Connect with us

3 dB bandwidth

Discussion in 'Electronic Design' started by [email protected], Jun 21, 2005.

Scroll to continue with content
  1. They all do. You are mistaken. A 3db change in volume is *easily*
    detectable by anyone. A figure of 1db would be more relevant as a rough
    guide to a "not always immediately detectable when drunk".

    Kevin Aylward
    SuperSpice, a very affordable Mixed-Mode
    Windows Simulator with Schematic Capture,
    Waveform Display, FFT's and Filter Design.
  2. Guy Macon

    Guy Macon Guest

    Yeeeeees! That's a 3dB drop. :)

    Actually neither one of us should be saying "drop." If I had
    specified power drop and you had specified voltage drop it
    would have been clear that we are both right.
    Try to make it work without transistors sometime. In the days
    when we all used vacuum tubes/valves and transformers, getting the
    maximum amount of power out of one transformer and into the next
    transformer was important, thus the matching, and thus the 3dB
    power drop.

    When semiconductors and especially op amps hit, it became feasible
    to have a low output impedance feeding a high input impedance
    (and occasionally the reverse, with trans impedance amps). When
    that hit, not only wasn't there a power drop, the power transfer
    was annoyingly close to zero, and everyone started thinking in
    terms of voltage rather than power. Once you start measuring
    voltage in a Low Z to high Z line, the drop when you add a load
    approaches the convenient number zero. Filters, however, still
    needed a -(something)dB point, and the old -3dB stuck.

    A common occurrence nowdays is for newly minted recording
    engineers to think that adding together two close to identical
    phased signals (think two microphones right next to each other
    in front of the bass guitar amp) add up the same way as adding
    together two random phased signals (think two microphones close
    micing two violins playing in unison). Loads of fun watching
    them trying to figure out why they don't add up the same way...
  3. Guy Macon

    Guy Macon Guest

    For sine waves and instant changes, yes. Try it with an orchestra with
    one minute of silence between the two playings. I believe that you will
    find that with changes of less than 3dB SPL, you will be wrong about
    which one was louder about half the time.
  4. Pooh Bear

    Pooh Bear Guest

    Hah ! Quite likely. Aren't we humans fantastic !

    The ear quite happily deals with immediate dynamic range issues by offering TTS (
    temporary threshold shift ). It's like somone on high figured we might need it and
    gave it to us !

  5. Guy Macon

    Guy Macon Guest

    Indeed. I don't think TTS accounts for 3dB SPL being the minimum
    detectable change under the test coditions above, though. Below
    65 or 70 dBA SPL there is no TTS. See

    One real-world application of the 3dB rule is buying a new stereo.
    All other things being equal, will you be able top hear the difference
    between a 200W amplifier and a 400W amplifier? Probably not.
  6. Sure you can, if you switch between them in an A/B test.

    Sure, the ear brain is not very good at absolute measurements.
    Remembering one level then another 1/2 hour later might well be a bit
    difficult, but that is irrelevant to the issue as to whether or not 3db
    is the minimum detectable as an accurate representaion of the ability of

    Kevin Aylward
    SuperSpice, a very affordable Mixed-Mode
    Windows Simulator with Schematic Capture,
    Waveform Display, FFT's and Filter Design.
  7. Rich Grise

    Rich Grise Guest

    Even though I'm no expert, it seems feasible that 1 db would be
    "just detectable"; 3 db is, after all, _twice_ as much power!

    I've also heard or read somewhere that it takes a full 10 db change to
    sound "twice as loud".

  8. Robin

    Robin Guest

    Especially if that change happens at "20kHz" :)

  9. My last slide rule is in the drawer for sentimental reasons,
    a gift from my father when I was 10 and such devices
    were used every day by many engineers. I used it for a
    few years before electronic calculators became ubiquitous.
    The "calculated by hand" I remember is my father at the
    kitchen table with a scratch pad and a thin-paged book
    full of 5 digit logarithms. That was where, with some
    amazement, I learned about interpolation. The long slide
    rule was just for approximate work, he explained. I just
    wish I knew what he was doing that required enough
    accuracy to justify all that effort.
    That reminds me of a gift from somebody faintly aware of
    my fondness for pendulum clocks. Its plastic pendulum,
    whose motion was derived from the electronic innards,
    went back and forth linearly, not sinusoidally. What an
    abomination! I still shudder thinking of it.
  10. Joel Kolstad

    Joel Kolstad Guest

    Have you seen those horrible radios that are styled -- in plastic! -- to look
    like vintage tube radios with their beautiful wooden cabinets? They even come
    with 100% fake plastic "tubes" even though internally they're regular
    transistor radios. :-(
  11. John Larkin posted a photo of one. IIRC, he said it cost all of $10.

    Just for fun I tracked down the source in China. ;-)

    Best regards,
    Spehro Pefhany
  12. Joerg

    Joerg Guest

    Hello Spehro,
    And they haven't even bothered to mimic a tube sound which would have
    cost next to nothing.

    Regards, Joerg
  13. Fred Bloggs

    Fred Bloggs Guest

    Hey retard- no one gives a damn about calculating anything by hand or
    any other of your pseudo-mathematical bs ( remember the math it took you
    50 years to learn by rote?). The practical significance is, in most
    cases, that the gain drops as F3db/F, or possibly a power thereof, at
    frequencies above 3dB, and it's flat up to 3dB. Get it? Quite sad you
    will go to your grave in such a confused and worthless state- too bad
    for you there is no such thing as a career in sitting on a park bench
    blithering to yourself.
  14. Fred Bloggs

    Fred Bloggs Guest

    Correction: you used it for *many* years before the calculator came on
    the scene, you old fart.
  15. Reminds me of the story of the forest ranger that was patrolling the
    campgrounds and came upon this pair of snakes that was making whoopee on the
    rustic picnic table.

    When confronted, they said that anybody should be able to multiply on log

  16. Guy Macon

    Guy Macon Guest

    Noah had a similar problem when he told the animals on the ark
    to "go forth and multipy." A pair of snakes stayed behind.
    They couldn't follow the order because they were adders.
  17. DERF transform applied.

    How many frequency responses have you seen that were
    flat up to the 3 dB cutoff frequency? Does this mean you
    are among the crowd that considers +/- 3dB to be flat?
    Yes, I'm afraid so.

  18. Fred Bloggs

    Fred Bloggs Guest

    In the olden days, 3dB was comparable to the IL . So yeah.
    Yeah, 1dB either way doesn't amount to a hill of beans to anyone who
    uses the Log scale of measurement. +/-3dB is an industry standard error
    band on signal level for RF.
    Well- of course you pretend...
  19. YD

    YD Guest

    Tube or SS?

    - YD.
  20. That's the whole point of the original joke.

    He made a table out of logs, because even _adders_ can multiply on a
    log table.
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