FM output looks like AM

Discussion in 'Wireless & RF Design' started by fasti, Dec 27, 2011.

  1. fasti

    Thread Starter New Member

    Dec 27, 2011
    4
    0
    Hi everybody,

    I want to learn about FM circuits from simulation applications (like Spice)
    so I installed "QUCS" (because I have Ubuntu-Linux at home,
    and Qucs is native on linux, and also it's free.. :D).

    My problem is that I simulated a basic FM transmitter circuit,
    and the output looks like AM. (i.e. the carrier only changes in amplitude
    and not in frequency).

    In the screenshot attached, I'm showing the circuit diagram with
    the components values, and the 4 graphs show the input to the transistor
    and the output from the oscillator (after removing the DC with the last cap)
    in 2 time scales (0 to 0.5 uSec, and 0 to 10 uSec).
    The modulating input is a 10 MHz signal and the carrier is 100 MHz.
    They both appear in the graphs with correct frequencies,
    but the modulation looks like AM, not FM... (not even AM really)

    DoesntLookLikeFM.png

    Is the FM transmitter circuit wrong (I copied it from some website)
    or is it the application that's not simulating it right?
    (I tried changing the values of all the elements and even different
    BJT models, but it didn't make an FM shaped output).

    And also, why is the output so weak?
    it's less than 10mV p-p... Is it supposed to be that small?

    I'm beggining to feel like I don't really understand
    why the transistor is supposed to modulate the frequency
    and not the amplitude...

    Thanx a lot for any explanations,

    fasti
     
  2. bertus

    Administrator

    Apr 5, 2008
    15,638
    2,344
    Hello,

    The source U is DC coupled, use a capacitor between source and the base of the transistor, like 1 nF.
    Also the capacitor from base to ground (10 nF) will almost short the signal from the source U, make it MUCH smaller, like 100 pF and try again.

    Bertus
     
  3. fasti

    Thread Starter New Member

    Dec 27, 2011
    4
    0
    Hi Bertus!

    Thanks a lot for the fast response!

    I did both of the adjustments you suggested, and the amplitude problems
    disappeared, but I still don't see any frequency modulation.

    I attached another screenshot that shows a time interval of 0.2 uSec
    (between 5 uSec and 5.2 uSec, when the output is steady)
    and the frequency doesn't seem modulated (it looks stable).
    (I tried different modulation frequencies, and they all give the same output).

    attachment=38209

    Any other thoughts?

    Thanx again,
    fasti
     
  4. bertus

    Administrator

    Apr 5, 2008
    15,638
    2,344
    Hello,

    You are trying to modulate a 100 Mhz signal with a 10 Mhz signal.
    You will not see the modulation as modulation frequency is to high.
    Try to simulate with a 10 khz modulation signal and let it run over the complete cycle of the modulation signal.

    There is a little project for an FM modulated transmitter in this thread:
    http://forum.allaboutcircuits.com/showthread.php?t=7130

    Bertus
     
  5. fasti

    Thread Starter New Member

    Dec 27, 2011
    4
    0
    Simulating 10 KHz modulation was for some reason tremendously slow on my pc,
    so I took the modulation signal down from 10 MHz to 500 KHz,
    and still, no FM to be seen.

    Here's a screenshot:
    attachment=38210

    Also, as you can see, the amplitude problems came back (look at the
    large scale graph on the top right).
    Enlarging the capacitor from the transistor base to the ground
    to 100 nF (big, but similar to the circuit in the project you linked to)
    fixes the amplitude changes, and didn't effect the frequency.

    There really is no frequency modulation after the circuit stablizes:
    attachment=38211 (screenshot with 200nF from the transistor's base).
    It should look like this! (link)

    What are we missing?
    Is it the circuit that's wrong or the application?
    What could be wrong with the circuit?

    I really do appreciate the help. Thanx again, fasti
     
  6. bertus

    Administrator

    Apr 5, 2008
    15,638
    2,344
    Hello,

    Try to reduce the modulation amplitude.(in the range from 10 to 100 mV).
    With 1 Volt as modulation amplitude the bias of the transistor will be influenced, resulting in amplitude changes at the output.

    Bertus
     
  7. crutschow

    Expert

    Mar 14, 2008
    12,977
    3,221
    It may be that the simulation transistor model does not properly model the transistor change in collector capacitance with bias, which I believe is what changes the oscillation frequency and thus generates the FM.
     
  8. fasti

    Thread Starter New Member

    Dec 27, 2011
    4
    0
    I only managed to reduce the modulation amplitude from ±5V to a little
    less than ±1, by reducing the 47 kOhm from the power supply to the base to 470 Ohm.
    Lower resistances don't reduce the modulation amplitude more.
    What else can I do to reduce it?

