I mean a decrease in the drivers Qts. More or less an increase in motor damping so you end up with less low frequency output without EQ. And the EQ is also a diminishing return rather rapidly as you will end up with power compression due to voice coil heating very quickly. Balance in the key to getting the greatest output from a system. A motor and the enclosure it sits in are a system. And that is inextricably linked together. A larger enclosure will almost always give you greater low frequency output when you are using a high BL motor. Larger enclosures are the area where a skilled DIY build can produce greater output than something available commercially.
Yes, I also mean decrease in Qts and increase in motor "damping". The quotes around damping clarify that it looks like a damping term in the mechanical equation of motion but the energy is not lost to heat as is the case with Qms or mechanical damping. Instead, the energy is transformed back to electrical energy in the coil and can be recovered by the amp, depending on the design. This actually allows one to make the system very power efficient. The EQ is needed to correct the response shape, but even with the necessary boost, actual power consumption will still be *lower*. You have to look at the impedance curve in addition to the frequency response curve.
Very high BL and thus very low Qts has a knee point where you are not getting anything for the greater motor strength versus the box size. You get no more output in SPL in the low end for the stronger motor. A very high BL driver is a poster child for that very situation. Too strong of a motor and no greater low frequency output for the outlay in cost to make the motor stronger. As with X-max BL has become a number to tickle the desire of people looking for something different. The truth is quite different from what is being marketed.
You keep using the term "low frequency output" where I think you mean "low frequency sensitivity". There is a big difference. Actual low frequency output depends on a lot of possible factors. Increasing BL or Qts could affect output in either direction depending on which factor is responsible for limiting performance. If the limiting factor is thermal power handling and compression, then the higher BL and lower Qts system will usually *increase* output. If, however, the limiting factor is amplifier voltage, then the reduced *sensitivity* (output per volt vs. frequency) around the resonance frequency (or the higher of the two impedance peaks in a ported system) will lead to *decreased* output there. As you say, it could go either way, and it depends on the system.
But to repeat what I already said. You can almost always increase sensitivity by reducing the coil resistance or running two or more coils in parallel in a MVC design. Or if you are also excursion limited, you can add drivers and run those in parallel. You just need an amp that can keep it together when presented with a lower DCR. Those spots where you need the EQ boost are typically where impedance is high and where not much power flows to the coils despite the higher voltages required.
As a simple math reality check. You have a great driver that has 25mm true X-max in a sealed enclosure so you can EQ the snot out of it. To get 3 db more what do you need? Another 25mm. If you had such a driver capable of the 50mm X-max in the same sealed enclosure you will not be able to get the full excursion before you massively exceed the drivers safe operating area. So to get the desired X-max you nuke the driver. Not so good for warranty purposes.
So buying a driver that for instance has 4mm more X-max looks enticing. But in reality you are looking at maybe another db in output.
Last but not least lets look at amp output.
Forgive me, but last I checked, doubling excursion gets you +6 dB not +3 dB. But anyway, I agree that excursion is also important, but I don't think this has much to do with motor strength and the efficiency benefit that it brings. If I can wring all the available excursion out of a driver with less power, that's a *good* thing. That means my driver will see less power throughout its operation than another driver with the same displacement capability that needed more power to get it there.
I'm aware of the limitations you speak of with real world amps. Indeed, they are all over the map in terms of what they can deliver and for how long, especially in the lowest subwoofer frequencies. If you haven't seen it already, a few members have done some tests to try to assess the real world performance of various amps. See, e.g. Luke's basic Amplifier Tests and Amplifier Comparison SpeakerPower SP2-12000 and Powersoft K20-DSP-Aesop. Both of the amps compared in the latter thread are serious performers, even if they can't sustain their rated power indefinitely. What's important is how long and to how low a frequency the amp can sustain high power output, and how impedance effects its behavior. Unfortunately, this kind of information is almost never available from the manufacturer, at least publicly.
Motor parts. Not just the magnet and the metal.
The spiders and the surrounds.
Few companies are actually tooling and doing the R&D for the spiders and the surrounds in their drivers. They simply buy what is available.
I've worked with a few companies that do exactly that.
The reality is that you need to do the math, and the engineering to get your drivers parts to work as a system.
An example. Many companies are using dual spiders. You stack two spiders spaced apart from each other at an arbitrary distance and everything should work out right. Or you could do a little thinking and design the top and bottom spiders as mirrors of each other. When one spider has the curve going up the other spider has the curve going down. So one is the reverse of the other in terms of rolls and valleys. The benefit? 2nd order distortion drops as you have a truly balanced restorative force. The spider on all larger high X-max drivers are the dominant part in the suspension of the voice coil and the centering of the voice coil.
TO get this to work properly you have to do design math that will prove out that you have the correct valley lengths and flats on the valleys and peaks. Grabbing open tooling and hoping that it will work is not what pushes the limits in driver design. It is the status quo. What everyone else is doing.
I just dropped a little over a grand in spider tooling. And it's not open! It's mine. Same goes for surround tooling. You want to push the limits in driver design you have to actually do the design work!
No doubt, these are all very important design details. To the extent that one may have to compromise suspension performance (or excursion) to accommodate a stronger motor, I can understand where "bigger" may not always be better. Of course, stronger motors effectively give the amp more control over the driver, so the relative influence of soft parts will tend to be reduced, at least as long as the driver is within its excursion limit. Small boxes also tend to reduce the influence of the suspension because the air spring is a lot stiffer than the suspension is. (Again, this is only true within the excursion limit.)
Anyway, I understand there are many practical situations in which one would want a driver with less BL than one with more. But high BL (all else the same) does generally improve efficiency, that is, output per watt of power. It just comes with side-effects that must be taken into account in the design of the overall system. The amp is a critical part of that system, and it's a shame that so little information is available concerning their behavior.