# The use of amplifier stages

I don't understand the need of amplifier stages if i want for exmple to amplify a small AC signal. Sure i get that it is used to improve the gain, but isnt an RC transistor amplifier made in such a way, that the collector voltage is equal to half Vcc in order to have maximum swing. If so, why do we need more than one amplifiyng stage? To get bigger current? Here's what i know about that: When im building a light sensor for example, i form a voltage divider with a photoresistor and another resistor to ground. The output is connected to the base. I bias it in such a way that the base current is 1/4th of the whole current through the photoresistor and sometimes this doesn't give large enough collector current. This is when i use something like amplifiyng stage. I either use another transistor to increase the gain by connecting them like darlington transistors, or i use the main weak signal to turn off a transistor, which in turn will allow for another transistor to turn on, giving me high enough output. And this is all i use amplifier stages for-to increase a weak current. Is there anything else i dont get here?

Svein
If you want the amplified signal to have little distortion, you need to linearize the stages - and that linearizing uses negative feedback, removing most of the gain. If you do not care about the relation between the input and output signal, an operational amplifier will usually give you a gain of 10 000 - which means that the output will spend almost all the time close to one of the supply voltages.

Harrison G
Averagesupernova
Gold Member
Too much gain in a single stage increases the chances of oscillation also.

Harrison G
davenn
Gold Member
Too much gain in a single stage increases the chances of oscillation also.
yes, exactly, one of the major reason amplification is done in stages. Keeps everything stable

D

Harrison G
Baluncore
Devices that amplify are usually specified by a “Gain Bandwidth Product”, GBP = Gain * BW.
If you want a gain of 10 and a bandwidth of 100kHz you will need one device with a minimum GBP of 1MHz.
If you want a gain of 100 and a bandwidth of 100kHz you will need one device with a minimum GBP of 10MHz.
Alternatively, it may be cheaper or more reliable to use two stages, each with a gain of 10 and each with a GBP of 1MHz.

Harrison G
Too much gain in a single stage increases the chances of oscillation also.
yes, exactly, one of the major reason amplification is done in stages. Keeps everything stable
D
Mmmmh...It is a common misconception to believe that reducing the open-loop gain of an amplifier using negative feedback would stabilize such a stage.
Exactly the opposite is true. The stability margin will be reduced correspondingly.
On the other hand, negative feedback enhances bandwidth and dynamic range and reduces harmonic distortions.
Hence, again we have to find a trade-off between some conflicting requirements.
This is normal for electronic design - each good design is a trade-off between several requirements.
Each advantage/improvement must always be paid for by some degradations of other parameters.

Last edited:
Harrison G
Averagesupernova
Gold Member
Mmmmh...It is a common misconception to believe that reducing the open-loop gain of an amplifier using negative feedback would stabilize such a stage.
Exactly the opposite is true. The stability margin will be reduced correspondingly.
On the other hand, negative feedback enhances bandwidth and dynamic range and reduces harmonic distortions.
Hence, again we have to find a trade-off between some conflicting requirements.
This is normal for electronic design - each good design is a trade-off between several requirements.
Each advantage/improvement must always be paid for by some degradations of other parameters.
There is certainly more to the story than this. There are certainly cases where a lot of gain in a small physical space can cause oscillation. Isolation of several stages can help prevent this. No one here is saying that the only reason for multiple stages is for reduced chances of oscillation.

Last edited by a moderator:
Harrison G and davenn
There is certainly more to the story than this. There are certainly cases where a lot of gain in a small physical space can cause oscillation.
I am sure that you will agree with me that oscillations can occur only in the presence of feedback.
A gain stage - even with very large gain like an opamp - will tend to instability only if there is feedback - and it does not matter if the feedback loop is designed or if it is an unwanted (parasitic) effects.

Let me explain a little the background of my comment in post#6:
I have several years experience with teaching analog electronics to students. And very often I have seen that there is a lot of confusion connected with the terms "stability" and "feedback". Why?
Because negative feedback
* IMPROVES DC stability (of the operating point against temperature effects and other uncertainties), and at the same time
* DEGRADES dynamic stability (against oscillations).

At least in one of the answers given in this thread the therms "opamp" and "gain reduction by feedback" are mentioned - and it was my only intention to make clear (and the OP seems to be a beginner) that negative feedback cannot improve dynamic stability at all.
However, negative feedback has many, many advantages - and therefore, as I have tried to explain, a trade-off is always necessary to find the best solution for a specific task (DC stability, AC stability, dynamic range, bandwidth, THD, cost, number of stages, ...).

