Would the curves in this be counterproductive? If so, why?
Thanks,
Eric
An externally hosted image should be here but it was not working when we last tested it.
Bluntly, yes. Therein is the difference between a back loaded horn and a front loaded horn. The baseline physics is the same but the objectives and application thereof are significantly different in the majority of cases.
Note that as GM points out, all horns are technically 1/2 wave types, but 99.9% of back-loaded horns have their expansion heavily compromised in order to reduce the proportions, resulting in a significant disparity between the 1/4 wave fundamental F0 and the impedance-matched frequency Fc, with resonant action supplying the gain across the F0 - Fc BW, with the usual caveats of harmonic pipe modes appropriate to the length & expansion.
Back horns are, as also noted above, only useful over a limited BW, with a practical ROT maximum being ~250Hz - 300Hz above which you ideally need to have a 2nd - 3rd order acoustic low-pass to prevent excessive group delay. The ideal is to have the bass horn supplying gain to the system's mass-corner frequency Fh (take as 2F0 / system Q including any series R and / or the amplifier output impedance). If the practical mass-corner is > 300Hz, then you either need to have a short front loaded midrange horn to supply additional gain between the horn's upper corner frequency & the system mass-corner, or accept a hole in the response in that region, the width & depth of which varies depending on the specifics of the implementation.
Smooth curves in a horn promote maximum efficiency across the full BW the basic horn design (inc. the low pass chamber volume & dimensions) possesses. However: in most back horns this causes more problems than advantages. Keep the above in mind: back-horns are only useful over a limited BW, above which you need to roll them off to avoid excessive GD. Low frequencies (which you want & are using the horn to produce / amplify) have too long a wavelength to be significantly troubled by bends & sharp edges in the horn. As frequency rises & wavelengths shorten, they are increasingly affected by the presence of bends, hard edges & the diffraction associated with them etc. Which, if you design it right, is inherently a good thing. Smooth curves basically promote efficiency at the top end of the horn's BW -right where you least want it / where it's least advantageous, because it's in that region you want to have your acoustic low pass. So, while smooth curves do look very pretty, the acoustic reasoning they're usually based on is essentially flawed, because while it's absolutely valid (vital) for most front-loaded horns, especially midrange or HF types, back-horns, especially those compromised in size to the ideal (almost all), have a different set of operating criteria.
Re Martin's excellent pages -what GM said. Required reading. Re his compact horn design, it's a really nice example & should work very well. Essentially he's used the internal Helmholtz resonator to create an acoustical Cauer (notched) low-pass filter. You sometimes see these internal Helmholtz resonators in other speaker types, but not often. Onur's 'Singular' large QW pipe is probably the best known example.
Note that as GM points out, all horns are technically 1/2 wave types, but 99.9% of back-loaded horns have their expansion heavily compromised in order to reduce the proportions, resulting in a significant disparity between the 1/4 wave fundamental F0 and the impedance-matched frequency Fc, with resonant action supplying the gain across the F0 - Fc BW, with the usual caveats of harmonic pipe modes appropriate to the length & expansion.
Back horns are, as also noted above, only useful over a limited BW, with a practical ROT maximum being ~250Hz - 300Hz above which you ideally need to have a 2nd - 3rd order acoustic low-pass to prevent excessive group delay. The ideal is to have the bass horn supplying gain to the system's mass-corner frequency Fh (take as 2F0 / system Q including any series R and / or the amplifier output impedance). If the practical mass-corner is > 300Hz, then you either need to have a short front loaded midrange horn to supply additional gain between the horn's upper corner frequency & the system mass-corner, or accept a hole in the response in that region, the width & depth of which varies depending on the specifics of the implementation.
Smooth curves in a horn promote maximum efficiency across the full BW the basic horn design (inc. the low pass chamber volume & dimensions) possesses. However: in most back horns this causes more problems than advantages. Keep the above in mind: back-horns are only useful over a limited BW, above which you need to roll them off to avoid excessive GD. Low frequencies (which you want & are using the horn to produce / amplify) have too long a wavelength to be significantly troubled by bends & sharp edges in the horn. As frequency rises & wavelengths shorten, they are increasingly affected by the presence of bends, hard edges & the diffraction associated with them etc. Which, if you design it right, is inherently a good thing. Smooth curves basically promote efficiency at the top end of the horn's BW -right where you least want it / where it's least advantageous, because it's in that region you want to have your acoustic low pass. So, while smooth curves do look very pretty, the acoustic reasoning they're usually based on is essentially flawed, because while it's absolutely valid (vital) for most front-loaded horns, especially midrange or HF types, back-horns, especially those compromised in size to the ideal (almost all), have a different set of operating criteria.
Re Martin's excellent pages -what GM said. Required reading. Re his compact horn design, it's a really nice example & should work very well. Essentially he's used the internal Helmholtz resonator to create an acoustical Cauer (notched) low-pass filter. You sometimes see these internal Helmholtz resonators in other speaker types, but not often. Onur's 'Singular' large QW pipe is probably the best known example.
Scottmoose,
Thanks, I actually understand that, I think.
Re Martin's pages, I actually found them before I found this forum, but since I was interested in BLHs I went straight there without reading the earlier stuff on TLs etc, where I probably should have started.
Also, your post gave me a new perspective regarding the use of Quarter wave or TL or BL enclosures with a single FR driver. Please tell me if this is essentially correct or not:
In such a system, we are effectively trying to create a two-way system, where the exposed front of the driver handles the HF range, and the back side of the driver, combined with the enclosure, produces the LF range. And the real challenge is to design the enclosure in such a way as to enhance the LF range but without introducing undesirable artifacts (strong resonances, noticeable delay, etc).
I'm sure this is very oversimplified, but is this a fair statement or not?
Eric
Completely fair, and perfectly accurate. That's exactly what you're trying to do with any back loaded cabinet, be it a horn, resonant pipe, BR or whatever.
Hi,
I'm trying to stop you overcomplicating things due to a
desire to make a "beautiful" complex curved cabinet.
I completely expect the MJK BLH design to better the BK12m
using the same Fostex FE126En driver, and work as claimed.
One of its major advantages is bass tuning via its corner placement.
Sometimes keeping it relatively simple build wise is a good idea *.
rgds, sreten.
* Having said that there may be some mileage in converting the MJK design into
tapered sections, easy enough, however he has not modelled it, as he states.
Last edited:
So would I. I designed the BK12m to be a very simple forward-firing scoop-type BLH enclosure that makes a convenient kit. A more refined box like Martin's or various others will have a sonic edge assuming, like all boundary loaded cabinets, that the wall construction is solid enough to provide decent loading / reflectivity without leaching LF frequencies.
That it is.
I wouldn't bet my last shilling on it, for the reasons I mention above. You might (might) get a touch more gain & a little smoothing, but with highly BW limited boxes it's generally marginal & needs to be set off against other factors. Win some, lose some. Constant expansion usually ends up needing a larger box to fit everything in too, which can be a consideration. YMMV as always of course.
- Status
- Not open for further replies.