Jim, I meant "Sonata" - sorry. I put one too many syllalables in the word.
In general, the greater the size discrepancy between the drives, the greater the radiation pattern discrepancy at the crossover frequency. The midbass drivers in an MTM array have a much wider horizontal than vertical radiation pattern, and in my opinion this is undesirable if you use a dome tweeter.
Let me try to explain why I think the radiation pattern matters. I'm evidently in the minority on this; most designers don't go out of their way to deal with it, so don't take my word a gospel.
Okay, the ears derive timbre not only from the first-arrival sound, but also from the reverberant sound within the room. At listening distances of more than about five feet, most of the energy that reaches your ears is reverberant sound. Now your ear/brain system suppresses directional cues from this reverberant sound, but you're still picking up loudness and timbre and ambience cues from it.
One characteristic components of live music is, a natural-sounding reverberant field. Live voices and instruments even sound natural from the next room, where there is no line-of-sight to the instruments and therefore no direct sound; only reverberant sound. In contrast, very few loudspeaker systems sound convincing from the next room. One reason is, their "power response" (summed omnidirectional response) is very uneven. I believe this is one of the major causes of the perceived difference between live and reproduced music.
Let's look at why the power response of most speakers is uneven. No doubt you're familiar with "beaming". As an example, let's take a 6.5" two-way speaker, crossed over to a 1" dome tweeter at 2.5 kHz. The woofer's radiation pattern is wide wide at low frequencies, but then it starts to narrow as the wavelengths shrink to where they're comparable in size to the diaphragm's dimensions. So the pattern has narrowed to maybe 90 degrees in the crossover region. Then the tweeter takes over, and the radiation pattern blooms out wide wide again. Eventually the tweeter also starts to beam, but usually not until the very top octave.
The problem region is the lower end of the tweeter's range, roughly 2.5 to 5 kHz. In this region the speaker may well be putting out 8-12 dB more energy into the reverberant field than at most other frequencies. And this just happens to be the region where the ear's sensitivity is the greatest. So what happens is, the extra reverberant field energy in this region makes female vocalists sound a bit harsh, and the eventual result is often listening fatigue (I could explain why if you want).
Now the MTM format's vertical radiation pattern on the M's side of the crossover will be comparable to that of an oval driver the width of one of the midranges but the height of the whole MTM stack. So in the vertical plane, the radiation pattern will be much narrower than in the horizontal plane. The result will be a significantly greater net radiation pattern discrepancy in the crossover region than if we just had a single midrange driver.
Of course there are advantages to MTM's. Much better pattern uniformity in the vertical plane with odd-order crossovers; higher power handling; and reduced floor-bounce reflection energy due to the narrowed vertical pattern (though the vertical pattern will usually be wide once again in the 4 kHz region, where the ear's sensitivity peaks, because that will be produced by the tweeter).
So it's a trade-off. Surprise, surprise.
I'll try to comment on impedance matching later.
Duke
In general, the greater the size discrepancy between the drives, the greater the radiation pattern discrepancy at the crossover frequency. The midbass drivers in an MTM array have a much wider horizontal than vertical radiation pattern, and in my opinion this is undesirable if you use a dome tweeter.
Let me try to explain why I think the radiation pattern matters. I'm evidently in the minority on this; most designers don't go out of their way to deal with it, so don't take my word a gospel.
Okay, the ears derive timbre not only from the first-arrival sound, but also from the reverberant sound within the room. At listening distances of more than about five feet, most of the energy that reaches your ears is reverberant sound. Now your ear/brain system suppresses directional cues from this reverberant sound, but you're still picking up loudness and timbre and ambience cues from it.
One characteristic components of live music is, a natural-sounding reverberant field. Live voices and instruments even sound natural from the next room, where there is no line-of-sight to the instruments and therefore no direct sound; only reverberant sound. In contrast, very few loudspeaker systems sound convincing from the next room. One reason is, their "power response" (summed omnidirectional response) is very uneven. I believe this is one of the major causes of the perceived difference between live and reproduced music.
Let's look at why the power response of most speakers is uneven. No doubt you're familiar with "beaming". As an example, let's take a 6.5" two-way speaker, crossed over to a 1" dome tweeter at 2.5 kHz. The woofer's radiation pattern is wide wide at low frequencies, but then it starts to narrow as the wavelengths shrink to where they're comparable in size to the diaphragm's dimensions. So the pattern has narrowed to maybe 90 degrees in the crossover region. Then the tweeter takes over, and the radiation pattern blooms out wide wide again. Eventually the tweeter also starts to beam, but usually not until the very top octave.
The problem region is the lower end of the tweeter's range, roughly 2.5 to 5 kHz. In this region the speaker may well be putting out 8-12 dB more energy into the reverberant field than at most other frequencies. And this just happens to be the region where the ear's sensitivity is the greatest. So what happens is, the extra reverberant field energy in this region makes female vocalists sound a bit harsh, and the eventual result is often listening fatigue (I could explain why if you want).
Now the MTM format's vertical radiation pattern on the M's side of the crossover will be comparable to that of an oval driver the width of one of the midranges but the height of the whole MTM stack. So in the vertical plane, the radiation pattern will be much narrower than in the horizontal plane. The result will be a significantly greater net radiation pattern discrepancy in the crossover region than if we just had a single midrange driver.
Of course there are advantages to MTM's. Much better pattern uniformity in the vertical plane with odd-order crossovers; higher power handling; and reduced floor-bounce reflection energy due to the narrowed vertical pattern (though the vertical pattern will usually be wide once again in the 4 kHz region, where the ear's sensitivity peaks, because that will be produced by the tweeter).
So it's a trade-off. Surprise, surprise.
I'll try to comment on impedance matching later.
Duke