BR-Horn

    Hello,

    Hope english is not a problem over here...

    I would like to examine a BR-horn in somewhat more detail. I am talking about a horn which ports the rear chamber back into the horn.

    If you look at the electrical equivalent schemes for a horn and a BR cabinet it seems to me it should be possible to gain SPL below the horns midband by making sure the added capicitance of the BR-port equals out the inductance formed by the rear chamber and the stiffness of the cone.

    What do you guys think about this idea? Probably there are people who have build this kind of construction. Did they find any rules of thumb?

    Best regards,

    Walt

    I am waiting anxiously to see the result of Martin King's next take on his model as his current sheets provide a more detailed take on what happens. The trick is the placement of the different nodes (for want of a better word of positioning of driver, throat, mouth, etc) and to optimize these for a given application.

    The phase shift of an offset alignment is said to improve driver dampening over a wider bandwidth than bass reflex, something which can be seen in a damped TL in ML's backloaded horn models.

    Manfred Schuster (aka Sound Klinik on this board) considers himself quite the expert on the matter, but is rather sparing with concrete design considderations as he does these things for a living. His implementations seem to have very good bandwidth allowing integration with smaller horn-tops.

    Now ... where is that freeware, simple to use FEA/BEA model which accepts a T/S input for driver models...

    Zitat von "Timber_MG"

    His implementations seem to have very good bandwidth allowing integration with smaller horn-tops ...

    The idea of the 18.BRH was simply allowing a higher x-over frequency and good efficiency by a (rather ...) small sized box.

    Hybrides: Many people "think" in formulas when developing a loudspeaker system (... some believe that developing means the use of simulation software ...). A theoretical model (that first must be discovered ... :wink: ) ist useful when calculating the right dimensions ... but the first step is always a good idea / basic plan.

    A vent that radiates directly into the room is ONE way (that can easily be modeled ...). Always imagine, what could happen to the acoustic wave when trying some different "ways" ... while looking at some "sinful" frequencies ...

    I will shoot from the hip (Manfred, your dissection is welcome, you may call a spade a spade, I for one don't have the resources to experiment much with more difficult to model enclosures): If one models the rear output of a horn in hornresponse (i.e. ignoring the front), one finds a peak at fc and another one a little bit higher up (i.e an imperfect low-pass)

    Now if one adds in the front speaker in hornresponse, one can only do so as a scoop (delay the front wave by the effective length of the rear horn) with perhaps a little less offset). However one cannot considder the loading on the front of the cone.

    Now the front horn (assumign a slightly asymetrical configuration meaning speaker not at the "port") should presumably minimise driver excursion at the second peak to prevent the rear output in the mid-bass (generally around 90Hz) from destructively interfering with the front wave OR some part of the system must modify the acoustic phase of the front/rear (but not both) wave (remember, a "horn" is a refractive device) in such a way as to re-inforce the rear-wave acoustically (such as the helmholz resonance of a BR)

    At high frequencies the front airpath will dominate, though there might be a couple fo resonant cases with teh rear chamber and 1/2 WL offset from the driver)

    Now this is a fair jungling act as the two systems are inter-dependant. The FEA would be nice to visualize the resonant behaviour of the system, especially at higher frequencies.

    Becasue of higher bandwidth, the system should be as un-resonant as possible. Now the problem is , that at high levels the speaker will produce more higher frequency material than in its input and if this excites a resonance (often seen as a null in the FR plot). This can be avoided either by significantly reducing the mechanical excursion of the speaker (Manfred claims significant broad-band loading as does Tom Danley in his tapped horn) or by moving these resonance far out of band (see Syrincs's BR with holes in the ports and high excursion woofers)

    Personally I just don't have the ersources to play around with this concept, so I would rather get a resonable idea of what's going on by simulation than build many boxes (as many have done apparently)

    Zitat von "Timber_MG"

    Becaue of higher bandwidth, the system should be as un-resonant as possible.

    ... regarding the time-domain: rather good idea ... :)

    Zitat von "Timber_MG"

    Now the problem is, that at high levels the speaker will produce more higher frequency material than in its input and if this excites a resonance (often seen as a null in the FR plot). This can be avoided either by significantly reducing the mechanical excursion of the speaker (Manfred claims significant broad-band loading as does Tom Danley in his tapped horn) or by moving these resonance far ...

    ... you have to increase the mechanical "load" that the speaker "sees" ... this will cause a "broadband loading" ...

    ... another question: Did you - in the meanwhile - build some of the "SK-Bastelhorns" (using here so-called MDF ...) as you might want to ?

    Hallo Manferd

    Ich bin mindestens fuer die naechsten 4 Monate in 600km Entfernung von zu Hause bei einem Praktikum und kann ausser sehr kleinen Projekten (bzw. Simulationen) leider wirklich nichts machen. Ich habe aber erstmal 6 Stueck bei einem Schreiner im Auftrag und werde hoffentlich vor Ende des Jahres berichten koennen.

    MDF oder pine(=tannenholz aus Botswana)-ply, je nach dem ob es mit dem Geld stimmt. (MPX ist hier so teuer, dass ich nich einmal auf die Idee komme, da gehe ich lieber gleich auf Verbundstoffe)

    Der Rest des Systems werden doppel 12"/1.4" an einem Oblate Spheroidal Waveguide wenn's interessiert. Hier ist alles komisch was Preise anbetrifft, Eminence ist sogar teurer als 18Sound.

    Martin

    Zitat von "Timber_MG"

    ... oder pine(=tannenholz aus Botswana)-ply, je nach dem ob es mit dem Geld stimmt. (MPX ist hier so teuer, dass ich nich einmal auf die Idee komme, da gehe ich lieber gleich auf Verbundstoffe)

    ... ist Tanne nicht etwas weich ?

    Zitat von "Timber_MG"

    Der Rest des Systems werden doppel 12"/1.4" an einem Oblate Spheroidal Waveguide wenn's interessiert. Hier ist alles komisch was Preise anbetrifft, Eminence ist sogar teurer als 18Sound.

    ... mit "Oplate" hatte ich bisher nicht "das Vergnügen" ... willst Du Dir eine Art Array bauen ?

    Ja, aber es ist leicht und vor allem billig (um ein 8-10 faches im Vergleich zu Birke Multiplex hier) und wenn es verstrebt wird.... Ist allemal besser als Chipboard (welches oft fuer Prototypen verwendet wird, spanplatte?)

    Die Idee ist, vorlaeufig zu sehen, ob ich mit einem einzigem Top mit 2 Baessen pro Seite zurecht komme. Evtl. vieleicht mit drei Modulen pro Seite bei hoher Trennung je nach Musikstil (es gibt verschiede Meinungen zu der Moeglichkeit, aber die Hornkontur ist nach Simu ja nicht optimal fuer den Bass (aber bei proportional geminderter Auslenkung welches unten rum bei den meisten Chassis sowieso das Limit darstellt) und die Reaktanz in hornresp gibt einige Hinweise im Vergleich zu hyperbolischen Konturen auf moegliche Gruende diesbezueglich.

    Dazu ein paar (erstmal 4-6) 12"/1" Mufus und evtl einen kompakten 15" Bass a la Achenbach.

    Die Waveguidekontur formt eine Sphaerische Wellenfront und hat wie Tractrix kaum einen Eigenklang, soviel dazu (Konstruktion aus "fibre glass - foam composite")

    Wir kommen aber ganz vom Thema ab ;)