Its a sad fact that this thread started off about measurements and moved past that to topics completely superfluous.
Its good to see that with the exception of the insult, people agree with me(cheers for the PM's), I don't always write in the best style but I certainly know what I'm talking about having large amounts of practical experience unlike the arm chair designers, especially in regards to ATC where they can only summise based on virtually zero measured data. Its a shame that my enthusiasm and knowledge is undermined by a couple of individuals.
The acts in correcting my misinformation peddling are commendable(and transparent) but I suggest SM move on.
Despite the drama I still think a good discussion can be had.
Dunc:
Good questions (some at last!)
Most of the ATC loudspeaker range, with the exception of the lower end, use SL or Super Linear drive unit technology. SL is basically a non-magnetic material in the bass driver's motor which helps decrease higer order distortion products caused by differences in the magnet force relating to the magnet material leading to an effect called hysteresis.
ATC use this driver exclusively in their own products, its not available to the DIY'er nor to OEM's. Instead what they have available is a range of drivers that are built on the same basket, cone material and voice coil(VC) but have a motor sans SL technology. Available to the DIY'er are two variations on the 9" and 15", the variation are differing VC topologies. The short coil version is an underhung design which offer a more linear BL product with varying excusion as opposed to the long coil version which is overhung and offers higher efficiency and output.
I use the 9" SB75-234SC which is the shortcoil/underhung design.
So ATC use a different bass driver that I have no measured data on and only subjective experience. I've said it before but one unlikely reason for the lower XO point on the bass/mid transition is that the ATC SL drivers don't go up high enough to be usuable in those regions. So it could have been a lesser of two evils - problems with the mid or problems with the bass driver. However given the measured performance I have on the 9" SC - that shows its quite linear upto 2.5Khz, I'd suggest that that not be the case as with 24dB/Oct filtering the response would be around 55dB down at that point.
Here's the measurements I took on the 9" SC late last year:
Conditions were 1w/1m.
In light of this and the fact that ATC wouldn't design a bass/mid driver for a 'high performance' loudspeaker intended to cross at 380hz that didn't actually extend a good way past that, I'd suggest that wasn't the reason for the lower XO point.
So its unlikely that the bass driver can't actually cross higher. What does that leave us with?
Another consideration is voicing and maximising the range that mid driver covers. Crossover points are often ugly things and having one right at 500hz is in one of the ears more sensitive regions so lowering it may have been subjectively better at the expense of linearity of the design as a whole since they're using the mid where distortion starts to sky rocket near driver resonance, However, given digital crossovers, phase problems at the XO point can be virtually eliminated on axis and the ATC also lends itself well to a superb power response which also means good linearity off axis too. This was the gist of my point all along, the digital XO's allow to move the XO point up, improve driver integration and allow more linear operation. Whilst we know that ATC use the 9" SL rather than the SC we can take it as a given on the performance of the mid because that is exactly the same as the one I've shown data on here and distortion is bad low down, so being sensible about things we'd like to avoid that rather than start to use it at 380hz with shallow filters.
I seriously doubt any ATC representative would touch this thread given what's gone before. A real shame because they'd be able to say exactly why they did what they did.
As it stands we have measured data on the ATC mid which shows where the driver is most linear and where its not. Someone designing the for the most linear operation will make sensible decision based on that data.
The debate isn't "ATC made a bad speaker but why?" But rather "hey this can improved for more linear operation". Of course I have radically different XO topologies to the ones ATC decided to use and these are both more flexible and costly. The ATC mid benefits from these more extensive crossover technologies.
