Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Why does a buffer is used in an audio amplifier?

  1. Nov 13, 2011 #1
    Let a simple concept of an audio amplifier is to be designed like this (in BJT) : CC *CE* CE * OUTPUT

    .... and CC acts as a buffer.

    Why does a buffer is required for the amplification? Wouldn't it be much more simpler if CC is being cancelled out?

    CC = Common Collector
    CE= Common Emitter

    Thanks in advance!
  2. jcsd
  3. Nov 13, 2011 #2


    User Avatar
    Science Advisor
    Homework Helper

    The purpose of the CC stage may be to give a high input impedance.
  4. Nov 13, 2011 #3
    Audio amp need to drive a speaker that is very low impedance ( 8Ω). You need a stage that can drive such a low impedance load. CE stage usually has very high output impedance to get the gain, so you need CC to buffer the high impedance CE stage in order to drive the speaker.

    Actually a lot of amp use Darlington......two CC stage in order to prevent the speaker load from loading the CE stage!!!
    Last edited: Nov 13, 2011
  5. Nov 13, 2011 #4


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    From what I understand of the OP, the CC in question is at the input and not at the output. In all amplifiers, it should be borne in mind that what is required is Power Gain (A simple passive transformer would give you as much voltage gain as you wanted but would not help when the input signal is low power.)

    There are many configurations of amplifiers and the choice of configuration will depend upon the impedance of the signal source and the signal level, the impedance of the load and the required driving level - obvious but it needs to be said.

    Many signal sources have a high impedance so they require a high impedance loading if you want to avoid losing most of the signal voltage. A CC (emitter follower) does the job well and, because of the 100% voltage feedback, the linearity is good. Although it gives no voltage gain, it has a low output impedance, which means that it can feed the next stage(s) with much higher signal power. (The input resistance of an emitter follower circuit is around βRe, which can be several MΩ - very useful as a buffer)

    Using a straight CE configuration for an output stage gives a fairly high output impedance but pairs of transistors (push - pull) with a lot of feedback can achieve a low output impedance and good linearity.
  6. Nov 13, 2011 #5
    I guess I missed the CC*CE*CE thingy, I read enough audio amp schematic and actually designed guitar amplifiers. Never ones I saw this configuration of CC*CE*CE. It just did not even register to me.

    If very high input impedance is needed, a JFET CS stage is the answer instead. If the source is so weak that even the miller cap might hurt, use a cascode configuration instead. For very low level signals, CC as the first stage will also add noise without any signal amplitude gain. It is just not a common practice.

    Also, I never see a CE as output stage also as what I explain in my post before. The only output stage similar to CE is only found in tube amp where it is the common Cathode configuration and an output transformer is used as load on the plate side to step the impedance down to match the speaker. But this is not in the context of this post.
  7. Nov 13, 2011 #6


    User Avatar

    actually a Darlington is very nearly the same thing as a CC leading a CE.

    and the idea is to greatly reduce base current which is equivalent to bumping up the input impedance.
  8. Nov 13, 2011 #7
    No, I meant darlington CC stage. If you google any BJT power amp and look at the output stage, most likely you'll see a darlington CC stage. This is because even the [itex]\beta=100\;[/itex], it will still present a 800Ω load to the preceding CE stage. This is still way too low. You need another CC stage before it to bring the input impedance to 100X800=80KΩ

    I should said that it is a darlington type of design, not exactly using darlington transistor. Usually they use one power BJT as CC to buffer the CE stage, and in turn drive multiple BJT that configured as CC in parallel to increase the current handling capability. It takes two stage of CC to be sure not to load the CE stage down. In this aspect, power MOSFET has the advantage, but then you have to deal with the high input capacitance of the MOSFET!!!! It's a double edge sword.
    Last edited: Nov 13, 2011
  9. Nov 14, 2011 #8
    i see... (*digesting it)
  10. Nov 14, 2011 #9


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Let's nor try to run before we walk. I don't think the OP contained information for serious circuit design. The general principle of producing a high input impedance is worth mentioning, though.
  11. Nov 14, 2011 #10
    Like Aleph said, only the usual purpose of a buffer is to take a signal from a high impedence output, and drive it out a low impedence output.
  12. Nov 16, 2011 #11
    oo, thanks for the info guys
  13. Nov 17, 2011 #12
    Should i consider using a CC for the OUTPUT or a class B power amplifier?
  14. Nov 18, 2011 #13
    A lot of the amplifiers use CC as output stage, that has nothing to do with what class it is operating. Class B only means the transistor is conducting in half of the cycle.....conducting only for 180 deg of the cycle.

    Say for a class B push pull using complimentary CC transistor ( NPN CC to pull up and PNP CC to pull down). You adjust the quiescent current very close to zero ( when the output is exactly at the mid point). So when the amplifier output want to swing +ve, only NPN is pulling up, the PNP is off. But when the amp want to pull down from the mid point, only the PNP is pulling down, NPN is off. Each transistor responsible for only one side.

    You don't use class B for audio amp as your tittle imply. Most likely the high end amplifiers configured to run in class A instead. Class B has too much cross over distortion.
  15. Nov 18, 2011 #14


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    If you actually want an amplifier to listen to then use an established design (even a kit). You are almost certain to be disappointed with a first, second or even third stab at your own design. otoh, if you are just doing this for self education, read some books on amp design and not just 'text book' stuff which is what you'll get here. If you do try to design and build anything for yourself, you will have all the problems of linearity and then self oscillation and squegging etc. (as soon as you introduce feedback) and Hummmmmmm.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook