A Friendly Nudge In Designing A Multi - Stage Bjt Amplifier

Discussion in 'The Projects Forum' started by Blomqvist, Dec 3, 2015.

  1. Blomqvist

    Thread Starter New Member

    Jun 10, 2015
    5
    0
    [​IMG]

    Hey all,

    I hope this is the right section!

    As a final project, a group of us are designing a Three Stage BJT amplifier using the 2N3904 transistor with an expected non-inverting gain of 15 V/V at a midband frequency of 15 Hz. The source is a 100mV 1Hz pulse with a Thevenin resistance of 200k. The two common emitter stages contribute the 15 V/V and the final stage is a simple emitter follower (voltage buffer). Each stage is separated by a coupling capacitor, which strips any DC component that may have become amplified during the previous stage. This allows the output to only realize AC amplication. The amplifier will output a peak voltage of 1.5 V to a load resistance of 100 ohms. Note, the capacitor crossed-out in red is considered removed. Since we want to realize a high input impedance at stage one, the capacitor would only work against this goal. However, an alternative solution is to include a small emitter resistance in series with R4//C2.

    We are having a hard time trying to find a "baseline" or a place to start for finding component values. So far, we have come up with expressions for Rin, Rout, gain, vpi, and Vc for each common-emitter stage (hopefully they are right). Obviously, we're not looking for any responses that solve the problem for us, but simply nudge us in the right direction. All we know is: (a) There is a high input impedance that we want to resolve with the first stage, (b) Due to (a), the gain will be fairly small compared to stage 2 where stage 1 will roughly contribute only 1.5 V/V, (c) we are willing to suggest that 50mV (50% of Vin) will drop across the 200k.

    Any thoughts on how to approach the resistor values? I feel like we're almost at the point where all the values will fall into place, but we are missing that important piece of information that will allow us to really pick up speed.

    [​IMG]
     
  2. crutschow

    Expert

    Mar 14, 2008
    12,977
    3,221
    A gain of 15 can readily be obtained with one stage. Why are you using three? :confused:
     
  3. Blomqvist

    Thread Starter New Member

    Jun 10, 2015
    5
    0
    Part of the original design specifications laid out for the project required that the amplifier be realized in three stages. Additionally, given that a non-inverting gain is sought, at least two common-emitter configurations are required to bring the output back into phase with the input.
     
  4. Bordodynov

    Active Member

    May 20, 2015
    637
    188
    That scheme, which is superior in all respects under consideration. This greater stability gain, high input impedance, lower noise, lower output impedance and less distortion.
    Excessive gain can be reduced by using a common, negative feedback. That's exactly what I did.
     
  5. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,984
    3,719
    You should use a design with feedback to control gain. A differential amp can be a great first stage, just like most op amps have.

    This allows you to get feedback from your last stage to insure that your output is a very good amplified version of your input. Without feedback, your output is very dependent on gain of individual transistors which vary significantly lot to lot and as temperature changes.

    Finally, an AB output (push/pull output) is much more efficient than a class A amplifier.

    Here is a differential amp. One input is your signal (+input), one is feedback (- input) that is connected to the final output stage - connected with a resistor. This resistor can be used to set gain as well. Only one output of the differential amp is used (just below the + input transistor).

    image.jpg
     
  6. Blomqvist

    Thread Starter New Member

    Jun 10, 2015
    5
    0
    I like what you have suggested, but as a student just finishing a class on electronics, I feel that I lack a particular understanding that would allow me to utilize a differential stage in this project effectively. For instance, with this approach, I am not sure what a sufficient second stage would be. The original idea was to use amplifiers studied in class (basic Common-E, CC, CB amplifiers) to build an amplifier that could realize a specific gain. I want to use a differential amplifier, but I don't know where I would start. Especially since I don't know the load on each stage.

    Is there a qualitative approach to analyzing stages without getting into the details of component selection? In other words, how can I look at a stage and know that there are particular qualities I would want or expect?

    Should I start at my output stage and work down or vice-versa?
     
  7. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,984
    3,719

    Close. Go two steps beyond that.
    Generally you start by deciding how powerful you want the circuit to be (1 watt, 3 watts, 10 watts, ...).
    Then decide which speakers you want to use (pick a power rating that meets your needs. Most come in 2, 4, 8 or 16 ohms - higher for headphones). 8 is the most common.

    Then calculate how much voltage your power supply will have to be if you use an AB output stage. You can do the math based on ohm's law.
    Or, even higher voltage needed if you select a Class A output stage. Here you will need ohms law and info on the maximum efficiency of a Class A amplifier.

    See, you are half way there.

    As for the amp design, the Differential amp generally connects to a class A mid-stage. That class A has one NPN transistor with a resistor and two diodes in series connecting the positive power supply to the collector of the NPN. The resistor is connected directly to the postive rail and the two diodes between the resistor and collector. The emitter is simply connected to ground.

    then, the push-pull is easy. the bases of each connect to the nodes between diode and resistor (NPN) and diode and collector (PNP) of the previous stage. Done.

