I'm saying the horn probably isn't loading the driver well enough below 1000 hz because of it's length, real mouth dimension and flare rate to cross it that low and expect a flat response without eq - like you use. Correcting a horn for a flat on axis response beyond where it's acoustically loading to me is heresy - but that's me 
I am afraid that I still have to disagree. The round-over will not have any significant effect on the "pattern flip" at the lower end of a waveguides pass-band, because it is just too small. The pattern flip will be just about the same with or without the round-over. The round-over is effective at higher frequencies and helps to minimize diffraction and reflection from the baffle edge. These are both good things to suppress and so a round-over is essential, but its effect on directivity is minimal.
I'm saying the horn probably isn't loading the driver well enough below 1000 hz because of it's length, real mouth dimension and flare rate to cross it that low and expect a flat response without eq - like you use. Correcting a horn for a flat on axis response beyond where it's acoustically loading to me is heresy - but that's me
To me, the whole concept of "horn loading" is a bunch of KraP. It is a hold over from Webster's approach and has no real validity in the context of what we are talking about. I "correct" the "axis response" of all my systems well below what you would call "cutoff" (another useless concept.) I see no problems - as long as you have the excursion capability, which I do. It's the polar collapse at the low end that is an issue to me and that cannot even be approached with Webster's Theory. Acoustic loading and polar control have nothing to do with one-another.
But IIR correctly you have stated that Webster's equations could be "valid" at low frequency.
I have to agree with Dr. Geddes. The measurements I've seen of the AH425 show little change between 1 and 3 meters, and fit Bjorn Kolbrek's BEM simulations pretty closely.
The AH425, at 16.5" diameter, is similar in overall size to the integral OSWG of the Summa enclosure. The Altec/GPA 416 and 515 are likewise similar in size to the Summa 15" woofer.
The 2-way Lansing Iconic dates back to the late Thirties, has drivers with similar radiating areas, and has similar crossover frequencies. The Altec Model 19, from the Seventies, is another variation on the 2-way medium-sized horn/waveguide and 15" direct-radiator bass-driver theme. It's a good concept that's been around for a long time.
P.S. Bjorn, congratulations on your new degree! You worked long and hard on that one!
The AH425, at 16.5" diameter, is similar in overall size to the integral OSWG of the Summa enclosure. The Altec/GPA 416 and 515 are likewise similar in size to the Summa 15" woofer.
The 2-way Lansing Iconic dates back to the late Thirties, has drivers with similar radiating areas, and has similar crossover frequencies. The Altec Model 19, from the Seventies, is another variation on the 2-way medium-sized horn/waveguide and 15" direct-radiator bass-driver theme. It's a good concept that's been around for a long time.
P.S. Bjorn, congratulations on your new degree! You worked long and hard on that one!
Last edited:
The measurements shown here by the AH425 users at different distances showed clear differences in the low frequency loading and response variations at different distances. Have you tried the horn crossed over at a higher frequency? I'm asking because maybe it would be easier to finalize your BTA speaker if you can actually get it to sound right.
The iconic used a multicell horn of a lower flare rate and much larger size and crossed at 800 Hz. The model 19 crossed at 1200 cycles. Your horn is actually small when the 2-3" outer lip ring is taken away - Maybe 11-12" diameter mouth - nothing wrong with that provided you don't run it too low. if it works for you than why is it such a lengthy process to finalize?
The Gedlee concept is not at all like you propose. They do not look at the mid horn as an acoustic loading device, acoustic transformer, they look at as a way of controlling the radiation pattern of the compression driver. If you look at the raw measurment of the driver and waveguide you will see the Gedlee speaker requires a broadband filter to get the response inline so it's listenable.
The iconic used a multicell horn of a lower flare rate and much larger size and crossed at 800 Hz. The model 19 crossed at 1200 cycles. Your horn is actually small when the 2-3" outer lip ring is taken away - Maybe 11-12" diameter mouth - nothing wrong with that provided you don't run it too low. if it works for you than why is it such a lengthy process to finalize?
The Gedlee concept is not at all like you propose. They do not look at the mid horn as an acoustic loading device, acoustic transformer, they look at as a way of controlling the radiation pattern of the compression driver. If you look at the raw measurment of the driver and waveguide you will see the Gedlee speaker requires a broadband filter to get the response inline so it's listenable.
Pooh, If you consider the on axis and the off axis response below the frequency the horn or waveguide is controlling directivity and add both together you see that there really is not this sudden change at what is called the cutoff frequency. The total output stays very consistent but it just isn't going where you want it to so we say it is falling in output which isn't truly correct.
I'm not saying that I'm saying the response fall off due to the horn not loading the driver in the bottom of it's range and it dosen't show up in close up measurements. Virtually all midrange horns don't control directivety below 1200 hz (surely not this little AH425) and I see the direction you want to take this but it is not what i am talking about. I'm saying why crossover so low if you get better performance crossing it higher?
Pooh,
I take it that Lynn and his co-designers are attempting to use the 15" as high as they think they can go and not get into the rough area of the 15". I don't see any great solution besides a midrange horn or third smaller driver to cover that range. That seems to be a no go in their minds so this is the situation that they have to deal with and make the compromise that works the best in this situation. You and I would use a midrange but that is just us I imagine.
