Is it possible to cover the whole spectrum, high SPL, low distortion with a 2-way?

In a cabinet of the right volume and bass reflex, I'd seriously doubt that ^

So did I, initially.
If you do some research, you'll find several threads in this and other forums, in which both drivers are discussed by people who've actually used both the TD15H and TD18H+.
Much to my surprise, the TD18H+ is generally regarded as a woofer and not a true subwoofer. Some people have used it up to 700Hz.
Sure, in a big enough cab it'll do <30Hz, which would suffice in most applications for which this woofer was designed.

If infrabass is the objective, the AE IB18HT-8 is the best option.
This beast will do 12.3 Hz -3dB in room, in a 1303.3L BR cab with Bessel alignment.
 
Last edited:
If cab size isn't important, but effiency is, the TD18H+ is the better option compared to the TD15H.
Camplo knew this already.

I have to water down my earlier comment somewhat. The TD15H is only a better option, if you don't care about efficiency, and at the same time want to use a cab of reasonable proportions.
 
Last edited:
Aren't you concerned that might help maintain the wavefront at the edges and allow it to widen, rather than dissipating it? (At the least a combination may serve)

Widening of the wavefront is not the purpose of an OSWG. I guess minimizing/elimination the narrowing, the on-axis null and interferences at the mouth were areas of concern.
Dr. Geddes has opted for an asymetrical mouth, whereas mabat optimized the throat to mouth transition.
Please, correct me if I am wrong.
 
Last edited:
diyAudio Moderator
Joined 2008
Paid Member
Look at LeCleach for a moment. It is a large radius moving smaller with distance. The diffraction effects a reconstruction of the wavefront that suits the widening. There becomes no better termination than the horn itself, and the room walls are going to be flooded.

Instead would it be better to establish the wavefront at all frequencies in a waveguide by making it large enough, then do something else, like turning the edge diffraction into entropy so it dissipates.

minimizing/elimination the narrowing
Can we take a small lesson from hyp/ex (higher T but less than conical).

By the way, I'm not certain that your reference to an asymmetrical mouth is entirely in context here?
 
Last edited:
You guys are alright by me....

(upload killed my res =(
This is with the flattened response checked and everything at ~116db
116 is pretty loud right? The td15m(yellow) makes it to 31hz before it goes south, xmax wise. The 18h+ is like the 18" version of the 15M but with more xmax, and the 15H actually leaves xmax 36hz to 25hz...Isn't leaving the xmax a bad thing ?
The 15H (4.607cbft) requires a larger box than the 18H (2.923cbft) according to winisd btw...(Ported)

Again, 'horses for courses'; you ideally want the Fs to be at/below whatever the lowest frequency it's likely to need to reproduce at some useful SPL. With a 29 Hz Fs I don't see how they can legitimately list it as suitable for HT, which always been referenced to 20 Hz nor low enough for a modern studio monitoring app, but if your music doesn't go below the lowest note on a piano [27.5 Hz] it's low enough.

Very potent comment, I get the feeling that the 18h+ is beating the 15H at everything above the 18H+'s fs.....and 30hz is about the lowest 808 I hear in music that I've analyzed. So though I want to take on mix/mastering projects geared towards cinema, someday I think, if I extend the thought, and add 2 certified subs next year to execute the multisub approach, I'll be in there...Good approach?
 
Last edited:
Sure, in a big enough cab it'll do <30Hz, which would suffice in most applications for which this woofer was designed.

If infrabass is the objective, the AE IB18HT-8 is the best option.
This beast will do 12.3 Hz -3dB in room, in a 1303.3L BR cab with Bessel alignment.

I think the cab. volume/vent freq. used by others were the limiting factors. ;)

..You could improve this further with an LCR for the driver's resonance (..with respect to subjective result).

The IB will certainly do louder lower, but I'm not sure that it would do so with the same clarity. That lower Qe for the "woofer" tend's to produce a very "tight"/"defined" result. :)
 
Last edited:
Widening of the wavefront is not the purpose of an OSWG. I guess minimizing/elimination the narrowing, the on-axis null and interferences at the mouth were areas of concern.
Dr. Geddes has opted for an asymetrical mouth, whereas mabat optimized the throat to mouth transition.
Please, correct me if I am wrong.

It is the need for smooth termination and I suppose that holds for every horn. There should not be an edge, IMO. Then we can talk about the degree of edginess. Curve such as clothoid offers virtually none - you won't find any order of "edge" there. There will still be a mismatch of curvatures at a (flat) baffle junction however. Ideally the baffle would match curvature around the horn as well, although I'm still not sure that unbaffled but rolled-back horn would be better. At least for an infinite baffle this final curvature mismatch doesn't seem to make much trouble. "Inside" the horn it will be much more important.
 
Last edited:
Pretty much, that's what it's designed for.

