Originally Posted by OllieS
Originally Posted by arnyk
Within reason the need for beefy amps is relieved by the relatively high crest factor of music. The crest factor of music is generally far more than 6 dB, which means that you can run 8 channels of music with a power supply and heat sinks that can only run 2 channels of sine waves.
Okay, but if music has a high crest factor then that means that dynamically you do need high current for dynamic portions in music. In sudden bursts you might need a lot more power, like 10x or 100x the power.
Like you said, music has a high crest factor but that doesn't mean you don't need a beefy amp.
Short term power doesn't require a fraction of the beef that is required by steady state power.
Let me give you a real world example. Once upon a time I experimnted with using amps designed for car audio at home. The obvious sticking point is the need for a 12-15 volt power supply with the ability to supply lots of current in short bursts.
I found that this power supply was required to avoid overloading ans shut down with musical peaks.:
It costs $150, and it weighs over 20 pounds. It will put out 36 amps all day long.
Laster on I was able to power the same amplifier through the same music at the same loudness with the same speaker with one of these:
Tiny, costs under $20, weighs less than two pounds, and has very fast acting current limiting that will shut it down in a heartbeat if you try to draw too much current.
No. But I didn't tell you the whole story. I also added one of these:
Actually, these are far larger than the one I use (500,000 uF versus the 40,000 uF)..
The stiffening cap about matched the weight and cost of the $20 power supply above, but my system cost and weight was still only a tiny fraction of the lab power supply I used at first. Furthermore no way would the smaller lighter system handle pure sine waves but that was not what I wanted to listen to. BTW, the speakers I was using to audition and test were KEF Q10s that have a broad region with about 3 ohms in their impedance curve..