Shop OBEX P1 Docs P2 Docs Learn Events
Considering LM4876 Audio Amp — Parallax Forums

Considering LM4876 Audio Amp

bradharperbradharper Posts: 64
edited 2011-06-16 12:48 in General Discussion
In my quest to move away from the TDA7052, I've narrowed the search down to a few candidates and currently focusing on the National LM4876.

I'm wondering if anyone has experience with this IC?

I need to produce quality 16-bit, 22khz mono PCM audio through an 8ohm 1W speaker (300hz~20Khz range), which the LM4876 is spec'd to handle @ ~5V. The "AUDIO POWER AMPLIFIER DESIGN" section in the datasheet features examples that are pretty close to my specs with the exception of a lower frequency bandwidth limit (100 vs. 300hz).

The reference schematic indicates only 5 small external components ( I have very little board space), and I *think* the addition of an RC circuit on the input would be all that I need to handle the PWM signal from the Propeller. Am I correct?

Anyone see a reason why this chip wouldn't be ideal for my requirements or have any opinions or insights to offer?

As always, your time is much appreciated.

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-06-10 10:21
    The S2 robot uses the LM4876's 2W cousin, the LM4990. Here's the schematic:

    attachment.php?attachmentid=82013&d=1307726444

    The input comes directly from a Propeller pin. The LP filtering is accomplished via the additional feedback cap. FYI, this amp does not seem to have excessive power for the S2's one-inch speaker.

    -Phil
    728 x 340 - 14K
  • bradharperbradharper Posts: 64
    edited 2011-06-10 10:42
    Thanks Phil. I have a question regarding what I referred to as an "RC" circuit, and your LP filtering achieved by the addition of C38. Is that LP filter (R5 --> C38 ?) acting in the same functional role (DA conversion) of the "RC" circuit that I mentioned?
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-06-10 11:03
    Apparently, yes, in that it limits thre frequency response of the amp. (I did not design the circuit, BTW.) I checked to make sure nothing else is connected to the Propeller pin, and there isn't. So the integrating cap, along with the native response of the amp itself, provides all of the low-pass filtering.

    -Phil
  • bradharperbradharper Posts: 64
    edited 2011-06-16 11:39
    (I did not design the circuit, BTW.)

    -Phil

    Do you know who did by chance? I'd like to understand how, and to what extent, the deltas from this schematic and the datasheet version effect the frequency response.
  • LeonLeon Posts: 7,620
    edited 2011-06-16 11:48
    The calculations are described in the data sheet.
  • bradharperbradharper Posts: 64
    edited 2011-06-16 11:55
    Leon wrote: »
    The calculations are described in the data sheet.

    The datasheet calculations are in reference to the "Typical Application" schematic - right? I'm looking to understand how (if so) the additional cap in the S2 schematic (C38) alters the frequency response.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-06-16 12:29
    The speaker circuit is basically an op-amp integrator with DC gain control. See this tutorial for more information:

    -Phil
  • bradharperbradharper Posts: 64
    edited 2011-06-16 12:36
    The speaker circuit is basically an op-amp integrator with DC gain control. See this tutorial for more information:

    -Phil

    Excellent. Much appreciated...
  • LeonLeon Posts: 7,620
    edited 2011-06-16 12:40
    Although Nat Semi deosn't have a SPICE model for the LM4876, it should be quite easy to model the circuits by making a few assumptions, and compare the frequency response.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-06-16 12:44
    My SPICE simulator is a proto board and soldering iron. :)

    -Phil
  • LeonLeon Posts: 7,620
    edited 2011-06-16 12:48
    That can be quicker, and you know that the results will be accurate. :)
Sign In or Register to comment.