how make your speker faster

08-26-12: Unsound
Keep in mind that the Quads that can produce a credible square wave do so due to the concentric wiring that disperses the wave form in a timely manner.

... and that's what's demonstrated by good step response. Perfect step response shows one impulse that then slopes down uniformly as the sound decays. Scroll down on this page for the step response of the Quad ESL-2805, followed by its square wave response.

Now here are measurements for the Dunlavy Signature SC VI, an all-dynamic loudspeaker. Notice how similar its step response is to the Quad.

For contrast, here's the measurements page for the highly regarded Revel Ultima Salon 2. Notice what a convoluted mess it is, and this is more typical of step responses of multi-way dynamic loudspeakers. Even the Wilson Maxx doesn't do that kind of step response. I think at a minimum it requires all drivers to be in phase and nothing but first order crossovers (6 dB per octave).

Other examples of good step response: Thiel SC-4 and Vandersteen Model Seven.

Almarg, isn't that a common misconception of what a square wave is? - 'A square wave is just a series of alternating positive and negative steps, with some amount of time between them.'

I always thought that a square wave was actually a sine wave with many higher harmonics added to the base sine wave, thus creating a continuous 'square wave'. Passing one of these square waves through a speaker intact is a test of the large bandwith of the speaker and its response time.

Inpep, both statements about what constitutes a square wave are correct. Your statement is expressed from the perspective of what is referred to as the frequency domain, and mine was expressed from the perspective of the time domain.

Regards,
-- Al

Thanks Al, I guess all that process control math (Laplace transforms etc.) has conditioned me (I have a Chemical Engineering degree)to only think in the frequency domain. :)

Salut, Bob P.

08-27-12: Inpepinnovations
Thanks Al, I guess all that process control math (Laplace transforms etc.) has conditioned me (I have a Chemical Engineering degree)to only think in the frequency domain. :)

And that, I think, is where some of the communication breakdowns occur in discussions of audio. So many of the audio tests are oriented toward *sound* and fundamental audiology irrespective of musical values such as timing and expressiveness. For example, since human frequency response disappears at around 20 Khz, many think that any engineering to widen the bandwidth out to 100-150 Khz is an unnecessary expense. But if you examine square wave response along with frequency response curves, you find that bandwidth (a frequency consideration) affects risetime (a timing and clarity consideration).

Another thing I've noticed is that there are often measurements of amplitude range (e.g., s/n ratio, headroom, sensitivity, power handling), but not measurements of amplitude *resolution*--how small the increments of source amplitude can be resolved. This translates into the musical values of expression and nuance.