Coaxial direct radiator/compression driver units...

[454Casull]

Many coaxial units are basically a compression driver firing through a hole in the middle of the motor structure of the larger direct radiator. Claims are made that they are point-source units... but isn't the acoustic center of the compression driver behind that of the direct radiator?

Something like [IMG]this.http://www.audioheritage.org/images/.../604-xsect.jpg[/IMG]

[GM]

Correct, as it should be to account for the XO's delay, though historically the offset/XO point/slope was not optimized.

GM

[454Casull]

So if the compression driver's signal was delayed just enough so that it would arrive at some point on-axis at the same time as the direct radiator, would the unit act as a point-source?

I asked a question in this thread... Check it out for yourself.

http://diyaudio.com/forums/showthrea...539&highlight=

Specifically, BillFitzmaurice's reply.

[dwk123]

There are probably a couple different ways to define 'point source' depending on what you're trying to prove.

One would be that the source exhibits no lobing, which requires that there be no path-length differences to generate cancellation. In other words, it behaves as though the acoustic centers are in the same spot.

A more theoretically correct definition would probably add the constraint that a 'point source' exhibit minimum-phase amplitude response. This is much tougher, and is only realizable with first-order acoustic xovers or via the use of non-minimum-phase digital xovers.

In terms of your original question, the coax drivers usually control the dispersion of the tweeter so that *when properly time aligned by the crossover*, the two units will act as a point-source over the main radiation area. I believe that many good coax setups satisfy this criteria - the Tannoy dual-concentric studio monitors for example claim to do this.

[GM]

Pretty much, though it's truly only phase aligned at the XO point, but then many so-called coherent point sources (FR drivers) can have numerous mechanical XOs, so their acoustic phase isn't all that perfect either. Anyway, a driver is no longer a point source once its radiator is > the WL it's producing, so in the BW we're discussing it's not a point source, just physically ~coherent.

Not sure what your point is WRT the other thread, but comparing an inline driver layout where the HF is behind (and a part of), the LF is way different than with the HF well in front and electronically delayed since the former will not suffer from the off axis comb filtering that the latter will have mass quantities of due to the LF's diaphragm acting as horn sufficiently large enough to confine the HF's output over most/all its BW.

GM

[454Casull]

http://tinypic.com/4me1g

How do they determine where the LF sources lie? Obviously the HF source is basically the compression driver (another question - why is the compression diaphragm convex in the Tannoy unit, instead of concave?), but why do they indicate the LF source to be so aft of the LF voice coils (not to mention the LF diaphragms)?

[GM]

Before it could be measured, it was assumed that the acoustic center of the motor was where all sound originated. We now know that it is frequency dependent.

Because of the way they chose to design the phase plug.

For all the components of a signal to arrive on time, the fundamental must lead the 1st harmonic, etc., so the driver's acoustic center marches backwards with increasing frequency. Measure a fairly coherent 'fullrange' driver and you'll find that the HF appears to be coming from some point behind the driver, ditto the compression driver.

I assume they are showing that the acoustic centers of the two drivers are one and the same at this physical point and that this is the XO point, whereas some (mythical?) coax has this much misalignment. Anyway, the pic is just typical marketing BS since it doesn't delve into the subject enough, like how the XO and polarity impacts it, ergo the coax can be made 'coincident' with the correct XO point/slope and polarity connection.

GM