I take it that Lynn and his co-designers are attempting to use the 15" as high as they think they can go and not get into the rough area of the 15". I don't see any great solution besides a midrange horn or third smaller driver to cover that range. That seems to be a no go in their minds so this is the situation that they have to deal with and make the compromise that works the best in this situation. You and I would use a midrange but that is just us I imagine.
I'm not seeing the BTA designs with a single 15" driver though! I see multiple 15.. using a good 8-12" in place of one of the dual 15's should make the crossover transition and overall design better performing - they don't have to horn load the low midrange driver but that's a completely different story and may require years of debate in this thread so i am trying to stay away from that.
Maybe forget the idea of the majestic 1940's 15" two ways and utilize the modern high performance midrange drivers with the modern high perfomance AH425? If not maybe consider copying an old altec design like the Great Plains Audio people do with their drivers? Then the old sound becomes new and better plus you can say it's proven? I don't know exactly what attracts me to participating in this thread but I do.
I know that sound well. Still having a set of older Altec Barcelona speakers sitting in a large family room, a 15" and horn. I can see the nostalgia for those designs but once I became aware of the midrange problems they never sounded the same again. Can't seem to unlearn that sound and that it just isn't quite right.
"Could" is a broad word, but I am talking about directivity and Webster can't give us any information about that. It "can" give reasonable results at LFs for the acoustic impedance, however, studies have shown that its results are most in error right at "cutoff".
But at these lower frequencies we are losing control of directivity, anyway.
Also, am I correct in stating that Webster's equations at cutoff do not present as good of a result, in theory, as you achieve in reality?
Pooge,
When you think about it almost all horns are typically 1/4 wavelength at the so called cut-off frequency. So no matter the shape, hyperbolic or conic at the lower frequencies you have no real control of the dispersion, it becomes somewhat an omni-directional radiation pattern. All this talk of controlling dispersion is poppy-**** at those low frequencies. Once you have a wave that no longer is contained within the boundaries of the horn or waveguide you essentially have lost pattern control. I day say look at the polar plots for any type of horn shape and you will see this. Now the conic sectional horns will have the most consistent pattern up to a set point and then it all falls apart, there is only one way to totally control that and that is an infinite length horn and that is only something that is theoretical anyway. I don't know anyone using full length horns for low frequency control even outdoors, they would just be ridiculous in size, save the crazies with concrete horns built into a home or outdoor application.
When you think about it almost all horns are typically 1/4 wavelength at the so called cut-off frequency. So no matter the shape, hyperbolic or conic at the lower frequencies you have no real control of the dispersion, it becomes somewhat an omni-directional radiation pattern. All this talk of controlling dispersion is poppy-**** at those low frequencies. Once you have a wave that no longer is contained within the boundaries of the horn or waveguide you essentially have lost pattern control. I day say look at the polar plots for any type of horn shape and you will see this. Now the conic sectional horns will have the most consistent pattern up to a set point and then it all falls apart, there is only one way to totally control that and that is an infinite length horn and that is only something that is theoretical anyway. I don't know anyone using full length horns for low frequency control even outdoors, they would just be ridiculous in size, save the crazies with concrete horns built into a home or outdoor application.
Yes. My point is that why criticize the Webster equations as being inferior when discussing this range. At the low end, resistive load is key, whether or not a horn or baffled driver. While an argument could be made that you could overcome a low resistive load with amplifier current and shear displacement, the question is whether or not it is an equally satisfying sonic character. Speaking for many people that have heard a good (properly construction) bass horn, I don't think it is.
Hi Pooge,
A proper measurement shows you what a given horn does but a couple thumb rules can tell you what ball park you are playing in beforehand.
For a simple horn, one can find the approximate point the pattern control is lost given it’s mouth size and mouth wall angle here;
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1975-05 AES Preprint) - Whats So Sacred Exp Horns.pdf
Also explained is the why of the second horn angle at the mouth and the cause of pattern flip when the Vertical and Horizontal angles are incompatible with the mouth shape .
Best,
Tom
But my focus is NOT pattern control here. It is the loading at the low end where Webster's equations were criticized as useless for showing directivy, in an area where loading is a more important than the use of OS coordinates at higher frequencies. No one is going to properly cross over a horn at a frequency high enough to where the polars collapse.
Pooge,
Ah, pattern control happens at a much larger dimension that the point where the radiation resistance stops changing.
Why bother with those equations at all?
These describe the simplest horns with ideal conditions, while they were a breakthrough of understanding in the day and are still referred to because of that, not because they are of much help designing a speaker that you will build (if that makes sense).
Instead find /read Wm Leach’s paper on low frequency horn loading and if you can work the math (It took me months to program my commodore 64 to run it in “commodore Basic” back then but when it ran and worked, I literally had tears in my eyes) , one can find the optimum driver parameters for a given horn or the optimum horn parameters for a given driver and bandwidth.
Then use a program like AKABAK or that horn response program to fine tune those things based on the dimensions of a finite, real horn.
AKABAK is a pain to learn but lets you investigate things with its tools.
I am not sure what you mean about polar’s collapse, polar’s are governed by the same part of the horn that is described in Don’s paper, it’s just that up high, that part is much closer to the throat and if the horns wall angles were narrow there, so is the pattern.
This was the down side of the exponential shaped horn where a constant beam width was desired. Up high they get narrow.
How low are you talking about and by loading do you mean resulting gain over the same driver, same voltage as a direct radiator?
Tom