TYVM for your thoughts. I believe I understand the gist of what you are saying, and thanks for keeping it KISS, because I'm sure it otherwise would get very long and technical, very quickly.
As I pretty much know nothing about speaker design, at least compared to an expert like you, I do not even know what you exactly mean when Don's drivers are "only 90 degrees apart". In phase if I had to guess, but, I wouldn't think any amount of degrees are good? I also do not know how small these angles can typically be in a speaker, nor how large. I, of course, do not have the slightest inkling on why certain angles provide better transient response than another.
Ok, looking at the graphs, you must mean the amplitude response?
It's fine with me if you don't explain, if it's the case that your time would be wasted on a neophyte. I'll let you discriminate where typing might benefit a neophyte.
I understand your thoughts about how I was thinking about this backwards. I suppose I meant that if one had free access to any specific drivers, if one would choose a certain set if only due to preference of xover slope necessitated. I'm sure that's still backwards, as drivers should be picked firstly based on their performance for the money. I was being hypothetical.
IIRC, persons that prefer steeper slopes might've pointed out off-axis response as the main reason.
Anyways, thanks.
No, I do not mean the amplitude response. It is the third graph, labeled phase response. If you look at the crossover frequency you will see that the tweeter (green) is 90 degrees ahead of the woofers (red).
Now you state that any number of degrees is bad, and you are correct. However with any analog crossover it is inevitable!
Lets take a first order crossover. The tweeter is fed by a cap. In a capacitative circuit current leads voltage by 45 degrees. The woofer is fed by an inductor, and in an inductive circuit current lags voltage by 45 degrees, so the combined phase angle is 90 degrees.
Now as you add orders, you are really cascading filters, so you add 90 degrees of shift per order. So in a fourth order filter the angle is 360 degrees, so the tweeter is a full cycle ahead of the woofer at crossover. It is worse than that as woofers have deep cones, and the tweeter is physically ahead also.
Now the time delay of the shift can be calculated by the formula: - The speed of sound (roughly 1100 ft/sec) = frequency X wavelength. So it is easy to work out the time shift as for a first order filter it is a 1/4 wavelength and for second order a half wave length and so on. Then you have to add to the time shift, the distance the tweeter is physically ahead of the woofer's acoustic center. Then there is the matter of the driver offset, because the drivers are physically separated on the baffle.
There have been attempts to set the tweeter back. However the awkward cabinetry causes reflections and on the whole does more harm than good.
Coaxial speakers are a solution to a degree, with the cone of the woofer acting as a wave guide. However there are still reflections from the cone, and dispersion can be limited. However in my view coaxials have a lot to offer as center channel speakers. I'm a little surprised the DIY community on this forum has not embraced the SEAS coaxial driver. It is one of the best coaxial units around.
So I think you can see that all things being equal, reducing the number of crossover points has to be a good thing.
The up shot is, that harmonics get physically separated from their fundamental in time and space.
Now the veteran British designer Ted Jordan was of the opinion that any speaker that did that, was bound to have finite limitations, and I'm inclined to agree.
Ted was chief of research for Goodmans loudspeakers back in the fifties.
When Peter Walker launched his Quad electrostatic loudspeaker at the Audio Fair at the Hotel Russel in 1957, other manufacturers were shaken.
The management of Goodmans loudspeakers charged Ted with coming up with a competitive product. He went to work on a full range loudspeaker.
As the product was nearing completion, the owner died. The company had to be sold by his widow to pay the onerous death duties prevailing in the UK at that time. The new owners were only interested in short term profit. They did not want to spend precious dollars on R & D. Just trade on the name and roll out any old junk. So you see the seeds of the West's self destruction goes a long way back.
So Ted Jordan and the companies financial controller, Leslie Watts were laid off.
They formed
Jordan Watts loudspeakers, and produced the module and introduced it at the Audio Fair Hotel Russel in 1961. I was there for the introduction of the Quads and the modular loudspeakers. You can find the story there, especially from a link to streaming audio of an interview I did with Leslie Watts in 1976.
Ted and Leslie parted company as you will find out if you listen to the audio.
Ted has been involved in a number of ventures, but has now handed his business interests over to some
Swedish investors.