    Well, almost done. Now you need to select the right size resistors and put a last resistor back from your output (push pull) to the - Input of the differential amplifier.

    Something like this - from Randy Slone...

    [​IMG]
     
    Last edited: Dec 3, 2015
  8. Bordodynov

    Active Member

    May 20, 2015
    637
    188
    Dear interlocutors you forget that it's not about sound and infrasound (1 Hz, 15 Hz). If you look at my chart, you'll see the output of the my amplifier a great capacity. Although it is not the Problem. There supercapacitors. Another argument in favor of the differential stage. Then, using a bipolar power supply (+ V and -V) can eliminate the output capacitor.
     
  9. Bordodynov

    Active Member

    May 20, 2015
    637
    188
    See
    Ampl15_diff.png
     
  10. dl324

    Distinguished Member

    Mar 30, 2015
    3,211
    619
    This looks like homework to me, and the OP only asked for a nudge...
     
  11. Blomqvist

    Thread Starter New Member

    Jun 10, 2015
    5
    0
    Nope. Not homework.

    It's a final project for a class as specified in the original post. If you read again, I even mentioned that I'm not looking for answers, but general hints that can get me through projects like these. If it was homework, are you suggesting perhaps one should not seek help on forums like these?

    To all that have helped,

    I appreciate your work, but the amplifier was never meant to compete with an operational amplifier. The Three Stage Amplifier for this project is designed for one task only. To provide a gain of 15 V/V to a resistive load modeling an audio indicator. The general idea was that we would amplify a heart beat pulse such that an audio indicator would beep on every up-beat.

    Introducing a differential amplifier and an AB class output stage made this project move away from what my group intended on in the first place. I have mentioned that this project is for my Electronics class. Ergo, the project was meant to reflect what we have learned -- not configurations that we have not covered.

    I have not messaged in a few days due to exams, but I'm closing this thread. Given from the helpful responses, my group and I have come up with solutions to the original project proposal. Thanks!
     
  12. dl324

    Distinguished Member

    Mar 30, 2015
    3,211
    619
    Your distinction between homework and class work is merely a matter of semantics. You're asking for help on something that you will receive a grade for. I don't think your instructor would appreciate giving you credit for work you didn't do.

    My comment was intended for those who were trying to redesign your project and/or feed you answers; neither of which will help you learn as much as helping you do it yourself, nudging as you put it. Posting in the Projects Forum instead of Homework Help is part of the problem. You consider this a project; I consider it homework. It's something you're doing because you have to, not because you want to... To your credit, you stated very clearly that it was a final project; but some apparently miss that point...
     
    Last edited: Dec 5, 2015
  13. Blomqvist

    Thread Starter New Member

    Jun 10, 2015
    5
    0
    You are correct, final projects for a class tend to have a grade attached to them. This does not mean I cannot ask for advice on a process I have never done. I can look for guidance just like anyone in their respective field would do. I have mentioned this thread to my professor so I am in no way taking credit for other's effort. Just like for any type of research, citations are expected.

    I fear that this thread went out of control with ideas that did not pertain to my project and hence did not directly help me, which is why I'm closing it.

    ...Can I do that? Can't find a button.
     
  14. dl324

    Distinguished Member

    Mar 30, 2015
    3,211
    619
    My hat is off to you. You have more integrity than many who come here for help on homework/projects. Most responders can tell the difference and guide Posters to solutions, which most will freely admit can be extremely tedious, instead of simply providing answers.
     
  15. k7elp60

    Senior Member

    Nov 4, 2008
    478
    69
    Some simple things may help you with your current design and schematic. For all practical uses the gain of a stage like Q1 or Q2 is the division of the collector resistor value of R3/(the value of an un bypassed resistor in series with R4) The first thing to do is determine a collector current.
    The choose a R3's value so at the collector current times R3 will make the collector or Q1 at near 1/2 the supply voltage. With 5mA of collector voltage and a 820 ohm resistor the collector will 9V-4.1V or 4.9V. The choose a resistor in series with R4 for gain. Say 3.5, so the resistor in series with R3/3.5=234 ohms....closest standard value is220 ohms.This call the 220 ohm resistor R4A. Next determine the total emitter to ground voltage. We know that the voltage drop acrosss R4A is 1.1V. Lets use a 100 ohm resistor for R4, it will drop 0.5V. As a result the base has to 1.6V +0.6V or 2.2V. Using the minimum beta of the 2N3904, divide the collector current 5mA by beta. Then multiply this current by 10. That is the current flow thru R2. The current thru R1 will be that value + the base current. R1 and R2 in parallel is very close to the input impedance of that stage.
    I hope this helps.
     
Loading...