Is It Possible to Predict a Driver's Distortion from Thiele Small Parameters?

[cerrec]

Title says it all. I'm shopping for subwoofer drivers, and I would like to have some idea how much harmonic distortion they will produce.
Is there any way to do this mathematically? Given the advanced state of speaker simulation software, I would hope there is.
Any thoughts and ideas would be appreciated!

 

[MarcelvdG]

I don't think so. Thiele introduced them to simplify his calculations of the small-signal response, he didn't consider distortion at all.

 

[Mark Tillotson]

They are parameters for a linear model of the driver, so can't know about non-linearities.

 

[DcibeL]

No, but impedance and frequency response provide some hints.

 

[Mark Tillotson]

How?

 

[DcibeL]

Low inductance implies a motor with shorting rings. High resonant peak often will be presented in HD lower in frequency. For example, metal cone resonating at 9kHz, will be presented with high 3rd order harmonic at 3kHz. Some hints..

 

[cerrec]

@DcibeL
This is the type of info I'm looking for! Could you elaborate. Please understand you're dealing with a noob here. When you say "low inductance", could you put a numerical range for that. What is "HD"? What "hints" would imply a driver with low distortion?

 

[lcsaszar]

There are other parameters of the loudspeaker electromechanical model in addition to the Thiele-Small parameters. Like Xmax, Bl variation over the gap lenth and beyond, Rdc variation as a function of temperature which is the function of instantenous power, spider and surround nonlinearity as the function of excursion (simplified in a single Xmax parameter), cone breakup vs. frequency vs. amplitude, and there should be more, I believe. Also the same T-S parameter triplet can be achived by many different construction. For example ferrite and AlNiCo magnet could yield same parameters, yet they sound different.
Also just like for amplifiers, harmonic distortion does not characterize the quality of a loudspeaker exclusively.

 

[Boden]

There are -little known amongst hobbyists- so called non linear loudspeaker parameters.

https://www.aes.org/e-lib/browse.cfm?elib=13881