Colpitts Oscillator

Discussion in 'General Electronics Chat' started by sudip, Mar 15, 2010.

  1. sudip

    Thread Starter New Member

    Mar 15, 2010
    Hello Experts!

    I've tried to design a Colpitts oscillator but it's not meeting the Barkhausen criterion and so failed to oscillate. Please take a look into my netlist.

    *******NETLIST using Orcad PSpice********
    X_Q1 N06431 N00139 N02241 awb2n2222 PARAMS:
    R_R1 N06431 N00127 50K TC=0,0
    R_R3 0 N00139 100M TC=0,0
    R_R4 N00139 N00127 100M TC=0,0
    L_L1 N00293 0 0.15e-6
    C_C2 N02241 N00293 0.01e-9 IC=7 TC=0,0
    C_C1 N00293 N00139 0.1e-6 IC=7 TC=0,0
    C_C3 0 N02241 100p IC=7 TC=0,0
    V_V_supply N00127 0
    +PWL 0 0 1n 5
    R_R5 0 N02241 100 TC=0,0

    I've also specified an initial condition for all capacitors =7 and made these changes into simulation profile RELTOL=0.0001 and ABSTOL=0.001p;

    Can you please help me out to identify the problem?

    Thanks in advance,
  2. bertus


    Apr 5, 2008

    Please post the pictures here.
    You can use the manage attachments button to do so.

  3. Audioguru

    New Member

    Dec 20, 2007
    RapidShare is not rapid, it is asleep. I gave up waiting for it to wake up and give me the schematic.
    If the schematic is posted here then it is seen immediately.

    EDIT: Horray. RapidShare woke up.
    The 100M base bias resistors have values that are about 10,000 times too high so the transistor is turned off as shown.
    Also, the ratio of the collector resistor and emitter resistor is too high so there is not enough positive feedback.
  4. kkazem

    Active Member

    Jul 23, 2009

    It looks to me like you've got several component values way too far off to work. The only reason it works at all in the beginning is due to your initial conditions, which shouldn't be needed. First of all, the emitter resistor needs to be much larger than the collector resistor. Next, your bias resistors of 100 MEG each are way too big. Your tapped capacitors that go to the emitter where they connect to each other should be the same value (i.e.: 100p/100p and not 10p/100p). There is absolutely no need for an initial charge of 7 volts on all the caps, these initial charges should be left out. Please refer to my SPICE schematic (attached as a pdf), and the two plots (1 & 2). Plot-1 shows the start-up from the instant that the 5V is applied, and Plot-2 shows a highly zoomed-in area of the steady-state and sinusoidal feedback and bias nodes. Admittedly, the actual output has some distortion in it, but then again, I made no attempt to optimize the circuit. Please feel free to reply with any other questions. Just to let you know, I used LTSPICE-IV, a free SPICE simulator with schematic capture and post simulation graphics (as you can see from the plots). It's better than the ORCAD PSPICE, which costs a lot of money, except for the student version that has very limited nodes (25, I think). The LTSPICE has unlimited nodes (actually, limited only by the amount of memory in your PC). Good luck.
    Kamran Kazem
  5. sudip

    Thread Starter New Member

    Mar 15, 2010
    Thank you all for such quick replies.

    Hello Kamran,

    Special thanks to you for your detail look into my oscillator. I have tried with new values and without any initial conditions. I'm uploading the simulation result, it looks not good :(. Is there any external circuitry that I can add to increase it's gain? and so α.A>1; where α=C2/C3 (in my schematic) and A=V_out/V_feedback. Difference between two simulation results (LTSpice and PSpice) shows there is major gap between these two simulators.

  6. flat5

    Active Member

    Nov 13, 2008
    kkazem's circuit as a gif file.