Levels- and spectra-pathing symmetrical arrays of monopoles - diyAudio
Go Back   Home > Forums > Blogs > Grasso789
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Rate this Entry

Levels- and spectra-pathing symmetrical arrays of monopoles

Posted 6th January 2017 at 05:23 PM by Grasso789
Updated 4th February 2017 at 08:33 PM by Grasso789 (Bits)

This is about spectra-pathing ("multi-way") loudspeakers made of several monopoles, in explanation omni-directional sound sources, but I really mean direct radiators as opposed to horns, which are more uni-directional. I think about home, rehersal and club loudspeakers, not stadium ones. While spectra-pathing enables large power at wide bandwidth, seams between paths are usually audible, altho that could have been avoided rite from the start, at least technically. But instead of fixing problems at source, one has relentlessly used power to paint over them.

Computers can cover up phase distortion, which has been introduced by common spectra-pathing, and improve beamwidth. But the natural solution is a symmetrical arrangement of lopass drivers embracing one hipass driver, an arrangement commonly referred to as D'Appolito arrangement, combined with a cross-over of first order.

When Germans started to loose ground in World War II, widerange drivers playing from Large E to 10 KHz were state of the art. Mister Eckmiller developed a co-axial two-spectra-paths driver consisting of hipass encircled by lopass piston. But for mass production just to lift treble power tweeters were added to existing designs. Due to interference between lo- and hipass, amplitude is rarely flat. If the manufacturer wants to publish a plot, he chooses the rare case occuring at only one microphone position. When stereo came up, it became even worse. A common stereo has just two loudspeakers placed far apart. If this is used to play mono content, of whom there is plenty even in most stereo recordings, then certain frequencies (slitely different ones for each ear) are canceled due to runtime difference between left with rite channel. But if such stereo is used in a chamber, AND if its loudspeakers are quite dispersive, then reflections off walls fill in missing frequencies. In cinemas and concerts power is important, too: Each recipient receives his random amplitude, but visuals and athmosphere fill in missing bits. The sum of all recipients' opinions is power, say success of the movie or record.

It took until the eigthies, that Joe D'Appolito, who had looked into spectra-pathing, pointed out, that a symmetrical woofer-tweeter-woofer arrangement may have phase difference between hi- and lopass as occuring in a cross-over of odd order and will still direct its beam up-front. In contrast to that, an insymmetrical woofer-tweeter arrangement, which has phase difference between hi- and lopass, tilts the beam at cross-over frequency. Another way to get around this beam tilt is an even-order also called Linkwitz-Riley cross-over. Practical guys favored Linkwitz-Riley, while idealistic ones kept being attracted by D'Appolito. I want a cross-over of first order due to its perfect time and transient behaviour AND simplicity.

The only well known serious application of the D'Appolito trick is JBL's CBT (for Constant Beamwidth Transducer) models, which were introduced AD2008, and for which the designers obtained a patent by treating aforementioned phase difference as delay, which in conjunction with physical dimensions and relative driver amplitudes determines beamwidth.

AD2000 Don Keele had thoroly researched, how a driver array has to be in order to produce a smooth beam of certain width. His CBT concept uses levels-pathing --he calls it shading-- but not spectra-pathing. All CBTs are symmetrical, even if only by means of using ground plane as mirror. The simplest CBT, omni-directional, is a ring made of twelve (one per hour) identical drivers mounted close together. (The more space is wasted in surrounds and basket rims, the worse acoustic response becomes. Ideally, pistons nearly touch each other.) The most powerful CBT built up to date, also omni-directional, may be an Atmende Kugel used to excite halls evenly.

I do not strive for a loudspeaker containing dozends of drivers and having very wide or narro beamwidth. I rather have a few drivers making a frontal beamwidth of cosine also known as dipole in the midrange, where it pays off. I strive to have stability and smoothness of convex baffle. I notice, that D'Appolito and CBT both use levels-pathing: In case of D'Appolito, each lopass driver is half as loud as hipass driver at cross-over point. A three-drivers CBT has each outer one approximately 6dB down.

Misters Keele with Horbach have combined symmetrical arrangement with computer-run linear-phase cross-overs, simplifying and straightening design --after the complexity and time delay of a Finite Impulse Response filter became accepted--, so beamwidth becomes a nifty function of driver sizes and relative levels only. I use computers for the design stage only, not within the running loudspeaker.

I have a three-spectra-paths "Krassolito" box, which I run as an intensity- also known as co-incident- or syncro-stereo loudspeaker in my narrow living room. Left and rite midrange are L and R, while woofers and tweeter are fed L+R from a third amplifier. Mésure-là one Bel channel seperation at 4 KHz at an angle 1h off axis. This means a less obvious but wider stereo sound, since it is told to a calm listener solely by reflections off walls. I am quite satisfied.
Posted in Uncategorized
Views 219 Comments 0
Total Comments 0

Comments

 
Hide this!Advertise here!

New To Site? Need Help?

All times are GMT. The time now is 08:07 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2
Wiki