Harrison G
davenn
Gold Member
Mmmmh...It is a common misconception to believe that reducing the open-loop gain of an amplifier using negative feedback would stabilize such a stage.
Exactly the opposite is true. The stability margin will be reduced correspondingly.
On the other hand, negative feedback enhances bandwidth and dynamic range and reduces harmonic distortions.
Hence, again we have to find a trade-off between some conflicting requirements.
This is normal for electronic design - each good design is a trade-off between several requirements.
Each advantage/improvement must always be paid for by some degradations of other parameters.

never mentioned any thing about feedback ... just gain
Its standard practice, particularly in RF gear to do amplification in stages for the reasons mention

I stand by my and ASN's comments

Harrison G
never mentioned any thing about feedback ... just gain
Its standard practice, particularly in RF gear to do amplification in stages for the reasons mention
I stand by my and ASN's comments
davenn - did I argue against several stages? In no way! Please read my explanation in post#8.
Primarily, it was my intention to shown that and why a trade-off between conflicting requirements is always necessary.
Perhaps I was not able to express myself very clearly - due to my limited knowledge of your language.

Harrison G
Because negative feedback
* IMPROVES DC stability (of the operating point against temperature effects and other uncertainties), and at the same time
* DEGRADES dynamic stability (against oscillations).
Can you show the math you're basing this on along with your definitions of 'DC stability' and 'dynamic stability'?

Harrison G
Can you show the math you're basing this on along with your definitions of 'DC stability' and 'dynamic stability'?
I am not sure what you are asking for.
* Do you ask me to proove why negative DC feedback defines and stabilizes the DC operating point of an amplifier?
* Do you ask me to demonstrate in this thread the relationship between the phase margin and the feedback factor ?

Harrison G
I am not sure what you are asking for.
* Do you ask me to proove why negative DC feedback defines and stabilizes the DC operating point of an amplifier?
* Do you ask me to demonstrate in this thread the relationship between the phase margin and the feedback factor ?
I can see from this what you'd like to show, but I don't think that'll address what you seem to be stating here in general.

If I can show you a system which, in open loop, has a highly oscillatory response to a step input, do you mean to tell me that it's impossible to improve upon its transient response using negative feedback?

Harrison G and Averagesupernova
davenn
Gold Member
davenn - did I argue against several stages? In no way!
so why quote me then ????

Mmmmh...It is a common misconception.......
this wasn't arguing against me ?

if you are not arguing against what I said ... then you shouldn't have quoted me !

Harrison G
I can see from this what you'd like to show, but I don't think that'll address what you seem to be stating here in general.

If I can show you a system which, in open loop, has a highly oscillatory response to a step input, do you mean to tell me that it's impossible to improve upon its transient response using negative feedback?
In case the system has a "highly oscillatory response" it has internal feeedback loops. Thats all.
More than that, I dont think that this case is relevant for our discussion here.

Harrison G
if you are not arguing against what I said ... then you shouldn't have quoted me !
davenn - with all respect, dont overrate my answers.
As I have stated already - perhaps I have chosen not the correct words or terms.
Didnt I try to explain my motivation?
Perhaps I was wrong to quote you in the context of my answer - sorry for that. OK?

Harrison G
More than that, I dont think that this case is relevant for our discussion here.
Then please define 'dynamic stability' and show, mathematically, how negative feedback degrades it.

Harrison G
Averagesupernova
Gold Member
It seems so often negative feedback gets a bad rap because while the intention may be to provide negative feedback, that feedback unintentionally becomes positive feedback at some frequency or frequencies.
LvW, you say here:
A gain stage - even with very large gain like an opamp - will tend to instability only if there is feedback - and it does not matter if the feedback loop is designed or if it is an unwanted (parasitic) effects.
You want to be careful with your words here. It looks like you are saying: If I lay out an op-amp on a breadboard and due to sloppy construction practices there was enough parasitic capacitance and no negative feedback that the thing would oscillate and applying negative feedback would not solve this oscillation. It is always assumed that there is always SOME positive feedback that is unavoidable due to parasitic affects. One of the reasons we apply negative feedback and scatter the gain out in stages is so the overall positive feedback is never large enough to start and sustain an oscillation. Try telling a control systems engineer that negative feedback in a PID loop will improve chances of oscillation.

