traditional midrange-tweeter separation and lobing vs. supertweeter

[JoshuaTechnomage]

When I first started learning about speaker design many years ago one of the issues often discussed was paying strict attention to the acoustic center separation of the 2 drivers vs. crossover frequency. It was said that both drivers should be kept as close together as possible, preferably less than than the wavelength of the crossover frequency.

However many speakers seem to ignore this "rule" and particularly in the case of supertweeters crossed over very high it is not possible. However in many cases it does not seem to be detrimental to sound quality according to reviews I have read.

I wonder if it's only a serious issue in a narrow passband directly related to the crossover frequency? Where the default separation of drivers is close to the wavelength of the crossover frequency creating odd lobing behavior? But is not a serious issue if the drivers are much further apart or much closer together?

[planet10]

The reason that one can often get away with this, is because the XO is way up where the ear is less sensitive.

It can be a real issue with a typicak 2-way where the XO is right in the ears most sensitive area (~1.5-4kHz)

dave

[JoshuaTechnomage]

Perhaps, but I think this may be interesting to experiment with when I have some drivers available. If I could find the right (identical) drivers that would allow me to test:

1. Coaxial and time-aligned (the woofer will effect the tweeter wave-launch unless the tweeter is mounted flush or forward of the apex [largest part] of the woofer cone in which case it will need to be time-aligned by other means.
2. Mounted separately so the acoustic centers are equal to the crossover frequency compared to slightly off (also time-aligned).
3. Mounted much further apart compared to the crossover frequency (again time-aligned).

Maybe a Neodymium Scanspeak tweeter and 8" Peerless or similar woofer would be good for this experiment with a crossover ~2khz.

[Drew Eckhardt]

Originally posted by JoshuaTechnomage
When I first started learning about speaker design many years ago one of the issues often discussed was paying strict attention to the acoustic center separation of the 2 drivers vs. crossover frequency. It was said that both drivers should be kept as close together as possible, preferably less than than the wavelength of the crossover frequency.

The only natural sounding speakers I've heard have avoided wide separation via some mechanism and/or had reduced off-axis output so you'd expect a smaller power response notch about the cross-over region, etc.

Low cross-over (1440Hz) dipole with exceptionally beefy tweeter, mid-tweeter (1KHz XO, 3-4" between acoustic centers), coaxial mounting, full range plus back-horn or bass driver.

However many speakers seem to ignore this "rule" and particularly in the case of supertweeters crossed over very high it is not possible. However in many cases it does not seem to be detrimental to sound quality according to reviews I have read.

Most reviews compare speakers to other speakers rather than to actual musical performances.

[JoshuaTechnomage]

Originally posted by Drew Eckhardt


The only natural sounding speakers I've heard have avoided wide separation via some mechanism and/or had reduced off-axis output so you'd expect a smaller power response notch about the cross-over region, etc.

Low cross-over (1440Hz) dipole with exceptionally beefy tweeter, mid-tweeter (1KHz XO, 3-4" between acoustic centers), coaxial mounting, full range plus back-horn or bass driver.



Most reviews compare speakers to other speakers rather than to actual musical performances.

Ahh yes, almost forgot about the dispersion characteristics between two different drivers at the crossover frequency. Testing the ears sensitivity to driver dispersion vs. driver separation for a given crossover would require a whole other set of experiments. I couldn't properly test a coaxial configuration with a small midwoofer (say 4" or 5") since a tweeter mounted in front of it would likely cause severe interference with the higher frequencies coming off the midrange/woofer. (At least more severe than with a much larger midwoofer)

Someone should make a flat diaphram woofer with a large diameter voice coil and mount a tweeter in the center. You can make the woofer cone out of honeycomb sandwich with kevlar or carbon fiber, maybe like Eton uses. Use a felt/foam ring around the coaxial tweeter so it doesn't reflect off the woofer diaphram so the woofer excursion doesn't alter the high frequencies.

[JoshuaTechnomage]

Someone should make a flat diaphram woofer with a large diameter voice coil and mount a tweeter in the center. You can make the woofer cone out of honeycomb sandwich with kevlar or carbon fiber, maybe like Eton uses. Use a felt/foam ring around the coaxial tweeter so it doesn't reflect off the woofer diaphram so the woofer excursion doesn't alter the high frequencies.

Someone could even market a specially designed 15" woofer like this with an 8" voice coil and design the central area to accept any of the popular 6" full ranges such as Lowther or Fostex for example with a special designed sealed chamber in the rear so the woofer backwave doesn't muck up the full range driver.