Connect with us

generating sine wave

Discussion in 'Electronic Design' started by pinku, Aug 9, 2006.

Scroll to continue with content
  1. pinku

    pinku Guest

    Is there any simple method to generate a sine wave of frequency 1MHZ by
    using LM324
    op-amp/by using filters /by using oscillators (please mention the
    specifications and schematic diagrams).
    please reply me with in 24 hours(urgent).
     
  2. We don't do 'urgent'. The LM324 isn't suitable for operation at 1 MHz.
    Choose a faster op-amp.
     
  3. Fred Bartoli

    Fred Bartoli Guest


    You're making the same mistake as Bill.

    It sure can be done: (but I wont recommend this at all)

    250p
    ||
    .--------||----------------.
    | || |
    | |
    100uH | |\ |
    ___ | |\ 6V >-|+\ |
    6V >-UUU-+-|+\ ___ | >--+
    L | >-+-----|___|-+-|-/ |
    .-|-/ | 1K | |/ |
    | |/ | | ___ |
    '------| '-|___|-+
    | 1K |
    .-. .-.
    | | | |
    | | 4K7 | | 4K7
    '-' '-'
    | |
    === ===
    GND GND

    For L, say Q>10 @ 1MHz.


    To the OP:
    It'll oscillate but I won't say you why.
    One thing I can tell you though: don't show this to your professor (unless
    he has a good sense of humor).
    What happens if you exchange L and C. Why?
     
  4. In message <44d9d9c0$0$14343$>, dated Wed, 9 Aug
    2006, Fred Bartoli
    No, I'm being realistic. Using the LM324 at 1 MHz is the rankest of bad
    design.
    This is with no DC supplies to the power pins? Evil!
     
  5. Fred Bartoli

    Fred Bartoli Guest


    Sure, but the design requirement to be completed with 24h always make me
    facetious.


    Where did you see power pins?
     
  6. In message <44d9ea62$0$14315$>, dated Wed, 9 Aug
    2006, Fred Bartoli
    They aren't shown, and no supply voltages are mentioned, but 6 V DC is
    applied to both op-amp + inputs. So I wondered if the power goes through
    parasitic diodes or something.
     
  7. Guest

    It isn't a mistake - if the OP asked the question, he isn't in state to
    make any sense of your kind of answer. And he still won''t get more
    than the 160mV out=put swing set by the 0.5V/usec slew rate as worked
    out by TuT - I haven't checked that calculation, but it looks to ba
    about right.
    My guess is that it would still oscillate at the same frequency,
    because you'd still have a 180 degree phase shift bewteeen the voltage
    drop across L and C, not to mention the 180 degrees across the two op
    amps.
     
  8. Tim Williams

    Tim Williams Guest

    If you want to use the LM324 you'll need a negative resistance from the
    parts bin.

    Tim
     
  9. Fred Bartoli

    Fred Bartoli Guest

    <> a écrit dans le message de

    It isn't a mistake - if the OP asked the question, he isn't in state to
    make any sense of your kind of answer. And he still won''t get more
    than the 160mV out=put swing set by the 0.5V/usec slew rate as worked
    out by TuT - I haven't checked that calculation, but it looks to ba
    about right.
    My guess is that it would still oscillate at the same frequency,
    because you'd still have a 180 degree phase shift bewteeen the voltage
    drop across L and C, not to mention the 180 degrees across the two op
    amps.


    No it won't. The frequency would be lower.
    This "oscillator" is less obvious than it first appears.
    A quick analysis should have revealed that the drawn one shouldn't
    oscillate, but...


    For the case of perfect 1MHz GBW opamps:
    1) phase shift isn't 180 degrees. The opamp has a 1MHz GBW, giving -45°
    phase shift for the follower at 1MHz.
    The inverter has a 500kHz loop GBW product, giving 180-atan(1M/500K) = 116°
    phase shift, and a total 71° shift for both amplifiers going asymptotically
    down to 0.

    2) the filters are 2nd order HPF and LPF and phase shift isn't 180°, but
    respectively +90° and -90° at, not resonance, but corner frequency.

    So for the LPF case, total phase shift is 71-90=-19 at the 1MHz corner
    frequency.
    The LPF phase shift ranging from 0 to -180°, there's clearly a solution
    slightly before 1MHz.
    Now loop gain would be: -3db (follower) -7 db (inverter) + 20 db(Q=10) =
    +10dB.
    So it'll oscillate sligthly before 1MHz.

    For the drawn HPF case, the filter phase shift ranges from +180 downto 0.
    For the amplifiers, phase shift ranges from 180 down to 0, and there's no
    solution to the total phase=0 condition. It won't oscillate at all.

    This for perfect 1MHz GBW opamps.


    Now, we use real opamps.
    The opamps are not perfect at all and have extra poles and excess phase
    shift too.
    Using a good LM324 model, at 1MHz the follower shows -90° and the inverter
    +69° for a -22° total (and still -10dB gain).
    Above, phase continues droping at a fast rate.

    Looping this through the HPF, which didn't oscillate with the perfect 1MHz
    GBW opamps, there's now a solution which has to be _above_ the HPF corner
    frequency. A quick AC sim indicates 1.06MHz.

    Looping this through the LPF, which phase shift is always negative, the
    frequency has to drop _below_ the LPF corner frequency to satisfy the
    Barkhausen criteria.

    Hence exchanging L and C will lower the frequency. A quick AC sim shows
    about 770kHz.

    When the amplitude rises from 0 at startup, it'll set where the slew rate
    limiting will make the average gain unity.
    The slew rate limiting introduces additional phase shift which will lower
    the oscillation frequency, down to about 650kHz and 1.03MHz.

    One funny side effect is that if you substitute the LM324 for a better
    opamp, or if the LM324 hasn't enough excess phase shift, the "oscillator" as
    drawn won't oscillate.
     
  10. joseph2k

    joseph2k Guest

    Why are you doing the student's homework assignment? Do you really want
    that kind of twit to graduate?
     
  11. Fred Bartoli

    Fred Bartoli Guest

    Did you really look at the answer?

    I doubt bulk copying such an answer does any good to his graduation.
    Now, if he's clever enough to understand it, explain it and argue about the
    pros and cons, he's probably well above the required level of this homework
    (which he isn't since he posted the dumb question here).
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-