DC coupled multistage amplifier questions

Discussion in 'The Projects Forum' started by UTEEstudent, Mar 10, 2009.

  1. UTEEstudent

    Thread Starter New Member

    Mar 10, 2009
    4
    0
    I have a project to build a multistage DC coupled amplifier with a gain of at least 100, an input resistance of 1MΩ, and Rout of 100Ω.

    For MOSFET's we have to use the ALD1106 or ALD1107 and for BJT's we have to use the 2n3904 or 2n3906.

    We can use as many stages as we want. β is approximated to be between 150-250, and Vtn = 0.7V.

    I'm trying to use a 2 BJT setup with the first stage being the npn and the second stage being a pnp.

    The input signal is a 20 - 50mV sine wave.

    My main questions are:
    1. Can this type of circuit create at least Av = 100 or do I need another stage?
    2. How in the world do I even begin analysing it? I need to figure out how to approach the problem so I figure out what kind of resistors to buy.
    3. Does anyone know of a good PSpice website for help? I have the circuit built but don't have any idea how to edit parameters or get graphs.

    This is my first post/visit here so I'm not sure how the site works quite yet, but my circuit is essentially the same as the one linked here (the cascaded one):
    http://forum.allaboutcircuits.com/showthread.php?t=5840&highlight=multistage+amp

    This is basically the design I sketched out with my R8 being 100Ω and my R1 being 1MegΩ. My capacitor values are going to be 22μF just because my profesor mentioned it in class. Other than that I am paralysed for the next step, and the project due date is approaching fairly rapidly, so I reach out to you all for a push of guidance.

    The book we are using has a one paragraph description of what dc coupled means, and that's it. No instuctions on exceptions you have to consider when direct coupling or anything like that. It does, however, have a fair amount of detail on ac coupled amplifiers.

    Thanks.
     
  2. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    With a load of 100 ohms then the PNP transistor will have a very high quiescent current and a low input impedance that will load down the NPN transistor.
    Try the PNP transistor as an emitter-follower and have all the gain from the NPN transistor. But it will be very distorted at high output levels (less than clipping levels).

    If R1 is 1M then the medium input impedance of R2 and R3 and the base of the NPN transistor will be an attenuator so the gain will be much less than 100. Bootstrapping of the input of the NPN transistor will increase its input impedance.
     
  3. UTEEstudent

    Thread Starter New Member

    Mar 10, 2009
    4
    0
    I forgot to mention that there is no criteria on distortion. He just wants to see the gain at the output.

    From your response I'm guessing the gain of 100 isn't going to happen with the current setup? Per your suggestin, I know the gain of the common collector will only provide unity gain, so the first stage would be all of my gain.

    Would removing R2 and R3 provide a much higher gain or should I just add another front stage to the circuit? Most examples I see in my book for AC coulped amps have a mos common sourse, then common emmiter, then common collector output stage.

    Back to the 2 Bjt setup here:
    If the gain of the first stage is Avt = -gmRL/(1+gmRe) where gm=40Ic, and I want to get a gain of Avt = 150 (just for some breathing room), I did some algebra and came up with the expression:
    gm(150Re + RL) = -150

    Would RL just be the input resistance of the common collector bjt? Working under this assumption I came up with this equation after substituting for gm and RL:
    Ic(6,000Re + 804,100) = -350

    Assuming this is actually right does anyone know of a 2nd equation so I can solve for these values?
     
  4. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Try this procedure.

    max input = 50 mv
    max output = 5 volts.

    If these are RMS values? then output = 2*5*1.4 peak to peak = 14 volts.
    If these are peak to peak then output = 5 volts peak to peak.

    Now consider both the power supply and the ouput transistor biasing to acheive this, driving into 100 ohms.

    Then work backwards to the input seeing what drive you need to supply the output and the next transistor back etc.

    Post again if you need more help with this.
     
  5. UTEEstudent

    Thread Starter New Member

    Mar 10, 2009
    4
    0
    I'm going to assume it's a peak to peak value for the gain because I don't remember him stating which to go with.

    I followed you up until you mentioned the transistor biasing and the power supply driving into the 100Ω load.

    Are you saying to analyize the pnp bjt with it's Rout as 100Ω, use the 5V value as the output voltage, and use this to find the Q point?

    What's giving me the most problem with this assignment is figuring out the collector and emitter resistors; I feel clueless on how to get started on them.
     
  6. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Do you have access to a copy of "The Art of Electronics" by Horowitz and Hill? It's a rather expensive book, similar price to a college textbook at a college bookstore, even when ordered from a major online bookseller like BN or Amazon.

    It devotes a couple chapters to the exact questions you are asking, as well as many other aspects of electronics. Might check your local library for a copy. I remember somebody mentioning they had it in PDF format, but I haven't found/seen it.

    As Studiot said, the first thing to decide, as output impedance and power are known, is the required supply voltage and current ability, then work biasing and gain stages "backwards" from the load to the input level.

    Current sources are your friend. "Static voltage divider" biasing often shown with AC Coupled amplifiers will not work. A DC Coupled amplifier from scratch is a tough assignment with reference materials. Don't feel confused!

    Here is a schematic/description for an unrelated DC coupled amp. It meets exactly zero of your requirements as far as parts go. Although it should give you an idea on how biasing works, they explain the function/schematic fairly well.
     
  7. studiot

    AAC Fanatic!

    Nov 9, 2007
    5,005
    513
    Bootstrapping is one method of creating high input impedance. Here is another.
     
  8. UTEEstudent

    Thread Starter New Member

    Mar 10, 2009
    4
    0
    I think I have the output stage figured out. I changed my setup from what I have posted previously to a common source (stage 1), common emitter (stage 2), common collector stage setup.

    I have found all the values for the final stage except the early voltage, VC, VCE, and ro (I think).

    How do I find these values assuming all my other values are correct?

    More importantly, I don't see how in the world I'm suppose to analyze the second stage. The only information I can get out of it is that it's my gain stage so I set the gain equation equal to 200, but the equation calls for multiple values I don't have yet (resistor values and transconductance).

    I also know that the node at the collector has a voltage of 7.7V and 464μA via the output stage's base current.

    Any help on this part and the eventual common source MOS input stage, or suggestions how I could reconfigure my transistor setup would be appreciated.
     
Loading...