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

Designing and setting the operating point of a single tuned transistor amp

  1. May 30, 2009 #1
    Hello everyone, I'm just getting started in studying radio frequency design and have a question about amplifiers with a tuned circuit as the load. When setting the operating point of a standard class A transistor amplifier with a resistive load, one uses the collector resistor to set the current so that the transistor collector sits at about 1/2 Vcc.
    My question is how one would go about setting the operating point when the load is a parallel tuned circuit consisting of a capacitor, inductor, and resistor. I know that the gain is going to be -gmZ where Z is the parallel combination of the resistance and reactance in the collector circuit, including the input impedance of the following stage. I also know that at the resonant frequency the collector load will appear basically resistive and the gain will be at maximum. However, I'm confused about the DC operating point - since the inductor is a short at DC as far as DC is concerned the collector will be sitting at about Vcc. Since the transistor load at DC is essentially a short, do you bias the transistor the way one would set up the bias on an emitter follower? Do you calculate the impedance of the collector load at resonance to put the collector at 1/2 Vcc for AC, or is this amplifier essentially operating class C and it doesn't matter? Any input would be appreciated.
     
  2. jcsd
  3. May 30, 2009 #2
    If you can get your hands on the February and March issues of QST magazine I think you would find it interesting. There is a two part article on designing and building transistor linear amplifiers.
     
  4. May 30, 2009 #3
    Assuming you have an NPN bipolar transistor output with a grounded emitter, do you have room for a 3/4 volt resistor IR drop between the emitter and ground? If so, take another NPN, tie the emitter to ground, the base to the emitter of the output transistor, and the collector to the base of the output transistor, Select the resistor value for a 3/4 volt drop at the desired current. This will regulate the output collector dc current. You might also bypass the emitter resistor with a capacitor(s) to reduce the RF impedance.
     
  5. May 30, 2009 #4

    vk6kro

    User Avatar
    Science Advisor

    You just set up the amplifier with forward bias to obtain a "reasonable" idling collector current.
    If it is to be class A, the whole input signal must produce changes in the collector current.
    The important thing is that the bias to achieve this must be very stable.
    Since transistors vary, this bias is normally adjusted after construction.

    Without an input signal, yes, the voltage on the collector will be the same as the supply voltage.

    As soon as you start varying the collector current, a voltage is developed across the tuned circuit if it is at the right frequency.

    It is the same with the transistor having a transformer load. The change in current through the primary causes a voltage to be developed across the primary and secondary of the transformer.

    Incidentally, you would never put a parallel tuned circuit in the collector of a transistor. The transistor would act as a low impedance across the tuned circuit and make its tuning very broad. The collector current should go through only part of the coil to minimise this effect.
     
  6. May 31, 2009 #5
    Thanks so much everyone for your replies - I plan on checking out those issues of QST. I'm not sure I completely understand VK6KRO's statement that the low impedance of a parallel tuned circuit as the collector load for a transistor would cause the tuning to be too broad - since the transistor acts like a current source isn't the collector impedance very high for AC? I think I have seen transformer coupled parallel tuned circuits as the load for transistors in cheap AM radios, is something different if they're transformer coupled? Or is the circuit in the collector load not parallel tuned at all? :confused:
     
  7. May 31, 2009 #6
    From an AC perspective V+ is at ground potential because there is a large capacitance from V+ to ground. This means that the collector - emitter is in parallel with the tuned circuit. Fortunately in most cases the operating bandwidth of the amplifier must be wide enough that that isn't a problem.

    The type of bias depends on the frequency of operation and the temperature stability you need. I used to design amplifiers at 900 MHz and the transistors for that frequency required that the emitter be grounded. We biased the linear stages by putting a 100 ohm resistor between the parallel tuned elements and V+. The junction between the 100 ohm resistor and the tuned elements was AC grounded by means of a capacitor series resonant at 900 MHz. Since our amplifiers needed a wide temperature stability, for bias we used a PNP with the emitter connected to the junction, the base through a resistor to V+ and the collector through a voltage divider to ground to the base of the RF transistor.
     
  8. May 31, 2009 #7

    vk6kro

    User Avatar
    Science Advisor

    The output admittance of most transistors is given in the data sheets.
    It is typically 100 uSiemens which is 10000 ohms. I've seen it as low as 1000 ohms.
    If you did want broad band response and low gain this would be OK.
    Putting a 10K resistor across a parallel tuned circuit would never be done otherwise.

    However, if you look at any transistor radio circuit (when they still used transistors !) you will see that they always tapped the transistor's collector down on the coil of the load tuned circuit. This has the effect of reducing the damping on the tuned circuit and also giving a voltage step-up due to auto-transformer action.
    Transistor radios always need as much gain and selectivity as they can get.

    Fortunately, transistors can drive a low impedance load very well. This way their voltage gain across the transistor is kept low and this helps with stability. The voltage gain is partly achieved by transformer action in the tuned circuit.

    FET RF amplifiers only work properly if you give them a low impedance load and have voltage stepup in the coupling circuitry. It is normal to see a parallel tuned circuit in the gate circuit of the next stage and just a low impedance link of a few turns of wire coupled to the tuned circuit from the drain of the previous stage.
    This is not a bad thing. I have built a two stage FET amplifier with a voltage gain of 10000 and most of that came from step-up action in the tuned circuits.

    There used to be a conception that FETs were just llike valves. They do have high input impedance, but their output is quite different and they have to have a low impedance load or they become unstable, just like bipolar transistors do.
     
  9. May 31, 2009 #8
    Ah, things are starting to make sense now. So if I wanted to imagine an analytical model of a transistor with a tuned LC load it would be something like a voltage controlled current source in parallel with the transistor collector impedance, the LC, and the input impedance of the following stage, neglecting for now the transistor capacitances and any coupling capacitance (which I'm sure for radio frequencies can be significant). {art of the problem was I wasn't thinking of the supply rail as being an AC ground. I can see how having the transistor collector impedance of 10k ohms in parallel with the LC would ruin the Q factor of the circuit - this must be part of why pentodes were used in the RF amps of tube superheterodyne radios; they had high plate impedance and didn't load down the circuit as much as a triode would.

    One further question for vk6kro: could you elaborate on why BJTs and FETs become unstable with high impedance loads? I remember reading somewhere that one of the reasons early transistor audio amplifiers sounded poor compared to tube amplification was that transistors couldn't provide as much gain per stage and thus required many gain stages and large amounts of negative feedback, but never read an adequate explanation as to why this was so. Is this related to their inability to handle high impedance loads? Thanks again everyone for your time.
     
    Last edited: May 31, 2009
  10. May 31, 2009 #9

    vk6kro

    User Avatar
    Science Advisor

    That is correct. The equivalent circuit is exactly as you describe.

    FETs have GMs of up to 30 mA per volt (much higher than almost any valve) but more drain to gate capacity than most triode valves.
    So, they are fundamentally unstable with a high impedance load.

    (GM is the change in drain current for a 1 volt change in gate voltage)


    WITH a low impedance load, they are the best amplifiers around. You drive a tuned circuit with a low impedance link and get a big step-up in voltage to drive the high impedance input of the next stage. Great stuff.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Designing and setting the operating point of a single tuned transistor amp
  1. OP-amp and transistor (Replies: 2)

Loading...