speaker drivers push/pull

[jrosenstein]

Hi Barefoot, this came up in a thread on homerecording.com, and I'd be curious to see what you have to say about it.

Do drivers behave much differently pushing vs. pulling? I'm probably not using the right terminology here, but I'm sure you know what i mean. People can't *hear* phase in itself. But if i sent in two identical signals with reversed polarity into a speaker, would i be able to measure a difference (besides for the polarity, of course).

It seems to me that as far as fidelity goes, a speaker should be able to do anything negative that it can do positive -- after all, it's going to be pulling in half the time.

Feel free to get technical on my ass.

- jacob

[barefoot]

Jacob,

I know exactly what you mean and this is an excellent question. Looking at the typical dynamic loudspeaker it's obvious that the thing is not symmetric about the motor. And this asymmetry does indeed affect the linearity. Symmetry of the driving force is one of the factors that determines the quality of a driver.

Whether the effect of reversing the polarity can be measured or not depends on the nature of the input signal. Sine and triangle waves, for example, have directional symmetry. If you reverse their polarity then shift their phase by 180°, you wind up with the exact same waveforms. These types of waveforms would not illuminate any driver polarity dependence. Sawtooths, on the other hand, are very different when the polarity is reversed. The velocity of the cone in either direction does depend on the polarity of the waveform and a phase translation doesn't change this fact. Most drivers would likely show at least some measurable difference in their distortion spectra depending on the input polarity of something like a sawtooth.

This "distortion asymmetry" doesn't say much, however, about the rightness or wrongness with respect to fidelity of an absolute polarity switch. A driver will need to reproduce waveforms all kinds and with every manner of inherent phase and polarity, so the driver's own asymmetric distortion characteristics will show up no matter what. The question of absolute polarity is in my opinion more an issue of the ear's ability to distinguish certain types of transient events. For example, can our ears (and bodies) distinguish between the polarity of a recorded implosion or explosion, snare crack, etc.? To a very limited degree I think we can distinguish. So, only in this respect do I think absolute polarity is important.... but certainly not crucial.

Thomas