Colpitts oscillator not working as expected

Discussion in 'General Electronics Chat' started by Robin Mitchell, Nov 20, 2014.

  1. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    Hi all,

    I have designed a simple Colpitts oscillator as shown below:
    [​IMG]
    In LTSpice it works great! I have built this in real life with all values the same. The only difference is that the transistor used is a BC548 instead of what is shown (This is because I know that the 2N3904 and the BC548 are interchangable).

    When I power the circuit I DO get an oscillation however it is very small in amplitude (less than 1vPP) and there is a large DC offset (about half the supply typically).

    What is going on?
    It should be noted that the circuit is built on breadboard so will parasitic capacitance affect anything?

    All the best,
    Robin
     
  2. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    Yes, especially at those frequencies. Breadboards are horrible above about 100kHz...
     
  3. bertus

    Administrator

    Apr 5, 2008
    15,645
    2,344
    Hello,

    The coil L1 is extremely small, try a value of about 100 X larger eq. 1 μH.
    The capacitors C2 and C3 can be reduced to about 200 pF.

    Also a breadboard will have parasitic capacitance.

    Bertus
     
  4. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,773
    1,103
    Take the output via a cap if you want to get rid of the DC offset.
     
  5. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    BR-549, Robin Mitchell and atferrari like this.
  6. atferrari

    AAC Fanatic!

    Jan 6, 2004
    2,645
    759
  7. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
  8. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    I think it was 1 nF capacitors and 5nH coil. Coil was made by wrapping standard 22 gauge insulated solid hookup wire around a pencil 6 or 7 times (maybe it was 9 times?). There are calculators on line. I was able to pick up the silence on an fm radio. Then stretch the coil a bit (just spring-like, not deforming permanently) so you can hear the fm receiver catch and lose the signal. The signal was quite strong - up to 2x supply voltage.
     
    Robin Mitchell likes this.
  9. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    I'll rebuild it. I really should have a notebook.
     
    Robin Mitchell likes this.
  10. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    So I built it on stripboard, breadboard and LTSpice. None of them oscillate. Instead they have a decayed oscillation that dies out after 45uS
     
  11. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    I think you need a smaller capacitor connecting the tank to the emitter if that happens (from memory).
     
  12. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    100nf tank capacitors and 50nF inductor

    Also, the load on the output must be very low or you will lose your signal.

    My output was the 10M ohm cap (I need to measure voltage across a component on my ipad simulator - no meter available).

    The output across the 10M resistor is 6.5 v peak to peak. 100 MHz.

    Also, you should expect 20 to 30 pF across each tank cap from board capacitance so use 68 or 82pF caps for prototyping.

    Coil could be 5 turns, 0.25" about 0.7" long.
    Calculate your own coil here. This site will post as 5.00E-002 microhenries
    http://www.66pacific.com/calculators/coil_calc.aspx


    Cheers

    [​IMG]
     
    Last edited: Nov 22, 2014
  13. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    Ok so I built this too. Nothing.

    I have used the BC548 as the active component (and tried multiple devices), and I am using a 20MHz oscilloscope (which is analoge).
    I used all the values you suggested but for some reason there is no oscillation :(
     
    Last edited: Nov 23, 2014
  14. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    You will not see a 100 MHz signal on a 20MHz scope.
     
  15. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    I know that but its analog so would the signal just look like a block? And how would I get for example a 1MHz sine wave from the circuit you showed?

    Many thanks :)
    Robin
     
  16. GopherT

    AAC Fanatic!

    Nov 23, 2012
    5,991
    3,736
    image.jpg Increase the size of the components in the tank.
    The must stay somewhat "balanced" (as you go up in size). Not linear but make sure your simulator sows a signal just above your supply voltage. Also, not all simulators will work with self-oscillating circuits without an initials perturbation (enter a starting value).

    Here is a simulation of a 1MHz oscillator. Note that the slower you go, the more difficult it is to keep it going. Use the fattest wire (lowest resistance) coils possible and lowest ESR capacitors possible since the simulators do not have resistance in the ideal components.
     
Loading...