Connect with us

Class D audio amp power rating

Discussion in 'Electronic Design' started by N Cook, Jun 22, 2007.

Scroll to continue with content
  1. N Cook

    N Cook Guest

    For a bi-polar D class amp ( 2 n-channel mosfets) being powered from +38V
    and -38V what would be the ball-park maximum instantaneous and continuous
    RMS wattage driving into 8 ohms and not demanding THD requirements and only
    40Hz to 18KHz handling.?
  2. Eeyore

    Eeyore Guest

    38 volts peak > 180W into 8 ohms ( 90 W rms).

    In reality it'll be less due to device on resistance.

  3. MooseFET

    MooseFET Guest

    At the 40Hz end, you can have a huge THD and get nearly a 38V peak
    square wave. Ignoring the drop in the output devices:

    P = 38^2 / 8 = 180W RMS

    For a not very distorted signal it will be

    180W / sqrt(2) = 127W
  4. N Cook

    N Cook Guest

    Thanks for that.
    What would it be for say 10 percent THD
  5. Bob

    Bob Guest

    That is correct if the amp does not have an output transformer.

    If an output transformer is used the power output will be limited by
    the amount of
    current that the MOSFET's can handle.
    I'm not going to do parametric searchs on MOSFET's right now but I'd
    guess 500watt would be achieveable. In practise paralleling FETs or
    an H bridge might make more sense than looking for the highest
    since MOSFET that can switch fast enough.

  6. N Cook

    N Cook Guest

    I was trying to establish whether the amps I was repairing could be
    correctly rated at 100W rms output , variously stated 150 , 180W "music
    power" into 8 ohm with a THD at maximum power of 10 percent.
    What would the likely wattage of thermal loss be for such a 100W rms rated
    amp at say 20 and 50 watt rms output? As part of dummy load testing I was
    monitoring the output vented air temperature in use at different power
    levels, for the next time I come across such amps.
    I appreciate they are thermally more efficient but I could not find what
    temperatures to expect in a given ducted and vented cooling airflow
  7. Eeyore

    Eeyore Guest

    How about you show me a Class D amp WITH an output transformer ?

  8. Eeyore

    Eeyore Guest

    Do you just make this stuff up and blow it out of your ass ?

    Maybe you'd like to show your calculations for these output figures and how you got
    there without any data whatever ?

  9. Eeyore

    Eeyore Guest

    About what he said.

  10. Eeyore

    Eeyore Guest

    Depends hugely on device on resistance.

  11. BobG

    BobG Guest

    Can you see 3% distortion on a scope? Someone told me you couldnt, but
    I never had a scope and a thd meter hooked up at the same time. Maybe
    10% thd would be a 'slightly clipped' sinewave?
  12. Tim Shoppa

    Tim Shoppa Guest

    Class D can be done quite efficiently - 80 percent in a lot of off-the-
    shelf designs. So I would not expect the power dissipation to be much
    more than 10 watts for 50 watts output. (Efficiency isn't flat over
    the entire power range but around half to full power you'll be in the
    sweet spot for efficiency).

    In some circumstances class-D efficiency can reach 90-95-98 percent in
    certain applications but most audio amps seem to max out around 80
    percent efficiency in real-life commercial products.

    Vented air temperature by itself cannot be directly mapped into power
    dissipation without a metric buttload of other parameters and thermal
    and hydrodynamic/convection modeling. But if you want, just put a
    power resistor in the same box and see how much power it's dissipating
    when the vent temperature matches. It's not a perfect comparison but
    there are just so many variables in play.

  13. If the 3% is caused by clipping or zero crossing distortion,
    almost certainly. If it is a combination of only second and
    third harmonics, not at all certain.
    I think 10% would be an obviously clipped sine wave.
  14. mpm

    mpm Guest

    Someone wrote (& I couldn't decide who..)

    My response: Blah, blah, blah...

    If you don't define what a term means, (i.e., "music power") it can
    mean anything.
    And that's usually whatever the Marketing department thinks it should

    Maybe it's the SPL developed by the system while it is accelerating
    down the center of a black hole!

    Oh wait, there wouldn't be any sound in space?
    (Let's just keep that tidbit to ourselves....)

  15. Bob

    Bob Guest

    There are plenty of class D amps with output transformers, they are
    mostly for
    100volt line public address system use rarther than 8 ohm though thats
    a matter of the transformer ratio, there are 100volt line class D amps
    with output transformers that can be paralleled to drive 8ohm loads.

    Googling for "100volt line class d" finds a few examples.

    The BV440M model on the page below will do 440watts with a single
    22volt supply.
    You can see the transformer in the picture.

    Here's a class D amp brochure that mentions the output transformer.
    Tony Smith from this company has ocasionally posted to s.e.d on the
    subject of
    class D amps.
    No need to be obnoxious. Perhaps you should look up the meaning of
    in a dictionary.
    The original poster specified the psu voltage and a single pair of
    fet's. Based on
    what I'v seen of switching FET's in equipment I estimated that a
    single FET can
    switch 15amps at a few hundred kilohertz. I said I though 500watts
    would be
    achieveable, I didn't say I think that is the maximum possible.

  16. Eeyore

    Eeyore Guest

    I rather doubt there are plenty actually. With modern devices it's more sensible to use
    higher voltage supplies and direct couple.

    Yes, so it's irrelevant to the OP's question isn't it ?

  17. Bob

    Bob Guest

    Systems that need battery backup generally use 24V or 48V single
    supplies and have
    output transformers.
    Output transformers can mitigate various practical problems in public
    address systems.
    A high power amp without a transformer will generally need one side of
    the output
    connecting to earth due to EMC and safety standards. In office
    buildings where hundreds
    of meters of PA cabling and lots of junction box's are thrown in by
    the contractor that
    bid lowest and the occupier moves the partiton walls around
    relatively often so
    Mr Important can have a larger office it is a common occourance to
    find one wire in a
    loudspeaker line is short to earth somewhere. With an output
    transformer sound still
    comes out of the loudspeakers, without a transformer it often dosn't.
    It was relevant to the original hypothetical question about what is
    possible, before
    the OP told us he is repairing an amp and it is relevant to your
    demand to be
    shown class D amps with output transformers.

    Your demand implied that you believe calss D amps never have output
    People will think more of you if you can accept when you are wrong
    instead of
    getting stroppy.

  18. Eeyore

    Eeyore Guest

    That wasn't the OP's question !

    You decided to answer a completely different question you made up for yourself.

  19. Eeyore

    Eeyore Guest

    Bob wrote:

    You're talking complete unabated drivel now.

    Point to a single IEC standard that would require doing that.

  20. Phil Allison

    Phil Allison Guest

    "John Popelish"

    ** With any respectable * analogue * scope - 3% waveform distortion
    with a sine wave is easily visible.

    If the sine shape has become asymmetrical ( ie 2nd harmonic ) - it shows up
    as one maxima being larger than the other.

    One can easily see slight peak amplitude compression or a slight inflexion
    near the zero line that when checked on a THD meter is about 2 or 3 %.

    Very small amounts of peak clipping can be seen - little as 0.5%

    Forget doing this with a digital scope where all sine waves look distorted
    and noisy anyhow.

    ........ Phil
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