XMAX and THD question

[LOUT]

After reading this;

1% distortion is irrelevant. That's -40dB. Keep in mind that when using Klippel analysis to determine xmax the figure used is when excursion results in 10% THD.

I've been meaning to ask, does that 10%THD @ XMAX tend to scale somewhat consistantly downward as XMAX is lessened even among different woofers?
For example; is it pretty common for a woofer (almost any woofer) measuring 10%THD @ XMAX to measure roughly 1%THD @ half-XMAX?
Or
Is THD @ half-XMAX drastically different between different woofer models/sizes/brands (still assuming their original XMAX figure was taken at 10%THD)?...for example; a particular 10" woofer may have THD fall to 1% at half XMAX while a different 6.5" woofer's THD only falls to 5% at half-XMAX.

[LOUT]

Also saw this while trying to find the quoted post above, so I'm going to pick on billfitzmaurice today. :D

Distortion happens when you get up around xmax. The 0.5mm xmax of this tweeter is very generous, indicating that it might work just fine to 1kHz, though I wouldn't do so with only second order high passing, otherwise excursion below 1kHz could limit it to low power before THD became severe.

Is excursion (nearing XMAX) typically the most prominent source of driver distortion (assuming high-frequency distortion is well controlled with the XO)?
In other words, can you usually expect any decent driver to have low distortion when it's kept below XMAX and isn't forced to play high-frequencies above its range?


Of course anyone with answers or ideas is free to comment. The quotes from billfitzmaurice were mostly about showing what got my brain marbles rolling around...not so much a direct inquiry for bill himself.

[wogg]

I think that will have a lot to do with the motor design. Klippel analysis gives you the BL linearity, which I believe will track with distortion in the output. Stuff like overhung, double gap or general gap thickness, coil length and all that will contribute to how distortion increases as you near Xmax.

I really don't know how large those differences would be in the distortion profile though, so I'm really just thinking out loud. Interesting topic for sure.

The "Transducer Non-Linearities" section of Klippel measurements is where this analysis would live IMO. Every single one of those is going to be a contributing factor to distortion vs. Xmax.
http://www.klippel.de/know-how/measu...-overview.html

[billfitzmaurice]

THD is simply when the waveform you get out isn't the same as what you put in. The main source of THD is when the cone has reached its travel limit at the lower end of its pass band. Additional power won't get as much additional output at that lower end of the pass band, if any, as it will at higher frequencies, and now what's coming out isn't the same as what's going in. The situation is similar to what happens with clipping in the electronics.

[LOUT]

THD is simply when the waveform you get out isn't the same as what you put in. The main source of THD is when the cone has reached its travel limit at the lower end of its pass band. Additional power won't get as much additional output at that lower end of the pass band, if any, as it will at higher frequencies, and now what's coming out isn't the same as what's going in. The situation is similar to what happens with clipping in the electronics.

That explanation makes a lot of sense. Does that mean a driver's naturally un-flat frequency response is also considered as THD or is THD more specifically the additional differences in signal apart from the driver's 2.83v@1m FR?
EDIT: nevermind, dumb question. If only the differences at higher V counted, then I'd expect most/all drivers to claim ~0%THD at 2.83v which doesn't make sense.

Is the distortion level (aside from inherent frequency-curve inaccuracy of the driver) generally MUCH much lower for most drivers below XMAX?...or do the different motor/suspension/build aspects like wogg mentioned above tend to cause different woofers to vary pretty wildly in that regard?

[billfitzmaurice]

Driver sourced coloration would also be considered distortion. There are circumstances where xmax has little to do with distortion, the most common being break up mode, if you were to run that high. One genre that does just that is electric guitar, which also uses very low xmax woofers because they want high THD.

[LOUT]

Driver sourced coloration would also be considered distortion. There are circumstances where xmax has little to do with distortion, the most common being break up mode, if you were to run that high. One genre that does just that is electric guitar, which also uses very low xmax woofers because they want high THD.