'808' ?

Unless it's changed, full size cinema uses a 120 Hz XO Vs HT's 80 Hz, so ~30 Hz Fs driver is fine, ditto adding true subs, but when I responded earlier I was just answering Qs in auto mode.

But seems I saw where you're planning to use a 700 Hz XO on a driver best suited to ~ 500 Hz and maybe only ~350 Hz if planning on using all its Xmax.

I've never had the luxury of > ~ 4 mm, so frequency modulation distortion [FMD] might be an audible issue, i.e. the cone/surround distortion Scott mentioned, especially in a quiet setup like you plan.

GM
 
GM did you just use sarcasm....lol.
I say 630hz but if I try 500hz and like it, I won't likely change it....you have to think, I really want the horn to play to 200hz, so of course I am going to push the limits or at least try. At the 70-90db of normal listening the xmax is about non existent. When I'm modeling at 116db, the sealed 15M hits 3xmax at 90hz, so to follow the 3xmax best practice for critical listening, I'd cross no lower, I thought 100hz made sense.

HXDXW
84x14x23

I intend to put the mid point between the horn and mid woofer at ear level which is 48" (sitting in my fav chair) and the width of the front baffle matches the 23" of the horn, and the depth has been minimalized, causing less resonance on the most important plane.
OMG is this really coming together, shipping as soon as next week you guys, then I'll stare at the drivers for another month before enclosures get done lol!

GM - yeah 808....you know, the bass drum sound from the roland 808, the booom booom you heard and just about every rap song ever?



So here's a thing, can you guys name off some reference material you like to use for home theater system appraisal??
 
Last edited:
I stated that wrong above. No linear function of x could become infinite in a finite x, so it would likely be something like a Tan function, but still pretty simple.

But I agree that this kind of waveguide termination makes a lot more sense than a simple radius - which has an instantaneous change in slope at the junction. I wish that I had thought of this before.

Such a modification of the horn contour is at least known since more than 50 years. Manfred Harsdorff, Zeiss Ikon, filed a patent about the modification of the exponential horn and imho in a more coherent way by modifying the original horn formula to receive a smooth termination into the mounting plane and to shorten the horn. He also provided a figure of the resulting impedance curve.
 
It's exactly a clothoid (aka Euler spiral), i.e. a curve that has linearly increasing curvature along its length. It is actually quite hard to compute with :)

Euler spiral - Wikipedia

I believe there are other options of suitable curves - even a whole classes of splines. I chosed clothoid. The first one who suggested it to me was Dr. Bohumil Sýkora, Czech audio guru (I think he has an account here as bobolix). It was the BEM simulation that helped me to realize how much of the horn contour this curve can really take.

It is actually quite easy to compute the Cornu spiral. For this you need to solve the Fresnel integrals which can be solved by a simple series expansion, so a single loop in the program code. The problem with this series expansion is that it tends to diverge for larger angles but I have found that this does not play a role for this use case as the determining function angle remains in a region where the series expansion converges very fast.

I am curious about the observation field view (wave fronts) of your big wave guide you presented from lowest to the highest frequencies. Can your wave guide be used as free-standing? It seems that the simulation uses the infinite baffle trick which hides many of diffraction issues, so we are able to assess this.

1) At the point of inflection the radius of curvature of both curves should match.
2) You can implement this profile, and others as well, with a single fairness curve. [1]
3) Check out the profiles generated by shape optimization regimens. [2]
4) The declining curvature in the throat, gently spreads a higher pressure wave front.
The increasing curvature at the month mitigates lower pressure wave reflectance.
WHG

Thanks for sharing this! The Bezier Curves look really promising for some uses I have in mind but I need to dig into this bunch f formulas :D
 
I also looked at how to incorporate the termination flares into one equation for the whole countour (and I'm talking about OSWG) but is it not really nedeed - it's only a matter of convenience. For me the OS contour is a must as I can't stand beaming sources anymore - the collapse of stereo image, etc. Once you get used to a properly done constant and narrow directivity, there's no way back. So the question for me is how to terminate OSWG the best possible way and I see no much better way than how I do it now. That's all.

I am curious about the observation field view (wave fronts) of your big wave guide you presented from lowest to the highest frequencies. Can your wave guide be used as free-standing? It seems that the simulation uses the infinite baffle trick which hides many of diffraction issues, so we are able to assess this.
It is not a "trick", it is an examination of the waveguide for that conditions, i.e. without the influence of a particular box. It's allways better to start with a better waveguide itself.

Of course you can use it free-standing and it will get much worse. No point about that. You could make a roll-back then, i.e. to let the spiral continue (and maybe even change the scale around the circumference which is quite easy to do). I would have to prepare a complete 3D simulation for that. I have it on my list but not quite on the top.
 
Last edited: