T/S parameters valid for normal listening levels.

[Scottmoose]

I would have thought that would be a clear indication how I an defining distortion

Right, but it depends on type of distortion as they're in two distinct fields -linear & non-linear. Linear (aka frequency response) is essentially invariant within the designed mechanical operating range, so there is no variation with level. Non-linear certainly does vary with drive level though, even within that intended operating window -now much depending on the quality of the motor design etc.

[Nick]

I would have thought that would be a clear indication how I an defining distortion

Right, but it depends on type of distortion as they're in two distinct fields -linear & non-linear. Linear (aka frequency response) is essentially invariant within the designed mechanical operating range, so there is no variation with level. Non-linear certainly does vary with drive level though, even within that intended operating window -now much depending on the quality of the motor design etc.

It doesn't matter at all, but as you say there are two types of distortion, linear and non-linear, and I think when I say

distortion, that is a result of it being somewhere non linear.

It would be clear which of those two I was talking about.

[Scottmoose]

I don't think that this is true
When you look at the Klippel report I posted, FS changes significantly with excursion from 63.4Hz to 43.0Hz
Unless its effect is cancelled out by similar changes in Qts, Rms, Cms, Vas etc.
It won't affect my speaker too much since I have a small sealed box, but it could show up in others
Brian

Not exactly. You're in the acceleration (rising response) region here, where as noted you can and do get variations with drive level due to the electrical & mechanical non-linearities / asymmetries in motor design, suspension[s] etc. Klippel is very good at revealing the reasons for those, but you can find the effects easily enough with a simple REW or DATS etc. measurement setup. As you transition into the mass-controlled (nominal 'flat') BW these cease to have influence assuming you're in the design operating window & you'll struggle to find any changes in linear distortion other than those caused by external / mostly external factors. Non-linear, sure.

[Scottmoose]

It would be clear which of those two I was talking about.

I prefer not to second guess. It usually lands me in a world of trouble when I do.

[brig001]

Not exactly. You're in the acceleration (rising response) region here, where as noted you can and do get variations with drive level due to the electrical & mechanical non-linearities / asymmetries in motor design, suspension[s] etc. Klippel is very good at revealing the reasons for those, but you can find the effects easily enough with a simple REW or DATS etc. measurement setup. As you transition into the mass-controlled (nominal 'flat') BW these cease to have influence assuming you're in the design operating window & you'll struggle to find any changes in linear distortion other than those caused by external / mostly external factors. Non-linear, sure.

I don't really understand all of this, but changing FS moves the transition to the mass controlled region with excursion, and in my case by over half an octave down. This has to change frequency response of the driver with excursion, doesn't it?

Brian

[Scottmoose]

Other than in the sense that you have a fairly minor (relatively) shift in the rising / flat BWs, no.

[brig001]

I guess we would have to put all of the large excursion Thiele-Small parameters in WinISD or similar to know the total effect

[Scottmoose]

On alignment? Yes, unless you want to do it the old fashioned way on graph paper over an hour or so (I don't either) This is one reason why if you have several sets of parameters, it's usually a good plan to design for whatever set requires the largest Vb so you have some room for manoeuvre.

That said -we can sometimes go a bit far down the rabbit hole. We often say 'measure T/S values at high voltage as it's more representative' -which is true to a point. But the small signal parameters are what they sound like: small signal, and the counter argument says that within the driver's intended linear operating window, they (including Fs) shouldn't be changing a whole lot, and if they are, it's because the driver is being pushed into a non-linear operating range. For e.g. the attached is from the DATS measurement overview of an RS100 at different drive levels. It's biased of course, but it's a reasonable point, which Small, d'Appolito etc. also share (in context). Sometimes manufacturers will provide values at higher drive levels knowing that it's potentially shifting into a non-linear operating region, but accepting that's likely to happen, and that it can look better on paper. They're both right really within the different contexts they're talking about; it's partly due to a lack of industry standards.

[rowuk]

I prefer to think of Thiele/Small as having nothing directly to do with optimising distortion or sound quality, rather just providing a desired level of “bass”. The inaccuracies of a proper alignment for given parameters are surely miniscule compared to room effects.
In theory, a bass reflex (Helmholtz) port outputs 100% distortion as the resonant mode is usually not based on driver linearity, rather excited air pressure - often at an unrelated frequency.

[brig001]

I think whether room or speaker dominates depends on their relative frequencies. My living room has a strong peak at 50Hz, a large dip at 40Hz, and there’s a big suck out on the left speaker at 70Hz. A bass reflex cabinet tuned to 80Hz would be speaker dependent, below that, the room definitely dominates

Brian

[rowuk]

I think whether room or speaker dominates depends on their relative frequencies. My living room has a strong peak at 50Hz, a large dip at 40Hz, and there’s a big suck out on the left speaker at 70Hz. A bass reflex cabinet tuned to 80Hz would be speaker dependent, below that, the room definitely dominates

Brian

Tuning a ported box to 80 Hz, means that there is NO restoring force from the box below 80 Hz and that means if there is content below that frequency, the woofer "flops" in a very unrestrained matter - dramatically raising distortion, chuffing noise and the ability of the woofer to mechanically bottom out. If we try and couple a speaker like this to a subwoofer, we have all sorts of integration issues. Plugging the port raises the -3dB point usually to over 120 Hz further making the subwoofer easy to locate - instead of invisible.

[Scottmoose]

I'd generally call that an example (and understanding it's purely meant as an example) of bad design rather than an issue with vented boxes per se.

Any vented enclosure will ultimately unload 4th order at some point below Fb -even a nominally aperiodic [open] TL. It's really just a question of at what frequency & how low an amplitude it's already hit. That's one reason we had / have rumble filters etc. But they load the driver quite efficiently around Fb and across its operating passband, so ideally, you should be taking the BW and main power-band of the material you intend to be listening to into account, and select drivers & alignments accordingly -at least as far as other considerations allow anyway.

[brig001]

I have a LR4 high-pass at 45Hz just to be safe
SWMBO likes karaoke with her friends, and you only need to drop a microphone once…

[Scottmoose]

That'll do it.

[rowuk]

One could argue that a bass reflex design with an 80 hz tuning is the huge mistake if we like bass - regardless of what subwoofer that we use.
Then would come the discussion of how to even cross over to an 80 Hz tuned bass reflex design - plug the port and cross at 120 Hz or whatever.