I think I learned a new word. Is driver-sourced-coloration (the kind that happens consistently even at low voltage) referred to as "linear distortion"?

To xover lower in tweeters, there is no replacement for displacement.

Do dome-tweeters behave pretty similarly to woofers where you can expect them to be relatively "safe" if you stay at/below XMAX and below thermal/wattage limits?
Is it typical for a dome tweeter to measure ~10%THD @ XMAX?

[billfitzmaurice]

Coloration would be linear distortion, since the reproduced signal differs from the original in the relative amplitude of different frequencies. Harmonic distortion is non-linear, as it's the creation of harmonics in the reproduced signal not present in the original signal. It can be difficult to tell which is which, since they tend to both occur at the same time. For a real world example of both consider using an equalizer. It can totally alter the relative amplitudes of frequencies, but it doesn't create signals not present in the original waveform. At the opposite extreme is a synthesizer. It starts with a basic waveform, adding to it harmonics that it artificially creates to produce an entirely new waveform.

[Wolf]

HD will rise with decreasing frequency in tweeters. The lower you xover, the higher the HD is, all depending on the tweeter in question. You have to xover ABOVE this area to keep them happy. The larger the dome/diaphragm, usually the lower it can play. However, this also means usually the higher they can also NOT play; the larger they get. Some tweeters also have varying levels of HD with different input voltages.
Sometimes tweeters can be pushed a smidge lower by shaping the rolloff knee or notching out the offending range below the xover point.

Under typical operation, tweeters should be preferrably low HD. I know some tweeters have an inherent level of HD in their spectrum and not all higher HD tweeters sound awful. The point is to keep the HD bloom as frequency decreases locked up and anesthetized.

Later,
Wolf

[LOUT]

HD will rise with decreasing frequency in tweeters. The lower you xover, the higher the HD is, all depending on the tweeter in question. You have to xover ABOVE this area to keep them happy. Some tweeters also have varying levels of HD with different input voltages.
Sometimes tweeters can be pushed a smidge lower by shaping the rolloff knee or notching out the offending range below the xover point.

I know some tweeters have an inherent level of HD in their spectrum and not all higher HD tweeters sound awful. The point is to keep the HD bloom as frequency decreases locked up and anesthetized.

Is that HD bloom sometimes above the FS, or is it almost always below FS (or at FS)?
EDIT:(I understand the distortion rise isn't a single point, but usually a steep ascent...I guess I'm asking if most of the higher bit is at FS, below FS or if it can often start rising significantly above FS)

Does that HD bloom often not have much relation with the tweeter's XMAX, or does keeping excursion under good control usually also lead to keeping the distortion under good control?

[billfitzmaurice]

HD with tweeters is similar to with woofers. It has little to do with Fs, other than the closer to Fs the tweeter is run the greater the excursion and the higher the THD. One way to go lower is with a higher order HP filter. You don't see high THD at the lower end of the pass band, you see it at the upper end of the pass band. With woofers that can result in a subjectively warmer tone, thus the very low xmax of guitar drivers. With tweeters it's a subjectively harsh tone. Along that line guitar amps have high THD, especially when distortion pedals are used, and because that sounds really harsh in the highs guitar speakers don't use tweeters.

[fpitas]

All I have to add to this discussion is that large excursion can cause intermodulation, so-called Doppler distortion, if the driver handles a wide bandwidth. The cone motion phase modulates higher frequencies.

[wogg]

That came up just recently. Rod Elliot's article is good on this: https://sound-au.com/doppler.htm

[billfitzmaurice]

http://www.readresearch.co.uk/loudsp..._article_1.pdf
http://www.readresearch.co.uk/loudsp..._article_2.pdf

[fpitas]

http://www.readresearch.co.uk/loudsp..._article_1.pdf
http://www.readresearch.co.uk/loudsp..._article_2.pdf

Oh, horns. Does anybody even use those anymore?

[Steve Lee]

They make the Tijuana Brass sound so good . . .