ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2]  All   Go Down

Author Topic: Gain Structure  (Read 5591 times)

Andrew Welker

  • Church and H.O.W. Forums
  • Newbie
  • *
  • Offline Offline
  • Posts: 22
  • Denver, CO USA
Re: Gain Structure
« Reply #10 on: April 14, 2011, 11:00:07 am »

Is there anything at all about a clipped signal that can contribute to driver failure?

A clipped signal will contain many more high frequency harmonics than the original signal. This is due to the sharp corners that get introduced into the signal when it is clipped. Also, the more the system gets pushed into clipping the higher the average power that is contained in the signal.

If the amplifier output itself is clipping, then this is after any low- or high-pass filters and will cause other issues. If it is the low frequency amplifier, then the speaker will most likely die from over heating, since the speaker is taking all those high frequency harmonics and turning them into heat in the voice coil. If it is the high frequency amplifier, then it'll probably be over-excursion, since those high frequency harmonics  could possibly be many, many times the rated power of the driver.

http://mathworld.wolfram.com/FourierSeriesSquareWave.html
The link above demonstrates some of these concepts pretty well. The sharper the corner, the more high frequency harmonics are included in the signal.

I'm sure JR or Bennett will correct me if I'm wrong.
Logged

John Roberts {JR}

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 16610
  • Hickory, Mississippi, USA
    • Resotune
Re: Gain Structure
« Reply #11 on: April 14, 2011, 11:16:45 am »



Is there anything at all about a clipped signal that can contribute to driver failure?

Compared to an unclipped version of the same signal played through a larger amplifier, not much. The dominant phenomenon IMO for sound reinforcement driver failures is average power.     

That said, in hifi circles, with puny tweeters, there are cases where the extra HF content caused by clipping can stress a marginal HF driver, and there's an old JBL white paper that somebody will probably trot out to support that position. 

 /edit  the difference between a clipped and unclipped waveform in the same amplifier, is that the clipped waveform is also turned up louder making more average power.  /edit

Music generally has short term transients that are much more powerful than the average or continuous power level. Speakers likewise can handle more power transiently or short term, than they can long term so this is all good, so far. This is almost a trifecta of goodness, since amplifiers also put out more power short term than long term, but unfortunately amplifiers short term to long term ratio does not taper off fast enough enough to protect loudspeakers. 

The problem is this ratio of short term transient power to long term average power in music is not constant but changes with music type, performance, and even operation of the system. An amplifier sized to put out adequate short term transient power, to take full advantage of the loudspeakers peak capability, will typically be capable of putting out more than enough long term power to damage the loudspeaker.

So it it impossible to protect a loudspeaker, simply by sizing the power amp smaller, unless you forfeit a great deal of it's short term output capability. This is further aggravated by the less than obvious characteristic of power amplifiers, that their output is rated as clean (un-clipped)  sine wave power. That same amplifier is capable of 2x it's rated output power, if that sine wave is hard clipped to resemble a square wave. So in practice a power amp can actually put out more average power than transient peak power. /edit- I know this sounds inconsistent with my earlier statement that amplifiers make more short term peak power than long term... it's complicated. /edit

Technology is catching up with smart speaker protection built into some DSP based amplifiers that can limit short term and average power independently, but this is still in early days and not filtered down to lower cost models. Note: even this short term power criteria is a bit of a simplification, since it also depends at what frequency that peak occurs.   

I expect the entry level market will see this first inside powered speakers, where the designers work to make them customer proof. When operating without a net, you need to learn to identify power compression (when the speakers start losing output because of overheating) and back off, instead of turning up more, and releasing the smoke. 
 
JR
« Last Edit: April 14, 2011, 11:23:29 am by John Roberts {JR} »
Logged
Don't tune your drums half-ass. Listen to what a properly "cleared" drum sounds like.   http://circularscience.com/

ProSoundWeb Community

Re: Gain Structure
« Reply #11 on: April 14, 2011, 11:16:45 am »


Pages: 1 [2]  All   Go Up
 



Page created in 0.025 seconds with 22 queries.