Colpitts Oscillator Problem

Discussion in 'The Projects Forum' started by Bill B, Apr 22, 2015.

  1. Bill B

    Thread Starter Active Member

    Nov 29, 2009
    Howdy all,

    I have a project in which I need to generate 4 different, very stable, frequencies. The amplitude of these signals needs only to be on the order of 10s of mV. The frequencies is need to generate are 150 kHz, 500 kHz, 1.4 MHz, and 30 MHz. I plan to build a separate circuit for each frequency. I was hoping I could use a single design for each circuit. I selected the circuit below and simulated it. It worked fine in simulation, but I get no output after building the circuit on a project board. This design is for 500 kHz. Does this design not work for low frequency applications?
  2. AnalogKid

    Distinguished Member

    Aug 1, 2013
    A few questions:
    What is the application?
    What is the required frequency tolerance for each frequency?
    What is the required amplitude tolerance for each frequency?
    Are sine waves required?

    As I'm sure others will point out, you can derive all 4 frequencies from a single crystal oscillator with some logic. Not exactly analog design, but way stable.

  3. Bill B

    Thread Starter Active Member

    Nov 29, 2009
    This is a device to verify the operation of a discontinuous interference analyzer, or click analyzer. Click is a CISPR 14 conducted emissions test. The analyzer monitors the 4 frequencies I listed above. I guess I should provide a little background on what I'm doing.

    At this point, I have a working prototype that accomplishes what I need with the aid of a DC power supply and function generator. Depending on how the circuit is switched, it produces a pulse of known duration (C<10ms, 10ms<C<20ms, and 20ms<C<200ms) at the 4 frequencies, at set intervals (20 sec). The idea is to stimulate the analyzer with these frequencies at amplitudes high enough to violate the limits as stated in the CISPR 14 standard. So, for example, a 10 minute test duration with pulses every 20 seconds for a period C<10ms at 150 kHz, the result should be 30 short clicks in the 150 kHz channel. This is simply for verification that the system is functional and counting correctly. Its not for calibration.

    I would like to eliminate the use of the external DC power supply and function generator and have a self contained device that I can simply plug in and run. At this point, obviously, I am trying to produce the test signals. I'm not opposed to using logic, but I would prefer to avoid the harmonic content that comes with it. Also, I would like to generate all 4 frequencies simultaneously to cut down on test time.

    I haven't taken the time to measure the bandwidth of each filter, but I know it is pretty tight, so the frequency tolerance will be tight. I'm sorry I don't have a better answer for you there. The amplitude tolerance isn't significant. I will use a pot driven voltage divider to control the amplitudes as I only need them to be in the 30 - 50 mV range.
  4. DickCappels


    Aug 21, 2008
    It might be that it is wired incorrectly or that your inductor has too low of a Q to allow oscillation.

    As an aside, it is a good idea to add decoupling capacitors to the power supply connection.
  5. Bill B

    Thread Starter Active Member

    Nov 29, 2009
    Thanks for tip Dick Cappels. I got the thing working for the 1.4 MHz oscillator. It was a matter of changing the emitter resistor from 2.2 kOhms to 1 kOhms to allow higher gain. I also took the output off of the emitter. I have a nice clean sine wave at 1.4 MHz and 90 mV. Perfect.