yet i CLEARLY gave you a base emitter voltage of 0.65 volts.

It's a simple question with a simple answer, the answer is a current value for Ic. I'll give a few examples which may or may not be the correct result:
0.1ma
1ma
10ma
100ma
1000ma
any of those could be the right one, or none of them, but that's the result we need here. A simple current level using the exponential formula.

If the voltage control method is that good then we should have had an answer by the first or second reply, but now we are up to 5 replies and you still havent provided an answer for the original question which was pretty clear. This tells me you dont know how to calculated it with the parameters given?

 

[MrAl]

Hello again MrAl and Ratch,

because we have met already in this thread, I take the opportunity to jump into this new discussion in order to contribute my opinion.

I know, my previous answer in post#34 regarding the resulting Ic value was not (could not be) really satisfying.
However, that`s the only answer that can be given if Is is unknown.
But we have, of course, a similar problem using the relation Ic=B*Ib.
In this case - assuming for example Ib=1uA and B unknown - we have Ic=B*1E-6 A .

However, such an exercise has nothing to do with the physical principle of Ic control. And this was - as far as I have understood - the start of the discussion, was it not? (See the quoted sequence above).

Here we have clearly to distinguish between „to control“ and „to determine“ the Ic value, respectively. And your question to Ratch concerns Ic determination!
According to Shockley`s equation, the amount of collector current is determined by three quantities (Vt, Vbe, Is), however, there is only one external parameter that can control (modify, vary) the current Ic - and that is undoubtly the base-emitter voltage Vbe.

To me, it seems to be logical that a particular value for Ic can be calculated only if all parameters that determine this value are known. And this applies to B as well as to Is.
I am aware that the uncertainty range for Is is much larger than for B - however, this cannot be a criterion for the discussion voltage vs. current control (physically!).
(By the way, the influence of the Is uncertainty range can be reduced drastically using Re feedback).

More than that, we shouldn`t confuse the physical principle of Ic control with the practical implementation. Of course, there are solutions for dc biasing a BJT with a current Ib.
However, a closer look reveals that we have, of course, not a „current source“. Instead, it is a simple voltage divider between the static DC base-emitter resistance Rbe of the BJT and a large external resistor Rb - driven with a voltage.
And one shouldn`t forget: For calculating Rb (based on a given Ib) we need a value vor Vbe ! I repeat: Even the so-called „current biasing“ needs an assumption for Vbe.

Regards
W.

Hello there Winterstone,

This new discussion is related to the previous one because Ratchit brought up the topic of doing the calculation using the voltage control model, and pointed to a thread on anther site which quoted at least two formulas, one for Ic and one for re. I asked him a simple question, to calculate Ic with known values with a known transistor using the formula quoted in the thread he pointed to. He agreed that the formula was correct, but after 5 replies he still hasnet calculated the value of Ic for that circuit and operating point.

But the main point was that if the voltage control method was so much better, then we should have easily obtained a result by now.

'Is' is not given yet, but the transistor part number was given, 2N2222A. I think we can find a range of Beta on the data sheet, but we cant find Is can we? Perhaps you can find it, but you seem to be at a loss for this parameter too and this is just one little circuit with one little resistor in the collector

And remember we are looking for a particular current here, not an algebraic expression that leads to the current, so it should be like 1ma, 10ma, 100ma, 1000ma, etc.

Your comments welcome as usual.

 

[Winterstone]

Hello there MrAl,

thanks for your reply.
I am afraid I couldn`t express myself clearly enough in post#36. It seems you didn`t understand my point (probably my fault).
Therefore, I will repeat the core of my previous contribution in the following:

*We clearly have to distinguish between „to control“ and „to determine“ the Ic value, respectively.
And your question to Ratch concerns Ic determination!

*According to Shockley`s equation, the amount of collector current is determined by three quantities (Vt, Vbe, Is), however, there is only one external parameter that can control (modify, vary)
the current Ic - and that is undoubtly the base-emitter voltage Vbe.
To me, it seems to be logical that a particular value for Ic can be calculated only if all parameters that determine this value are known. And this applies to B as well as to Is.
You certainly will agree that it is impossible to calculate an Ic value without knowing Is.

*And - most important - this question (the value for Ic you are asking for) has nothing to do with the physical principle of Ic control.
It seems you are confusing "control" with "determine". Are YOU able to calculate the value of Ic (provided Ib is given) if B is unknown?

*As a real-life example: You can control the speed of a car with a corresponding pedal position - but you cannot derive from this information (position) the momentary speed of the car.
The speed is determined - in addition - by some other parameters (wind, slope of the street, etc.).

Do you know now what I mean?
That are my comments you were asking for.

Regards
W.

 

[Ratchit]

MrAl,

Ok, so this is your fifth post after the question and you still havent answered it. You're really being confusing now because you said if no voltage is applied to the base then no current will exist, yet i CLEARLY gave you a base emitter voltage of 0.65 volts.

Sorry, I was thinking of the applied voltage to RE after the Vbe drop. My apologies. The voltage bias method relies on a significant RE resistance and a Vb above 0.65 to set the emitter current. Once the emitter current is known, then other values can be calculated.

I asked him a simple question, to calculate Ic with known values with a known transistor using the formula quoted in the thread he pointed to. He agreed that the formula was correct, but after 5 replies he still hasnet calculated the value of Ic for that circuit and operating point.

That is because the current cannot be calculated without a resistance (RE) in series with the emitter-base diode and a voltage higher than 0.65 volts, at least not without knowing the Is first.

He agreed that the formula was correct, but after 5 replies he still hasnet calculated the value of Ic for that circuit and operating point.

Yes, the formula is correct provided that the conditons are correct for it to work.

But the main point was that if the voltage control method was so much better, then we should have easily obtained a result by now.

It is is a better method, but the result can only be obtained if the conditons stated above are right.

'Is' is not given yet, but the transistor part number was given, 2N2222A. I think we can find a range of Beta on the data sheet, but we cant find Is can we? Perhaps you can find it, but you seem to be at a loss for this parameter too and this is just one little circuit with one little resistor in the collector

Who needs beta for the voltage method of bias? Don't need Is either if Vb is high enough and RE is significant.

And remember we are looking for a particular current here, not an algebraic expression that leads to the current, so it should be like 1ma, 10ma, 100ma, 1000ma, etc.

You are looking for a method that sets the reasonable current you desire.

In conclusion, just because the method does not work when conditions are marginal, does not denigrate the method.

Ratch