    And I still don't see a difference in the output.
    (I kept the base capacitor at 200nF because otherwise the AM comes back,
    even with the reduced modulation amplitude).
    Here's a screenshot of the reduced modulation amplitude: (looks the same)
    attachment=38232

    Crutschow, thanx for your response!
    I looked in the model that the simulator has for the transistor and it looks
    quite specified (link to a screenshot of the simulator's BC549 model).
    (It's even more specified than the BC549 datasheet (link) :)).
    Specifically, the "base-collector zero bias depletion capacitance" and
    the "forward bias depletion capacitance coefficient" which both have values in the model.
    So it seems to have the collector capacitance values, I'm just not sure how
    the capacitance changes with bias according to their model.

    I noticed the parameter "zero bias collector-substrate capacitance" is set to 0,
    so I tried changing it, but it didn't effect the modulation.

    I also tried changing the transistor to different models - 2N2905 and
    2N2222, which are both fitting for this circuit (in reality) and very popular,
    but the output looks roughly the same.

    Do you think the problem has to be with the simulation program?
    Does the circuit (diagram and values) look like it should work?

    I'll try simulating other designs for FM transmitters and see if
    the output looks like FM.

    I still appreciate any ideas anyone has to offer,
    and also any theoretical explanations for the purpose of each of the elements
    in an FM transmitter circuit (the capacitors and resistors before and after the BJT, not the LC tank.
    Here's information about the LC tank for those of you who are interested:
    http://en.wikipedia.org/wiki/LC_circuit
    http://wiki.4hv.org/index.php/LC_Oscillator
    and one with a nice animation: http://www.greenandwhite.net/~chbut/lc_oscillator.htm).
    It's difficult to find information about the role that all the componants have,
    so I'm sure the readers of the forum would appreciate any resources you can share
    or any explanations you can give us.

    Thanks to everyone.
    fasti
     
  9. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    You will never see FM on a 'scope because the frequency deviation of 75kHz peak max is so small. On a 98MHz carrier the 75kHz (0.075MHz) is only 0.00077.
     
  10. xylon89del

    New Member

    Dec 28, 2011
    17
    0
    Hi,

    it looks to me that your circuit is just a normal LC tank tuned amplifier.

    For FM, you need a voltage controlled oscillator. normally there is a varactor in the VCO circuit.
     
  11. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Please post your QUCS project as a tarball (tar.gz) as an attachment. Or ZIP if board doesn't support .gz files
     
  12. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    I corrected the circuit so that it is the same as hundreds of simple transmitter circuits on the internet.

    The collector to emitter capacitance of the transistor changes when the modulation causes the transistor to conduct more and less. The changing capacitance produces FM.
    Most FM radios ignore the small amount of AM.
     
  13. xylon89del

    New Member

    Dec 28, 2011
    17
    0
    Hi, there is still no sign of varactor , which shud looks something like this
    http://en.wikipedia.org/wiki/Varicap

    Furthermore, if it is oscillator, be it a tuned oscillator or voltage controlled oscillator, there will be a gain(transistor) and a feedback network from the output to the input. The feedback network can be implemented using various configuration, the most popular configuration would be Colpitts, Clapps, Harley and etc...

    I cant see any feedback in the circuit, which means it is not an oscillator at all. Since it is not an oscillator, it is impossible to be voltage-controlled oscillator.

    I am not very familiar with this forum, somehow i try to post a reference link , it always fails. But, you can find this reference "Foundation of oscillator circuit design" in scribbs.
     
    Last edited: Jan 6, 2012
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Didn't you hear me? The collector-emitter capacitance of the oscillator transistor is the varactor. Look at the datasheet for a 2N3904 transistor that shows the capacitance changing when the collector-emitter voltage changes.

    Look at my corrected FM transmitter circuit. It is a modified Colpitts oscillator with the capacitor from the collector to the emitter creating positive feedback. The collector-emitter capacitance of the transistor is changed by the audio signal then the oscillator frequency changes making FM.

    There are hundreds or thousands of these simple FM transmitter circuits on the internet.
     
    • FM.GIF
      FM.GIF
      File size:
      11.6 KB
      Views:
      38
  15. xylon89del

    New Member

    Dec 28, 2011
    17
    0
    ok. may be you have your points too. because I never see this type of configuration in electronics official text book..
     
  16. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Maybe the extremely cheap and poor quality "toy" FM transmitter circuit is not good enough to be discussed in a text book.
     
Loading...