Harrison G
Svein
Mmmmh...It is a common misconception to believe that reducing the open-loop gain of an amplifier using negative feedback would stabilize such a stage.
Exactly the opposite is true. The stability margin will be reduced correspondingly.
On the other hand, negative feedback enhances bandwidth and dynamic range and reduces harmonic distortions.
Hence, again we have to find a trade-off between some conflicting requirements.
This is normal for electronic design - each good design is a trade-off between several requirements.
Each advantage/improvement must always be paid for by some degradations of other parameters.
Sort of right: Check out http://cds.linear.com/docs/en/application-note/an148fa.pdf.

Harrison G
Then please define 'dynamic stability' and show, mathematically, how negative feedback degrades it.
Perhaps you have overlooked that in post#8 I have mentioned "dynamic stability" together with "against oscillation"; more than that, in post#12 I gave an answer using the term "phase margin". I suppose, this should be sufficient to explain what I mean with "dynamic stability" in contrast to "DC stability".

Here is a book chapter (colorado university), which explains to you the term in detail:

Last edited:
Harrison G
One of the reasons we apply negative feedback and scatter the gain out in stages is so the overall positive feedback is never large enough to start and sustain an oscillation. Try telling a control systems engineer that negative feedback in a PID loop will improve chances of oscillation.
Averagesupernova - I have tried to respond to a question from the OP (Harrison G).
As it seems, he is not very well experienced - and his question was about the justification of several (lower) gain stages if compared to one single (high) gain stage.
We all know that we have good reasons to distribute the desired overall gain among a number of gain stages (bandwidth, dynamic range, THD, design aspects).
However - for my opinion, in general we cannot say that such an approach would improve stability ("keeps everything stable"). Dont you agree?
This was the background and the justification of my contribution in post#6 and post#8.
Please correct me if anything is wrong in my text.

Now you are trying to bring a more complex aspect into the discussion (control systems, PD controllers).
However, I am not sure if such a discussion will the questioner help to find an answer to his question, OK?

I suppose, this should be sufficient to explain what I mean with "dynamic stability" ...
No, it's really not. There's all kinds of mathematical definitions for 'stability' and you haven't given one.

Because negative feedback
* DEGRADES dynamic stability (against oscillations).
Reading this statement, you must be assigning some metric to 'dynamic stability', a measure of it, and you state that this metric gets degraded by applying negative feedback.

If you consider this metric to be 'phase margin', then qualify your statement mathematically. Show what you mean, precisely, so your words have some meaning. Maybe show a small example, using math, that illustrates your point? It shouldn't be that hard, with all your experience, and it will illustrate, to the OP especially, what point you're trying to make.

As a physicist doing electronics engineering for the past 4 years... this thread is pretty sad. As far as I can tell, LvW is exactly right.

The main thing causing confusion here, is that he has failed to mention (or I've missed) that even supposedly "open loop" amplifiers have negative feedback, through the parasitic capacitances inside the device.

I would also point out the following with regards to multi-stage high frequency amplifiers:
- Increasing gain means decreasing the "degenerative resistor" at the emitter of a bjt. This makes the gain of the bjt dependent on current, temperature, and supply voltage. In other words, it will drift around uncontrollably. This behavior is suppressed by reducing the gain.
- Gain stages multiply, which makes it easy to reach insanely high gain values with just a few stages.

Svein
No, it's really not. There's all kinds of mathematical definitions for 'stability' and you haven't given one.
.
I do not intend to repeat at this place definitions - unless the questioner will ask for more clarification. If you would read all the posts again you will notice that several contributors have used the term "stable" or "stability" without repeating definitions. And thats OK. While speaking about stability of amplifiers everybody who is familiar with the problems which might occur during design of amplifiers will have no problems. The only point which - in my opinion - has required some clarification was the discrimination between DC and AC stability. That`s what I have done in my former posts.
Reading this statement, you must be assigning some metric to 'dynamic stability', a measure of it, and you state that this metric gets degraded by applying negative feedback.
If you consider this metric to be 'phase margin', then qualify your statement mathematically. Show what you mean, precisely, so your words have some meaning. Maybe show a small example, using math, that illustrates your point? It shouldn't be that hard, with all your experience, and it will illustrate, to the OP especially, what point you're trying to make.

(Off the record: May I ask you: What is your problem with me resp. my contributions? Do you consider all the other contributions as "self-explaining" (post#3 and #4)? Your wording sounds a bit "strange" in my ears).

Nevertheless, with regards
LvW

What is your problem with me resp. my contributions?
You make vague statements that could mean any number of things, most of which are outright wrong. I don't consider that any sort of contribution.

If you really had a contribution to make, I would think you would be thrilled to show the math supporting it, since it would remove any doubt as to what point you're trying to make. It doesn't inspire a lot of confidence in your words if you're unable or unwilling to do that.