Connect with us

Speaker driver question

Discussion in 'General Electronics' started by TY, Apr 12, 2004.

Scroll to continue with content
  1. TY

    TY Guest

    I have recently designed a 4-speaker driver circuit using a simple NPN
    darlington (TIP120).

    Each speaker is driven individually by a TIP120 and is connected with
    one terminal on 24Vdc and the second terminal to the collector of the
    TIP120. The emitter of the TIP120 is connected directly to ground with
    the base driven by the output of a 555 via a 1k resistor.

    I have four TIP120s driven by a single 555.

    The aim of the circuit is just to make a LOT of noise... speficially
    around 30Wrms of 'noise' with the waveform generated by the 555. The
    problem is right now, this circuit does not seem to be loud enough but
    my calculations and simulations show me that there should be around
    30+W dissipated by the speakers right now.

    Can point me to the right direction on how to increase the volume of
    the sound generated?

    Thanks in advance.
  2. Robert Baer

    Robert Baer Guest

    Speaker impedance is speced at 1KHz, so use P=E*E/R wher R is the
    speaker impedance, and E is the RMS voltage across the speaker.
    Use a scope or a true RMS reading DVM/voltmeter (not common).
    You failed to mention the frequency, but for maximum perceptual
    intensity, thepeak (i am guessing) is roughly 4KHz (look up the
    Fletcher-Munsen curve for better info).
    Now, if you also desire maximumirritation, then you want to use two
    frequencies, with the difference being small, and the two "tones" to be
    a !discord!.
    Perhaps the transistors are not being fully turned on.
    It is a fact that standard darlingtone cannot be fully saturated.
    A complimentary darlington, however can.
    And if made with discretes, the first (PNP) transistor can have high
    beta and thus low BV, and the second (NPN) transistor can have a high BV
    and thus low beta.
    Whereas standard darlingtons are an IC, where the BV must be low to
    get very high "beta", or hav a lousy "beta" if the BV is decent.
    Try a FET instead.
  3. soundman

    soundman Guest

    First, each speaker is being drive by DC rather than AC. When the
    transistor is off, there ie no current, when the transistor is on, the
    current is one direction only. This limits the speaker to only half of its
    possible movement backwards and forwards. As a speaker aproaches the end of
    its travel, it simply won't move any further, although there is a real
    possibility of damage to the speaker. There are many ways to drive a
    speaker so that there is no DC component across them. For example, Use a
    TIP125 as well as the TIP120 and switch one on when the other is off. The
    node between the transistors will move from 0 to 24V and with a big
    capacitor in series with the speaker, you can drive the speaker from this
    point with the other speaker terminal attached to earth. If you split the
    power supply to make a = / - 12, then you can dispence with the capacitor.
    Make sure that the two transistors are never on at the same time. Better
    still, look up some information on audio amplifer design. There is a thread
    running right now about books for audio amplifers. They will tell you how
    to build amplifiers that are rather kinder to the speakers and can put out
    better signals than a square wave.

    Second, the impedance of the speakers will have a direct bearing on the
    output power. You haven't mentioned the impedance of the speakers.

    Third, the sensitivity of the speakers will affect the output. The more
    sensitive they are, the louder they sound.

    Finally, speakers REALLY don't like square waves as the 555 is putting out.
    See if you can find a better shape of signal to poke at the speakers. With
    the circuit you have described, they must be in considerable pain right now.
  4. If I understand your description well, the speaker is nor driven by AC but
    by pulsed DC. I guess a lot of the power is not converted to noise but to
    heat. To make a better use of the available power use a bridge output

  5. Robert Baer

    Robert Baer Guest

    Pulsed DC *IS* AC (with a bias).
  6. Rich Grise

    Rich Grise Guest

    The first thing is, most of that power dissipation is going straight
    to heat, as well as putting a constant deflection on the speaker cone,
    such that it could bottom out on peaks.

    You need to drive the speakers with AC, not pulsed DC. A BMF cap
    might work, with a totem-pole (complementary-symmetry) output,
    or an H-bridge.

    Good Luck!
  7. JeffM

    JeffM Guest

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