Active noise-cancelling silencers.

Discussion in 'UK Motorcycles' started by Krusty, Sep 24, 2010.

  1. Krusty

    Krusty Guest

    I remember when they were going to be the next big thing - silencers
    with speakers & sound generators that sent out an opposite sound wave
    to cancel out the exhaust noise. They disappeared without a trace for
    whatever reason.

    But stumbling across an article about them t'other day got me thinking.
    The ins & outs are way beyond my brain, but is it as simple as
    recording the exhaust noise, loading that into something like Audacity
    & using some functions in there to generate an 'opposite' sound file?
    And if so, would that only work for the RPM the recording was done at,
    or does the exhaust sound wave stay roughly the same frequency at all
    revs (or a band of n RPM)?
     
    Krusty, Sep 24, 2010
    #1
    1. Advertisements

  2. Krusty

    Paul - xxx Guest

    LOL, you're bored then ...

    Thinking back to when I had a Suzuki X7 with expansion chambers I doubt the
    exhaust note stays the same frequency, the X7 certainly high-pitch screamed
    at high revs and quietly-ish burbled (fsvo burbled) at low revs.
     
    Paul - xxx, Sep 24, 2010
    #2
    1. Advertisements

  3. Krusty

    Krusty Guest

    I wish... I've been battling with an extremely complex problem for the
    last couple of weeks, & have to keep kicking it out of my brain in the
    hope I have a eureka moment when it pops back in.
    Yeah I thought as soon as I posted that it was a stupid question. The
    first part still stands though.
     
    Krusty, Sep 24, 2010
    #3
  4. Krusty

    Dylan Smith Guest

    That wouldn't work - it's not just the frequency which must be exactly
    right, but the phase that must be exactly right also. Even if you
    had a continuous frequency sound you wanted to cancel out, a recording
    wouldn't be sufficient because it's doubtful you'd be in phase (well,
    precisely 180 degrees out of phase). At best you'd end up with a
    WwwwwwooOOONNGGgggwwwwoooONNNNGGggggwwwwwwoooONNNNGGGgggg sound (if you
    see what I mean) because the two sound waves wouldn't be exactly at the
    same frequency and so you'd end up with a "phasing" sound louder than
    what you're trying to cancel for some of the time. (Think of the sound
    you hear on a turboprop airliner when they don't bother to keep the
    props precisely in sync, that really annoying LOUD noise).

    Even when canceling a fixed frequency sound, you still need to actively
    keep the sound wave you're using for cancelation precisely 180 deg.
    out of phase with the sound you're trying to cancel, and this simply
    cannot be done passively.

    All bets are off trying to do it with a simple recording for a simple
    fixed frequency sound, all bets are definitely off for a bike exhaust
    which will run a large range of frequencies (and different wave forms).
     
    Dylan Smith, Sep 24, 2010
    #4
  5. Krusty

    M J Carley Guest

    Not really, even in so far as there is such a thing as `phase' on
    these signals. You can't cancel the noise everywhere so you aim to
    reduce it at selected points without increasing somewhere else that
    people might hear it.
    I suspect that is more likely to be due to a gust, than loss of sync.
    You don't: you take a feed off the reference sigal (on aircraft, you
    use the prop shaft rotation pulse). You don't have a feedback signal
    in this system anyway, so you do some cancellation at source, just
    past the end of the can.
    The way I would do it is put a ring speaker on the end of the can (not
    inside it) and drive it with a feed from the rpm signal through a
    smart bit of electronics.
     
    M J Carley, Sep 24, 2010
    #5
  6. Krusty

    Adrian Guest

    (M J Carley) gurgled happily, sounding much like they
    were saying:
    I'd just lob some money at Lotus. They've been doing it in production for
    20yrs.
     
    Adrian, Sep 24, 2010
    #6
  7. Krusty

    Jim Guest

    We are actually in the building next to where Southampton University
    used to have their automotive research labs: there's a big department
    called ISVR (Institute of Sound and Vibration Research) who partnered
    with Lotus for some of the work that you mention.

    Unfortunately they got shut down because the science park people felt
    that running JCBs etc all day whilst pointing microphones at them didn't
    fit in with the general ambience they were aiming for. Which I reckon is
    a bit sad.
     
    Jim, Sep 24, 2010
    #7
  8. Krusty

    crn Guest

    Which could set up backpressures, equivalent to blocking the pipe.
    If it were that simple it would have been done already.

    Active noise reduction only works for a small point in space no larger
    than wavelenth/10 of the sound to be reduced. Noise cancelling
    headphones only need to create a null point at the ear entrance so
    they work (give or take some leakage) so they are the simple solution.
     
    crn, Sep 24, 2010
    #8
  9. Using the patented Mavis Beacon "Hunt&Peck" Technique,
    typed
    <pulls up chair>

    crn teaches TBC about acoustics.

    This should be fun.
     
    Wicked Uncle Nigel, Sep 24, 2010
    #9
  10. Krusty

    M J Carley Guest

    Which is why I said to put on the end, not inside it. As for
    backpressure, you are cancelling the radiating mode on the pipe
    termination, not the full modal system inside the pipe, which requires
    much less energy with no need to stop the flow (the mean flow doesn't
    make noise).
    Best tell NASA:

    http://www.nasa.gov/centers/glenn/about/fs03grc.html
    Best tell Saab their Saab 2000 system doesn't work, then.
     
    M J Carley, Sep 24, 2010
    #10
  11. Krusty

    crn Guest

    Different case. This is similar to the method Bombadier use in the Q400
    turboprop. If you know that noise is being transmitted by vibration
    through a panel or bulkhead you would normally add a layer of absorbant
    padding between the panel and the interior lining and/or stiffen the panel.
    It is possible to add sensors and actuators to the interior lining to cancel
    the transmitted vibration, thus reducing both noise and padding weight.
    This is not the same as cancelling airborne noise, stopping the lining
    vibrating prevents the airborne noise being _created_ rather than trying
    to fix it later for a listener which is much more difficult.
     
    crn, Sep 24, 2010
    #11
  12. Krusty

    M J Carley Guest

    Saab's noise team say they used speakers to cancel airborne noise
    inside the cabin, which is a lot bigger than one tenth of a
    wavelength:

    www.a2acoustics.com/publikationer/AIAA_98_2231.pdf


    As for exhaust noise, this is a very sweet idea:

    http://www.mech.kuleuven.be/mod/other/topic_03_01
     
    M J Carley, Sep 24, 2010
    #12
  13. Krusty

    72degrees Guest

    72degrees, Sep 24, 2010
    #13
  14. Krusty

    M J Carley Guest

    M J Carley, Sep 24, 2010
    #14
  15. Krusty

    Dylan Smith Guest

    No, if it's continuous in still air, it's because the props aren't
    synced. (Most large turboprops should have some kind of automatic
    sync but it may not be perfect. Small twin engine aircraft have
    to be synced by the pilot manually (example, the elderly Piper Apache
    I did my multiengine rating in).
    You've just defined an active system :) Using a feed off a prop shaft
    to keep in sync or using a sound wave to keep in sync, effectively
    you're having to do something actively to keep the cancellation signal
    180 degrees out of phase with the sound you're trying to cancel.
     
    Dylan Smith, Sep 29, 2010
    #15
  16. Krusty

    M J Carley Guest

    Or propeller asymmetry.
    You can't keep your signal 180 degrees out of phase. ANC does not give
    you full cancellation. In aircraft active systems, the ANC uses
    feedback microphones to adjust phase and amplitude but doesn't aim for
    zero noise.
     
    M J Carley, Sep 29, 2010
    #16
    1. Advertisements

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.