Specs Part 1: How Much Power Do I Need? (geek/tech thread) ;)

[tombowlus]

Originally Posted by fdeck

I will take a stab at this, just for fun.

You could measure the performance of an amp in many ways: The industry standard is a pure sinewave into a resistive dummy load. But you could also compare amps based on how they reproduce a real bass signal.

My proposal is to perform two measurements:

1. Traditional peak amplitude of sinewave.

2. Peak amplitude of real bass signal.

Take the ratio and convert into dB. That's your headroom according to my proposed definition. If you really want to have fun, you could perform the second measurement using a real speaker load.

Of course this does not settle the issue, because we could argue forever over what constitutes a typical bass signal and a typical speaker load.

Gentlemen, start your oscilloscopes. I am getting ready to build my own power amp. When this happens, I will think about concocting a headroom test based on recording my own bass and looking at the waveforms. It might be possible to come up with some generalization of bass signals that allows the design of an idealized signal for performing the headroom test.

That's what I am talking about!

Now, half of you measure the empirical differences, and the other half sit back and tell us what you hear!

Seriously, though. This sounds like an excellent idea.

Tom.

[drurb]

Originally Posted by fdeck

I will take a stab at this, just for fun.

You could measure the performance of an amp in many ways: The industry standard is a pure sinewave into a resistive dummy load. But you could also compare amps based on how they reproduce a real bass signal.

My proposal is to perform two measurements:

1. Traditional peak amplitude of sinewave.

2. Peak amplitude of real bass signal.

Take the ratio and convert into dB. That's your headroom according to my proposed definition. If you really want to have fun, you could perform the second measurement using a real speaker load.

Of course this does not settle the issue, because we could argue forever over what constitutes a typical bass signal and a typical speaker load.

Gentlemen, start your oscilloscopes. I am getting ready to build my own power amp. When this happens, I will think about concocting a headroom test based on recording my own bass and looking at the waveforms. It might be possible to come up with some generalization of bass signals that allows the design of an idealized signal for performing the headroom test.

Well, I wouldn't call this a headroom test. Call it something else, perhaps the fdeck test. Actually, you might want to equate the sine wave and "bass signal" for rms. Anyway, what you might want to do is characterize parameters such as the crest factor of your typical bass waveform that is equated in rms to the sine wave and then see how the amp handles the peaks. This thread has become technical enough so I won't go any further. Maybe we can pick this up later via PM.